ISSN 0024-3922 LINGUISTICA XLVI Ljubljana, 2006 LINGUISTICA XLVI IO ANNI OREŠNIK SEPTUAGENARIO IN HONOREM OBLATA II Ljubljana, 2006 Revijo sta ustanovila t Stanko Škerlj in tMilan Grošelj Revue fondee par fStanko Škerlj et tMilan Grošelj Zbornik so uredili - Melanges rediges par Mitja Skubic - Pavao Tekavčić - Martina Ožbot - Stojan Bračič Natis letnika je omogočila JAVNA AGENCIJA ZA RAZISKOVALNO DEJAVNOST REPUBLIKE SLOVENIJE Sous les auspices de 1AGENCE NATIONALE SLOVENE POUR LA RECHERCHE SCIENTIFIQUE Hans Goebl Universite de Salzbourg UDK 811.133.1"15/19" : 81-1123 SUR LE CHANGEMENT MACROLINGUISTIQUE SURVENU ENTRE 1300 ET 1900 DANS LE DOMAINE D'OIL. UNE ETUDE DIACHRONIQUE D'INSPIRATION DIALECTOMETRIQUE 1. Remarques preliminaires La finalite de la presente etude est essentiellement d'ordre diachronique. II s'agit d'evaluer l'ampleur des modifications survenues entre 1300 et 1900, et partant dans un laps de temps de six siecles, sur le plan de l'amenagement linguistique1 de l'espace du domaine d'Oil ä proprement parier (c'est-ä-dire ä l'exclusion du domaine francoproven-Qal). Evidemment, la precision de nos constatations en la matiere depend avant tout de la qualite des sources empiriques ä notre disposition, ä savoir de Гatlas scripturaire d'Anthonij Dees d'un cote (valable, toute somme faite, pour le dernier tiers du 13e siecle), et de l'ALF de Jules Gillieron et d'Edmond Edmont de l'autre, dont les donnees ont ete collectees entre 1897 et 1901. Bien que les deux sources utilisees par nous, disposent de la meme structuration formelle en N points d'atlas et p planches (ou cartes) d'atlas - qui est d'ailleurs celle de tous les atlas linguistiques - il se creuse neanmoins, entre elles, une difference tout autre qu'insignifiante, ä savoir celle qui existe entre 1'ecrit (Dees) et l'oral (ALF). II faut y ajouter la question de la fiabilite documentaire des materiaux (editions de textes non litteraires) depouilles par A. Dees et son equipe. Comme, dans cette etude, il s'agit aussi d'une comparaison basee sur des donnees de masse et, par consequent, sujette ä la loi des grands nombres, il est egalement indique de se demander dans quelle mesure, dans la masse des donnees analysees par A. Dees, le nombre des transcriptions correctes (surtout quand il s'agit de graphies micro-regionales ou carrement locales) depasse celui des erreurs de lecture ou de transcription. Cependant, les resultats de l'analyse globale des donnees de l'atlas de Dees sont plu-tot rassurants et suggerent que leur fiabilite est suffisamment grande pour une analyse quantitative globalisante. Ceci dit, il va de soi que, pour une analyse qualitative et philologique au sens strict du terme, les donnees de l'atlas de Dees pourraient se presenter sous un autre jour: cf. ä ce sujet Völker 2003, 57-66. Mis ä part son caractere eminemment diachronique, cette etude represente aussi une contribution ä la linguistique quantitative et, plus particulierement, ä la dialectometrie. Ce sont lä deux domaines fortement impregnes de rigueur methodique et de precision 1 II s'agit d'un terme technique que nous utilisons depuis un certain temps. Celui-ci sert ä designer Taction semiotique qu'exercent les humains sur l'espace par l'utilisation conjointe de leur faculte langagiere pour la communication d'un cote et la demarcation de l'autre. Ce faisant, nous reprenons une idee chere ä Jean Sćguy (cf. Seguy 1973, passim). terminologique. C'est pourquoi la lecture des pages qui suivent ne sera pas facile et leur comprehension immediate ne sera pas garantie d'entree de jeu. Nous esperons pourtant que la qualite iconique des 5 figures et des 24 cartes ä couleur (toutes en annexe) rendra la täche de lecture et de comprehension moins ardue ä nos lecteurs. 1.1. L'«Atlas des formes et des constructions des chartes fran^aises du 13e siecle» d'Anthonij Dees (1980) L'atlas de Dees repose sur le depouillement de quelque 3300 chartes oiliques du 13e siecle, toutes publiees et dont la localisation a pu se faire avec suffisamment de precision. Elles appartiennent, du point de vue textologique, ä la categorie des documents non lit-teraires (cf. Dees 1980, X)2. Quant ä leur provenance regionale (ou «localisation»), Dees les a repertoriees dans 85 unites spatiales dont 28 figurent aussi sur les 282 planches de l'atlas publie. Ces 28 unites spatiales peuvent etre considerees comme des centres scrip-turaires «majeurs»« (voir la colonne Macro (Ma) sur la figure 1) alors que les autres en constituent des subdivisions diatopiques et ne represented, de ce fait, que des centres scripturaires «mineurs» (voir la colonne Micro (Mi) sur la figure 1): cf. Dees 1980,307312 et la liste respective presentee ici en appendice (figure 1). Les 3300 chartes prises en consideration ont ete analysees ä 1'aide d'un questionnaire phonetico-graphique et morphologique dont les items correspondent exactement aux en-tetes des 282 cartes de l'ouvrage publie (Dees 1980). On en trouve une liste detaillee aussi ä la fin de l'atlas de Dees («Inventaire des formes trouvees», 322-369). Une autre liste («Formes et constructions etudiees», ibidem 313-321) renseigne sur la structuration interne de l'atlas dedie, aux termes de son titre, ä l'etude «des formes et des constructions» de la scripta non litteraire du 13e siecle. Pour 1'analyse de la variabilite graphique bien connue des documents medievaux, Dees a decide de saisir, pour chaque item, les «alternances» entre deux groupes de graphics (ou de constructions morphologiques) concurrents. Evidemment, la definition de ces «alternances» ne saurait jamais etre absolue et depend de certaines presuppositions individuelles. Ceci implique qu'un autre chercheur, dans une situation analytique simi-laire, aurait pu adopter une solution differente3. En voici un exemple relatif ä la carte 7 de l'atlas de Dees. II s'agit du pronom personnel «moi» et de l'alternance entre la forme mi d'un cote et les formes me i, mey, moi et may de l'autre. Les formes en mi constituent, dans l'atlas de Dees, le groupe 1 alors que les autres formes sont rangees dans le groupe 2. Seuls les scores du groupe 1, moyennant six paliers ä hachures en noir-et-blanc, sont T Signaions encore l'existence d'un autre atlas scripturaire de Dees, publie en 1987 et relatif aux textes litteraires des 12e et 13e siecles, dont, malheureusement, les fichiers electroniques originaux utilises au cours de la publication, ne subsistent plus. Sa valeur documentaire a 6te mise en doute par beaucoup de specialistes de la tradition textologique de la litterature medievale franjaise: cf. ä ce sujet le rapport detaille de völker 2003, 57-66. Neanmoins, je persiste ä croire qu'une analyse dialectometrique de cet atlas, comparable ä Celle que nous presentons dans cet article, sera egalement fort interessante: cf. aussi Goebl 1993. 3 Dees est d'ailleurs conscient de ce fait lui-meme. Quant aux variantes «dichotomisees» de Particle contracte «du», il s'exprime de la maniere suivante: «D'autres dichotomisations de la meme serie des variantes de «du» auraient pu etre envisagees; en general la dichotomie qui a ete finalement retenue (parfois apres de longs täton-nements) est celle qui donne le meilleur resultat cartographique en termes de coherence des aires et qui per-mettra souvent de se faire une idee de l'histoire des formes etudiees.» (Dees 1980, XI). mis en carte alors qu'une mise en carte similaire n'a pas ete faite pour les scores du grou-pe 2: voir aussi la figure 3 (en annexe). Quant au bilan numerique des deux groupes en question, l'on en trouve, en bas des 282 cartes de l'atlas de Dees, les listes avec les scores respectifs (toujours en chiffres absolus et relatifs), si bien que des lecteurs assoiffes de statistiques peuvent controler la correction de tous les calculs effectues et aussi de la mise en carte des graphies apparte-nant au groupe 1. En ce qui concerne la mensuration de la variation scripturaire inherente ä son corpus, Dees a choisi de retenir, pour les chartes de chacun des 85 centres scripturaires considered, le pourcentage de celles qui disposent de graphies appartenant ou bien au groupe 1 ou bien au groupe 24. En voici un exemple de calcul relatif ä la carte 1, dediee au pro-nom personnel »je«, et au centre scripturaire (majeur) 01 (Charente, Charente-Maritime). La liste sur la «provenance des chartes» (Dees 1980,307-312) nous dit que cette macro-unite territoriale dispose d'un effectif total de 118 chartes. La moitie de gauche de la legende numerique au pied de la carte 1, nous renseigne que, de ces 118 chartes, 111 peuvent etre utilisees selon la logique de l'«alternance» sur laquelle repose la carte 1: groupe 1 (visualise): g,ge, gie, etc. groupe 2 (non visualise): i, ie, iou,je,jo, etc. Or, 1'introspection de ces 111 chartes a montre que l'ecrasante majorite - ä savoir 105,245 - fournissent des graphies en g, ge, gie, etc. (groupe 1) alors que le reste - c'est-a-dire 5,76 chartes - contiennent des graphies retenues comme «alternatives» (ou com-plementaires) en i, ie, iou,je,jo, etc. La mise en pourcentage successive fournit les resultats suivants: 105,24 : 111 = 94,81% et 5,76 : 111= 5,19%. On trouve ces valeurs dans la moitie de droite de la legende numerique de la carte 1. Or, ce sont elles que nous avons utilisees pour l'etablissement de notre matrice de donnees dialectometrique (voir la figure 3). Du point de vue quantitatif, l'atlas de Dees constitue done, tel qu'il se presente sous sa forme publiee, une matrice bidimensionnelle avec 28 entites territoriales et 282 attribute scripturaires dont les scores occupent le niveau de l'echelle metrique. Cet agence-ment formel le predestine, en quelque sorte, ä etre depouille par voie dialectometrique6. Precisons que la matrice de donnees utilisee dans l'atelier informatique d'A. Dees ä Amsterdam (et dont la direction etait assuree, jadis, par Pieter van Reenen), disposait, 4 «Ce probleme a ete resolu dans ce sens que nous avons systematiquement ramene les variations ä etudier ä des dichotomies. Cette dichotomisation, qui consiste ä repartir l'ensemble des formes rencontrees sur deux classes, dont la premiere est caracterisće par une ou plusieurs marques que la deuxieme ne comporte pas, peut etre illus-tree par Г exemple relativement simple des variantes de «tous», comprenant entre autres les graphies tous, touz, tos, toz. Ces quatre graphies se ramenent ä deux alternances, ä savoir -ou-: -o- et -i-: -z-.» (Dees 1980, XI). 5 Precisons que les effectifs non entiers (105,24 et non pas 105 ou 106) s'expliquent par la presence conjointe, dans une Charte donnee, de graphies appartenant aux deux groupes mis en contraste: «Si la chaite comporte des variantes appartenant ä l'une et ä l'autre de deux classes opposees, le temoignage est divise proportionellement aux frequences absolues des deux variantes dans ce document.» (Dees 1980, XII). 6 Dans notre compte rendu de l'atlas de Dees, nous avons dejä evoque l'utilite d'une dialectometrisation de ce dernier (Goebl 1982, 283). du cote des entites territoriales, non seulement de 28, mais bien plutöt de 85 unites spa-tiales qui, evidemment, couvraient l'espace du domaine d'Oil d'une fagon beaucoup plus detaillee; voir ä ce sujet, une fois de plus, la figure 1 (en annexe) Ce n'est qu'en 1996 que la coincidence de deux evenements propices a fini par dega-ger la voie pour une veritable «dialectometrisation» des donnees de l'atlas de Dees: 1) la remise des fichiers electroniques de l'atlas de Dees par Pieter van Reenen ä l'auteur de cet article, 2) l'arrivee d'un nouveau collaborateur dialectometrique ä Salzbourg, Guillaume Schutz7. Evidemment, nos activites dialectometriques ne se sont pas arretees lä et ont continue ä se developper dans deux directions: a) ä partir de 1998, par la creation d'un nouveau logiciel dialectometrique ultra-puissant appele VDM («Visual DialectoMetry») de la part d'Edgar Haimerl, b) ä partir de 1997, par la dialectometrisation de presque la moitie des 1421 planches originales de l'ALF. Les analyses qui suivent, reposent done, du point de vue informatique (VDM) et documentaire (disponibilite electronique conjointe des donnees de l'atlas de Dees et de l'ALF), sur des antecedents plutot complexes et echelonnes sur un laps de temps relati-vement long (1996-2005). Precisons en outre qu'une des täches de Guillaume Schiltz consistait ä eliminer du corpus original de Dees les cartes jugees trop lacunaires. C'est qu'en matiere de taxo-metrie, la presence exageree de lacunes (ou de «donnees manquantes») dans les donnees prises en consideration, cree inevitablement, dans les resultats finaux, des distorsions numeriques fort desagreables. II fallait done se defaire des cartes les plus lacunaires. C'est ainsi que Guillaume Schiltz a fini par retenir 203 des 283 cartes en question qui, quant ä leur appartenance aux differentes categories linguistiques, se repartissent comme suit: Categorie linguistique sigle Nombre des cartes (attributs scripturaires) de l'atlas de Dees vocalisme VOC 119 consonantisme CONS 66 morphologie MORPH 18 corpus integral TOT 203 1.2. L'»Atlas linguistique de la France« (ALF) Cette etude, d'inspiration comparative par excellence, se compose de deux volets: le premier se refere ä l'atlas de Dees et le second ä l'ALF. Evidemment, nous entendons par «ALF» aussi la version dialectometrisee de l'ouvrage original teile qu'elle figure dans la matrice de donnees issues des travaux de «taxation» qui se sont deroules ä Salzbourg entre 1997 et 2000. La dialectometrisation de l'ALF - dont le premier bilan a n Cf. les deux contributions (Goebl/Schiltz 2001 et de Schiltz 1997), issues de cette cooperation. ete publie dans cette revue meme (cf. Goebl 2000) - avait comme but d'analyser (ou de «taxer») un nombre aussi grand que possible de cartes originales de l'ALF appartenant ä differentes categories linguistiques (lexique, vocalisme, consonantisme etc., avec leurs sous-categories respectives). Le travail de taxation avait ete realise par une equipe de plusieurs jeunes romanistes salzbourgeoises qui se sont dediees ä cette täche meticuleuse (et aussi harassante) pendant trois ans (1997-2000)8. C'est parallelement ä la saisie taxatoire des donnees de l'ALF que M. Haimerl a developpe le logiciel VDM d'oü la possibilite d'appliquer aux donnees taxees - aussitot qu'elles ont ete enfournees dans l'ordinateur - la gamme entie-re des calculs dialectometriques desormais facilement accessibles9. Pour une comparaison directe des donnees de Dees et de l'ALF, il fallait encore pro-ceder ä la reduction de la densite du maillage de l'ALF. Ce travail, ainsi que la confection de tous les graphiques de cet article, a ete realise par Slawomir Sobota, un autre collaborates fidele et competent du laboratoire dialectometrique de Salzbourg. La structuration interne10 des materiaux dialectometrises de l'ALF tels qu'ils ont ete utilises dans cette etude, se presente done de la fagon suivante: categorie linguistique sigle Nombre des cartes "de travail"11 tirees de l'ALF lexique LEX 471 vocalisme voc 612 consonantisme CONS 505 corpus total TOT 1687 La masse considerable de 1687 cartes de travail disponibles - issues de l'analyse taxatoire de 626 planches originales tirees par voie aleatoire de l'effectif total des 1421 cartes de l'ALF - confere ä nos analyses dialectometriques de l'ALF une tres grande fia-bilite statistique. Ceci vaut egalement pour les sous-categories du vocalisme et du consonantisme. En ce qui concerne la fiabilite statistique des donnees de 1'atlas de Dees par contre, la situation n'est pas aussi rassurante, certes, mais encore suffisamment solide pour permettre l'analyse comparative envisagee. 13. Conventions (carto)graphiques. Pour une bonne intelligence du message iconique (et aussi quantitatif) des 24 cartes de cet article, les lecteurs sont invites ä se familiariser, d'entree de jeu, avec les conventions (carto)graphiques utilisees par nous. Q ° II s'agit de Barbara Aigner, Irmgard Dautermann, Hildegund Eder, Susanne Oleinek et Annette Ida Schatzmann. Qu'elles rejoivent ici l'expression de notre profonde reconnaissance! y Pour d'autres details (techniques et logistiques) relatifs ä la dialectometrisation de l'ALF, cf. aussi Goebl 2002, 67, 2003, 61-62 et 2005. 334. 10 Nous ne parlons ici que des categories linguistiques prises en consideration dans cette etude. 1 Nous entendons par «carte de travail» le resultat d'une analyse taxatoire appliquee ä une planche originale de l'ALF. Du point de vue formel, chaque carte de travail correspond ä un vecteur d'attribut (horizontal) de la matrice de donnees (voir aussi la figure 5). 13.1. Preparation cartographique des reseaux de Dees et de l'ALF Pour les deux projets dialectometriques (Dees et ALF) en question, il y a un reseau unique. Celui-ci dispose de 85 unites spatiales (appelees aussi «points»). Quant aux points de l'atlas de Dees, on en trouve une liste dans la figure 1 (voir la colonne Micro (Mi)). Pour ce qui est du reseau de l'ALF par contre, nous avons du l'amputer de ses parties non fran9aises et non oi'liques d'abord et l'elaguer convena-blement par la suite. L'etape suivante consistait dans l'application de la geometrie de Delaunay-Voronoi (triangulation, construction des mediatrices des cotes de triangles respectifs, fusion de ces dernieres et construction des polygones de Voronoi (appeles aussi polygones de «Thiessen» ou de «Haag»)12. L'utilite du reseau polygonal ainsi constitue est double: il permet la confection tant de cartes choroplethes (c'est-ä-dore ä plages coloriees) que de cartes isarithmiques (c'est-ä-dire munies de cotes de polygone ä epaisseur et coloriage variables). 13.2. Mise en carte du bilan numerique des calculs dialectometriques Evidemment, la dialectometrie n'a pas pour but de se gargariser de chiffres, mais vise en premier lieu ä les visualisier en bonne et due forme pour les faire parier ensuite aux yeux des linguistes. Pour ce faire, elle se sert, depuis longtemps, de tous les gadgets de la cartographie statistique internationale parmi lesquels figurent, entre autres: a) 1'utilisation de la gamme (et de l'agencement naturel) des couleurs de l'arc-en-ciel (ou du spectre solaire) pour la visualisation claire et distincte de la variabilite numerique discutee, b) l'usage de 2 ä 8 paliers chromatiques pour le rendement iconique de la variabilite numerique consideree, c) le recours ä differents algorithmes d'intervallisation (ici: MINMWMAX13 et MEDMW14) pour la constitution «objective» de classes (ou paliers) iconiques dif-ferentes, d) la construction d'histogrammes (ici: toujours ä droite, en bas de page) pour la visualisation adequate des proprietes statistiques des differentes distributions de frequence considerees15. 12 Pour une presentation detaillee de ces procedures, cf. Goebl 1981, 363-365 et 1984 i, 90-92. 13 Les scores ä visualiser sont d'abord repartis des deux cotes de la moyenne arithmetique. Ensuite les ecarts entre le minimum et la moyenne arithmetique et le maximum et la moyenne sont divises - pour une visualisation en six classes - par trois. C'est ä l'aide des valeurs ainsi determinees qu'il est possible de definir, toujours des deux cötes de la moyenne arithmetique en question, les seuils numeriques (inferieurs et superieurs) des six classes respectives. 14 Les scores ä visualiser sont repartis des deux cotes de la moyenne arithmetique d'abord et repartis par la suite, separement des deux cotes de la moyenne arithmetique, dans des classes egales en nombre d'objets (elements, points d'atlas etc.). 15 Pour plus de details cf. Goebl 1981, 363 et 1984 1,97-98. Les courbes gaussiennes (dites aussi «normales») qui recouvrent les histogrammes, ont ete calculees ä partir de la moyenne arithmetique et de l'ecart-type de la distribution de frequence respective, et permettent d'en controler rapidement certaines proprietes statistiques (telle la symetrie, entre autres). Le choix du nombre des paliers iconiques et de l'algorithme de visualisation (facile-ment realisable ä l'aide du logiciel VDM) se fait toujours en fonction de la finalite explo-ratoire de 1'analyse en question. Les modalites de la visualisation doivent done toujours rester au service de la demarche scientifique du linguiste et l'aider dans le developpe-ment et la verification de ses vues theoriques sur le sujet de recherche en question. 1.4. De la matrice de donnees ä la matrice de similarite (ou de distance) Toute recherche dialectometrique est organisee sous la forme d'une chaine dont les differents maillons constituent autant de »rond-points« methodiques. Evidemment, la chaine dialectometrique ä proprement parier commence par l'etablissement de la matrice de donnees. Les donnees de cette derniere se situent, quant ä l'atlas de Dees, au niveau de l'echelle metrique alors que les donnees de la matrice de donnees de l'ALF occupent le niveau de l'echelle nominale (ou qualitative): voir la moitie de gauche de la figure 5. Cette difference d'ordre metrologique influe directement sur le choix des indices de similarite (de correlation ou de distance) dont on a besoin pour calculer la similarite (linguistique) reciproque entre les 85 «points» scripturaires (Dees) ou dialectaux (ALF). Or, les manuels de taxometrie modernes16 en offrent une panoplie tres nourrie dont nous nous sommes servis des le debut de nos travaux dialectometriques dans les annees 70 du siecle dernier. Pour 1'analyse dialectometrique des donnees de l'atlas de Dees, nous avons utilise les trois indices suivants: a) le coefficient de correlation (de Bravais et Pearson) [r(BP)jk]17, b) la »Similarite Euclidienne Moyenne« (SEMjk): voir la figure 4, c) la »Distance Euclidienne Moyenne« (DEMjk)18: voir la figure 4. La relation entre SEMjk et DEMjk se definit comme suit: SEMjk + DEMjk = 100. Pour ce qui est des donnees de l'ALF par contre, dont le niveau metrologique est nominal (ou qualitatif), nous nous sommes servis de trois indices couramment utilises en matiere de dialectometrie, ä savoir: a) de l'«Indice Relatif d'Identite» (IRIjk)19 qui, ä cause de son usage frequent, consti-tue, en quelque sorte, l'indice de similarite standard de la dialectometrie, puis b) de I'«Indice Relatif de Distance» (IRDjk) dont la relation face ä I'IRIjk est regie par la formule suivante: IRDjk + IRIjk = 100, et ,finalement, c) de l'«Indice Pondere d'Identite (avec le poids 1)» (IPI(l)jk)20 qui permet de capter mieux que ne le pourrait faire lTRIjk, des structures diatopiques «plates» enfouies dans la profondeur des donnees ä analyser. 16 Cf. Bock 1974 (en allemand), Chandon/Pinson 1981 (en frangais) et Sneath/Sokal 1973 (en anglais). 17 Cf. Chandon/Pinson 1981,65 ou Sneath/Sokal 1973,137-140. 18 Pour une description detaillee de la demjk cf. Bock 1974,36-39, Chandon/Pinson 1981,58-59 et Sneath/Sokal 1973, 121-125. 19 Cf. Goebl 1981, 357-361 et 1984 I, 75-78. 20 Cf. Goebl 1987,70-79. Pour la selection d'un de ces indices, il n'y a pas de criteres fixes ou preetablis. Leur finalite est comparable ä celle d'un outil dont on se sert pour achever tel ou tel ouvrage. Comme, en matiere de taxometrie (et, partant, aussi de dialectometrie), il s'agit de decou-vrir (et non pas d'inventer ou d'y projeter!), par voie quantitative, des structures cachees au fond des donnees analysees, l'utilite et le rendement taxometriques de tel indice s'ave-rent avant tout ä travers sa capacite de detecter, avec suffisamment de clarte, de telles structures de profondeur. Ceci signifie que, pour ce faire, le dialectometricien doit se livrer ä un certain nombre d' experiences qui, elles, sont d'ailleurs facilement realisables ä l'aide du logiciel VDM. C'est que celui-ci dispose, dans sa panopolie methodique, d'un grand nombre d'indices de similarite et de distance. Ceci n'empeche que, pour des raisons purement formelles (ou mathematiques), la nature metrologique des donnees ä analyser (susceptibles d'occuper, alternativement, les echelles nominale, ordinale ou metrique) impose le choix de certains indices et en deconseille l'emploi d'autres. 2. Six coups d'oeil comparatifs sur les structures de profondeur21 de l'atlas de Dees et de l'ALF Nous procederons par la suite, ä titre d'exemple, ä six comparaisons (ou «plate-formes») typodiagnostiques dont 1'orientation methodique et la finalite gnoseologique seront tres variees. Du point de vue formel, il s'agit d'un cote de comparaisons purement visuelles (basees, bien sür, sur des prerequis formels et statistiques strictement iden-tiques), et de correlations spatiales au sens strict du terme de 1'autre (transposees, elles aussi, en visualisations adequates). Le but de ces rapprochements comparatifs est de per-mettre aux lecteurs de juger eux-memes des etonnantes ressemblances (mais aussi de certaines dissemblances non negligeables) qui existent entre l'amenagement geolinguis-tique du domaine d'Oi'l ä la fin des 13e et 19e siecles22. Chacune des 12 pages de l'an-nexe oü figurent les 24 cartes illustratives, sera arrangee de l'acon identique: le reseau de l'atlas de Dees y figure toujours ä gauche, celui de l'ALF ä droite. 2.1. Plateforme comparative 1: la mensuration de la similarite Voir les cartes 1-8. L'outil dialectometrique de la mensuration de similarite permet de saisir les rapports de similarite (ressemblance, identite etc.) linguistiques qui existent entre les N elements (dialectes, scriptae etc.) d'un reseau geolinguistique donne. Pour un reseau d'atlas de N points et p attribute, il est possible de dresser, au cours de la mensuration de similarite appliquee aux N vecteurs d'objets (verticaux)23 du reseau, une matrice de similarite (car- La consideration quantitative de donnees de masse presuppose qu'il y ait, a l'interieur des donnees analysees, un grand nombre de structurations non aleatoires et hierarchisees entre elles. L'analyse quantitative permet de les decouvrir, quitte d'ailleurs ä operer, suivant le cas, dans differents secteurs des donnees ä analyser. II en resulte la necessite de differencier les structures detectees suivant le degre de leur insertion dans la totalite des materiaux analyses: d'oü la distinction entre structures de «profondeur» et de «surface». 22 La precision des analyses dialectometriques et la perfection des graphiques presentees dans cet article depas-sent Celles de nos contributions anterieures (Goebl 1998 et 2001 ainsi que Goebl/Schiltz 2001). Ceci est dü avant tout au logiciel VDM et ä ses capacites numeriques et visualisatrices. 23 Se referer, pour une bonne intelligence de ce processus, ä la figure 5. ree) aux dimensions N fois N, dont les vecteurs contiennent 1'ensemble de 1'information (quantitative) susceptible d'etre reunie par le precede de la mensuration de la similarite en tant que telle. Or, chacun des N vecteurs de la matrice de similarite dispose de N-l scores situes au-dessous de la valeur maximale (de 1 ou 100%) et d'un seul score (sjj) toujours equivalent ä 1 (ou 100%), issu d'une mensuration reflexive appliquee au point de reference j lui-meme. Par la visualisation d'un tel vecteur, Ton obtient une carte de similarite laquel-le, par definition, se compose de N-l plages (ou polygones) reparties dans l'espace et coloriees en fonction de leur rendement numerique, ainsi que d'un point de repere (ou de reference) j laisse toujours en blanc. La lecture d'une carte de similarite est facile. Elle nous renseigne, par la distribution, dans l'espace etudie, d'un nombre pair de couleurs «chaudes» (p. ex. rouge, orange, jaune) et de couleurs «froides» (p. ex. vert, bleu clair, bleu fonce), sur la «position» du dialecte (parle ou scripturaire) en question parmi les dialectes restants du reseau examine. Soit dit entre parentheses, la carte de similarite constitue le pilier central de la char-pente heuristique (et aussi theorique) de la dialectometrie entiere. Pour l'etablissement des cartes 1,3,5 et 7 (relatives aux donnees de l'atlas de Dees), nous avons utilise le coefficient de correlation de Bravais et Pearson (r(BP)jk)24 alors que les scores des cartes 2,4,6 et 8 (relatives ä l'ALF) ont ete calcules ä l'aide de l'lndice Pondere d'Identite (avec le poids 1) (IPI(l)jk)25. C'est par l'application de ces deux indices particuliers que la comparaison dialectometrique des deux niveaux atlantogra-phiques acquiert le plus de precision et de nettete. Les points de reference, situes aux quatre coins du domaine d'Oi'l (dans le Poitou, en Picardie, en Ile-de-France et dans la Franche-Comte), ont ete choisis en fonction de leur diversite geographique. Leur «profil choroplethe» (appele aussi «profil de similarite») nous renseigne sur la localisation de - metaphoriquement parlant - leurs «meilleurs amis» (voir les polygones en rouge) et aussi de leurs «antipodes (ou pires ennemis)» (voir les polygones en bleu fonce). A remarquer en outre le decalage bien agence des polygones en position intermediate, c'est-ä-dire pourvus des couleurs orange, jaune, vert et bleu clair. D'une comparaison visuelle des quatre paires de cartes, il ressort avec nettete que les amenagements linguistiques de l'espace du domaine d'Oi'l tels qu'ils se presentent sur les quatre cartes de similarite scripturaires et leur pendants relatifs aux dialectes modernes, ont dü etre sensiblement tres proches les uns des autres en 1300 et en 1900. Evidemment, il s'agit d'un bilan global qui repose sur la synthese quantitative d'un nombre tres grand de distributions spatiales particulieres (de nature qualitative et quantitative). Neanmoins, cette constatation (ou plutot decouverte) est quelque peu surprenante - pour ne pas dire 24 Les valeurs du r(BP)jk oscillent entre -1 (pour une correlation [lineaire] absolument negative) et +1 (pour une correlation [lineaire] absolument positive). Precisons que, dans ce contexte, la notion de «correlation» rem-place celle de «similarite (dialectale, scripturaire etc.)» et aussi qu'ä la fin de son atlas (1980, 370-371), Dees lui-meme s'est servi deux fois du r(BP)jk pour le calcul de la similarite scripturaire entre plusieurs documents medievaux. La meme remarque vaut pour Dees 1984 et pour la contribution de Huber/van Reenen-Stein (1988) oü l'on trouve une autre application du r(BP)jk aux donnees de l'atlas de Dees. 25 Pour une explication detaillee de cet indice cf. Goebl 1987, 70-79. troublante - surtout ä cause du fait qu'elle nous suggere que la structuration de base du domaine d'Oil, telle que nous la connaissons ä partir des donnees de l'ALF, a dejä dü exister - mutatis mutandis - 600 ans plus tot26. Ajoutons encore - avant de passer au paragraphe suivant - une petite remarque sur la relation qui existe entre la diminution de la similarite linguistique et de la proximite geo-graphique (ou euclidienne) par rapport ä un point de reference prealablement determine. La relation entre l'amenagement linguistique et euclidien de l'espace est extremement complexe. Bien loin d'etre uniquement un phenomene lineaire, cette relation connait des realisations spatiales tres bigarrees. Sur chacune de nos huit cartes de similarite, eile est d'une autre nature et revet des configurations geographiques tres diversifiees. Evidemment, il serait interessant de l'etudier de plus pres comme ceci a ete fait par Jean Seguy dans son premier article d'inspiration dialectometrique intitule «La relation entre la distance spatiale et la distance lexicale», publie en 1971, tout en remplacant d'ailleurs la recherche de «courbes» pratiquee jadis par Seguy, par la consideration synoptique spatiale des correlations quantitatives respectives: voir surtout notre article de synthese de 2005 (passim). 2.2. Plateforme comparative 2: la correlation spatiale entre les amenagements linguistique et euclidien de l'espace du domaine d'Oil Voir les cartes 9-10 et la figure 5. A l'aide de la fameuse formule de Pythagore, il est facile de calculer, entre les 85 points de notre reseau combine (Dees et ALF) et ä l'aide des coordonnees cartesiennes (en X et y) y ayant trait, leurs distances (d) ou proximites (prox)27 reciproques. C'est ainsi que l'on obtient une matrice (carree) de proximite aux dimensions 85 fois 85 oü se trou-vent stockes les scores de proximite entre tous les 85 points de notre reseau. Or, cette matrice de proximite peut etre correlee, vecteur par vecteur, avec la matrice de similarite (ou de correlation) linguistique comme le montre la figure 5. Le resultat de ces calculs de correlation - executes, une fois de plus, ä l'aide du coefficient de correlation r(BP)jk - est un ensemble de 85 scores de correlation qui, eux, doivent etre mis en carte par la suite: voir les cartes 9 et 10. Le profil diatopique qui se degage de cette mise en synopse est tres eloquent. L'on y discerne une nette bipartition du domaine d'Oil. Alors que 1'Est est caracterise par une tres mauvaise coincidence entre les amenagements linguistique et euclidien de l'espace, l'inverse est vrai pour l'Ouest et le Centre. Precisons qu'une «mauvaise» coincidence entre les amenagements linguistique et euclidien de l'espace signifie que, dans la realite empirique, la similarite linguistique decroit plus vite (ou plus lentement) que ne le suggere le modele (theorique voire «ideal») de revolution diatopique parallele des deux variables. Or, il est vrai que de if* De telles convergences spatio-temporelles ont ete constatees, pour des faits linguistiques isoles, dejä ä plu-sieurs reprises: cf. Remacle 1992,157-169, Goebl 1989,97-98 et Besch 1985, 1793-1798. La relation entre distance (d) et proximite (prox) est la suivante: d + prox =100. telles divergences se manifestent toujours lä oü de fortes tensions politiques, cultu-relles ou sociales accelerent (ou ralentissent) le deploiement diatopique «naturel» de l'activite langagiere de l'HOMO LOQUENS. L'inverse est egalement vrai: ceci signi-fie que les zones marquees en rouge et orange renvoient ä une synchronisation spa-tiale tres marquee entre la configuration euclidienne de l'espace et l'etalage diatopique des similarites interdialectales. L'agencement iconique convergent des profils choroplethes des cartes 9 et 10 nous suggere done que, dans l'optique de notre analyse dialectometrique, les prerequis spatio-linguistiques de base du domaine d'Oi'l n'ont pratiquement pas bouge entre 1300 et 1900 et que, par consequent, ils doivent remonter ä une epoque anterieure ä 1300. 23. Plateforme comparative 3: les cartes ä cloisons (ou: cartes ä interpoints en fonction discriminatoire) Voir les cartes 11 et 12. Les cartes ä cloisons (appelees aussi «cartes ä interpoints en fonction discriminatoire») constituent un des outils les plus simples (et fort eloquents ä la fois) de la dialecto-metrie. Du point de vue heuristique, elles correspondent ä la methode bien connue du tra-cement combine d'isoglosses. En voici les differentes etapes28: a) triangulation du reseau (Dees-ALF): sur 85 points, il en resulte 225 cotes de triangle. b) construction, pour chacun des 225 cotes de triangle, de la mediatrice respective. Les 225 mediatrices, düment fusionnees entre elles, finissent par foumir le parquet polygonal du reseau-dees-ALF. c) Chacun des 225 cotes de polygone servira de support iconique pour la visualisation non plus de similarites, mais de distances interponctuelles, e'est-a-dire situees entre deux points d'atlas contigus. d) La visualisation des scores de distance se sert de deux variables iconiques: de l'epaisseur (elle augmente avec la valeur numerique de la distance linguistique interponctuelle) et de la couleur (la couleur bleu fonce est associee au maximum de la distance linguistique interponctuelle, et la couleur rouge au minimum de cette derniere). C'est ainsi que la syntaxe iconique des deux cartes repose sur le jeu optique dif-ferentiel de 225 cotes de polygone d'epaisseur et de couleur variables. II en resulte un compartimentage alveolaire specifique du reseau polygonal qui est tres eloquent. L'on y distingue en effet la presence alternative de zones fortement cloisonnees (et sillonnees, de ce fait, de beaucoup de cötes de polygone epais et colories en bleu fonce) et d'autres oü predominent les polygones minces et colories en rouge, orange et jaune. 28 Cf. aussi Goebl 1983 (passim) et 1984 I, 183-196. Alors que la carte 11 laisse entrevoir, par le biais de l'effet-cloison, quelques-unes des provinces historiques du Nord de la France (Poitou, Maine, Normandie, Artois, Picardie etc.), il n'en est rien pous la carte 12 (relative ä l'ALF et, partant, ä la date de 1900) qui est caracterisee par une bipartition diatopique tres claire: concentration de l'effet-cloison ä la peripheric (Sud-Ouest, Nord, Nord-Est, Est) d'un cote, et reduction de l'effet-cloison dans les regions centrales du domaine d'Oi'l de l'autre29. D'une comparaison visuelle des deux cartes il appert que le changement macrolinguistique survenu entre 1300 et 1900 a fortement enraye l'autonomie reciproque des provinces historiques, surtout au centre du domaine d'Oi'l, tout en repoussant la plupart des differences interponctuelles majeures encore subsistantes aux marges peripheriques (Picardie, Wallonie, Lorraine, Franche-Comte). Cette constatation cadre fort bien avec ce que nous savons de Taction pluriseculaire de Г irradiation (circulaire) du type linguistique francien aux depens des autres types linguistiques d'Oi'l. 2.4. Plateforme comparative 4: la synopse des «coefficients d'asymetrie de Fisher (CAF)» Voir les cartes 13-18. La synopse des coefficients d'asymetrie de Fisher (CAF)30 constitue - ä cause de son excellente utilite en matiere de diachronie - une des pieces fortes de la panoplie dialec-tometrique entiere. C'est qu'elle permet de saisir, par voie quantitative, une des proprie-tes centrales de n'importe quel reseau interactif, ä savoir le degre d'interaction avec lequel chaque element (ou membre) du reseau respectif, participe aux flux d'interactions (ou d'echanges) qui, inevitablement, sillonnent le maillage du reseau en question. Evidemment, il y aura, dans un reseau donne, toujours des elements dont la participation aux dynamismes internes sera grande alors que d'autres elements resteront ä l'ecart et feront figure - metaphoriquement parlant - d'«abstentionnistes» ou de «trouble-fete». Du point de vue linguistique, les phenomenes d'interaction saisis par le CAF correspondent ä ce qu'il est convenu d'appeler, en matiere de linguistique, «brassage» (ou aussi: compromis) linguistique« (ou «Sprachausgleich» en allemand). L'intensite du brassage linguistique sera considerable lä oü la plupart des attribute linguistiques locaux appartiennent ä des aires dialectales tres grandes ou en voie d'elargis-sement, alors que l'inverse sera vrai pour les elements du reseau dont les attributs linguistiques locaux font partie d'aires dialectales plutot restreintes ou en voie de retrecis-sement. Dans cette situation, le defi central que le dialectometricien doit relever, consis-te ä mesurer le taux quantitatif de l'enchevetrement (ou de l'imbrication) d'aires dialectales majeures et mineures. yg Pour des cartes ä cloisons relatives au reseau integral de l'ALF (et visualisees suivant des modalites differentes) voir Goebl 2002, 36; 2002, 40; 2003, 98-99 et 2004, 280. 30 Pour la formule du CAF et d'autres details statistiques y ayant trait, cf. Goebl 1981, 394-401; 1984 I, 150153 et 2003, 81-84. Les scores du CAF se calculent ä l'aide de deux parametres classiques (moyenne arithmetique et ecart-type) de la distribution de frequence (ou de similarite) respective et se situent des deux cötes de la valeur zero. Alors que les valeurs negatives (symbolisees, sur les cartes 13-18, en bleu fonce) correspondent ä des zones ä forte interaction (ou connectivite) linguistique, les valeurs positives (coloriees en rouge sur les cartes 13-18) renvoient ä des regions oü 1'interaction et le contact linguistique avec le reste du reseau sont relativement faibles. 2.4.1» Interpretation des cartes 13-14 II s'agit de deux analyses reposant chacune sur le corpus integral respectif (Dees ou ALF). Sur la carte 13, situee ä gauche et relative ä la fin du 13e siecle, les zones rouges31 (et, partant, «isolationnistes», c'est-ä-dire disposant encore d'une certaine individualite regionale) se situent toutes ä la peripheric du domaine d'Oil: l'on y dis-cerne, en effet, le Poitou, le Maine, la Normandie, le Hainaut et la Wallonie, la Lorraine et la Franche-Comte. Les plages bleues par contre, caracterisees par une connectivite linguistique beaucoup plus intense, occupent les parties centrales du domaine d'O'il tout en formant des zones-tampon emergeant entre les provinces his-toriques mentionnees ci-dessus. Sur la carte 14, relative aux dialectes paries vers 1900, la zone bleue (ouverte ä Гinteraction et riche en contacts linguistiques de toute sorte) a fait tache d'huile tout en repoussant ä l'extreme peripheric ses contre-parties ä faible interaction linguistique. A remarquer, dans le Poitou, le poly gone rouge isole qui, des pay sages scripturaires occi-dentaux du Moyen Age, constitue la derniere survivance. 2.4.2. Interpretation des cartes 15-16 et 17-18 Evidemment, il serait interessant d'etudier le phenomene d'expansion rendu visible sur les cartes 13 et 14, aussi ä l'aide de corpus partiels relatifs, par exemple, au vocalis-me ou au consonantisme respectivement. Ce faisant, l'on pourrait determiner la difference qui, eventuellement, existerait entre la rapidite avec laquelle l'espace du domaine d'Oil aurait ete homogeneise, au cours de six siecles, sur le plan du vocalisme d'un cote et sur celui du consonantisme de l'autre. Or, la lefon des cartes 15-16 et 17-18 est tres claire ä cepropos. Alors que, pour le seul vocalisme (voir les cartes 15 et 16), 1'effet ega-lisateur, jailli au centre du domaine d'Oil et localise aux points de jonction entre plusieurs provinces historiques, correspond de tres pres ä ce qui vaut pour le corpus integral (voir les cartes 13 et 14), tout autre est la situation pour le consonantisme dont la fortune dia-chronique semble avoir ete quelque peu differente. C'est que les cartes 17 et 18 nous suggerent que l'amenagement consonantique de l'espace du domaine d'Oil n'a pratiquement pas change entre 1300 et 1900 et que, de ce TI Pour mieux faire ressortir le message central des profils choroplethes des cartes 13-18, nous avons choisi une visualisation avec seulement deux paliers chromatiques. fait, la vitesse de revolution diachronique du consonantisme est restee de loin inferieu-re ä celle du vocalisme (voir les cartes 15 et 16). 25. Plateforme comparative 5: la correlation spatiale entre les amenagements phonetique et consonantique du domaine d'Oil Voir les cartes 19 et 20. La question de 1'evolution diachronique inegale du vocalisme et du consonantisme peut etre encore approfondie par l'application de la dialectometrie correlative. C'est qu'il est possible de calculer, pour les corpus de Dees et de l'ALF et ä partir de sous-corpus prealablement definis, deux matrices de similarite, l'une relative au vocalisme et l'autre au consonantisme, de les correler par la suite et d'en visualiser les resultats: voir ä ce sujet, une fois de plus, la figure 5 et les remarques methodiques dans Goebl 2005 (321338 et 349-350). La mise en carte de ces calculs32 est tres pertinente; elle confirme plei-nement 1'interpretation que nous avons donnee, dans le paragraphe precedent, des deux vitesses avec lesquelles le changement linguistique s'est deroule en matiere de vocalisme et de consonantisme. La structuration choroplethe de la carte 19 (relative aux donnees de Г atlas de Dees) est tres claire: les polygones en bleu fonce - equivalents aux zones oü le vocalisme et le consonantisme divergent le plus - occupent les endroits ou, au 13e siecle - au moins dans l'optique de 1'atlas scripturaire de Dees - , les sous-categories linguistiques du vocalisme et du consonantisme n'allaient pas encore «de pair». L'agencement iconique de la carte 20 par contre, dont la structuration est completement opposee ä celle de la carte 19, nous signale que la «mise au pas» du vocalisme et du consonantisme est chose faite dans les centres propulseurs situes au coeur du domaine d'Oil, et que les ultimes divergences majeures ne subsistent qu'ä l'extreme peripherie (Normandie, Hainaut, Wallonie, Franche-Comte). Cette reorientation spatiale (ou «synchronisation») des deux categories linguistiques de base, operee entre 1300 et 1900, ne contredit aucunement les grandes lignes de nos connaissances traditionnelles sur l'histoire de la langue frangaise. 2.6. Plateforme comparative 6: l'analyse dendrographique Voir les cartes 21-24. Du point de vue historique, l'analyse dendrographique (dite aussi analyse «gene-tique» ou «genealogique») a, en matiere de linguistique historique, une tres longue tradition. Celle-ci debute avec l'indo-europeaniste allemand August Schleicher, qui, en 1863, a propose un schema dendrographique pour la classification genealogique des langues indo-europeennes tout en se servant expressement de visees et it Le choix de quatre paliers chromatiques pour la visualisation des cartes 19 et 20 s'explique ä nouveau par la recherche d'une expressivite iconique aussi grande que possible. methodes empruntees ä la biologie du temps. Depuis lors, de telles experiences ont ete reiterees d'innombrables fois, le plus souvent de la part de comparatistes, lexi-costatisticiens et de specialistes de glottochronologie qui, en l'occurrence, se sont servis tres souvent aussi de methodes quantitatives33. L'utilite heuristique d'un schema dendrographique est double: c'est qu'il concerne tant la diachronie que la synchronic. II peut done servir d'un cote ä la reconstruction (theorique) d'etapes dia-chroniques anterieures ä celle du relevement des donnees analysees, et fournir de l'autre une classification hierarchique en synchronie, qu'elle soit spatiale (comme c'est le cas ici) ou non. Les arborescences generees par la taxometrie (le plus souvent d'inspiration biolo-gique) ont les proprietes formelles suivantes: a) II s'agit d'une hierarchie exclusivement binaire de classes (clusters, groupes etc.) disjointes et non recouvrantes. b) Les clusters du graphe disposent d'une heterogeneite quantitative interne differente suivant leur position ä l'interieur de l'arbre: plus tel cluster se rapproche de la racine (ou du tronc) de l'arbre, plus son heterogeneite quantitative interne augmente. La taxometrie moderne dispose, depuis quelque 50 ä 60 ans, d'une riche panoplie d'algorithmes classificatoires parmi lesquels nous avons choisi l'algorithme propose par le statisticien americain Joe Ward jr. (voir les cartes 21 et 22) et l'algorithme dit «Complete Linkage» (voir les cartes 23 et 24) tout en les appliquant aux donnees de 1'atlas de Dees et de l'ALF34. Les deux algorithmes servent ä l'etablissement d'une «classification ascendante hierarchique» (CAH). En recourant aux donnees numeriques stockees dans la matrice de similarite, les algorithmes-CAH commencent par fusionner, parmi les N elements sujets ä classification, ceux dont la similarite quantitative (i. e. dialectometrique) est la plus grande tout en engendrant - au cours de l'avancement, dans la direction des feuilles vers la racine de l'arbre, de ces fusions effectuees toujours par paires - un graphe hierarchique binaire disposant de N «feuilles», de N-l bifurcations (ou ramifications, embranche-ments) et d'une racine (ou d'un tronc) unique. Pour 1'interpretation dialectometrique des arbres calcules (et düment visualises par la suite), il faut en transposer le rendement classificatoire dans l'espace respectif. Ce processus est appele «spatialisation». En regardant les cartes 21 et 22, l'on constate qu'ä la premiere ramification des deux arbres fait echo, dans l'espace, une bipartition du domai-ne d'Oi'l dans la direction est-ouest, et aussi que les embranchements successifs creent des classes (ou «dendremes»)35 dans le «feuillage» de l'arbre dont les equivalents spa-tialises («choremes») constituent toujours des aires disposant d'une coherence diato-pique surprenante. Voir a ce sujet l'excellent volume-recueil de Hoenigswald/Wiener (1987). 34 Pour l'algorithme de Ward cf. Sneath/Sokal 1973,204 s„ Bock 1974,407 s. et Chandon/Pinson 1981,123124; pour l'algorithme dit »Complete Linkage« cf. Bock 1974, 392 s„ Sneath/Sokal 1973, 222 s. et Chandon/Pinson 1981,106s. IS Les termes «dendremes» et «choremes» designent des groupements juges importants (ou «significatifs») situes, respectivement, ä l'interieur de l'arbre ou de sa projection dans l'espace (spatialisation). Pour l'histo-rique de cette terminologie combinee cf. Goebl 2003, 85 (note 13). II va de soi que ni pour les deux algorithmes utilises ni pour les deux stocks de donnees examines, l'on n'est en droit de s'attendre ä des resultats classificatoires par-faitement identiques. L'on constate pourtant que les classifications obtenues obeis-sent, en ligne generale, toujours aux memes tendances (hierarchiques et spatiales), ce qui prouve, une fois de plus, que les differentes structures de profondeur identifiables dans I'atlas de Dees et dans 1'ALF - outre d'etre solidement ancrees dans les donnees respectives - se ressemblent de tres pres. II est egalement interessant de voir que l'in-dividualite diatopique de certains paysages linguistiques traditionnels, telles la Picardie, la Wallonie, la Lorraine et les provinces du Sud-Ouest, ressort fort bien aussi par le biais d'une C AH. 3. Conclusion. Des analyses dialectometriques que nous venons de presenter, il appert avec nettete que les deux stocks de donnees examines (Гatlas de Dees et l'ALF) sont riches en struc-turations diatopiques de toute sorte et que celles-ci se manifestent soit «en surface» soit «en profondeur». Jetons, pour illustrer encore une fois la difference entre les termes metaphoriques de «surface» et de «profondeur», un dernier coup d'oeil sur les dimensions des matrices de similarite utilisees et sur Г importance variable de leur apport aux diverses analyses dialectometriques presentees. Selon la formule combinatoire N/2 (N-l)36, nos matrices de similarite disposent - pour N = 85 - de 3570 scores de similarite differents. Ceci signifie que, pour une seule carte de similarite basee sur 85 scores nume-riques, le taux d'exhaustion est de 2,85% (= 85 : 3570) alors que le meme taux d'ex-haustion s'eleve ä 6,30% (= 225 : 3570) pour une carte ä cloisons qui repose sur 225 valeurs particulieres. La meme analyse nous dit cependant que le calcul du coefficent d'asymetrie de Fisher (CAF), des cartes ä correlations ou des arbres genealogiques ne peut se faire qu'ä l'aide du total des 3570 scores disponibles. II en resulte qu'il existe, parmi les differentes especes de 1'analyse dialectometrique, une certaine hierarchie taxo-metrique ä laquelle nous avons applique les concepts metaphoriques de «surface» et de «profondeur». Cet etat de choses est d'ailleurs bien loin d'etre l'apanage exlusif de la dialectome-trie. II est courant egalement dans toutes les disciplines quantifiantes oü, pour 1'analyse exploratoire des donnees empiriques et pour Г acceptation ou le rejet de certaines hypotheses ou de theories de recherche, l'on fait recours ä des precedes mathematiques ou sta-tistiques. Toujours est-il qu'ä la lumiere de telles analyses, notre conjecture de depart selon laquelle l'amenagement linguistique du domaine d'Oi'l des 13e et 19e siecles repose sur des prerequis tres similaires, a pu etre confirmee ä plusieurs reprises. II serait hau-tement souhaitable de reiterer cette experience, soit pour le domaine d'Oi'l et ä l'aide de donnees nouvelles, soit pour un autre domaine de la Romania et ä l'aide de donnees dejä disponibles. Pour une matrice de donnees avec N elements (et p attributs), une comparaison de ses elements par paires aboutit ä une matrice de similarite carree avec les dimensions N2. Comme la diagonale de cette matrice carree se compose toujours de scores avec la valeur 1 (ou 100%) et que les deux moities de la matrice carree figurant des deux cötes de la diagonale ne contiennent que des scores parfaitement identiques, le nombre de scores utilisables ä des fins classificatoires se calcule comme suit: (N2 - N ): 2 = N/2 (N - 1). 4. Remerciements La redaction de cette contribution ainsi que tous les travaux preparatories y ayant trait presupposent la cooperation concertee d'une equipe bien rodee. C'est pourquoi je tiens ä remercier tres cordialement tous ceux qui, de pres ou de loin, ont apporte leur contribution ä la recherche que nous venons de presenter. 4.1. Confection des figures et des cartes de cet article, entretien permanent du logi-ciel et du materiel de VDM: Slawomir Sobota (Salzbourg), 4.2. Creation et mise ä jour periodique du logiciel VDM37: Edgar Haimerl (Blaustein, Allemagne), 4.3. Taxation des donnees de l'ALF: Barbara Aigner, Irmgard Dautermann, Hildegund Eder, Susanne Oleinek, Annette Ida Schatzmann (Salzbourg), 4.4. Toilette stylistique de mon texte fran§ais: Lily Ditz-Fuhrich (Salzbourg). 4.5. Appui financier de nos recherches dialectometriques: projets 12 414 et 13 349 du Fonds national autrichien pour la recherche scientifique FWF «Fonds zur Förderung der wissenschaftlichen Forschung in Österreich». 5. References bibliographiques et abreviations techniques souvent utilisees. 5.1. Abreviations techniques souvent utilisees. CAF: coefficient d'asymetrie de Fisher (cf. le paragraphe 2.4.) CAH: classification ascendante hierarchique (cf. le paragraphe 2.6.) DEMjk: Distance Euclidienne Moyenne (voir le paragraphe 1.4. et la figure 4) IPI(l)jk: Indice d'Identitć Pondere (avec le poids 1) (cf. le paragraphe 1.4.) irijk: Indice Relatif d'Identite (cf. le paragraphe 1.4.) r(BP)jk: coefficient de correlation de Bravais et Pearson (cf. les paragraphes 2.1. et 2.2.) SEMjk: Similarite Euclidienne Moyenne (voir le paragraphe 1.4. et la figure 4) VDM: Visual DialectoMetry, logiciel dialectometrique crće par Edgar Haimerl (cf. le paragraphe 1). Consulter aussi notre site internet bilingue (allemand-anglais) relatif a la dialectometrie et au logiciel VDM: http://ald.sbg.ac.at/dm/. 5.2. References bibliographiques ALF: Gillieron, Jules/Edmont, Edmond (eds.) (1902-1910): Atlas linguistique de la France, Paris: Champion, 10 vol. (reimpression: Bologne: Fomi, 1968). Besch, Werner (1985): Die Entstehung und Ausformung der neuhochdeutschen Schriftsprache/Standardsprache, in: Besch, Werner/reichmann, Oskar/SoNDEREGGER, Stefan (eds.): Sprachgeschichte. Ein Handbuch zur Geschichte der deutschen Sprache und ihrer Erforschung, Berlin, New York: Walter de Gruyter, vol. 1,1781-1810. Bock, Hans Hermann (1974): Automatische Klassifikation. Theoretische und praktische Methoden zur Gruppierung und Strukturierung von Daten (Cluster-Analyse), Göttingen: Vandenhoeck & Ruprecht. Chandon, Jean-Louis/plnson, Suzanne (1981): Analyse typologique. Theories et applications, Paris, New York, Barcelone, Milan: Masson. Dees, Anthonij (1980): Atlas des formes et des constructions des chartes frangaises du 13e siecle, Tübingen: Niemeyer (Beihefte zur Zeitschrift für romanische Philologie, vol. 178). Dees, Anthonij (1984): Regards quantitatifs sur les variations regionales en ancien frangais, in: Goebl, Hans (ed.): Dialectology, Bochum: Brockmeyer (Quantitative Linguistics, vol. 21), 102-120. DEES, Anthonij (1987): Atlas des formes linguistiques des textes litteraires de Vancien frangais, Tübingen: Niemeyer (Beihefte zur Zeitschrift für romanische Philologie, vol. 212). Goebl, Hans (1981): Elements d'analyse dialectometrique (avec application ä l'AIS), in: Revue de linguistique romane 45, 349-420. Goebl, Hans (1982): compte rendu de Dees 1980, in: Zeitschrift für französische Sprache und Literatur 92,280-283. Goebl, Hans (1983): Parquet polygonal et treillis triangulaire: les deux versants de la dialectometrie interponctuelle, in: Revue de linguistique romane 47, 353-412. Goebl, Hans (1984): Dialektometrische Studien. Anhand italoromanischer, rätoromanischer und galloromanischer Sprachmaterialien aus AIS und ALF, Tübingen: Niemeyer (Beihefte zur Zeitschrift für romanische Philologie, vol. 191-193). Goebl, Hans (1987): Points chauds de l'analyse dialectometrique: ponderation et visualisation, in: Revue de linguistique romane 51, 63-118. Goebl, Hans (1989): -ORE(A) statt -ORIU(A) in der Normandie: einem Suffixwechsel mit skriptologischen Mitteln auf der Spur, in: Ernst, Gerhard/stefenelli, Amulf (eds.): Studien zurromanischen Wortgeschichte. Festschrift für Heinrich Kuen zum 90. Geburtstag, Stuttgart: Steiner, 92-102. Goebl, Hans (1993): compte rendu de Dees 1987, in: Zeitschrift für französische Sprache und Literatur 103 (1993) 185-187. Goebl, Hans (1998): Zu einer dialektometrischen Analyse der Daten des Dees-Atlasses von 1980, in: Werner Edeltraud/liver, Ricarda/storck, Yvonne/nlcklaus, Martina (eds.): Et multum et multa. Festschrift für Peter Wunderli zum 60. Geburtstag, Tübingen: Narr, 293-309. Goebl, Hans (2000): La dialectometrisation de l'ALF: presentation des premiers resultats, in: Linguistica 40,209-236. Goebl, Hans (2001): La strutturazione geolinguistica del dominio d'oil nel 13° secolo e alia fine dell'Ottocento. Un raffronto dialettometrico, in: Zamboni, Alberto/del Puente, PatriziaA'igolo, Maria Teresa (eds.): La dialetto-logia oggi fra tradizione e nuove metodologie. Atti del Convegno internazionale, Pisa 10-12 Febbraio 2000, Pisa: Edizioni ETS, 11-43. Goebl, Hans (2002): Analyse dialectometrique des structures de profondeur de l'ALF, in: Revue de linguistique romane 66,5-63 (avec 24 cartes ä couleurs). Goebl, Hans (2003): Regards dialectometriques sur les donnees de Р Atlas linguistique de la France (ALF): relations quantitatives et structures de profondeur, in: Estudis Romanics 25, 59-96 (avec 24 cartes ä couleurs). Goebl, Hans (2004): Sprache, Sprecher und Raum: eine kurze Darstellung der Dialektometrie. Das Fallbeispiel Frankreich, in: Mitteilungen der österreichischen geographischen Gesellschaft 146, 247-286. Goebl, Hans (2005): La dialectometrie correlative. Un nouvel outil pour Г etude de l'amenagement dialectical de l'espace par l'homme, in: Revue de linguistique romane 69, 321-367 (avec 24 cartes ä couleurs). Goebl, Hans/schiltz, Guillaume (2001): Der »Atlas des formes et des constructions des chartes frangaises du 13e siecle« von Anthonij Dees (1980) - dialektometrisch betrachtet, in: Gärtner, Kurt/Holtus Günter/Rapp, Andrea/völker, Harald (eds.): Skripta, Schreiblandschaften und Standardisierungstendenzen. Urkunden -sprachen im Grenzbereich von Germania und Romania im 13. und 14. Jahrhundert. Beiträge zum Kolloquium vom 16. bis 18. September 1998 in Trier, Trier: Kliomedia, 169-221. Hoenigswald, Henry M./Wiener, Linda F. (eds.) (1987): Biological Metaphor and Ciadistic Classification. An Interdisciplinary Perspective, Philadelphia: University of Pennsylvania Press. Huber, Оппо/van Reenen-Stein, Karin (1988): Correlations et groupements dans l'Atlas des formes et des constructions des chartes frangaises du XIIIe siecle, in: van Reenen, Pieter/van Reenen-Stein, Karin (eds.): Distributions spatiales et temporelles, constellations de manuscrits. Etudes de variation linguistique offertes ä Anthonij Dees ä l'occasion de son 60me anniversaire, Amsterdam, Philadelphia: Benjamins, 93-101. Remacle, Louis (1992): La differentiation dialectale de la Belgique romane avant 1600, Geneve: Droz. Schiltz, Guillaume: Die Anwendung künstlicher neuronaler Netze zur Klassifikation altfraiKÖsischer Dialekte [Amsterdam, Second International Congress of Dialectologists and Geolinguists, 1997], telechargeable sous l'adresse suivante: http://pages.unibas.ch/anglist/schiltz/data/ams_nn.pdf Schleicher, August (1863): Die Darwinsche Theorie und die Sprachwissenschaft. Offenes Sendschreiben an Herrn Dr. Ernst Häckbl, a. o. Professor der Zoologie und Director des zoologischen Museums an der Universität Jena, Weimar: Döhlau. Seguy, Jean (1971): La relation entre la distance spatiale et la distance lexicale, in: Revue de linguistique romane 35, 335-357. Sćguy, Jean (1973): La fonetion minimale du dialecte, in: Straka, Georges/GARDETEE, Pierre (eds.): Les dialectes romans de France ä la lumiere des atlas regionaux (Colloque de Strasbourg, 1971), Paris: Editions du CNRS, 27-37. Sneath, Peter AVSokal, Robert R. (1973): Numerical Taxonomy. The Principles and Practice of Numerical Classification, San Francisco: Freeman. Völker, Harald (2003): Skripta und Variation. Untersuchungen zur Negation und zur Substantivflexion in altfranzösischen Urkunden der Grafschaft Luxemburg (1237-1281), Tübingen: Niemeyer (Beihefte zur Zeitschrift für romanische Philologie, vol. 305). Povzetek JEZIKOVNE INOVACIJE MED 14. IN 20. STOLETJEM NA OZEMLJU FRANCOŠČINE DIAHRONIČNA ŠTUDIJA Z VIDIKA DIALEKTOMETRIJE Študija skuša biti primerjava dveh korpusov iz jezikovne sfere severne Francije, torej območja, ki ga navadno označujemo z izrazom langue d'oi'1. Temelji na dialektometričnih vidikih, je diahro-nična, saj primerja stanje v 14. in 20. stoletju. Korpus iz srednjega veka je objavil leta 1980 Anthoni Dees, in sicer na podlagi kakih 3000 neliterarnih dokumentov iz zadnje tretjine 13. stoletja. Zaobjema 85 središč za rokopise in zajema 285 jezikoslovnih vprašanj, ponajveč fonetičnih. Narečni korpus kolikor toliko naše dobe pa je vzet iz znamenitega francoskega jezikovnega atlasa, Atlas linguistique de la France (ALF). Avtorja sta Jules Gillieron in Edmond Edmont. Objava tega jezikovnega atlasa je prav iz prvih let preteklega stoletja. Naša študija sledi Deesovi s 285 eksploriranimi kraji v ALF in 1687 narečnimi jezikovnimi problemi (fonetičnimi in leksikalnimi). Vzporedna primerjava zajema tudi dialektometrične poglede. Dobljene rezultate primerjave podajamo z dialektometričnimi predstavitvami v barvah, ki pokažejo, da so se strukture obdržale skozi šest stoletij v glavnem nespremenjene, vendar pa se kažejo jezikovne inovacije v tem, da so se posamezne dialektalne regije območja stare francoščine (jezika d'oi'1) močno približale jeziku province lie de France (torej normi Pariza). Ugotavlja se hitrejši razvoj jezikovnih sprememb v samoglasniškem sistemu, počasnejši v soglasniškem. Legendes des 5 figures: Figure 1: Liste des unites spatiales (centres scripturaires, points) majeures (Macro) et mineures (Micro) de l'atlas de Dees (1980). Figure 2: Liste des cartes de l'atlas de Dees (1980) retenues pour notre analyse dialectometrique. Figure 3: Tableau explicatif de l'etablissement de la matrice de donnees de l'atlas de Dees (1980). Voir aussi la figure 4. Figure 4: Calcul de la similarite quantitative entre deux centres scripturaires j et k ä 1'aide de l'indice SEMß (Similarite Euclidienne Moyenne). Voir aussi la figure 3. Figure 5: Tableau explicatif de l'etablissement de deux matrices de similarite, de leur correlation successive (ä l'aide du coefficient de correlation de Bravais et Pearson [r(BP)jjJ) et de l'etablissement de la carte ä correlations. Legendes des 24 cartes: Carte 1: Carte choroplethe de la distribution de similarite (г(ВР)з j^) relative au point de reference 3 (DEES 3: Vendee, Deux-Sevres). Corpus: Dees 1980, 203 cartes (donnees integrales). Voir aussi le paragraphe 2.1. Carte 2: Carte choroplethe de la distribution de similarite (IPI(] )зjj relative au point de reference 3 (ALF 429: Saint-Germam-le-Princay, Vendee). Corpus: ALF, 1687 cartes de travail (donnees integrales). Voir aussi le paragraphe 2.1. Carte 3: Carte choroplethe de la distribution de similarite (г(ВР)зд j^) relative au point de reference 30 (DEES 30: Pas-de-Calais, sud-ouest). Corpus: Dees 1980, 203 cartes (donnees integrales). Voir aussi le paragraphe 2.1. Carte 4: Carte choroplethe de la distribution de similarite (IPI(1 )зој<) relative au point de reference 30 (ALF 298: Nort-Leulinghem, Pas-de-Calais). Corpus: ALF, 1687 cartes de travail (donnees integrales). Voir aussi le paragraphe 2.1. Carte 5: Carte choroplethe de la distribution de similarite (rfBP)^^ ^) relative au point de reference 56 (DEES 56: Paris). Corpus: Dees 1980,203 cartes (donnees integrales). Voir aussi le paragraphe 2.1. Carte 6: Carte choroplethe de la distribution de similarite (TPICI^j,) relative au point de reference 56 (ALF 226: Le Plessis-Piquet; Hauts-de-Seine). Corpus: ALF, 1687 cartes de travail (donnees integrales). Voir aussi le paragraphe 2.1. Carte 7: Carte choroplethe de la distribution de similarite (rfBPj^c) ^) relative au point de reference 79 (DEES 79: Haute-Saöne). Corpus: Dees 1980, 203 cartes (donnees integrales). Voir aussi le paragraphe 2.1. Carte 8: Carte choroplethe de la distribution de similarite (№1(1)79 ^) relative au point de reference 79 (ALF 46: Conflans-sur-Lanterne, Haute-Saone). Corpus: ALF, 1687 cartes de travail (donnees integrales). Voir aussi le paragraphe 2.1. Carte 9: Carte choroplethe de la correlation spatiale (г(ВР)оеоРгох SEM^ entre 'es am®~ nagements euclidien et scripturaire du domaine d'Oi'l vers 1300. Voir aussi le paragraphe 2.2. Carte 10: Carte choroplethe de la correlation spatiale (r(BP)QeoProx TP K I p entre les amenagements euclidien et dialectal du domaine d'Oi'l vers 1900. Voir aussi le paragraphe 2.2. Carte 11: Carte ä cloisons (ou carte ä interpoints en fonction discriminatoire: isoglosses quantitatives). Corpus: Dees 1980, 203 cartes (donnees integrales). Indice de distance: DEM;^. Voir aussi le paragraphe 2.3. Carte 12: Carte ä cloisons (ou carte ä interpoints en fonction discriminatoire: isoglosses quantitatives). Corpus: ALF, 1687 cartes (donnees integrales). Indice de distance: IRDj^. Voir aussi le paragraphe 2.3. Carte 13: Carte choroplethe de la synopse de 85 coefficients d'asymetrie de Fisher (CAF). Corpus: Dees 1980, 203 cartes (donnees integrales). Indice de similarite: SEMj^. Voir aussi le paragraphe 2.4. Carte 14: Carte choroplethe de la synopse de 85 coefficients d'asymetrie de Fisher (CAF). Corpus: ALF, 1687 cartes de travail (donnees integrales). Indice de similarite: №1(1)^. Voir aussi le paragraphe 2.4. Carte 15: Carte choroplethe de la synopse de 85 coefficients d'asymetrie de Fisher (CAF). Corpus: Dees 1980,119 cartes (vocalisme). Indice de similarite: SEMj^. Voir aussi le paragraphe 2.4. Carte 16: Carte choroplethe de la synopse de 85 coefficients d'asymetrie de Fisher (CAF). Corpus: ALF, 612 cartes de travail (vocalisme). Indice de similarite: №1(1)^. Voir aussi le paragraphe 2.4. Carte 17: Carte choroplethe de la synopse de 85 coefficients d'asymetrie de Fisher (CAF). Corpus: Dees 1980, 66 cartes (consonantisme). Indice de similarite: SEMj^. Voir aussi le paragraphe 2.4. Carte 18: Carte choroplethe de la synopse de 85 coefficients d'asymetrie de Fisher (CAF). Corpus: ALF, 505 cartes de travail (consonantisme). Indice de similarite: №1(1)^. Voir aussi le paragraphe 2.4. Carte 19: Carte choroplethe de la correlation spatiale (r(BP)sEM-VOC SEM-CONS) entre les amenagements vocalique et consonantique du domaine d'Oil vers 1300. Coefficient de correlation: rfBPjj^. Voir aussi le paragraphe 2.5. Carte 20: Carte choroplethe de la correlation spatiale (r(BP)iRi(i)-VOC IRI(l)-CONS) entre les amenagements vocalique et consonantique du domaine a'Oil vers 1900. Coefficient de correlation: KBP)^. Voir aussi le paragraphe 2.5. Carte 21: Classification dendrographique de 85 points-DEES. Algorithme agglomeratif: methode de WARD. Corpus: Dees 1980, 203 cartes (donnees integrales). Indice de similarite: SEMj^. Nombre des choremes et dendremes colories: 6. Voir aussi le paragraphe 2.6. Carte 22: Classification dendrographique de 85 points-ALF. Algorithme agglomeratif: methode de WARD. Corpus: ALF, 1687 cartes de travail (donnees integrales). Indice de similarite: IPI(l)^. Nombre des choremes et dendremes colories: 6. Voir aussi le paragraphe 2.6. Carte 23: Classification dendrographique de 85 points-DEES. Algorithme agglomeratif: Complete Linkage. Corpus: Dees 1980,203 cartes (donnees integrales). Indice de similarite: SEMj^. Nombre des choremes et dendremes colories: 5. Voir aussi le paragraphe 2.6. Carte 24: Classification dendrographique de 85 points-ALF. Algorithme agglomeratif: Complete Linkage. Corpus: ALF, 1687 cartes de travail (donnees integrales). Indice de similarite: 1Р1(1)^. Nombre des choremes et dendremes colories: 5. Voir aussi le paragraphe 2.6. Macro (Ma) nom Micro (Mi) Macro (Ma) nom Micro (Mi) 1 Charente, Charente-Maritime 1 15 Hainaut 42 La Rochelle et environs 2 Toumai et environs 44 2 Vendee, Deux-Sevres 3 16 Wallonie 45 Vendee 4 Liege (province) 46 Deux-Sevres 5 Luxembourg 47 3 Vienne 6 17 Ardennes 48 Mirebeau et environs (= Poitiers) 7 Ardennes nord 49 Loudun et environs 8 Ardennes sud 50 Chätellerault et environs 9 18 Marne 51 4 Indre, Cher 10 Marne ouest 52 5 Orleanais 11 Marne est 53 Eure-et-Loir 12 19 Region pansienne 54 Loir-et-Cher 13 Val d'Oise 55 6 Indre-et-Loire 14 Paris 56 Tours et environs 15 Seine-et-Marne 57 Loches et environs 16 20 Yonne 58 7 Maine-et-Loire 17 21 Aube 59 Maine-et-Loire sud + Angers 18 22 Haute-Marne 60 8 Mayenne, Sarthe 19 Langres et environs 61 Le Mans et environs 20 Chäteauvillain et environs 62 9 Bretagne 21 Chaumont et environs 63 10 Normandie 22 Haute-Marne nord 64 Manche 23 Haute-Marne nord-est 65 Eure 24 23 Meuse 66 Seine-Maritime 25 Verdun et environs 67 11 Somme, Pas-de-Calais 26 Meuse sud-ouest 68 Somme centre + sud 27 Meuse sud-est 69 Somme nord-ouest 28 24 Moselle, Meurthe-et-Moselle nord 70 Pas-de-Calais sud-est 29 Nancy et environs 71 Pas-de-Calais sud-ouest 30 Moselle sans Metz 72 Pas-de-Calais centre + sud 31 Meurthe-et-Moselle nord 73 12 Oise 32 Metz et environs 74 Oise sud-ouest 33 25 Vosges 75 Oise nord-ouest 34 Neufchäteau 76 Oise nord-est 35 26 Franche-Comte 77 Oise sud-est 36 Doubs 78 13 Aisne 37 Haute-Saone 79 Saint-Quentin et environs 38 Jura 80 14 Nord 39 27 Bourgogne 81 Lille et environs 40 Dijon et environs 82 Douai et environs 41 Cote-d'Or nord 83 Cote-d'Or sud + Saone-et-Loire 84 28 Nievre, AUier 85 Figure 1: Liste des unites spatiales (centres scripturaires, points) majeures (Macro) et mineures (Micro) de l'atlas de Dees (1980). 1) Vocalisme: 119 cartes cartes no. 2 48 96 97 98 100 5 50 101 a/b 102 103 104 6 51 105 111 114 115 7 53 120 124 125 136 9 61 138 140 142 143 11 62 144 146 148 149 12 65 153 155 156 158 13a/b 68 162a/b 163 164 165 16 69 167 168 170 171 18 70 173 174 177 178 20 74 181 182 185 187 21 77 188 189 190 195 24 78 196 200 202 203 29 79 204 205 210 212 35 83 213 215 222 224 38 87a/b 225 227 228 232 41 89a/b 235 238 243 246 42 92 247 248 250 253 45 93 254 255 257 259 46 94 262 264 267 2) ConsonaHtisme: 66 cartes cartes no. 1 8 15 23 25 27 28 33 49 52 56 58 59 60 63 67 72 73 80 85 86 88 90 95 99 113 117 118 119 126 127 133 134 135 137 151 154 157 161 166 191 197 198 199 201 209 214 217 218 219 220 223 231 233 239 241 244 245 251 252 258 261 263 265 266a/b 268 3) Morphologie (nominale et verbale): 18 cartes cartes no. 14 19 34a/b 37a/b 39 40 64 75 76 81 84 91 122 123 186 206 207 208 La numerotation des cartes correspond ä celle utilisee par DEES 1980. Figure 2 : Liste des cartes de l'atlas de Dees (1980) retenues pour notre analyse dialectometrique. Matrice de donnees (DEES 1980) (echelle metrique, en % i k DONNEES FICTIVES) Q: ^ to . CD XL CQ q 8 5 8 12 10 15 13 7 23 2 3 7 7 6 6 V 0 13 2 8 2 CO -Ш (Л 5 5 3 10 3 V 8 ffi zs 4 4 1 2 0 0 0 V) 3 0 22 X3,3 X4,3 19 5 13 ■j— 2 6 V 3 3 2 7 4—' ^ < 1 2 12 2 1 2 6 1 2 3 4 5 6 Centres scripturaires (elements) [1.. . j k ... /V/ V donnees manquantes Commentaire: Pour chacun des 298 attribute de son atlas de 1980, Anthonij Dees a defini deux ensembles de graphies complementaires dont un seul a ete mis en carte. En voici un exemple relatif ä la carte (attribut graphique) 1 de son atlas: il s'agit des differentes realisations de I'etymon latin EGO: graphies visualisees: g, ge, gie, etc. graphies complementaires non visualisees: /, /e, iou, je, jo, etc. Pour le calcul de la frequence (relative) des graphies visualisees ä tel endroit du reseau examine, A. Dees a etabli le quotient entre le nombre des chartes disposant I'attribut visualise (1) et la somme des chartes disposant I'attribut visualise (1) d'un cote et I'attribut complementaire non visualise (2) de I'autre. Quant ä l'exemple cite ci-dessus, les scores respectifs retenus dans la matrice de donnees se calculent comme suit: ^ chartes (du centre scripturaire j) avec des occurrences Zd de I'attribut visualise (1) % = 100 ^ chartes (du centre scripturaire j) avec des occurrences Zu de I'attribut visualise (1) et de I'attribut non visualise (2) Les chartes du centre scripturaire j qui ne disposent d'aucun succedane de I'etymon latin EGO, restent exclues du calcul. Figure 3: Tableau explicatif de l'etablissement de la matrice de donnees de l'atlas de Dees (1980). Voir aussi la figure 4. Calcul de la SEM _ ЈК (Similarite Euclidienne Moyenne)1' SEM = 100 - 100 * /'= 1 P Explication des symboles: SEM Similarite Euclidienne Moyenne entre les centres scripturaires j et k i index pour un attribut graphique j,k index pour deux centres scripturaires compares entre eux pourcentage repertorie dans la matrice de donnees ä l'endroit du centre scripturaire j et de Гattribut / p nombre des attributs graphiques disponibles (co-presents) dans les vecteurs j et k de la matrice de donnees (ä l'exclusion de donnees manquantes) 11 Par rapport ä la SEM^ la „Distance Euclidienne Moyenne" (DEM;(t) se calcule comme suit: SEM., + DEM, = 100. Verification, ä l'aide de la matrice de donnees de la figure 3 du calcul de la SEM;2 (entre les centres scripturaires 1 et 2) /=1 (2 - 12)2 + (0 - 22)2 + (4 - 1)2 + (5 - 3)2 + (23 - 2)2 + (5 - 8)2 = 1047 «• A remarquer les donnees manquantes dans / = 2 et 6, d'oü p = 6. • SEM12 = 100 - 1320,98 =-1220,98 Figure 4: Calcul de la similarite quantitative entre deux centres scripturaires j et k ä l'aide de l'indice SEMt (Similarite Euclidienne Moyenne). Voir aussi la figure 3. 6 • Calcul du terme (xr,-x2i): |2- • SEM, = 100 -100 Matrice de donnees A (echelle nominale) P г 1 i ™ j i j i i J h h h i X i e e f X 9 g a a a b c d I 2 3 4 5 S Points d'atias (elements) II . . . j - k . . - N] Matrice de similarite s(a|. A (symetrique) x donnees manquantes 100 66 66 25 25 25 66 100 50 33 33 33 66 50 100 33 33 33 25 33 33 100 75 75 25 33 33 75 100 100 25 33 33 75 100 100 6 5 4 3 2 1 Points d'atias (elementsJ IN . . . j ... 1 1 Matrice de similarite s(b)jk B (symetrique) 100 100 100 100 100 100 1 2 3 4 5 Matrice de donnees £ X X r 1 N 1 6 5 4 3 2 1 X donnees manquantes [N ■ ■ ■ k . j . . . 1] Carte ä correlations Figure 5: Tableau explicatif de l'etablissement de deux matrices de similarite, de leur correlation successive (äl'aide du coefficient de correlation de Bravais et Pearson [г(ВР).Ј) et de l'etablissement de la carte ä correlations. Cartes dialeetometriques to DEES 1980 Hainau! WALLONIE ALF Hainaut WALLONIE FRANCHE-COMTE fee. HAIMERL. KECK c. GOCĐL a.d. 1999 te de THIESSEN/HAAG/VORONOI 85 points d'ettos 222 segments de polygene fee. HAIMERL KECK cono, G a.d. 1999 carte de THIESSEN/HAAG/VORONOI 85 points d'al/as 222 segments de polygene Algorithme d'intervallisation MINMWMAX6-tuple Distribution de frequence (similarity) MINMWMAX 12-tuple 1 -0.24 - -0.13 (13) 2 -0.01 (21) 3 m 0.10 (19) 4 0.33 (15) 5 Hj 0.55 (11) 6 - 0.77 (5) 2= 84 Algorithme d'intervallisation Distribution de frequence (similarite) MINMWMAX 6-tuple MINMWMAX 12-tuple (12) 1 13.45 - 17.63 2 - 21.81 3 ■ - 25.99 4 - 34.40 5 Ü - 42.81 6 - 51.22 r(BP)3k( TOT) 1 = 84 №4(1). Л TOT) DEES 1980 Hainau! WALLONIE Hainaut WALLONIE PICARDIE FRANCHE-COMTE BOURGOGNE fee HAIMERL KECK cone. GOEBL ad. 1999 le de THIESSEN/HAA G/VORONOI 85 points d'atlas 222 segments de polygene Algorithme d'intervallisation MINMWMAX 6-tuple Distribution de frequence (similarite) MINMWMAX 12-tuple 1 -0.14 - -0.01 (16) 2 - 0.13 (8) 3 ■1 i ■•■ 1 LZ] 0.26 (23) 4 - 0.44 (19) 5 :j 0.62 (8) 6 - 0.81 (10) S = 84 FRANCHE-COMTE BOURGOGNE fee. HAIMERL. KECK conc GOEBL ad. 1999 carte de THIESSEN/HAAG/VORONOI 85 points d'atlas 222 segments de polygone Algorithme d'intervallisation MINMWMAX 6-tuple 1 ШЛ 14.86 2 Li ] 4[Z3 5 iZ ] Distribution de frequence (similarite) MINMWMAX 12-tuple r(BP)30,(J OT) - 18.69 (13) - 22.53 (21) - 26.36 (17) - 38.06 (24) - 49.76 (6) - 61.47 (3) Z = 84 21 24 27 30 33 36 39 42 45 4B 51 54 57 IPKVsoJ TOT) OJ Meurthe-et-Moselle FRANCHE-COMTE fee. HAIMERL, KECK CMC. GOEBL a.tf 1999 carte de THIESSEN/HAAG/VORONOI 86 points d'atlas 222 segments de polygene Algorithme d'intervallisation MINMWMAX 6-tuple Distribution de frequence (similarite) MINMWMAX 12-tuple 1 L.J -0.14 - 0.00 (8) 2 - 0.13 (20) 3 - 0.26 (19) 4 □ - 0.47 (20) 5 m - 0.68 (12) 6 - 0.89 (5) S = 84 _ FRANCHE-COMTE BOURGOGNE fee. HAIMERL. KECK oonc. GOEBL ad 190 9 called0 THIESSEWHAAG/VOROHOI 85 polnili d'ellas 222 segments de polygone Algorithme d'intervallisation MINMWMAX 6-tuple Distribution de frequence (similarite) MINMWMAX 12-tuple r(BP)x.(tot) 1 17.77 - 24.72 (14) 2 - 31.68 (23) 3 ■ - 38.64 (7) 4 □ - 48.57 (17) 5 Ш - 58.50 (15) 6 - 68.43 (8) 1 = 84 18 21 24 27 30 33 36 38 42 45 48 51 54 57 60 63 IPI(VS6A TOT) ARTOIS WALLONIE ARTOIS Hainaut WALLONIE Somme, Pas-de-Calais FRANCHE-COMTE BOURGOGNE c. HAIMERL KECK one. COEBL a.d. 1999 lede THIESSEWHAAGMOROHQt littoral limites linguistiques pćiiphćrigues Algorithme d'intervaliisation MINMWMAX 6-tuple Distribution de frequence (similarite) MINMWMAX 12-tuple 1 -0.06 - 0.02 (10) 2 - 0.11 (27) 3 ■ - 0.19 (15) 4 - 0.35 (13) 5 z: - 0.50 (12) 6 - 0.66 (7) S = 84 r(BP)79k(TOJ) LORRAINE FRANCHE-COMTE BOURGOGNE fee HAIMERL KECK conc GOEBL ed. 1999 carte de THIESSEN/HAAG/VORONOI 85 points tfatlas 222 segments de polygene -littoral — limites linguistiques piripheriques Algorithme d'intervaliisation MINMWMAX 6-tuple 1 H 1! 2.....E 3| 4! I 5fSB еШ Distribution de frequence (similarite) MINMWMAX 12-tuple - 19.40 (14) - 22.92 (6) - 26.44 (24) - 35.73 (34) - 45.01 (5) - 54.29 (1) 1 = 84 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 1Р1(1)„Л TOT) On FRANCHE-COMTE FRANCHE-COMTE BOURGOGNE fee. HAIMERL. KECK conc QOEBL e.d 1999 carte de THIESSEWHAAOVORONOt 05 paints d'Btles 222 segmenta de polygene - Iimitesl>nguis!lgiiesp6ript,šn fl diachronique Marek Gawelko Krakow UDK 811.1/2'0/'06 SUR UN ASPECT DES ETUDES HISTORIQUES En se fondant sur le principe qui dit que la science doit tendre vers le general, on est conduit ä etablir non seulement des changements concrets mais aussi des tendances evolutives. Nous partageons l'opinion selon laquelle il est possible d'etablir des prin-cipes generaux de changements linguistiques, definis de differentes fa5ons: principes, tendances, universaux, etc. Les arguments en faveur de cette these sont nombreux, ne serait-ce que ceux rappeles par Haspelmath (1999: 87). Ci-dessous nous nous occu-pons d'une methodes d'investigation de tels principes ou tendances. On a interet ä suivre le developpement des formes et fonctions dans un long espa-ce de temps, de preference ä partir de l'indo-europeen commun jusqu'aux langues les plus evoluees, telles que l'anglais ou le frangais (cf. Gawelko 2001a:40). Le terme »evoluee« veut dire ici »eloignee de la proto-langue«. Dans cette optique, on pourrait operer au moins trois coupes horizontales: la premiere serait representee par une langue ancienne, le sanskrit par exemple, la deuxieme par une langue slave, la troisieme par l'anglais ou le frangais. On sait que les langues slaves se sont eloignees moins de l'indo-europeen commun que les langues germa-niques ou romanes, en particulier l'anglais. Un inconvenient saute aux yeux: les langues modernes peuvent etre comparees sur la base d'un corpus satisfaisant quanti-tativement et qualitativement, les langues anciennes non. Le corpus defectueux est souvent une source d'erreurs, cf. l'objection faite par Disterheft ä tous ses predeces-seurs: »The confusion which is prevalent among previous descriptions of the Rigveda infinitive is consistent: small data collections have led to erroneous descriptions.« (1980: 34) Plus encore que l'ampleur du corpus, c'est sa representativite insuffisante qui constitue le danger pour le chercheur. Par exemple, Disterheft (1980: 13) parle du passif en avestan en se fondant sur des textes poetiques. La frequence de cette forme dans la langue parlee pouvait etre sensiblement differente. Les difficultes ne s'arretent pas au niveau du corpus, il faut ajouter differents pro-blemes resolus normalement par les locuteurs natifs, differents problemes de phone-tique, etc. Etant donne l'insuffisance du champ d'investigation offert par les vieilles langues, il y a tout interet ä suivre les tendances evolutives des langues accessibles ä une analyse multilaterale complexe et se trouvant ä des etapes evolutives differentes. En ce qui concerne les langues modernes, on dispose non seulement de textes nombreux et varies, y compris des textes paralleles, mais aussi de descriptions typologiques de langues particulieres (par ex. Albrecht 1970, Criado de Val 1972, Gak 1977, Hawkins 1986, Malblanc 1966, Manczak 1991, Vinay/Darbelnet 1958, etc.). L'etude des langues modernes peu evoluees et de langues modernes hautement evoluees, fon-dee sur un corpus representatif, est ä meme de completer 1'analyse historique excessi-vement sommaire des vieilles langues indo-europeennes. En principe, la reorganisation de la grammaire des langues indo-europeennes va dans la meme direction des le debut jusqu'ä nos jours. Ainsi l'etude des formes et fonc-tions apparaissant dans les langues modernes se trouvant ä differentes etapes evolu-tives permet de mieux determiner la specificite des tendances particulieres. La reorganisation de la grammaire peut etre reduite ä quelques elements fonda-mentaux: 1° la disparition ou la diminution, 2° 1'apparition ou 1'augmentation de 1'importance de certaines formes ou fonctions. Pour le premier cas on peut signaler le duel, l'impersonnel ou l'ordre de base OV, pour le second - l'infinitif, le passif ou l'ordre de base VO. Pour 1'instant un seul exemple suffit: l'infinitif ne parait pas avoir existe en indo-europeen commun (mais le probleme n'est pas definitivement resolu, cf. Jolly 1873:229, Garcfa-Ramön 1997:167). II est atteste dans les vieilles langues. II augmente sa frequence jusqu'ä 1'etape des langues les plus evoluees. Ainsi il presente un emploi modere dans les langues slaves, il est plus frequent dans les langues romanes et germaniques. Cependant la täche du linguiste consiste non seulement ä constater les faits mais aussi ä les expliquer. Aujourd'hui l'opinion predomine que ['explication des change-ments survenus laissait ä desirer tant qu'on negligeait l'importance du passage du type de langue OV > VO (cf. Lehmann 1974: 14). Prenons l'exemple du passif. En ce qui concerne 1'indo-europeen commun, nous pouvons accepter 1° avec Bauer (2000: 25) que l'opposition actif/passif n'etait pas denotee grammaticalement et 2° avec Lehmann (1974: 151) que le terme de passi-vum n'est pas bien fonde vu que les exemples cites n'ont jamais d'agent exprime. Dans les langues modernes il est denote au moyen d'une periphrase qui admet un complement agentif. En ce qui concerne son evolution, il convient de rappeler la constatation de F. Parker: »the change to VO word order acted as a causal agent in the development of the passive« (1976: 449). Cependant, un doute surgit: s'agit-il certainement, chez Parker, d'un lien causal entre les deux phenomenes ou peut-etre seulement d'un lien chronologique? Lehmann (1974: 183) constate seulement que ce n'est qu'apres le passage de OV ä VO que »the central meaning of the middle become that of passive voice«. Une autre remarque. La constatation de Parker, qui parle du lien entre le develop-pement du passif et le passage de OV ä VO, se rapporte ä des langues plus ou moins anciennes. Elle peut eventuellement expliquer la faiblesse du passif des vieilles langues (de type OV) par rapport au passif des langues slaves (de type VO), mais ne dit rien sur la faiblesse relative du passif slave par rapport au passif des langues romanes et plus encore par rapport au passif anglais. Nous pensons que c'est la tendance analytique qui caracterise les langues indo-europeennes des le moment de la proto-langue jusqu'ä Гetape des langues les plus evoluees, surtout anglaise et francaise. Le premier probleme ä etablir est celui de savoir si en effet les langues indo-euro-peennes s'acheminent vers l'analytisme. Differentes opinions ont ete avancees ä ce sujet (cf. Lehmann 1974:248, Harris 1978: 119-120, Wright 1983:15, Schwegler 1990:190, etc.). Pour notre part, nous avons essaye ä plusieurs reprises de montrer que les exemples fournis par differents chercheurs en vue de contester le progres de l'analytisme dans l'histoire des langues indo-europeennes s'expliquent par 1'action d'autres tendances, le plus souvent la tendance ä l'economie (par ex. Gawelko 2001a: 27-29, 2001b: 12-14). L'existence de differentes tendances n'elimine pas l'importance particuliere de la tendance analytique. On peut admettre que dans la langue differentes tendances se manifestent, souvent contradictoires, definies en anglais par des termes tels que conflicting goals. La question reste de savoir quelle est l'importance qu'on devrait assigner ä chacune d'elles dans le developpement des langues indo-europeennes. La tendance analytique permet d'envisager les trois coupes horizontales envisagees ci-dessus (les vieilles langues - les langues slaves - Г anglais et le frangais). Differents criteres motivent cette repartition: critere de Greenberg (la quantite des morphemes divisee par la quantite des mots), contingent des mots dans un texte parallele, contingent des mots ä fonction grammaticale, contingent des sujets pronominaux, contingent des prepositions, contingent des mots auxiliaires, proportion des mots generiques par rapport aux mots specifiques, etc. Ainsi, la tendance analytique est un facteur dont le pouvoir explicatif semble etre plus grand que celui du passage de OV ä VO: il explique, au contraire du passage de OV ä VO, la raison de la faiblesse relative du passif slave (de type VO) par rapport au passif anglais (aussi de type VO). Cependant, le meme doute persiste pour les deux facteurs: s'agit-il d'un lien causal entre ces facteurs et le developpement du passif? On connait bien la motivation de la frequence considerable du passif dans les langues depourvues de declinaison: dans les langues slaves on dit correctement, comme en pol. Matka kupila sukienkg »la mere a achete une robe« et Sukienkg kupila matka, le sujet et Гobjet y etant definis par les desinences. Dans les langues sans declinaison, on commence la phrase par l'objet en employant un passif ou une dislocation: La robe a ete achetee par la mere ou La robe, la mere I'a achetee. Un argument en faveur du lien causal entre la tendance analytique et le developpement de certaines formes et fonctions est apporte peut-etre par le fait qu'un certain nombre de ces elements evolue au fur et ä mesure qu'evolue la tendance analytique. Nous presentons quelques exemples, fondes tantot sur des travaux dejä publies tan-tot sur un corpus dont les details sont indiques ci-dessous. Le corpus I (presque 400 000 mots graphiques pour chaque langue prise en compte: allemande, anglaise, espagnole, frangaise, italienne, polonaise, portugaise et roumaine) se compose de 6 ouvrages entiers (4 ouvrages frangais: CH, E, P, V, 1 ouvrage anglais: A et 1 ouvrage latin: IC), des 3 premiers chapitres d'un ouvrage portugais (ML) et de fragments de deux ouvrages polonais (CD et Q). Pour chaque ouvrage, nous disposons de l'original et de traductions en sept langues (pour IC - en huit langues). Une exception: le texte CD n'a pas de version portugaise, ce qui nous a force de prendre en compte un fragment de la version portugaise d'un autre ouvrage (BA) aux dimensions äquivalentes ä Celles du texte manquant. Trois de ces ouvrages sont rediges dans une langue proche de la langue parlee: A, ML et CD. La langue de IC est quelque peu archai'sante. Le corpus II se compose, pour chaque langue, de fragments des neuf ouvrages indi-ques plus haut. De chaque ouvrage, nous avons retenu le fragment initial et le fragment final, chacun comportant, dans la version francaise, 2500 mots graphiques, ä ceci pres que nous ne coupons pas les phrases. Le corpus frangais comporte ainsi 45057 mots graphiques, au lieu de 45000 si les phrases avaient ete coupees. Le corpus francais une fois etabli, nous avons etabli aussi le corpus des autres langues qui est ä chaque fois compose des fragments equivalents. En principe, nous opposons le polonais, hautement synthetique, ä 1'anglais et au frangais, analytiques. Cependant profitant de notre corpus, nous determinons egalement si les differences typologiques secondaires qui caracterisent les cinq langues romanes presentent un rapport avec les tendances evolutives de certaines formes et fonctions. Les trois exemples qui suivent sont destines ä montrer le rapport entre quelques tendances evolutives et le degre de synthetisme/analytisme de la langue oü elles se mani-festent. L'infinitif. II provient d'une forme nominale figee (cf. p.ex. Jeffers 1975:135). Cependant, au cours de son evolution, il s'approprie graduellement certaines particu-larites verbales telles que l'expression de l'aspect (fr .faire - avoir fait, pol. robić - zro-bić), de la voix (ix. faire - etre fait), la presence d'un objet (fr. lire un livre, cf. il lit un livre). Sporadiquement il peut aussi etre accompagne d'un sujet syntaxique au nomi-natif (ex. prt. Chegou, por jim, o tempo de ela partir para Lisboa. - ML 86). Sa ver-balite augmente avec la possibility de se combiner avec un auxiliaire en formant des temps composes, etc. (cf. aussi Sala, ed., 2001: 277). Un infinitif evolue est un infini-tif qui comporte des categories grammaticales bien developpees propres au verbe fini. Le entere de base c'est le nombre global des infinitifs actifs et passifs etabli dans notre corpus II: 1 prt. 1512 occurrences, 2 esp. 1441, it. 1431, fr. 1404, 3 ang. 1105,4 all. 1018, 5 pol. 903, 6 roum. 185. Conformement ä la pratique courante, ces donnees ne contiennent pas l'infinitif formant le futur et le conditionnel. Les infinitifs passifs sont beaucoup plus rares que les actifs. Leurs occurrences ont ete etablies sur la base du corpus I, presque 10 fois plus grand que le corpus II: 758 exemples anglais, 280 allemands, 276 frangais, 269 portugais, 263 italiens, 176 espa-gnols, 101 polonais et 91 roumains. Ces chiffres presentent un certain rapport avec la vitalite du passif dans ces langues (cf. Gawelko 1999: 14). Un exemple tire de IC: lat. De ardenti amore et vehementi affectv svscipiendi Christvm. - fr. Du desir ardent de recevoir Jesus-Christ. - prt. Do grande e ardente desejo de receber a Jesus Cristo. -esp. Abrasado amor y vivas ansias de recibir a Cristo. - it. L'ardente amore e l'intenso desiderio di ricevere Cristo. - roum. Iubirea fierbinte §i dorul mare de a primi pe Hristos. - ang. On Ardent Love and Eager Desire to Receive Christ. - all. Die brennende Liebe und starke Sehnsucht, Christus zu empfangen. - pol. Zarliwa mitošć i gwaltowne poiq-danie przyjqcia Chrystusa. Seuls le latin et le polonais n'emploient pas l'infinitif. II y a deux conclusions ä tirer. En premier lieu, les grandes differences typologiques qui se manifestent entre le polonais et 1'anglais ou le franc ai s presentent un rapport etroit avec la vitalite de l'in-finitif: celui-ci est beaucoup plus faible en polonais que dans ces deux langues. On peut dire plus: il est beaucoup plus faible en polonais que dans n'importe quelle langue romane et, parait-il, aussi dans n'importe quelle langue germanique. En deuxieme lieu, les differences limitees qui se manifestent dans le degre de syn-thetisme/analytisme caracterisant les langues romanes presentent un rapport lache avec la vitalite de l'infinitif. Ce n'est pas le frangais, le plus analytique, mais le portugais qui jouit de la plus grande vitalite de son infinitif. Ce fait confirme l'importance de fac-teurs secondaires. On peut tenter d'expliquer la position de l'infinitif frangais en rappelant que la marche du frangais vers l'analytisme, comme celle de n'importe quelle langue romane, n'est pas homogene: il se separe des langues soeurs dans certaines des categories, tandis que dans d'autres il reste une langue typiquement romane. L'infinitif portugais non seulement s'avere le plus frequent mais aussi il subit le plus la tendance ä la grammaticalisation propre ä l'infinitif indo-europeen: il prend relativement souvent un sujet au nominatif (cf. Gawelko 2005), qui est plus rare dans les autres langues romanes, et tres rare ou inexistant dans les autres langues. L'infinitif portugais prend aussi des desinences, ce qui est un phenomene tres rare. Par contre, la derniere etape du developpement de l'infinitif indo-europeen: la substantivation au moyen d'un article n'y est pas aussi bien representee qu'en italien (cf. Gawelko 2004). Bref, les donnees ci-dessus confirment notre these de cette fagon que l'infinitif polonais s'avere beaucoup plus pauvre que l'infinitif de n'importe quelle langue romane (le roumain excepte; son appartenance ä l'union des langues balkaniques explique suffisamment bien la grande faiblesse de son infinitif) ou germanique. L'infinitif polonais ne prend pas un sujet au nominatif, il n'est substantive qu'en exergant une fonc-tion, celle de sujet par exemple. Par contre les differences typologiques limitees entre les langues romanes ont un rapport plutot lache avec le degre de developpement de l'infinitif. Bien entendu, dans les vieilles langues l'infinitif etait beaucoup plus faible que l'infinitif slave: provenant d'une forme nominale figee il etait en train de devenir infinitif. Le passif du type: etre + participe. Nous rappelons des donnees d'apres notre travail sur le passif roman (Gawelko 1999: 14): ang. 1301 occurrences, all. 928, fr. 785, roum. 776, it. 671, prt. 626, esp. 514, pol. 426. Ces chiffres ainsi que les constatations faites par Bauer, Lehmann et Parker, rappelees plus haut, conduisent ä quelques conclusions generates. Le passif n'a pas ete exprime grammaticalement en indo-euro-peen commun. Son developpement a ete accelere avec le passage du type de langue OV au type VO. II est done plus developpe dans les langues slaves, de type VO, que dans les vieilles langues. Dans les langues romanes et germaniques il augmente sensiblement sa frequence par rapport au passif slave. La position privilegiee du passif anglais ainsi que la position du passif frangais s'expliquent par l'analytisme de ces langues. C'est surtout la position de l'allemand qui devrait etre expliquee. Une constatation analogue ä celle faite au paragraphe precedent: les differences typologiques relativement secondaires ont un rapport plutot lache avec la frequence du passif. Un exemple: fr. On lui avait fait un lit dans la piece commune. (P, p. 167) - prt. Tinham-lhe feito uma cama no compartimento comum. - esp. Le hicieron una cama en la habitation comün. - it. Gli avevano fatto un letto nella stanza in comune. - roum. Iifäcu-serä un pat in odaia comunä. - ang. He was given a bed in the living-room. - all. Im gemeinsamen Aufenthaltsraum war ihm ein Bett aufgeschlagen worden. - pol. We wspölnym pokoju zrobiono mu posianie. Seules les langues germaniques emploient un passif. L'impersonnel. II a existe dans la proto-langue. II apporte un argument en faveur de la these que celle-ci etait, ä une epoque ancienne, une langue de type actif (cf. Bauer 2000: 335). Cependant, au cours de revolution des langues indo-europeennes, il dimi-nue son importance. L'argument connu en faveur de cette opinion est que, dans les langues analytiques, le modele phrastique de base SVO connait de moins en moins d'exceptions. Un autre argument est qu'on a releve pour certaines langues une ou des epoques oü une grande quantite d'impersonnels ont ete remplaces par des formes per-sonnelles (pour l'anglais, on se fonde en general sur van der Gaaf, par ex. Moessner 1984). Un autre argument est apporte par des indo-europeistes. Ainsi, B. Bauer constate: »Impersonal verbs are found in all Indo-European languages; the more one goes back in time, the more one will find.« (Bauer 2000: 97). Un exemple type: fr. J'avais chaud sous mes vetements sombres. (E, p. 26) - ang. I was hot under my dark clothes. - all. Ich schwitzte in meinem dunklen Zeug. - esp. Tenia calor con mi traje oscuro. - it. Ave. v o caldo sotto i miei vestiti scuri. - prt. Tinha calor, com o meu fato escuro. - roum. Imi era cald in hainele mele de culoare inchisä - pol. W ciemnym ubraniu byto mi gorqco. Seuls le roumain et le polonais emploient un impersonnel. Ä la rigueur, nous pourrions dejä tirer une conclusion ä partir de ces remarques, qui serait de nature ä appuyer notre these. Cependant notre corpus I est susceptible d'ap-porter une certaine contribution ä la determination de la vitalite de l'impersonnel. Nous presentons les constructions du type etre + predicat adjectival/adverbial qui ne se rap-portent pas aux phenomenes atmospheriques, comme dans ang. It is so good to be with Him! (ML). Le corpus offre 184 exemples fran to execute successfully. The main task of the players who are not 'it' is to . The players who are not 'it' also have other tasks active in parallel, such as monitoring. This involves constantly paying attention to who is 'it' at any given moment. The conditional role-part property of being 'it' is a dynamically changing one, frequently alternating its value throughout the game. The parallel task of monitoring who is 'it' needs to be active at all times. The participant who performs the activity of tagging is referring to the other participant's expectation procedures which are subsequently triggered. That is why we speak of referring behavior rather than just reference. First, a certain set of tasks is executed in one participant; their execution then results in triggering a set of tasks in another participant. The two sets of tasks in the two participants have to be compatible in the sense that they relate to the same domain of control (the tradition calls this the context of situation). The act of tagging triggers the relevant set of expectation procedures in the tagged participant on condition that the relevant set of expectation procedures is ready to be triggered and the related tasks ready to be activated, in which case the participant's behavior would change accordingly. By referring to a participant's domain of control and relevant expectation procedures, the communicating individual is assuming and at the same time validating his assumption that the other communicating individual has a corresponding domain of control with similar concepts. In the process of communicating, individuals are constantly monitoring for validation of each other's behavior (the other participant's nods, responses and other expressions of agreement and understanding) which gives them reassurance that similar concepts are being accessed in the plexes^ of their partners in communicating, forming a similar domain of control. Therefore human communicating is based on expectation procedures in participants involved in a relevant communicative situation which are that the other participants have the concepts forming the relevant domain of control. How do we account for successful referring? Although people share some common ground, they come from different backgrounds, which brings us to the question of how it is at all possible for any human communication to be successful, since our personal histories differ and thus contribute different input into the context of situation. Other linguists have expressed this fear for the process of making linguistics scientific. Bloomfield, for instance, feared that 'the occurrence of a speech [...] dependfs] upon the entire life-history of the speaker and of the hearer' (Bloomfield, in Yngve 1996: p.170). But the HL theory eliminates these fears because it finds that communicative behavior is lawful, which means that it is possible to discover the laws related to any given piece of communicative behavior, and thus achieve a scientific understanding of it (Yngve, 1996:p.l71). People expect 6 In HL, plex is a representation of the structure of a communicating individual as a long list of procedures all interrelated by categorial and conditional properties in a complex dynamic causal network (Yngve 1996:171). conformity to norms and easily overlook deviations or actually hear them according to their expectations (cf. Burazer 2004, SLE, Norway), otherwise human communication would in most cases result in failure. Success of referring behavior is conditioned by the domain of control in the participants in a particular linkage. This does not imply that participants have the same life history, which is of course impossible, but we are rather counting here on the principle of equivalent componential histories relating to the particular domain of control (Yngve 1996:169). People leacrn by observing and discovering patterns in human behavior. On the basis of the observed behavior in the tag linkage, for instance, the basic principle of the game of tag can in HL be formalized as (2): (2) [X] X [Y]<-it> X [X](Y) :: [X]<-it> X. [Y] The above is a formalization of the basic procedural property of the tag linkage. It represents the participants' activities in the process of tagging, which is a repetitive pattern of communicative behavior in the tag linkage. The participants X and Y have the property of being it/-it at a certain value at a certain point in the game. The participant who is 'it' touches the other participant, who is '-it', which sets the participant X's property of being 'it' to change to '-it', and the participant Y's property of '-it' to 'it'. The discovery of patterns of communicative activities can also be sensed behind Austin's and later Searle's approach to analyzing human communication in terms of speech acts. Speech acts are a theoretical representation of patterns of human behavior underlying interpersonal communicative activities. The Speech Act Theory approach In line with the Speech Act Theory, the activity of tagging can be interpreted as the player performing a perlocutionary (performative) speech act: 'you're it' really means 'I tagged you'. This behavior is rule governed, since there are certain rales of the game that the players observe. Searle's hypothesis was that speaking a language is a matter of performing speech acts according to systems of constitutive rales (1969:38). Prior to stating this hypothesis he compares human behavior in games such as football, baseball and chess (1969:36; this analogy has been used prior to Searle by Saussure and Witgenstein) to human behavior in certain social situations. Violating the rales of a game results in negative attitude towards the particular player or team. Generally, in social behavior such violations or adherence to the rales (of etiquette) would represent the basis for appraisal of behavior, such as 'he was rude', 'he was immoral', 'he was polite', which Searle tentatively poses as evidence for existence of rule governed human behavior. We can claim with some certainty that there exists evidence from everyday examples that human behavior is 'rule governed'7. These 'rules' are forms of socially 7 Note the use of inverted commas to mark that this expression in itself represents a domain confusion and as such is not a part of terminology used in Human linguistics. accepted behavior. Apart from the criminal behavior defined by the law, people are not legally bound to behave in a certain way, but society as such dictates certain ways of accepted behavior, and consequently condemns deviations from these ways as unacceptable. It is human nature to have the need to belong and be accepted by the society, therefore we can note in human behavior conformity to norms. The rules of socially accepted behavior undergo considerable changes through time. Numerous examples from everyday situations support this statement. In the past, before the existence of modern telecommunication systems, such as telegraphs, telephones and the electronic media, people used to communicate indirectly through messengers or through the postal service. So, for instance, arranging to meet someone took quite a while if one wanted to come announced, in addition to observing certain norms of social behavior. Then, with the invention of the telegraph, getting a message through to someone became considerably easier and quicker, and even more so with the invention of the telephone. Showing up at someone's doorstep all of a sudden became incredibly rude - at least in my experience. The social norm became ' You should call first.' If one did happen to show up at someone's doorstep, one would have noticed a change in communication from 7 'm so glad you're home' to suddenly apologizing 'Sorry, I know I should have called first.' Nowadays, in the time of mobile telephones and electronic mail communication, it might be unimaginable to just show up at someone's doorstep, even if you did just happen to pass their house. The rules of social behavior dictate that it is polite and expected of you to call first, even if you're standing right in front of the person's doorstep. The availability of modern means of communication has changed people's perception of what can be expected of the other person in terms of making interpersonal communication easier, so a certain level of predictability of human behavior is expected. Similarly, expectations have changed in terms of decoding the meaning of sounds such as music heard in the middle of the street or in a classroom situation, or a person 'talking to themselves' in a coffee shop. Some time ago, these instances would have probably been understood as some sort of deviations from the social norms of behavior. Nowadays, the first expectation when music is heard seems to be that it must be someone's mobile telephone. And if one hears someone walking behind them and talking loud, one expects they are talking to someone on the phone rather than addressing them or talking to themselves. Validation and appraisal of our past behavior results in forming expectation procedures in an individual's plex which are then triggered in a particular situation, accessing appropriate related concepts and further triggering appropriate task procedures, in accordance with the individual's social and cultural experience. Conclusion The words reference and referring constitute a part of terminology in many different fields of research and we, the readers/listeners, manage to adjust their specific meanings in accordance with the field of research in question. Definitions from philosophy imply that it is words that refer to or point out objects and other worldly phenomena. But in human linguistics it is not words but people who refer or point to something. In human linguistics, participants do not refer to elements or words in the text. It is concepts in the hearer's or reader's domain of control that are accessed. They are concepts of items in the external reality or in the imagined external reality or abstract concepts. A certain given communicative behavior may trigger access to different concepts at different times depending on the domain of control. But in human linguistics, referring behavior is concerned with the triggering of procedures in dependence on the context rather than with people using words (Yngve 1996:277), as in traditional linguistics. In human linguistics, communicating individuals are the participants in the process of communicating, and their linguistic, social, and emotional properties are involved in their understanding of referring behavior. Traditional analysis of grammatical structures and lexis reveals the surface layer of the message. It does not, however, explain the phenomena behind successful communication in cases such as two friends using unfinished sentences and cue words, or in cases where the message carries double meaning, yet the receiver extracts the correct one, suitable to the particular context of situation. In order to be able to account for how people communicate, a much wider framework is needed—one that reaches beyond surface grammatical structures, or even the context of situation—one that takes into account all factual elements playing a role in successfully communicating. This requires that we move from the logical to the physical domain of investigation, and Human Linguistics is the theory equipped with the proper means to do so. Bibliography Austin. J.L. 1962. 'How to Do Things with Words'. 2nd ed. (first 1955) Cambridge: Harvard UP. Beaugrande, Robert-Alain de, and Dressier, Wolfgang. 1981. Introduction to text linguistics. London; New York: Longman. Burazer, Lara. 2004. 'Handling Context Dependence in a Human Linguistic Framework'. Paper at the 37"1 annual SLE meeting, ''Effects of mobility on language'. Kristiansand, Norway, 27th-31st July. Burazer, Lara. 2004.'Reconstituting Notions of Reference'. In Hard-Science Linguistics, Yngve and Wasik, eds. Continuum: London, New York. Cook, Guy. 2003. 'Applied Linguistics'. Oxford: Oxford University Press. Drake, Stillman. 1961. 'GALILEO'. Past masters series, Hill and Wang, New York, A division of Farrar, Straus and Giroux, 1980:71 Haliday, M.A.K. & Hasan, Raqaiya. 1976. 'Cohesion in English'. London: Longman. Halliday, M.A.K. 1985. 'An introduction to functional grammar'. London ; Baltimore, Md., USA : Edward Arnold. Hamilton, E. and Cairns, H., eds. 1961. 'The Collected Dialogues of Plato including the letters'. Bollingen Series LXXI, Pantheon Books. Bollingen Foundation, New York, N.Y. Saussure, Ferdinand de. 1974. 'Course in General Linguistics' (trans. Wade Baskin). London: Fontana/Collins. Schiffrin, Deborah. 1994. 'Approaches to discourse'. Oxford: Blackwell. Searle, John R. 1969. 'Speech Acts - An Essay in the Philosophy of Language'. CUP: London, New York. Shakespeare, William. 1987. 'Romeo and Juliet'. UK: OUP. Todorov, Tzvetan. 1970. 'Introduction ä la litterature fantastique', Paris : Editions du Seuil. Yngve, Victor H. 1996. 'From Grammar to Science- New Foundations for General Linguistics'. John Benjamins Publishing Company: Amsterdam - Philadelphia. Yule, George. 1996. 'Pragmatics'. Oxford: Oxford University Press. Povzetek POJMOVANJE REFERENCE (NANAŠANJA) IN SITUACIJSKEGA KONTEKSTA S STALISČA JEZIKOSLOVJA ČLOVEKA Referenca oz. nanašanje je pojem, ki se pogosto pojavlja tako v filozofskem kot v jezikoslovnem diskurzu. V preteklosti je bila referenca definirana kot odnos med besedo ali izrazom in predmetom v resničnosti oz. med besedo in pojmom v abstraktni resničnosti. Kljub nekaterim porajajočim se dvomom v to filozofsko osnovo jezikoslovne znanosti, je pojmovanje, da je jezik predmet obravnave v jezikoslovju, ostalo zakoreninjeno. Victor H. Yngve s svojo teorijo jezikoslovja človeka (Human Linguistics, tudi HL) predstavlja korak izven okvirov jezikoslovja jezika, kjer je predmet znanstvene obravnave s stališča trdih znanosti neotipljiv. Za predmet znanstvene obravnave postavlja človeka in njegove odzive v različnih komunikacijskih situacijah. Obravnava reference s stališča jezikoslovja človeka na primeru preproste otroške igre lovljenja omogoča vpogled v zapleteno mrežo človeškega razumevanja v situacijskem kontekstu. Rezultati obravnave kažejo, daje človeško medsebojno komuniciranje, v vsej svoji zapletenosti in sočasnemu vplivu številnih dejavnikov, moč znanstveno formalizirati v okviru orodij, ki jih ponuja HL. Kljub nesporni raznolikosti izkušenj, si ljudje delimo številne družbeno sprejete vedenjske vzorce, ki predstavljajo skupni imenovalec človeškega delovanja in zagotavljajo uspešnost komunikacije. Martina Ožbot University of Ljubljana UDK 81'25 :81'27 ON THE POSSIBILITIES AND IMPOSSIBILITIES OF CROSS-CULTURAL COMMUNICATION 1. Introductory remarks As is well known, the second half of the last century has witnessed an unprecedented increase in cross-cultural communication at a practical level as well as a remarkable development of research on various aspects of translation as cross-cultural communication par excellence. Such an interest in the study of translation appears to be directly linked with the expansion of translational activities and reflects the importance attributed to them in the society at large. At the same time, the burgeoning growth of translation studies is to be explained within the context of the expansion of the discipline of linguistics over the past half-century, an important part of which is the development of various text-oriented branches in which attention has been given to previously largely unstudied phenomena of the functioning of language in real communicative situations. The aim of this contribution is to present an overview of the development of translation studies and also to show that it was originally made possible by function-centred approaches to research on language. An attempt will be made to sketch out the principal differences between traditional thinking about translation - which was concerned almost exclusively with the dichotomy between literal and free translation on the one hand and with the problem of translatability between natural languages on the other - and some contemporary theories which have to a large extent overcome the traditional conceptual oppositions. This is true in particular of the functional approaches to translation, in which categories such as translatability, literalness and freeness have acquired only a very relative value and are considered in terms of the concrete purpose which a given translation is expected to serve in the target culture. The new perspective has also implied a redefinition of some fundamental translational issues such as the relation between translation and text production, the question of the translator's skills, and the concepts of translational "equivalence" and of translation error. 2. Translation: a growing area of research and an expanding practical activity For centuries, translation has attracted the attention of numerous thinkers who have been interested in various aspects of it.1 But it has happened only relatively recently that systematic studies with solid theoretical underpinning have been dedicated to translation as an object of empirical study and not merely as an object of reflection A variety of monographs and readers dealing with translation theories are available in English. See, for instance, Lefevere (ed.) (1992) for an overview of traditional ideas and Gentzler (1993) and Venuti (ed.) (2000) for presentations of modern approaches. based on intuition and experience. Apart from the growing interest in translation as a research field, the activity of translation itself is becoming more and more prominent in today's world. Everywhere, and especially perhaps in small cultures and in the non-English speaking world, translated texts can be found, directly or indirectly, in the greatest majority of communicative situations. The increasing dependence on translations in our every-day lives may at first sight appear paradoxical, since parallel to the growth of translational activities there is also an increase in foreign language learning, a consequence of which is that the number of people who have a working knowledge of at least one foreign language is becoming higher and higher. However, the two phenomena, i.e. the growing amount of translational activities and the expansion of foreign-language learning, are to be considered complementary rather than oppositional, since both are related to the ever-greater intensity of international (and therefore intercultural and interlingual) communication. The importance attributed to (high-quality) translation in today's world is also reflected in the growing number of translator- and interpreter-training university departments and institutes. However, the increase in translation research and in translational activities does not necessarily imply that any substantial change in the quality of translated texts has taken place. In actual fact, a lot of inappropriate translations are still being produced, and the reasons for their poor quality have remained largely the same for centuries: they concern, among other things, the translators' insufficient source-language and/or target-language competence, their insufficient knowledge of the textual world, their not taking into sufficient consideration the addressee and their lack of methodological preparation for their task. Likewise, the theoretical reflection on translation may also appear not to have taken any big steps forward over time. 3. Translation theory through history The opinion that during the two millennia of its existence translation theory has largely been in a state of stagnation is not an isolated one and has been expressed by many scholars. One of them, George Steiner, in his well-known work After Babel observes that thinking about "the art and theory of translation" has been moving in a vicious circle, since a lot of attention has been dedicated to the dichotomy "literal translation vs free translation" on the one hand and to the similarly useless and unproductive question of whether interlingual translation is at all possible (Steiner 1975: 238 ff.). Let us at this point have a closer look at the two issues. 3.1. Literal translation vs free translation The basic opposition between literal translation and free translation2 goes back to the ancient dichotomy verbum vs sensus, which can be encountered in authors such as j The opposition "literal translation vs free translation" can be seen as vaguely related to the distinction between formalism and functionalism in linguistics. In actual fact, however, it has been re-elaborated in such a variety of ways, in which issues other than those of form and function have played a central role, that the analogy does not appear crucial either for our understanding of the development of the translational dichotomy through history or for the applications which are made of it today. As is the case with the majority of the dichotomies used in the study of language, this one too has a merely provisional value and does not reflect the reality of language in its actual functioning (cfr. Barbe 2001). Horace and St Jerome and which is analogous to the older distinction between interpres and orator introduced by Cicero. Over the centuries, the dichotomy "literal vs free translation" was recycled a number of times and still appears to have preserved some of its explanatory potential. Through history, different periods and different authors displayed different preferences for either of the two oppositional terms and put forward different interpretations of the dichotomy. For instance, in ancient Rome, literary translators tended to favour free rather than literal translation, which can be explained by their emulation of Greek authors. The free-translation principle was carried to extremes in the 1 8^-century French culture where elegance and compatibility with target-culture taste were the most important criteria of a good translation. This gave rise to radical shifts from the source text and to the production of the so-called belles infideles. New versions of the free-translation principle have also been proposed by contemporary scholars like, for instance, Eugene A. Nida, who has mainly investigated the translation of the Bible3 (see Nida 1964). Different views were dominant, for instance, in the German culture at the end of 18^ and the beginning of 19 centuries. According to Johann W. Goethe, Wilhelm von Humboldt and Friedrich D. E. Schleiermacher, to name only the most representative thinkers, translations should make the target readers aware of the difference between the source language and culture on the one hand and the target language and culture on the other, and to achieve this objective structural adaptations should be made by virtue of which the target language will be a reflection of some characteristics of the source language. In the process of translation, the target language and culture are thus enriched by the input from the source language and culture, by means of which new linguistic and cultural models are imported. This line of thought was continued by a variety of theoreticians such as Walter Benjamin - most notably in his essay "The task of the translator", ("Die Aufgabe des Übersetzers", 1923) - and more recently by Jacques Derrida, the author of the now classic treatise "Des Tours de Babel" (1985). Today, a major representative of the foreignizing principle in translation is Lawrence Venuti (see Venuti 1998), the author of a source-oriented approach, who suggests that translation be non-fluent and non-ethnocentrical in order to convey a sense of foreignness of the source text and of the source culture. Related to the dichotomy "literal translation vs free translation" - which allows endless possibilities of interpretation - is also the concept of the translator's fidelity, which is extremely flexible and slippery, since a translator who is faithful may translate either literally or freely, depending on what exactly one takes the concept of fidelity to mean; in principle, it is equally applicable to those who translate according to the principle of "literal translation" (in any of its possible forms) and to those who follow the principle of "free translation" (again, in any of its possible forms). What is more, in spite of substantial differences between the single versions of the original dichotomy, all the oppositions are ultimately based upon the idea that, unavoidably, in translation either the A thorough and systematic presentation of issues involved in Bible translation is offered in Krašovec (ed.) (1998). source or the target side is favoured. In other words, it is either the reader that makes an effort to come close to the target text, in which case the latter remains more firmly rooted in the source culture, or it is the translation that "gets close" to the reader, in which case the link of the target text with the source side is weaker. Although in the past decades translation theory has often centred upon other issues, the opposition between literal translation and free translation seems to have preserved some explanatory and heuristic value. It has been used over and over again - often also by theoreticians who do not favour either of the given terms - under old labels (faithful translation VS unfaithful translation; translation according to the letter vs translation according to the spirit) as well as under new ones such as innovative translation vs conservative translation, philological translation vsfree translation, source-orientated translation vs target-oriented translation, domesticating translation vs foreignizing translation, naturalization vs exotization (van Leuven-Zwart 1990: 75), static equivalence vs dynamic/functional equivalence (Nida 1964), overt translation vs covert translation (House 1977: 188-204), non-illusory translation vs illusory translation (Levy, quoted in Pym 1992: 178), documentary translation vs instrumental translation (Nord 1991a: 105-106; 1991b: 11,72-73; 1997: 47-52), direct translation vs indirect translation (Gutt 1990: 149 ff.), semantic translation vs communicative translation (Newmark 1982: 22-23), adequate translation vs acceptable translation (Toury 1995: 56-57), opaque style vs transparent style (Snell-Hornby, quoted in Vannerem/Snell-Hornby 1986: 191), observational reception vs participative reception (Pym 1992: 178), etc. If, however, one takes a look at the reality of concrete translations and at their actual functioning in communicative situations, one can soon realize how unproductive it is to insist either on the principle of free translation or on the principle of literal translation. It is obvious that neither relatively literal nor relatively free translations are able to give a realistic idea of the source culture, language and text. Any target text can never be more than an approximation of source text, which implies that "equivalence", at any level from that of the lexeme to that of the text, is no more than an illusion.4 3.2. The (im)possibilities of translation The second issue which frequently preoccupied those who dealt with questions of translation in the past - and is to an extent also present in contemporary research on translation - concerns the problem of the very possibilities of interlingual translation. The main argument against the possibility of translation are the differences between natural languages in the ways in which they conceptualize the extralinguistic reality, which, according to some, implies that a message in language A can never be reproduced in language B. Behind such beliefs is the concept of "linguistic relativity"5 put forward by American anthropological linguists, most notably by Benjamin Lee Whorf in the 4 The concept of "equivalence" has been dealt with extensively. See Snell-Hornby 1986: 13-16, 1988: 13-22, 1992: 21, Kußmaul 1986: 225 and Nord 1991b: 22 ff.; a different view is offered in Koller 1995. 5 The problem of linguistic relativity and translation is examined in Pedersen 1988: 30-46. See also Kay 1996: 110-111. 1940s, although similar ideas are much older, as is generally known; they were debated by exponents of the German classical tradition such as Johann Gottfried Herder and Wilhelm von Humboldt as well as by numerous earlier philosophers like Francis Bacon, John Locke, Giambattista Vico, Gottfried Wilhelm Leibniz, the authors of the Port Royal grammar and others, and can be traced down to Plato's dialogue Cratylus. In more recent times, the question of translatability has been posed again; for example, Willard van Orman Quine in his well-known essay "Meaning and Translation" (1959) speaks about the "indeterminacy of translation", which is due to the lack of immediate conceptual correlations shared by individuals intra- and interlinguistically. Indirectly, a tiny shadow of the untranslatability concept may perhaps also be present in modern translation theories inasmuch as they consider the idea of equivalence totally inadequate in terms of the actual reality of translation. Certainly, the principle of linguistic relativity - which may hold equally for intra-and for interlingual communication - cannot easily be discarded: if it has not yet been fully proved, it has not been disproved either, nor is it likely to be in the future. But in spite of all the impossibilities of our understanding each other, interpersonal communication does seem to function, however imperfect it may be, and our messages do get across day by day, intra- and interlingually. Therefore, linguistic relativity may be considered a phenomenon which has to be taken into serious account in the study of human communication in general, it is, however, by no means an obstacle which would invalidate the feasibility of translation as such. 4. Beyond possibilities and impossibilities The two issues discussed in the previous section may be regarded as major impediments which, over the centuries, have slowed down the development of translation studies to a considerable degree. On the other hand, it is necessary to realize that important progress in the study of translation could not be made before a theory of language became available which was broad enough to encompass the reality of the functioning of language in general, including all kinds of interlingual and intercultural communication. Once such a theory was formulated with the advent of text- and function-centred approaches to the study of language, it was possible for translation research to start to develop rapidly. As a consequence, one can now notice considerable differences between traditional reflection on translation and modern theoretical approaches to it, which can be summarized as follows: 1. Although modern approaches to translation are based in some measure on older ideas about translation, they are characterized by coherence and systematicity by virtue of which they have the nature of proper theories, i.e. frameworks with explanatory and predictive value. 2. Traditional reflection on translation was concerned almost exclusively with literary texts, besides, of course, Biblical texts. Modern translation theory, on the other hand, takes into consideration texts in general, literary as well as non-literary. This makes it possible for translation theory to be of use in translator-training curricula and, consequently, to have an impact upon the practice of translation itself. 3. Instead of dealing with the two above-mentioned issues (i.e. the question of translatability and the opposition between literal and free translation") modern translation theory, and in particular its functional branches, concentrate upon questions of the communicative appropriateness of translations in concrete target-culture situations, in which they are expected to function, as well as upon the role of the translator as a cross-cultural text producer. As a result, it appears that we are now in a position to understand better what translation is actually about and what exactly are the principles upon which the work of good translators is based. As has been pointed out, the growth of translation studies over the past decades can only be explained in the light of the development of the study of language as communication. It is understandable that in pretextual linguistics - just as in traditional literary studies - the phenomenon of translation as a genuinely textual activity could not be of primary concern to the researcher,6 whereas at a practical level "translation" was for centuries used (and still is) as a method of foreign-language teaching and learning. As such, however, it cannot really be compared to the activity of cross-cultural communication, since in the language classroom the method of "translation" is typically applied at levels below that of the text and usually no consideration is given to contextual factors and to the pragmatics of the actual communicative situations, which play such a central role in real-life translation (cfr. Kvam 1992: 196, Snell-Hornby 1986: 25-26). Since the functions of the two activities differ starkly, it is no exaggeration to say that they hardly have in common anything more than the name (cfr. Jakobsen 1994: 144, Snell-Hornby 1992: 18). With the advent of suprasentential linguistics in the 1950s and 1960s the situation changed radically, since more and more attention began to be dedicated to the actual language use ("parole") as opposed to language as a system ("langue"; cfr. de Beaugrande 1978: 8, 95,98; de Beaugrande 1991: 21-22 passim; Kußmaul 1986: 224; Ivir 1996: 151-156). Not surprisingly, the "pragmatic turn" ("die pragmatische Wende", Kvam 1992: 193; cfr. Hönig/Kußmaul 1982: 34) opened up new possibilities for translation research. Within this framework, translated texts, literary and non-literary alike, were immediately considered an interesting object of investigation. It was in such an atmosphere that the functional approach to translation started to develop in the early 1970s, mostly at some German centres for the training of translators and interpreters. The approach is best represented by the so-called skopos theory ("Skopostheorie")7 which was formulated by Hans J. Vermeer and Katharina 6 An exception were, for instance, early efforts in machine translation made in the 1950s by a group of American scholars, among whom were also Noam Chomsky and Victor Yngve (cfr. Yngve 1996: 38-39, 44-45). 7 Apart from skopos theory, functionalism is also represented in some other approaches to translation which are independent of it. Among their authors are, for instance, Roger T. Bell, Basil Hatim and Ian Mason (see Bell 1991, Hatim/Mason 1990, Hatim/Mason 1997). However, these scholars do not offer their own general theories of translation which would be as broad and axiomatic as the skopos theory. Mention must also be made of Gideon Toury, who has developed a genuinely function-centred translation theory, which is devised almost exclusively for the study of the translation of literary texts and does not have any practical aims in the field of translator training. For a presentation of the differences between Vermeer's theory and his own see Toury 1995: 25-27. A complex function-centred framework, which has various points in common with the skopos theory and which, unlike Toury's approach, aims to be applicable to the training of translators, is Justa Holz-Mänttäri's "theory of translational action" (see Holz-Mänttäri 1984). Reiß and from which a number of theoretical models were derived. The basic premise of the skopos theory - which can be seen as part of a broadly conceived action theory of human behaviour according to which human beings always act in a purposeful way - is that in an act of translation the translator's decisions are, pro to typically, determined by the function the target text is expected to have in the target culture,8 which implies that her/his work is guided by all kinds of contextual and pragmatic considerations (see Ožbot 2001: 139-140). If a translator is about to translate a text such as, for instance, Shakespeare's Romeo and Juliet or Hamlet, he/she will proceed differently when the translation is intended for the stage and when it is meant to be used, as a radical adaptation, for school children in their textbooks, to mention just two out of a virtually infinite number of options9. The skopos theory has been modified by various researchers, most notably perhaps by Christiane Nord (Nord 1991b: 28-30; 1997: 124-125), who, besides functionality, also postulates the category of loyalty. Nord agrees that it is important to produce target texts on the basis of functional considerations (target function, target culture, target readership), but, she adds, translators must always be loyal to "their partners in translational interaction", at "the source and the target sides" (Nord 1997: 125), otherwise any target function could be attributed to any text-to-be-translated. At an applied level, the skopos theory has also been developed further by other scholars, like, for instance, Hans Hönig and Paul Kußmaul, who, as teachers of translation, have dedicated themselves primarily to research on the translator's skills, and have emphasized in particular the importance of his/her self-confidence and self-awareness, as well as of ethical issues in translation (see Hönig/Kußmaul 1982, Hönig 1995, Kußmaul 1995). As an eminently functional approach to translation, the skopos theory is target-oriented, i.e. it is centred upon the text-to-be made. This can to some extent account for its great potential for being applied at a practical level. Since an act of translation always involves an attempt to produce a target text which can fit into a given target situation, it is the target rather than the source side that is necessarily of primary importance in translational practice. Translation, like text production in general, is essentially determined by contextual and pragmatic factors relevant in the concrete situation in which it is supposed to function. As a consequence, the relationship obtaining between the source text and the target text can display various degrees of closeness, but, in principle, the former is no more than a blueprint for the construction -of the latter; it is always the target linguistic and extralinguistic situation rather than the source text itself which determines what the final product of the translation process o The essence of skopos theory is expressed in the sentence "Die Dominante aller Translation ist deren Zweck" (Reiß/Vermeer, 1991:96). 9 Such an example shows that in order to perform well as cross-cultural communicators translators must possess a variety of different skills, among which source-language competence and target-language competence should not be overestimated. Equally important are also source-culture competence and target-culture competence, knowledge of the source textual tradition and of the target textual tradition, knowledge of the textual world and, last but not least, translation competence in a narrow sense, i.e. professional knowledge about what it actually means to translate and how to do it (Ožbot 1997). will be like. It is not surprising that within such a concept of translation, translation error is not identical to language error: the translator commits an error when he/she produces a target text which is functionally inappropriate either as a whole or in one or more of its parts. In view of all this, the two issues presented above, which occupied a central position in traditional translation theory, lose virtually all their significance when considered in the light of a functional approach to translation. The opposition between literal translation and free translation appears trivial unless it is considered in terms of the function a given translation is supposed to perform in the target culture and even then the two options are no more than the translator's "working techniques", with a merely relative value. Likewise, the translatability question is not regarded as theoretically pertinent at all. If the process of translation, which essentially consists in message mediation, is taken as a primarily cultural and not linguistic phenomenon, what is being translated are texts or textual contents, not words as such; or, to say it with Coseriu, it is what is extralinguistic or, better, "non language-specific" ("übereinzelsprachlich"; Coseriu 1978: 20) that is conveyed in translation. 5. Translation studies and linguistics By way of conclusion, let us examine briefly the nature of the relationship between the study of translation and linguistics, which is a complex one. This is due to a number of interconnected reasons, among which the following two seem to be particularly important: first, if it is true that the development of translation studies was made possible, as this paper has tried to show, only after new research perspectives on language had began to establish themselves, it is equally true that at the time when such new perspectives were being developed langue-centred traditional linguistics was still dominant. This is why early endeavours of translation scholars to have their subject accepted as an academic (sub)discipline often met with opposition, which created an impression of an irreconcilable discrepancy between the study of language on the one hand and the study of translation on the other. This antagonism lasted for decades and the 1980s and 1990s saw a heated debate on the epistemological status of translation studies (see Snell-Hornby 1988: 7 ff., 1996: 18 passim, Honig 1995: 16; cfr. also Ožbot 1998) which has only recently began to calm down (Baker 2001). Second, given that in any translational activity language is fundamental as a means to which translations as texts own their existence and that, at the same time, translation involves much more than "just" language, translation studies may indeed appear to be a discipline which transcends linguistics. However, the presence of other elements besides language does not determine only translational activities, but is a fundamental characteristic of the production of texts in general. Therefore it seems reasonable to conclude that the study of translations as texts necessarily falls within the scope of a broadly conceived discipline concerned with the study of texts (cfr. Reiß/Vermeer 1991: 1-2, Coseriu 1978), both at a theoretical and at an applied level. References Baker, Mona (2001) The Pragmatics of Cross-Cultural Contact and Some False Dichotomies in Translation Studies. CTIS Occasional Papers I, 7-20. Barbe, Katharina (2001) The dilemma with dichotomies. Language & Communication 21, 89-103. Beaugrande, Robert-Alain de (1978) Factors in a Theory of Poetic Translating. Assen: Van Gorcum. Beaugrande, Robert-Alain de (1991) Coincidence in Translation: Glory and Misery Again. Target III/1, 17-53. Bell, Roger T. (1991) Translation and Translating: Theory and Practice. London/New York: Longman. Coseriu, Eugenio (1978) Falsche und richtige Fragestellungen in der Übersetzungstheorie. In: L. Grähs et al. (eds), Theory and Practice of Translation. Bern/Frankfurt am Main/Las Vegas: Peter Lang, 17-32. Gentzler, Edwin (1993) Contemporary Translation Theories. London/New York: Routledge. Gurr, Ernst-August (1990) A Theoretical Account of Translation - Without a Translation Theory. Target II/2, 135-164. Hatim, BasilAan MASON (1990) Discourse and the Translator. London/New York: Longman. Hatim, Basil/Ian MASON (1997) The Translator as Communicator. London/New York: Routledge. Holz-mänttäri, Justa (1984) Translatorisches Handeln: Theorie und Methode. Helsinki: Suomalainen Tiede-akatemia/Finnish Academy of Science. House, Juliane (1977) A Model for Translation Quality Assessment. Tübingen: Narr. Ivir, Vladimir (1996) A Case for Linguistics in Translation Theory. Target VIII/1, 149-157. Jakobsen, Arnt Lykke (1994) Starting from the (other end): integrating translation and text production. In: C. Dollerup/A. Lindegaard (eds), Teaching Translation and Interpreting 2. Insights, Aims, Visions. Amsterdam/Philadelphia: John Benjamins, 143-150. Kay, Paul (1996) Intra-speaker relativity. In: J. J. Gumperz/S. C. Levinson (eds), Rethinking linguistic relativity. Cambridge: Cambridge University Press, 97-114. Koller, Werner (1995) The Concept of Equivalence and the Object of Translation Studies. Target VII/2,191-222. Krašovec, Jože (ed.) (1998) Interpretation of the Bible. Ljubljana/Sheffield: Slovenska akademija znanosti in umetnosti/Sheffield Academic Press. Kubmaul, Paul (1986) Übersetzen als Entscheidungsprozeß. Die Rolle der Fehleranalyse in der Übersetzungsdidaktik. In: М. Snell-Hornby (ed.), 206-229. Lefevere, Andre (ed.) (1992) Translation/History/Culture: A Sourcebook. London/New York: Routledge. Leuven-zwart, Kitty M. van (1990) Translation and Original. Similarities and Dissimilarities. II. Target II/l, 69-95. Newmark, Peter (1982) Approaches to Translation. Oxford: Pergamon. Nida, Eugene A. (1964) Toward a Science of Translating: With Special Reference to Principles and Procedures Involved in Bible Translating. Leiden: E. J. Brill. Nord, Christiane (1991a) Scopos, Loyalty, and Translational Conventions. Target III/l, 91-109. Nord, Christiane (1991b) Text Analysis in Translation: Theory, Methodology, and Didactic Application of a Model for Translation-Oriented Text Analysis. Amsterdam/Atlanta: Rodopi. Nord, Christiane (1997) Translating as a Purposeful Activity: Functionalist Approaches Explained. Manchester: St. Jerome. Ožbot, Martina (1997) Ustreznost sporočila v slovenskem prevodu Zgodovine italijanske književnosti A. Momigliana. (Magistrska naloga.) Ljubljana: Univerza v Ljubljani, Filozofska fakulteta, Oddelek za romanske jezike in književnosti. Ožbot, Martina (1998) Linguistics and the Study of Translation. In: B. Caron (ed.), Proceedings of the 16th International congress of linguists. Oxford: Pergamon, paper no. 0446. Ožbot, Martina (2001) Making sense across cultures: the establishment of coherence in translated texts. In: Janet Cotterill/Anne Ife (eds), Language across boundaries. London: British Association for Applied Linguistics/Continuum Press, 135-150. Pedersen, Viggo Hj0rnager (1988) Essays on Translation. K0benhavn: Nyt Nordisk Foriag Arnold Busck. Pym, Anthony (1992) The Relations Between Translation and Material Text Transfer. Target IV/2, 171-189. Reib, Katharina and Vermeer, Hans J. (21991) Grundlegung einer allgemeinen Translationstheorie. Tübingen: Niemeyer. Snell-hornby, Mary (ed.) (1986) Übersetzungswissenschaft - eine Neuorientierung. Tübingen: Francke. Snell-hornby, Mary (1986) Übersetzen, Sprache, Kultur. In: M. Snell-Hornby (ed.), 9-29. Snell-hornby, Mary (1988) Translation Studies: An Integrated Approach. Amsterdam/Philadelphia: John Benjamins. Snell-hornby, Mary (1992) The professional translator of tomorrow: language specialist or all-round expert? In: Cay Dollerup/Anne Loddegaard (eds), Teaching Translation and Interpreting. Training, Talent, Experience. Amsterdam/Philadelphia: John Benjamins, 9-22. Snell-hornby, Mary (1996) Translation und Text. Wien: WUV-Universitätsverlag. Steiner, George (1975) After Babel: Aspects of Language and Translation. Oxford: Oxford University Press. Toury, Gideon (1995) Descriptive Translation Studies and beyond. Amsterdam/Philadelphia: John Benjamins. Vannerem, Mia/Mary SNELL-HORNBY (1986) Die Szene hinter dem Text: 'scenes-and-frames semantics' in der Übersetzung. In: M. Snell-Hornby (ed.), 184-205. Venuti, Lawrence (1998) The Scandals of Translation: Towards an ethics of difference. London/New York: Routledge. Venuti, Lawrence (ed.) (2000) The Translation Studies Reader. London/New York: Routledge. Vermeer, Hans J. (1996) A skopos theory of translation. Heidelberg: TEXTconTEXT-Verlag. Yngve, Victor H. (1996) From Grammar to Science: New Foundations for General Linguistics. Amsterdam/Philadelphia: John Benjamins. Povzetek O MOŽNOSTIH IN NEMOŽNOSTIH MEDKULTURNE KOMUNIKACIJE V članku je predstavljen nastanek prevodoslovja kot raziskovalnega področja, ki ga je omogočil razvoj nadstavčnega jezikoslovja v zadnjega pol stoletja. Nakazane so poglavitne razlike med tradicionalno prevodno refleksijo - v kateri sta zavzemala osrednje mesto dihotomija med dobesednim in svobodnim prevajanjem in vprašanje o tem, ali je prevedljivost med naravnimi jeziki sploh mogoča - in sodobno teorijo prevajanja funkcijske usmeritve, ki so pogojene prav s spremembami v jezikoslovni raziskovalni paradigmi. Ko je jezikoslovje začelo proučevati poleg jezikovnega sistema tudi jezikovno rabo in s tem besedilno delovanje jezika v realnih komunikacijskih situacijah, so postala raziskovalno relevantna tudi vprašanja o prevodih kot besedilih oz., natančneje, o značilnostih in zakonitostih njihovega tvorjenja in funkcioniranja, medtem ko so se z nove perspektive problemi tradicionalne prevodne refleksije, kakršna sta omenjena dva, izkazali za povsem obrobne in so zadobili le relativno, predvsem hevristično in razlagalno vrednost. Članek se zaključi z razmišljanjem o razmerju med jezikoslovjem in prevodnim raziskovanjem, pri čemer je slednje umeščeno v okvir široko zasnovane vede o besedilih. Karmen Teržan Kopecky Universität Maribor UDK 81'25 : 8Г23 TRANSLATE ALS KOGNITIVE ENTITÄTEN ODER DIE NATÜRLICHKEITSTHEORIE IM KONTEXT DER SPRACHLICHEN UND PSYCHOSOZIALEN PARADIGMATA DER MEHRSPRACHLICHEN KOMMUNIKATION 1.0 Einleitung Jahrtausende alt ist das Gefühl bzw. - gepaart mit Überlegungen und Reflexionen - die spätere Überzeugung von Philosophen, Literaten und Denkern, dass der ins Wort gefasste Gedanke den Gedanken selbst beeinflusst, verändert, ihm zu einer neuen, wenn auch vergleichbaren Existenz verhilft, aber dennoch den intendierten Sinn in entsprechender Kombination mit allen möglichen - auch nichtsprachlichen - Faktoren zu vermitteln und an Menschen zu transportieren vermag, die uns schon dadurch mehr oder weniger nahe stehen, dass sie die natürliche, soziale und kulturelle Umwelt mit uns teilen, aber unter gewissen Voraussetzungen auch an jene, die anderen Kulturkreisen, Religionen, ja Welten (und sei es nur im Sinne von Generationen) angehören. Transferiert und mutiert durch Gedanken, Gefühle und Erfahrungen einzelner Menschen, sozialer Gruppen und Gemeinschaften oder ganzer Generationen verbreitet sich der Gedanke im ständigen Wandel von Konventionen, Sprachen und Kulturen. Vergleichbar mit einem realen oder fiktiven Universum ist er in Sprache gefasst in seinem Wesen allen zugänglich (universell), jedoch gleichzeitig von Fall zu Fall unterschiedlich als Eigentum der jeweiligen Kultur, Glaubens- oder Sozialgemeinschaft, ja sogar des einzelnen Menschen, gefärbt von dessen eigener, persönlicher Lebensgeschichte, die jedoch von der genetischen Geschichte des Menschen aufgenommen wurde als das kollektive, in die genetische Struktur gefasste Gedächtnis der Menschheit, das sich durch die Evolution von Beginn an bis heute entfaltet hat. Kein Erlebnis, kein jemals gedachter Gedanke geht je endgültig unter; vielmehr bildet er die Basis für weitere Veränderungen, für Wandel, Ausbreitung, Bereicherung, aber auch Schwund und Untergang, wenn er sich im Sinne Darwins als nicht lebensfähig erwies bzw. für den ewigen Kreislauf des Lebens - des Wortes - unter Umständen sogar störend hätte sein können. Mit dieser etwas poetisch anmutenden Einleitung in die nüchterne Materie der Linguistik wollte ich eigentlich nur darauf verweisen, in welch vielfältiger Weise das komplexe System Sprache an alle Modalitäten des menschlichen Lebens - sei es seiner Geschichte oder seiner Gegenwart - verpflichtend gebunden ist. Diese komplexen Verbindungen ergeben sich schon unter Angehörigen einer Sprach- und Kultur-gemeinschaft, geschweige denn, wenn es darum geht, die in einer Sprache festgehaltenen Intentionen in andere Sprach- bzw. Kultursysteme umzusetzen bzw. - in den Worten Andrew Chestermans (Chesterman, 2000) - in den Kulturkreis des Zielpublikums auszubreiten. 1.0 Damit habe ich nun die mittelnde oder vermittelnde komplexe Tätigkeit des translatorischen Handelns (Übersetzen/Dolmetschen) zur Sprache gebracht und gleichzeitig Platz eingeräumt für jene Sprachtheorien, die über das sprachliche System hinaus denken, es in die natürliche, kulturelle und soziale Umwelt einbinden sowie mit den psychosozialen Gegebenheiten seiner Benutzer in Relation bringen. Funktionalismus, kognitive Theorien und die Natürlichkeitstheorie (konkret: die slowenische Theorie der sprachlichen Natürlichkeit: STSN) bilden in der Tat jenen theoretischen Rahmen des gegenwärtigen Aufsatzes, der die Grundlage darstellen wird für die Suche nach Berührungspunkten und gemeinsamen Erklärungsansätzen, aber auch Divergenzen zum Thema mehrsprachige Kommunikation. 2.0 Über den gemeinsamen Nenner der Informationsmittlung wird auf die Kernfrage der sprachlichen Verarbeitung und Speicherung von für kommunikative Zwecke verwendbaren Propositionen eingegangen. Um das Thema weiter zu führen und die Arbeitshypothesen aufstellen zu können, werden vordergründig die Erklärungsmodelle der Systemtheorie, wie sie in der kognitiven Linguistik Anwendung findet, und des STSN bemüht. Gibt es Überlappungen, wenn ja, worauf sind sie zurückzuführen und inwiefern verschärft sich durch das oben angedeutete spezielle Wissen die Optik für den Beobachter des sprachlichen / kommunikativen Handelns? Kann auf diese Weise auch in translatorische Probleme mehr Klarheit gebracht werden? 3.0 Auf die o.a. Fragenkomplexe wird in der weiteren Folge unter Berücksichtigung konkreter translatorischer Handlungen innerhalb des slowenischdeutschen und deutsch-slowenischen Sprachenpaares eingegangen. 4.0 Den ersten Berührungspunkt zwischen den obgenannten Theorien finden wir in der ihnen gemeinsamen Behauptung, dass es sich bei allen die Kommunikation mitgestaltenden Ebenen um natürliche und interagierende Systeme handelt (Kognition, Sprache, Kultur). 5.0 Die kognitive Linguistik (Langacker1991, 1999; Rickheit-Strohner 1993; Karpf 1990) sieht ihren Ausgangspunkt in der Informationsverarbeitung. Im Mittelpunkt steht der Mensch mit seinem (auch sprachlichen) Eingreifen in die Umwelt zum Zweck seines (besseren) Überlebens, vor allem im Sinne seiner sozialen und kulturellen, aber auch biologischen Bedürfnisse. Indem der Mensch seine Umgebung mit seinen Sinnen wahrnimmt, verändert (modifiziert) er sie zugleich auch mit seinem bereits vorhandenen Weltwissen und seiner sprachlichen Kompetenz (Wahrnehmung als Erfindung = KONSTRUKTIVISMUS). Somit greift er auch mit der Sprache gestaltend in seine Umwelt ein, indem er die kommunikativen Ziele verwirklicht. Die Welt, auf die sich die beiden natürlichen Systeme Sprache und Kognition beziehen, ist demnach prinzipiell für alle Benutzer natürlicher Sprachen in ihren wesentlichen Zügen vergleichbar - in dieser inhärent gegebenen universellen Komponente liegt die Übersetzbarkeit von Sprachen überhaupt begründet -, gleichzeitig ist sie aber typologisch Sprachgruppen, Sprachfamilien) und individuell (Einzelsprachen) unterschiedlich. 5.1 Auch die Natürlichkeitstheorie (NT) geht vom Universellen, Prototypischen und Einzelsprachspezifischen aus, wobei alle Ebenen sowohl mit der Umwelt als auch untereinander interagieren, um neue, effiziente Kommunikationsmöglichkeiten zu schaffen, sich anzupassen und/oder zu verändern. Dabei zeichnen sich vor dem Hintergrund der Umwelt universelle Eigenschaften beider Systeme ab. 5.2 Die NT (Orešnik 1994, 1999, 2000, 2003, 2004; Mayerthaler 1983, 1987; Wurzl 1987, Dressler 1987,1990; Teržan Kopecky 1997,1998,2001) arbeitet mit folgenden Postulaten: * Hoher kognitiver Aufwand resultiert in hohen Markiertheitsstufen seiner Verbalisierang ((morpho)syntaktischer Konstraktion). Die natürliche und erwartbare Folge einer solchen Konstellation ist eine niedrige Token- und Typen-Frequenz und im Vergleich zu weniger markierten Konstruktionen, die auch weniger komplexe funktionale (kognitive) Komponenten transportieren, ein niedrigerer Grad der Natürlichkeit. * Evolutionsmäßig, im Sinne der kategorialen Genese und den dabei durchlebten Grammatikalisierangsphasen, kann die Steigerung der Token- und Typen-Frequenz einer M+Konstraktion ein Hinweis sein auf die Verringerung ihrer Markiertheitswerte bei gleichzeitiger Steigerung ihres Natürlichkeitsgrades (die Konstraktion wird üblich und ist bereits eine typische Konstituente des jeweiligen grammatischen Systems). * Der Natürlichkeitsgrad von Konstraktionen steht in einer Abhängigkeitsbeziehung zum jeweiligen Alter der Konstruktion. In diesem Sinn spricht man von Grammatikalisierangsphasen, wobei die späte oder Postgrammatikalisierungsphase durch Frequenzsteigerung und Markiertheitsreduzierung signalisiert wird. Die Letztere bezieht sich im STSN auf die Markiertheitsgrade des mikrogrammatischen Umfeldes von Konstruktionen. Eine spezielle Ausprägung der STSN ist als Resultat statistischer, im Sinne der Korpuslinguistik angelegter Auswertungen umfangreicher deutscher und slowenischer Korpora (darunter auch zahlreiche Übersetzungstexte) als Repräsentanten unterschiedlicher Entwicklungsperioden von untersuchten grammatischen Kategorien (Grammatikalisierangsphasen) aber auch Textsorten erarbeitet worden. Diese Auswertung hat die Generierung des folgenden Postulats herbeigeführt: STSN: M+ = MGK+ (= relativ hohe Markiertheitsgrade von Konstruktionen werden in der Regel /im natürlichsten Fall von relativ hohen Markiertheitsstufen des mikrogrammatischen Kontextes begleitet (.Reaktion des mikrogrammatischen Kontextes auf die Komplexität/Natürlichkeit der Konstruktion)) * Frequenzsteigerung und Markiertheitsreduzierung können auch vom Auftreten der komplexen syntaktischen Konstraktion in für sie natürlicheren Textsorten ausgelöst werden. Auch hier ist, vor allem in frühen Grammatikalisierangsphasen, ein Ausgleich der Markiertheit des Kontextes und der Konstraktion der Normalfall. Erst die Steigerung der Token-Frequenz und das zunehmende Alter der Konstruktion ebnen den Weg zu weniger komplexen textuellen Welten. Das oben Vorgeführte zeigt, dass auch die angesprochene STSN die Sprache als ein natürliches, mit der sozialen und kulturellen Umgebung interagierendes und stark an die kognitiven Prozesse gebundenes System betrachtet. 6.0 Die kognitive Theorie geht davon aus, dass für die Erkennung/Verarbeitung von Informationen ein integrales System zur Verfügung steht, dessen Effizienz sich aus dem Zusammenwirken seiner Komponenten ergibt, wodurch die zwei Welten in ein unteilbares Ganzes verschmelzen: die philo-genetische (Erbmasse; durch neuronale Vernetzungen vollzogene Filterung von eingelangten Informationen, ihre Analyse und Gruppierung zu Modulen; genetisch bedingte Arbeitsverfahren des Neocortex als des zentralen Prozessors) und die onto genetische (Interaktion mit der Umwelt, Ein-knüpfung von psychosozialen Paradigmen in die Erkennungsprozesse). Bei den angesprochenen Komponenten (vgl. Petri 1994; Kess 1992; Karpf 1990, Karpf, Dringel-Techt 1995) des kognitiven Systems handelt es sich um a/ die Information und den zentralen Prozessor (Hirnrinde mit ihren integrierenden Funktionen, die bereits von den internalisierten, aus der sozialen Umwelt aufgenommenen Informationen mitbestimmt sind und das s.g. Weltwissen konstituieren: unter anderem Kommunikationskonventionen, Eigenarten des Textproduzenten und des Textrezipienten, ihre Erwartungshaitun -gen, Verhaltensverpflichtungen, usw.) b/ Uniwelt der beiden Zentralkomponenten (Objekte, die mit ihnen funktional verbunden sind) c/ Struktur (Relationen zwischen den Komponenten) d/ Funktion (Beziehung des Systems zur Umwelt) Die oben angeführten Komponenten des natürlichen Systems Kognition erzeugen das Verhalten des Systems, in welchem sich Zustände, vor allem aber auch konstante Veränderungen abzeichnen. 6.1 Die Veränderungen zeigen sich als Dynamik des kognitiven Systems, was gleichzusetzen ist mit Veränderung von Zuständen in der Zeit unter gleichzeitiger Auswirkung auf das Verhalten des Systems (Rückwirkung) (Langacker 1991, 1999; Rickheit-Strohner 1993 u.a.). Die die Dynamik erzeugenden Systemkomponenten zerfallen im Wesentlichen in zwei Gruppen: Die autonomen Teilsysteme erzeugen ihre Dynamik bzw. die Veränderung des (K)Systems ohne Abstimmung mit der Umwelt; die interaktiven Teilsysteme stimmen sich hingegen mit der Umwelt ab. Beide kann man als Funktionen des ARBEITSGEDÄCHTNISSES (Aktivierung nach Wichtigkeit und Häufigkeit der bereits erfolgten positiven Aktivierung) sehen. Somit kann man auch sprachliche Veränderungen zumindest teilweise als Folgen der spezifischen Arbeitsweise des Gedächtnisses auffassen, welches selbst als eine phylogenetische Gegebenheit entscheidend an der Gestaltung der natürlichen Systeme (wie Sprache und Kognition) mitwirkt, indem es an der so genannten Selbstorganisation (vgl. Karpf 1990), in deren Rahmen es als zentraler Prozessor in ständiger Wechselwirkung mit der Umwelt steht, maßgeblich mitbeteiligt ist. 6.2. Entscheidend für die Erzeugung der Systemdynamik sind folgende Prozesstypen: A/ Die Steuerung als Basisprozess der Informationsverarbeitung bedeutet umgesetzt in die Kommunikationssituation, dass der Textproduzent die Zustände und das Verhalten des Textrezipienten steuert. B/ Die Regelung sorgt für die Wiederherstellung eines Zustandes bzw. seine Beibehaltung trotz Einwirkungen der Umwelt. Es handelt sich dabei um ein besonders ausgeklügeltes Gestaltungsprinzip, das eigentlich für die GEGENWEHR DES SYSTEMS und seine STABILITÄT zuständig ist. Es verhindert die „Überdosierung" der Einwirkungen der Umwelt auf das K-System und funktioniert als eine Art Kontrollsystem zur Erhaltung des minimalen Gleichgewichts der Zustände des sprachlichen und des kognitiven Systems. 6.2.1 Hier ist zum Beispiel jene „Notbremse" angesiedelt, die verhindert, dass z.B. die Abschwächung der Endsilben die Wurzelsilben „angreift", was Wörter zu Silben zurückführen würde, und die meines Erachtens unter anderem auch dafür sorgt, dass der sprachliche Innovationsprozess nicht ausschließlich die Richtung Markiertheitsabbau einschlägt, sondern des öfteren auch die Markiertheitsumkehr beobachtbar ist, was für das systeminterne Gleichgewicht sorgt. Aus diesem Grund wird diesem Prozesstyp im gegenwärtigen Aufsatz die meiste Aufmerksamkeit gewidmet, da zum einen gerade hier ein relevanter Verknüpfungspunkt mit der STSN angesiedelt ist, und zum anderen üblicherweise dieser Verarbeitungsebene weniger Aufmerksamkeit geschenkt wird, weil sie ausgesprochen diskret wirkt. 6.2.1.1 Die folgenden Beispiele aus empirischen Untersuchungen zur o.a. natürlichen/universellen Gesetzmäßigkeit, die man bei L-, bei K-, aber auch bei T(ranslations)-Prozessen beobachten kann, veranschaulichen auch die unter Punkt 5.2. vorgestellten Postulate. Sie beziehen sich speziell auf die Markiertheitsverhältnisse zwischen Ausgangs- und Zieltexten im Rahmen der aufgenommenen und ausgewerteten sprachlichen Handlungen des konsekutiven Dolmetschens (mehrerer von der Autorin aufgezeichneter und statistisch bearbeiteter Ereignisse zu unterschiedlichen Zeiten und Themenschwerpunkten). Gleichzeitig geben die unten angeführten statistischen Daten Aufschluss über die übereinzelsprachlichen (universellen), aber auch prototypischen grammatischen Parameter, die sich auch von den sonstigen kulturell und sozial bedingten L-spezifischen Erscheinungen, die gerade bei Translationsprozessen besonders gravierend in den Vordergrund treten, dadurch abheben, dass die Unterschiede durch die regelnde Tätigkeit des Sprachsystems größtenteils getilgt werden -explizit an jenen Stellen des grammatischen Systems, wo sich universelle Prinzipien der NT geltend machen. 6.2.1.2 Von den aufgenommenen und transkribierten Texten, entstanden bei unterschiedlichen Handlungen des Konsekutivdolmetschens mit folgenden Variablen (Thema, ZT-Dt.; ZT-Sl; AT-Produzenten), wurden folgende, übereinzelsprachlich als konstant angenommene (= Konstituenten der universellen sprachlichen Ebene) Parameter untersucht. Die Vergleiche des Verhaltens der mikrogrammatischen Konstellation zeigten bei untersuchten grammatischen Parametern mit Ausnahme der typologisch bedingten Distributionseigenheiten keine relevanten Abweichungen in folgenden Fällen (AT(sl/dt) - ZT(sl/dt): • Zahl der gebrauchten Nebensätze (ist offensichtlich eine nicht typologische oder L-spezifische Eigenschaft) • analytische Konstruktionen (40%sl : 60% dt) (Die geringere Abweichung zu Gunsten des Deutschen ist offensichtlich L-spezifisch bedingt (= das Dt. als bevorzugt analytische, das SI. eine vorwiegend synthetische Sprache = hypothesenfreundliches Resultat.) • Auftretenshäufigkeit des Präsens 70% (übrige T 30%) M+ Kollok. = 35% (Es handelt sich um eine hypothesenfreundliche Gegebenheit im Sinne der STSN, weil das im Tempus-Paradigma weniger markierte Präsens generell (übereinzelsprachlich) höhere Token-Frequenzen aufweist, jedoch relativ niedrigere Markiertheitsstufen der kollozierenden grammatischen Parameter bevorzugt.) Vergangenheitstempora 15% (davon 40% in NS) (Auch hier zeigt sich eine hypothesenfreundliche Gegebenheit im Sinne der STSN, weil die im Tempusparadigma relativ markierteren Vergangenheitstempora erwartungsgemäß eine niedrigere token-Frequenz haben als die präsentischen Tempora, dafür aber in stärkerem Maße die Umgebung von komplexeren grammatischen Parametern präferieren, wie etwa die Nebensätze (NS).) W+I-Futur 7% 60% in NS (Auch beim analytischen Futur zeigten sich hypothesenfreundliche Resultate im Sinne der STSN (sowohl in den Ziel- als auch in den Ausgangstexten). Sie weisen nämlich als markiertere Konstruktionen noch niedrigere token-Frequenzwerte auf als die Vergangenheitstempora, dafür steigt ihre Kollokationstendenz mit den komplexeren Nebensätzen um einen beträchtlichen Prozentsatz an.) PF 2% 45% inNS (Noch niedriger ist die token-Frequenz bei stark markierten Belegen des Präsens pro Futuro, das als nicht-präsupponierte Struktur (vor allem im Slowenischen) im Schnitt beider Sprachen die niedrigste Frequenz aufweist, dazu eine relativ hohe Kollokation mit markierteren Nebensätzen. Man würde in diesem Fall zwar noch höhere Präferenzen des Nebensatzes erwarten, was aber m.E. durch den diesbezüglichen typologischen Unterschied zwischen dem Deutschen (im Deutschen sind die PF-Belege weitaus üblicher als im Slowenischen) und dem Slowenischen verursacht wird.) 62.12.1Untersucht wurden auch die Auftretenshäufigkeiten des PASSIVS als einer komplexen syntaktischen Konstruktion, die relevante Abweichungen zwischen AT (Ausgangstext) und ZT(Zieltext) ausweisen: ZTsl = 1,5% ZTdt = 3,0% (Die Ursache dieses Ergebnisses ist sprachtypologischer Art, da das Deutsche die Passivkonstruktionen typologisch bedingt bevorzugt.) Die token-Frequenz des Passivs bzw. seine Distribution auf Nebensätze (NS) in den Ausgangs- und Zieltexten zeigt das folgende Bild: AT + ZT = 50% in NS (Den Grand für eine solche Ausgeglichenheit zwischen AT und ZT sehen wir im oben explizierten Prinzip der NTS, wonach eine universelle Affinität von komplexen (M+) Kategorien zur komplexeren grammatischen Umwelt besteht (vgl. Orešnik 2004). Nebensätze sind demnach in beiden Sprachen, der slowenischen und der deutschen, gleichermaßen beliebte (grammatische) Mikro-Umwelten für das kognitiv komplexe Passiv). Man kann davon ausgehen, dass der Mechanismus der Regelung tilgend eingreifen würde, wenn dieser Grundsatz aus irgendeinem Grund gestört wäre. 6.2.1.2.2 Untersucht wurde auch die Anzahl der gebrauchten Sätze im Allgemeinen und die Zahl der gebrauchten Nebensätze in den AT und den ZT. Folgendes Resultat konnte eruiert werden: ZT=M+ = mehr Sätze (27%) + mehr komplexere NS als in AT (Durch die allgemein relativ höhere kognitive Markiertheit der ZT (Anspannung, Konzentration, mehr Verarbeitungsoperationen) werden dort mehr markiertere Nebensätze gebraucht. Der Grundsatz „M+ präferiert M+" hat sich auch in diesem Fall bestätigt. Dass sich die Waage zu Gunsten der ZT geneigt hat, hat eine pragmatische Begründung (Umstände der Kommunikation, Lampenfieber, Überbrückungszeit, usw.) Eine pragmatische Gegebenheit, die sich beim Dolmetschen besonders deutlich zeigt, ist die Frequenz der PAUSENGESTALTUNG. Erwartungsgemäß sollte die Dichte der Pausen zu Gunsten der Zieltexte (längere Prozesszeiten, Anspannung usw.) umschlagen, vor allem, wenn diese nicht die für den Translator natürlichere Muttersprache sind. Die statistischen Ergebnisse zeigen folgendes Bild: ZT = generell um 74% mehr als AT (Dieses Ergebnis bestätigt die Vorhersage im Sinne der Gebundenheit an die kognitiven Verarbeitungsprozesse)) die wichtigsten I-Träger: 1/ vor NP = 45% 2/ vor Satzanfang = 30% 3/ vor VP = 15% 4/ vor sonstigem = 7% (Zwischen AT und ZT waren keine relevanten Abweichungen feststellbar, was auf den universellen Charakter der Pausensetzung hindeutet, die sich nach der Relevanz der Wörter im Aktivierungsprozess richtet. Dabei scheinen die NPs als die stärksten Informationsträger zu fungieren, dicht gefolgt vom Satzanfang, wo die Proposition, nachdem sie innerhalb des AT erkannt wurde, in die neue textuelle Umwelt richtig eingebettet werden muss. Fast um die Hälfte weniger wird vor den VPs gezögert (das entsprechende Verbum wird durch die Wahl des Nomens mitaktiviert), und noch einmal halbiert wird die Pausensetzung vor allen anderen Satzkonstituenten.) Im Bereich der systeminternen Regelung haben sich die oben explizierten Prinzipien der STSN durchweg auch iibereinzelsprachlich als geltend ausgewiesen. Weitere Komponententypen sind noch: C/ Die Handlung ist die Anpassung an die Umwelt zur Erreichung oder Aufrechterhaltung von Sollzuständen der Umwelt (d.h. aus internen Modellen der Umwelt werden erstrebenswerte Zustände der Umwelt gebildet). Diese enge, interaktive Verknüpfung der internalisierten Sollwerte und der tatsächlich erfolgten Reize aus der Umwelt ist auch als die oben bereits erwähnte Selbstorganisation bekannt und teilt darüber hinaus viele Erklärungsansätze mit dem Konstruktivismus, der Wahrnehmung als Erfindung versteht, wodurch die Kreativität des menschlichen kommunikativen Handelns noch besonders stark hervorgehoben wird). Dadurch wird der Handlungscharakter der sprachlichen und kognitiven Verarbeitung verdeutlicht als die weitere gemeinsame Eigenschaft beider Entitäten. Das erstrebenswerte Resultat ist die KOMMUNIKATIVE TÄTIGKEIT, die eine regulierende, objektivierende und abstrahierende Funktion unter gleichzeitigem Zusammenwirken der Teilprozesse von Handlungen (also von Motiven, Zielen, Folgen von Handlungen und Operationen) zu leisten hat. Auf diese Weise entsteht die MIKROSTRUKTUR der mit Hilfe der Sprache vollzogenen kommunikativen Tätigkeit. (Langacker 1999; Givon 1995; Chesterman 2000.) D/ Die Aushandlung setzt die Integration mehrerer KOMMUNIKATOREN mit Hilfe von Handlungen voraus. (Dies kommt bei den T(translatorischen) Prozessen noch in besonderer Weise zum Tragen.) Die Erreichung eines gemeinsamen Zieles durch die Partnerinteraktion ist auf dieser Verarbeitungsebene der ultimative Sollzustand. Ziele, Situationen und Informationen aus der inneren und äußeren Welt regen zu kooperativen Handlungen an. Somit führen Verständigung und erfolgsorientierte Einstellung zur AUSHANDLUNG bzw. Festlegung des intendierten Themas (Einigung zwischen dem Textproduzenten und dem Textrezipienten bezüglich Thema und Interpretation der Information unter Einbeziehung des kommunikativ relevanten Wissens). (Reickheit-Strohner 1993; Langacker 1991, 1999.) Im Sonderfall der T-Prozesse muss dieses Aushandeln zwischen dem Ausgangstextproduzenten und dem Ausgangstextpublikum stattfinden, des Weiteren zwischen dem Translator und dem Zieltext sowie dem Zieltextpublikum (= maximale Komplexität der Verarbeitungsprozesse). Das Ziel ist die Herstellung relevanter Ähnlichkeiten zwischen dem AT und dem ZT, bzw. die Mimesis des gleichen Textes in eine andere textuelle Umwelt (vgl. Wilss 1988; Chesterman 2000). Einige interessante Fälle von missglückten Aushandlungen zwischen Translator, Ausgangstext und Zieltext bzw. seinen Adressaten (bei denen sprachliche, zumeist semantische Asymmetrien nicht entsprechend getilgt wurden) möchte ich hier als Beispiele anführen (Als Übersetzer fungierten in den angeführten Beispielen Studenten.): (1) Wie beurteilen Sie den Preisverfall ... - Kako ocenjujete propad cen ... (Die missglückte Aushandlung beruht auf mangelnder ausgangs- und zielsprachlicher Kompetenz, auf fehlendem Wissen über ausganssprachliche und zielsprachliche Konventionen) (2) Sie ist bekennende Muslimin ... - Ona je znana muslimanka (Die missglückte Aushandlung beruht auf mangelnder ausgangs- und zielsprachlicher Kompetenz; es fehlt entsprechendes Wissen über die Kommunikationskonventionen in der Ausgangssprache und auch die Diskurskompetenz in der Zielsprache/Zielkultur weist Mängel auf) (3) Markenforscher .... über den schleichenden Tod der Marke - Raziskovalec znamke ... o plazeči se smrti blagovne znamke (Die missglückte Aushandlung beruht auf mangelnder ausgangs- und zielsprachlicher Kompetenz; es fehlt entsprechendes Wissen über die Kommunika -tionskonventioncn in der Ausgangssprache; die Diskurskompetenz in der Zielsprache/Zielkultur weist gravierende Mängel auf) Die angeführten Beispiele zeigen, dass die entsprechende Aushandlung vor allem bei translatorischen Prozessen die meisten Hindernisse auf den Weg stellt. Diese konzentrieren sich vor allem auf außersprachliche Gegebenheiten wie Diskurskompetenz, Kommunikationskonventionen, Kulturkonventionen, mangelhaftes Weltwissen und Ähnliches. 7.0 Versuchen wir nun noch näher auf die Berührungspunkte zwischen Translation und Kognition einzugehen, indem wir die o.a. Systemkomponenten als Vergleichselemente hinzuziehen. Es wird davon ausgegangen, dass ein derartiger Erklärungsversuch gleichzeitig die Frage zu beantworten vermag, warum eine solche Fragestellung für Untersuchungen von inter- und intralingualen Kommunikationssituationen und Überwindungen von Kommunikationsproblemen überhaupt produktiv sein könnte. 7.1 Die neurolinguistische Realität der Sprache gründet sich auf spezifische neuronale Verknüpfungen (Muster von Synapsen und neuronalen Verknotungen und ihre Aktivierungsmodalitäten), die im Langzeitgedächtnis der Sprachbenutzer angesiedelt sind und somit, wie bereits oben dargelegt, sowohl die universelle Komponente als auch die individuelle Komponente des sprachlichen Systems ausmachen. Jedoch sind diese biologischen Gegebenheiten nicht als fertiges Produkt vorgegeben (mit Ausnahme ihrer Beschaffenheit und Funktionsweise), sondern werden in der Interaktion mit der sozialen und kulturellen Umwelt in ihrer konkreten Ausführung fertiggestellt, jedoch nur um durch das ständige Interagieren der phylo- und ontogenetischen Komponenten ständigem Wandel/An -passung an neue Umstände und Bedürfnisse unterworfen zu sein. Die Folge davon ist: 7.2 Zwischen einer Sprache Ll(x) und einer anderen Sprache L2(x) entstehen mehr oder weniger große Abweichungen (biologisch, sozial, kulturell, geschichtlich bedingt). Auf diese Weise verbinden sich in jeder L universelle, prototypische und L-spezifische Komponenten/Merkmale. Dadurch sind Sprachen überhaupt übersetzbar, denn sie teilen sich die universellen Merkmale, die aus vergleichbaren Realitäten generiert werden. Die prototypischen Merkmale verbinden Sprachfamilien und basieren auf ähnlichen bzw. verwandten Realitäten. Letztlich gibt es aber Eigentümlichkeiten, die nur einer einzigen L angehören und sie von allen anderen in der ihr typischen Weise abheben, weil sie von der ihr eigenen Kultur, von sozialen Konventionen, Normverstoßen, Diskurskonventionen usw. hervorgebracht wurde. 7.2.1 Zwischen einer beliebigen Sprache Ll(x) und einer anderen beliebigen Sprache L2(x) liegen unter Umständen auch sozialgruppenbedingte Abweichungen (soziolektale Komponenten), des Weiteren interpersonelle Unterschiede (persönliche Erfahrungen, Erlebnisse, Gedanken, Vorstellungen, Gefühle) oder idiosynkratische Komponenten. (Hier finden wir oft die Ursachen für die geringe Kommunikationseffizienz, was im Extremfall sogar zum innersprachlichen Übersetzungsbedarf führen kann.) 7.2.2 Für die T(ranslatorischen) Prozesse sind Konvergenz und transnationale Kommunikation die charakteristischen Züge, basierend auf ähnlichen Überzeugungen aller involvierten Kommunikationssubjekte über die jeweils intendierten Informationen/Propositionen (Chesterman 2000; Wilss 1988; Malmkjaer 1993). 7.2.3 Das T(System) und das L(System) werden in allen Diskursen von zu erreichenden Zielen und zu realisierenden Zwecken gesteuert, wobei T u.a. dann erfolgreich ist, wenn den kognitiven Gegebenheiten (KS) des AT-Publikums und des ZT-Publikums entsprochen werden kann. Dies gelingt durch Überbrückung kognitiver Barrieren, kulturspezifischer Präsuppositionen usw. Aus dem Angeführten kann das folgende Axiom abgeleitet werden -tränSlAxioNSprözeSSe — L(System) + K(System) + T(System)+ Kult( System) = Interdependenz Nur das Zusammenwirken aller Teilsysteme führt zur erfolgreichen intersprachlichen/interkulturellen Informationsmittlung 7.3 Ein komplexes dynamisches Netzwerk kann man aber nicht nur auf dem Makroniveau der T-Prozesse beobachten, sondern auch auf allen Mikroniveaus der mitwirkenden Teilsysteme: L selbst ist ein Netzwerk von Subsystemen (kognitiven, kulturellen, sozialen ...); K(ultur) ist ein Netzwerk von (Sub)systemen (semiotischen, linguistischen, kulturellen, sozialen). So werden die Komponenten der Systeme unter den Systemen geteilt; sie bilden ihre dynamische Basis, von der aus alle typologischen und spezifischen Merkmale der besprochenen lebendigen Systeme generiert werden. (Systeme sind nicht statisch sondern dynamisch, d.h. dem ständigem Wandel unterworfen.) So sind Eigenschaften, wie beispielsweise der mehrfache Gebrauch von gleichen Parametern (Kultur, soziale Konventionen, kommunikative Kompetenz, usw.), sowohl in sprachlichen wie auch in den mit ihnen interagierenden nichtsprachlichen Systemen präsent. Dieser Umstand sorgt dafür, dass die Systeme entsprechend offen und dynamisch sind und dass sie gleichzeitig miteinander kommunizieren können (vgl. Steuerung, Handlung, Aushandlung), um neue kommunikative Werte zu erzeugen: T, L, K, Ku = lebendige Systeme, die vergleichbaren Gesetzmäßigkeiten unterliegen; sie breiten sich aus und verändern sich dadurch gleichzeitig (erzeugen Mutationen) = Translation Meme: „ideas spread, develope and replicate like genes" (Chesterman 2000) 7.4 Relevante Ähnlichkeiten gibt es demzufolge zwischen den natürlichen Systemen Sprache, Kognition, Kultur und Translation, erzeugt durch ihre gemeinsamen (Sub-) Komponenten: 7.4.1 L ist ein integraler Sub-Bestandteil der mentalen Welt des Benutzers. Durch Interaktion und Koordination wird KOMMUNIKATION als Transfer konzeptualer Inhalte (Informationen) erzeugt. Das wiederum setzt eine enge Korrelation zwischen L-Prozessen und K-Prozessen und -SYSTEMEN voraus. 7.4.2 Die K(ognitive)-Infrastruktur dient zur Verarbeitung und Verwendung des Welt - und Situationswissens (Wissen von Objekten und Ereignissen in der physischen Welt, von Strukturen und Konventionen in der sozialen Welt, von eigenen und allgemeinen mentalen Strukturen und Operationen und auch Emotionen, Absichten, Standpunkten und Beziehungen). Auf diese Weise werden einerseits - basierend auf Per-zeption und Interpretation der Realität - Urteile generiert, andererseits -basierend auf kreativer Imagination - Fiktion. RELEVANTE ÄHNLICHKEITEN mit T-Prozessen sind offensichtlich. 7.4.3 Die Ähnlichkeitsbeziehungen zwischen L/K und G(Grammatik) basieren auch auf Kompatibilität und Konvergenz, was im Einzelnen etwa folgendermaßen funktioniert: Das KS setzt das GS (grammatisches System) darüber in Kenntnis, welche konzeptuellen Elemente (Weltwissen) für bestimmte Kommunikationszwecke lexikalisiert werden müssen. 7.4.3.1 Das GS hat Rückwirkung auf die Identifikation und Interpretation von konzeptuellen Elementen. Somit wird deutlich, dass die G-Kategorien als SEMANTISCHE OPERATOREN fungieren, wodurch der KREISLAUF dieser natürlichen Systeme ermöglicht wird. Durch Veränderung und Anpassung entsteht ein wechselwirkendes, lebendiges, der Kreativität fähiges Gebilde: die menschliche Kommunikation (= Koordination des gesamten Wissenserwerbs und Versorgung aller Typen der vom Wissen aktivierten Handlungen). Sprachliches Wissen und nichtsprachliches Wissen werden dynamisch gekoppelt, was eigentlich die Kohärenz des gesamten menschlichen Verhaltens ausmacht. Einige Bespiele aus den von Studenten produzierten Übersetzungsversuchen sollen die enge Verbundenheit zwischen Grammatik und Konzeption veranschaulichen und gleichzeitig die negativen Folgen von Unzulänglichkeiten auf dem einen oder anderen Gebiet unterstreichen: (4) Manfred Schmidt ... warnt Markenunternehmen ... - Manfred Schmidt je posvaril podjetnike (mangelnde AS-Kompetenz, unzulänglicher Grammatik-Transfer, fehlendes fachsprachliches Spezialwissen > fraglicher Kommunikationserfolg) (5) Dann wird draußen noch immer beworben, was drinnen schon wegrationalisiert wurde. - Na trgu tako podjetja navzven konkurirajo s stvarmi, katere pa so že naredili odvečne pri racionalizaciji, (mangelnde AS/ZS-Kompetenz, mangelnde K-Kompetenz; unzulänglicher Grammatik Transfer, einschließlich Stilebenen) (6) Cerkev je bila v siloviti krizi - die Kirche ist in heftige Krise geraten (Perfekt statt Prät. = temporale Inkompetenz.; perfektiv statt imperf. = mangelhaftes Wissen im Bereich des Aspektes > fraglicher Kommu -nikationserfolg, vor allem textsortenspezifisch) (7) Ki naj bi bila popravila - die wieder in Ordnung gebracht haben konnte (mangelhaftes grammatisches Wissen im A-T-M-Bereich > fraglicher Kommunikationserfolg) (8) je očka mirno odracal - watschelte Vater ruhig (imperf. statt perf. = mangelhaftes grammatisches Wissen im A-T-Bereich > fraglicher Kommunikationserfolg, vor allem textsortenspezifisch) (9) v nekaterih pokrajinah še danes pravijo - in einigen Landschaften wird noch heute gesagt (perf. statt imperf. = mangelhafes grammatisches Wissen im A-T-Bereich > fraglicher Kommunikationserfolg, vor allem textsortenspezifisch) (10) ...vom Genfer Institut...- iz Instituta v Genovi (mangelnde ZS-Grammatik, unzulängliches Wissen über Übersetzungsnormen) 7.4.4 Kognitive Strukturen sind ihrerseits, genau wie die Sprache, unzertrennlich mit der Umwelt verbunden, denn sie werden durch die strukturellen Eigenschaften des Verhaltenssystems, in welchem sie kodiert sind, kanalisiert (= Einfluss der soziokulturellen Umwelt (Ku) auf L, K und folglich auch auf die T-Prozesse, die eigentlich als dynamische Summe der vorerwähnten Entitäten und noch einiges mehr aufzufassen sind. In diesem Umstand liegt nämlich die Begründung für das Entstehen der kulturell und sozial bedingten Asymmetrien zwischen unterschiedlichen L/K-Sy steinen. Einige Beispiele aus der Übersetzungspraxis von Studenten sollen die Ku/K-induzierten, nur mit guten translatorischen Erfahrungen überbrückbaren konzeptuellen Hindernisse veranschaulichen. (Neben dem sprachlichen Wissen ist für erfolgreiche T-Prozesse auch jenes über die kulturellen Eigenheiten und Konventionen und die jeweils ritualisierten Kommunikationsmuster und Diskurskonventionen erforderlich): (11) »Nur abspecken, wo Fett da ist« - Shujšati samo, kjer je maščoba. (mangelnde kommunikative Kompetenz/Konventionen in der ZS; fehlende Diskurskompetenz in der AS und ZS; nicht erfolgter Sprach-Kulturtransfer) (12) Wie beurteilen Sie den Preisverfall ... - Kako ocenjujete propad cen ... (mangelnde ZS -Kompetenz/Konventionen; unzulängliche Diskurskompetenz in der Zielkultur) (13) ... dass sie oftmals persönlich wegen ihres Engagements bedroht worden ist. ... - ... da je bila pogosto osebno ogrožena zaradi svoje vpletenosti (mangelnde ZS -Kompetenz/Konventionen; unzulängliche Diskurskompetenz in der Zielkultur) (14) ... um gleich hohe Ergebnisse zu erzielen ...-če želimo dobiti enako velike rezultate (mangelnde kommunikative Kompetenz/Konventionen in der ZS; nicht erfolgter Sprach-Kulturtransfer) 8.0 Schlusswort und Zusammenfassung Die Parallelitäten und unabdingbaren Wechselwirkungen zwischen den natürlichen Systemen Sprache, Kognition, Kultur und Translation kann man wie folgt zusammenfassen: 8.1 Alle verbindet der Umstand, dass sie offene, dynamische und natürliche Systeme sind, ausgewiesen durch Eigenschaften von natürlichen Systemen (=Vererbung von genetischem Material; Ausdehnung und Wechselwirkung mit anderen Systemen; Metamorphosen, Mutationen). 8.2 Eine weitere gemeinsame Eigenschaft ist die Kommunikabilität (= Neigung zur Kommunikationsoptimierung (Mitberücksichtigung von sozialen Normen; Aushandlung als Überwindung von kognitiven Hindernissen usw.). 8.3 Die natürlichen Systeme L, K, Ku und T wirken ineinander, ergänzen sich. Nur als Gesamtheit sind sie kommunikationsrelevant. Unter Punkt 7.4.4 wurde jedoch erwähnt, dass die T- Prozesse nicht nur die dynamische Summe von K/L und Ku Systemen sind, sondern noch um einiges mehr. Das Letztere soll nun noch kurz expliziert werden. 8.3.1 Integraler Bestandteil des T-Systems sind nämlich neben den o.a. auch weitere außersprachlich verankerte Entitäten wie Loyalität, Integrität und andere ethische, kommunikative, stilistische u.a. Normen (dynamischer Aspekt) oder der systematische Bruch dieser Normen, bekannt als Stilem. Weitere Merkmale sind zudem noch die Relationalität zwischen AT und ZT (in Abhängigkeit von Texttyp, Anforderungen des Kunden, Intentionen des AT/Produzenten, Erwartungen des ZT/Rezipienten), die Art der Adaptiertheit an die ZT/Kultur (= Berücksichtigung der Kultureme als Elemente/Bausteine des natürlichen Systems Kultur), AT und ZT/ Relation im Sinne von Hinzufügen oder Auslassen von Informationen (Textausdehnung in neue kommunikative Welten). 8.3.2 Alle erwähnten natürlichen Systeme, besonders aber das T-System zeichnen sich durch den Einsatz von nichtsprachlichen (auch prosodischen) Mitteln aus, was bei T-Prozessen noch besonders stark zum Tragen kommt, womit ihre ausgesprochen hohe Komplexität ausgewiesen wird. Wenn eine der zum Erfolg führenden Komponenten der T-Prozesse aussetzt oder Mängel aufweist, können in Extremfällen fatale Folgen entstehen. Nicht gerade fatal, dennoch überlegenswert ist das folgende Beispiel: (15) „Mejne prehod je dovoljen: z izkaznici za „ mali mejni prehod". Z potnim listom ali z osebni izkaznici za avstrijske državljane in tudi za slowenskedrzavljane in tudi za taksne osebe, katere nimajo vise, peske ljudi, kateri se vozijo z bicikelnom in z malim motornim vozil do 125 ccm." (So lautet die Aufschrift auf einer offiziellen Tafel, angebracht an einem Grenzübergang für den kleinen Grenzverkehr zwischen Österreich und Slowenien. Eine qualitätsbewusste translatorische Lösung könnte in diesem Fall positive diplomatische Konsequenzen haben ...) 8.4 Ähnliche Beschreibungs- und Erklärungsmodelle des natürlichen Systems Sprache werden auch von der NT angeboten. Die NT wie auch die STSN gehen davon aus, dass sprachliche und nichtsprachliche Elemente Wechsel wirkend offene, dynamische sprachliche/kommunikative Systeme ohne rigide Abgrenzungen zwischen ihren Bestandteilen (Paradigmen, Modulen) bilden. Dadurch sind sie von Fall zu Fall einzigartig und somit finalistisch (= abhängig von situativen, kulturbedingten und anderen eine konkrete Kommunikationssituation prägenden Merkmalen). Gleichzeitig berührt die Tatsache, dass sie (L, K, Ku und T) gleichen generellen Gesetzmäßigkeiten unterliegen, ihren universellen Charakter, wodurch sowohl Kommunikation als auch Übersetzbarkeit möglich gemacht werden. 8.5 Auch das als Regelung bekannte Korrektiv, wodurch Markiertheit, Frequenz, Systemangemessenheit etc. als universelle Prinzipien übereinzelsprachlich wirken, konnte in allen im Aufsatz behandelten natürlichen Systemen als wirksamer Mechanismus konstatiert werden. 8.6 Konkrete Berührungspunkte zwischen Kognition, Sprache, Kultur und letztlich Translation mit den Erkenntnissen der STSN wurden im Rahmen der hier diskutierten Pilotstudie lediglich für die Systemkomponente der Regelung näher erläutert. Die Ergebnisse bestätigten die Arbeitshypothese, wonach in allen Systemen die MikroVerhältnisse durch die Regelung im Gleichgewicht gehalten werden. Im Konkretfall wurden die Markiertheitsund Frequenzverhältnisse zwischen AT und ZT als stabile sprach- und kulturübergreifende Uni Versalien ausgewiesen. Weitere Untersuchungen sollen auch auf andere Komponenten des kognitiven Systems und auf fehlende oder bestehende Parallelitäten zur L im Sinne der Erklärungspotenziale der STSN näher eingehen. 8.7 Abschließend bleibt festzuhalten, dass Sprache, Kognition, Kultur und Translation komplexe systembezogene Tätigkeiten sind, rekurrent und ineinandergreifend, vom Menschen interaktiv, multidisziplinär (selbst)kon-struiert. Sinngemäß fasst der vom Konstruktivismus hervorgebrachte Satz "Sprache ist nicht, sie geschieht" das beobachtete sprachliche, kulturelle, kognitive und translatorische Geschehen/Handeln des Menschen zusammen. Literatur: Butzkamm, W. (1989). Psycholinguistik des Fremdsprachenunterrichts. Francke Verlag Tübingen. Chesterman, A (2000). Mernes of Translation. John Benjamins, Amsterdam, Philadelphia Dressler, W.U.; Mayerthalet, W.; Panagl, O.; Wurzel, U. (1987). Leitmotifs in natural Morphology. Amsterdam. John Benjamins. Dressler, W.U. (1990). »The cognitive perspective of »naturalist« linguistic models«. In: Cognitive Linguistics 1-1; 75-78. Givon, T. (1995). Functionalism and Grammar. Amsterdam. John Benjamins. Hatch Marcussen, E. (1983). Psicholinguistics. A second Language Perspective. Cambridge. Newbury House Publishers. Karpf, A.(1990). Selbstorganisationsprozesse in der sprachlichen Ontogenese: Erst- und Fremdsprache(n). Tübingen. Narr. Karpf, A. & Dringel-Techt. (1995). »The Role of Morphological Naturalness in Second Language Development«. In: Dressler, W. U. & Burani, C. (eds.) (1995), Crossdisciplinary Approaches to Morphology. Wien: Verlag der Österreichischen Akademie der Wissenschaften. 131-147. Kess, Joseph F. (1992). Psicholinguistics. Psychology, Linguistics ans the Study of Natural Language. Amsterdam/Philadelphia. John Benjamins. Langacker, W. Roland (1991): Concept, Image, and Symbol. Mouton de Gruyter, Berlin, New York. Langacker, W. Roland (1999): Grammar and Conceptualization. Mouton de Gruyter, Berlin, New York. Malmkjaer, K. (1993). "Underpinning translation theory". In: Target 5(2): 133 - 148. Mayerthaler, W., Morphological Naturalness, Karoma, Ann Arbor. 1985. Orešnik, J.(1999). Krepke in šibke dvojnice v skladnji. Ljubljana. Slovenska akademija znanosti in umetnosti. Orešnik, J. (1994). Slovenski glagolski vid in univerzalna slovnica. Ljubljana: SAZU. Orešnik, J. (2000). "Naturalness. The scale formats > sem (+/-A, -A) and > sem (+/-A, +A). In: Linguistica 40, 237-62. Orešnik J. (2000). Naturalness: Some Slovenian (morpho)syntactic examples. Slovenski jezik -Slovene Linguistic Studies 3, 2001a, 3-31. Orešnik, J. (2003). "Naturalness. Some Norwegian (morpho)syntactic examples". In: Norsk Lingvistik Tidsskrift, 21,57-70. Orešnik, J. (2004). Naturalness in (morpho)syntax: English examples = Jezikovna naravnost v (obliko)skladnji: angleški zgledi. (Dela 61). Ljubljana: Slovenska akademija znanosti in umetnosti. Petri, H.L.; Mishkin, M., (1994). »Behaviorism, cognitivism and the neuropsychology of memory«. In: American Scientist, 92. Rickheit G., Strohner H.(1993). Grundlagen der kognitiven Sprachverarbeitung. Tübingen, Basel; Francke. Teržan-Kopecky, K. (1997). »Die Typologie der Abfolge von Spracherwerbsprozessen«. In: Sprachtypologie und Universalienforschung. Berlin. 50, Nr.. 4,348-357. Teržan-Kopecky K. (1998). „Zur natürlichen Syntax". In:Sammelband des II. internationalen Symposions zur Natürlichkeitstheorie. Redaktion: Karmen Teržan Kopecky. Maribor. Teržan-Kopecky, K. (2001). Psihološke dimenzije jezikovnih ravnanj. Maribor, Pedagoška fakulteta. Wilss, W. (1988). Kognition und Übersetzen. Zur Theorie und Praxis der menschlichen und der maschinellen Übersetzung. Tübingen: Niemeyer. Povzetek TRANSLATI KOT SPOZNAVNE ENTITETE ALI TEORIJA NARAVNOSTI V KONTEKSTU JEZIKOVNIH IN PSIHOSOCIALNIH PARADIGEM VEČJEZIKOVNEGA SPORAZUMEVANJA Osrednja tema prispevka so naravni sistemi oz. teorija o kognitivnem sistemu (njegove komponente, dinamika), kot jo predlagajo sodobne spoznavne teorije (Langacker, 1999, Rickheit-Strohner 1993 in drugi). Teorija o ustroju in aktivnosti spoznavnega sistema pri človeku je upo -rabljena kot temelj za preverjanje skupnih (naravnih) lastnosti, ki si jih slednji deli še z drugimi naravnimi sistemi, zlasti z jezikom, ki je v prispevku obravnavan v posebnih okoliščinah medkulturnega sporazumevanja (prevajanje/tolmačenje slo ne). Prikazan je tudi proces konfrontacije dveh teoretskih pristopov, ki najdeta skupno podmeno v opisovanju mehanizmov spoznavne (in/ali jezikovne) predelave informacij in kasnejšega ravnanja z njimi {sporočanje/sporazumevanje). Pri prvem gre za že omenjeno spoznavno teorijo, pri drugem pa za teorijo naravnosti, natančneje slovensko teorijo jezikovne naravnosti (STJN). Na mikrojezikovnem nivoju medjezikovne stičnosti (slo-nem prevodna besedila) je prikazano delovanje sistemske komponente uravnavanje, ki med dragim skrbi zato, da jezikovno univerzalni pojavi, kot so zaznamovanost (spoznavna in jezikovna) in pogostnost rabe delujejo sorazmerno neodvisno od značilnosti konkretnih sporočevalnih okoliščin (doživljajskega okvirja). V kontekstu medjezikovnega/medkulturnega pretoka jezikovnih informacij so seveda v ospredju še druge sistemske komponente, ki pa so se izkazale kot izjemno odvisne od širšega in ožjega doživljajskega okvirja (kulturemov, sporočevalnih modelov in konvencij, idr.). Teodor Petrič Universität Maribor UDK 8Г27 VORSCHLÄGE ZUR ERWEITERUNG DER SLOWENISCHEN VARIANTE DER NATÜRLICHKEITSTHEORIE 1. Einleitung Im Aufsatz werden Vorschläge zur Erweiterung des in Slowenien entstandenen Modells der Natürlichkeitstheorie gemacht. Im weiteren Verlauf des Aufsatzes verwende ich die Akronyme NT für die Natürlichkeitstheorie in der Ausprägung ihrer bekanntesten Vertreter (Dressler, Mayerthaler, Wurzel) und SNT für das slowenische Modell zur Erweiterung der Natürlichkeitstheorie, dessen Grundidee wohl mit Orešnik (1986) erstmals international veröffentlicht wurde und seitdem in einer Reihe von Arbeiten überprüft und weiterentwickelt wird (mit Orešnik & Teržan & Trobevšek & Snedec 1990 als wichtigem Meilenstein). 2. NT 1981 und NT 1987 Die NT gemäß Dressler & Mayerthaler & Panagl & Wurzel 1987 geht von folgender Bewertung der Natürlichkeit einer sprachlichen Kategorie aus: (1) Eine sprachliche Kategorie ist weniger natürlich (markierter), falls sie (Schmid 1997: 340): (a) nicht uniform enkodiert (universales Prinzip), (b) nicht morphotaktisch und/oder morphosemantisch transparent (universales Prinzip), (c) nicht ikonisch / diagrammatisch sind (universales Prinzip), (d) nicht typangemessen ist (typologisches Prinzip), (e) nicht mit den charakteristischen Struktureigenschaften des betreffenden Sprachsystems übereinstimmen (einzelsprachliches Prinzip) (f) oder zu einer weniger stabilen oder weniger umfangreichen Flexionsklasse gehören (einzelsprachliches Prinzip). Die NT in dieser Ausprägung wurde vor allem auf phonologische und morphologische Kategorien angewandt. In späteren Arbeiten (z.B. Dressler 1989, Dotter 1990, Schmid 1997, Mayerthaler & Fliedl & Winkler 1998 u.a.) wurde die Natürlichkeitstheorie auch auf andere Ebenen der Sprachbeschreibung (z.B. Textlinguistik, Syntax, Zweitspracherwerb) ausgeweitet. In der NT 1981 und NT 1987 werden die systemunabhängigen Prinzipien der Transparenz, Uniformität und des konstruktionellen Ikonismus (Diagrammatismus) als auch der Begriff der Natürlichkeit (bzw. das Gegenstück: Markiertheit) verwendet. Diese Prinzipien spielen auch in der SNT 1990 und SNT 1999 eine grundlegende Rolle zur Bewertung der Natürlichkeit: (2) Ein sprachliches Element ist gemäß Mayerthaler (1987: 49) maximal transparent, wenn es sich nach dem Fregeschen Prinzip der Kompositionalität richtet (morphosemantische Transparenz) und wenn Silben- und Morphemgrenzen übereinstimmen (morphotaktische Transparenz). (3) Die Symbolisierung bzw. Enkodierung einer sprachlichen Form ist gemäß Mayerthaler (1981: 34) maximal uniform, wenn sie eineindeutig ist (d.h. wenn einer Funktion genau eine Form entspricht und umgekehrt) und wenn sie allomorphiefrei ist. Ist eine sprachliche Form nur allomorphiefrei, jedoch polyfunktional, liegt schwache Uniformität vor. (4) Wenn eine semantisch markiertere Kategorie merkmalhaft kodiert wird, ist sie gemäß Mayerthaler (1987: 48-49) konstruktionell ikonisch, sonst nicht-ikonisch oder sogar kontraikonisch. (5) Mit Hinblick auf die universalgrammatischen Präferenzen ist eine sprachliche Form gemäß Mayerthaler (1987: 49) maximal natürlich (maximal unmarkiert), wenn sie den Prinzipien der Transparenz, Uniformität und Ikonizität entspricht. Mayerthaler (1981: 46) legt folgende Prinzipienhierarchie fest: konstruktioneller Ikonismus > (»vor« bzw. »über«) uniformer Symbolisierung > Transparenz. Prinzipien, die hierarchisch untergeordneten Bereichen angehören, sind im Falle eines Natürlichkeitskonflikts die Verlierer (Mayerthaler 1981: 46). 3. SNT 1990 Orešnik & Teržan & Trobevšek & Snedec (1990: 5) postulieren die Existenz von verstärkten und nicht-verstärkten Konstruktionen einerseits (bei synchronischem Bezug auch starke und schwache Varianten genannt) sowie geschwächten und nichtgeschwächten andererseits: (6) Verstärkte Konstruktionen sind ausdrucksseitig aufwendiger, semantisch eindeutiger, nicht so sprechereffizient, jedoch hörerfreundlicher als die entsprechenden nicht-verstärkten Konstruktionen. Eine verstärkte Konstruktion erleichtert somit die Dekodierung der vom Textproduzenten intendierten Bedeutung. Für die geschwächten bzw. nicht-geschwächten Konstruktionen gilt das Umgekehrte. Die Grundannahme von Orešnik & Teržan & Trobevšek & Snedec (1990: 6) besagt folgendes: (7) Verstärkte Konstruktionen behaupten sich im Zeitraum, in dem sie als Variante einer entsprechenden nicht-verstärkten Konstruktion auftreten, vorzugsweise in relativ komplexen grammatischen Umgebungen, können sich jedoch in einem späteren Stadium auch in anderen Umgebungen verbreiten. Für die sogenannten geschwächten Konstruktionen soll die umgekehrte Aussage gelten. Im Rahmen der SNT 1990 sind wohl bislang die meisten Monographien entstanden, z.B. Teržan 1988,Teržan 1991, Trobevšek 1987, Trobevšek 1991,Snedec 1988, Petrič 1990, Petrič 1995a, Marinčič 1990. Die Grundannahmen der SNT 1990 wurden auch im Rahmen zahlreicher Konferenzbeiträge und wissenschaftlicher Zeit -schriftenartikel geprüft und die Ergebnisse international bekannt gemacht. Nicht zuletzt haben wohl auch die beiden internationalen Natürlichkeitssymposien in Maribor 1993 und 1996 (Boretzky & Dressler & Orešnik & Teržan & Wurzel 1995; Teržan 1998) zur Profilierung des in Slowenien entstandenen linguistischen Ansatzes beigetragen. 4. SNT 1999 Die SNT 1990 und die SNT 1999 stellen eine auf die Syntax angewandte Erweiterung der systemunabhängigen NT dar (vgl. Mayerthaler 1981 und Dressler, Mayerthaler, Panagl, Wurzel 1987). Gemäß Orešnik (1999: 192) wird die Aussage in (7) präzisiert, denn es wird behauptet, dass für die Variante eines Variantenpaars und ihre Umgebung die folgenden alternativen Eigenschaften gelten: (8) Alternativen für eine Variante und ihre Umgebung (nach Orešnik 1999:172 - 175) Variante: verbindet sich mit Merkmal: Verknüpfung Merkmal: sym sem sym + + und / oder - sym - - und / oder + sem + - und / oder + sem - + und / oder - Symbole: [+sym]: natürlicher hinsichtlich Enkodierung (Transparenz, Uniformität, Ikonizität); [-sym]: weniger natürlich hinsichtlich Enkodierung (Transparenz, Uniformität, Ikonizität); [+sem]: natürlicher hinsichtlich semantischer Komplexität; [-sem]: weniger natürlich hinsichtlich semantischer Komplexität Anmerkung: statt der sonst üblichen Größer- und Kleiner-Symbole (<>) verwende ich m.E. visuell leichter interpretierbare Plus- und Minuszeichen (9) Zumindest ein [+sym]-Wert zeigt die Tendenz, sich zumindest mit einem zusätzlichen [+sym]-Wert und/oder mit zumindest einem [-sem]-Wert zu verbinden; (10) Zumindest ein [-sym]-Wert zeigt die Tendenz, sich zumindest mit einem zusätzlichen [-sym]-Wert und/oder mit zumindest einem [+sem]-Wert zu verbinden; (11) Zumindest ein [+sem]-Wert zeigt die Tendenz, sich zumindest mit einem zusätzlichen [+sem]-Wert und/oder mit zumindest einem [-sym]-Wert zu verbinden; (12) Zumindest ein [-sem]-Wert zeigt die Tendenz, sich zumindest mit einem zusätzlichen [-sem]-Wert und/oder mit zumindest einem [+sym]-Wert zu verbinden. Der metasprachliche Ausdruck »sich mit... verbinden» bezieht sich sowohl auf die beobachtete sprachliche Einheit als auch auf ihre unmittelbare Umgebung. Im Rahmen der SNT 1999 sind mittlerweile mehrere wissenschaftliche Bücher (Orešnik 1999, Orešnik 2001, Dobro volje 2005) und eine Reihe von Zeitschriften- und Konferenzbeiträgen veröffentlicht worden. 5. Fragestellungen zur Erweiterung der SNT Als Ansatz, eine Erweiterung der SNT vorzuschlagen, soll die Beschreibung der Grundannahmen in (8) bis (12) aus Orešnik (1999: 172 - 175) sowie die Beschreibung der Sem- und Sym-Skalen in Dobrovolje (2005: 28) dienen. Dobrovoljc (2005: 26 - 30) beschreibt, wie die die Sem- und Sym-Skalen aus Orešnik (1999) und Orešnik (2001) aufeinander bezogen werden. Ein Fall ist die folgende Sem-Skala (Dobrovoljc 2005): (13) >sem (weniger transparent, transparenter) / syntaktische Einheit Im Hinblick auf die kognitive Komplexität sind weniger transparente syntaktische Einheiten natürlicher als transparentere. - Im Interesse des Sprechers ist es, weniger umfangreiche syntaktische Einheiten zu äußern, weil ihm dies weniger Mühe bereitet. Setzt man voraus, dass diese Annahme stimmt, erhebt sich die Frage, welche Faktoren bewirken, dass der Sprecher in wirklichen Textzusammenhängen statt der ausdrucksökonomischeren syntaktischen Einheit eine aufwendigere (also etwa eine transparentere) präferiert? Eine weitere Fragestellung betrifft einzelsprachliche Unterschiede. In Orešnik (1999), Orešnik (2001) und Dobrovoljc (2005) werden verschiedene grammatische Paradigmen behandelt und hinsichtlich ihrer Natürlichkeit bewertet. Die Natürlichkeitsbewertung ist in vielen Fällen nur auf bestimmte Sprachen beziehbar. Welche Faktoren bewirken nun, dass in einer Sprache beispielsweise eine weniger transparente Konstruktionsweise geläufiger ist als in einer anderen? 6. Zur Rangfolge der SNT-Prinzipien in Texten Sprachliche Konstruktionen sind das Produkt von Optimalisierungsprozessen in der Kommunikation, bei möglichst geringen kognitiven Kosten soll eine möglichst große Wirkung erreicht werden (Minimax-Prinzip). Zwischen den sprachlichen Bedürfnissen des Textproduzenten und des Textrezipienten besteht ein Konflikt: Geringer Kodieraufwand hat einen höheren Dekodieraufwand zur Folge. Die in Texten verwendeten sprachlichen Konstruktionen stellen mehr oder weniger erfolgreiche Kompromisslösungen zur Bewältigung des kommunikativen Konflikts dar. Die vom Textproduzenten intendierte Bedeutung (d.h. der propositionelle Gehalt und/oder der illokutive Inhalt einer Äußerung bzw. eines Textsegments bzw. eines Textes) kann vom Textrezipienten besser verstanden werden, wenn sie transparent, uniform und ikonisch übermittelt wird, außerdem wenn sie mit typangemessenen und/oder systemangemessenen Mitteln kodiert wird (d.h. mit prototypischen sprachlichen Mitteln, häufiger verwendeten Ausdrücken,...) übermittelt werden. Der kognitive Aufwand beim Kodieren und Dekodieren kann oft nur indirekt bewertet werden. Dazu können folgende heuristische Hilfsmittel verwendet werden (vgl. auch Mayerthaler 1981, Orešnik (2003: 59- 60), Dobrovoljc (2005: 32 - 33, 36, 38): 1. das phylogenetische Alter (Sprachgeschichte) 2. das ontogenetische Alter (Spracherwerb) 3. Sprachpathologie 4. Versprecher 5. Prototypizität 6. Strukturprinzipien (Symmetrie, Integration) 7. Gebrauchsfrequenz 8. Sprachtypologie (Universalien: Verbreitung in den Sprachen der Welt) 9. Paradigmengröße (einzelsprachliche Musterbildung) 10. Spezieller Gebrauch (Irregularität) 11. Kategoriengebrauch und Prozessdurchführung (Transformationen) 12. Textualitätskriterien (Kohärenz, Kohäsion, Situationalität, Akzeptabilität, ...) 13. ... Die Art und Weise der sprachlichen Kodierung und Dekodierung ist von vor- und außersprachlichen Umständen abhängig. Ausgangspunkte für den Textproduzenten: 1. Bewertung des außersprachlichen Situationszusammenhangs und der Teilnehmer 2. vorsprachliche Zielsetzung (Illokution, Sprechhandlung) 3. intendierter propositioneller Gehalt I Textsegment I Ausgangspunkte für den Textrezipienten: 4. Bewertung des außersprachlichen Situationszusammenhangs und der Teilnehmer 5. Erschließung des vorsprachlichen Ziels des Sprechers (Illokution, Sprechhandlung) 6. Erschließung des intendierten propositionellen Gehalts Der außersprachliche Situationszusammenhang und die Teilnehmer können im Rahmen eines Textvorkommens meist als Konstanten betrachtet werden. Ziel der folgenden Darstellung ist eine Einschätzung der Kodiervor- und -nachteile von sprachlichen Konstruktionen. Zu diesem Zweck wird eine Tabelle erstellt, in der ein Vorteil einer Konstruktion auf einer bestimmten Werteskala durch einen höheren Rang ausgedrückt wird. Als Input in einer solchen Tabelle können folgende Größen gewählt werden: (a) der vom Sprecher intendierte propositionelle Gehalt - prop (b) der vom Sprecher intendierte illokutive (modale) Inhalt - ill (c) der vom Sprecher geäußerte sprachliche Ausdruck - con Als Output einer solchen Tabelle können folgende Größen gewählt werden: (a) theoretisch mögliche sprachliche Konstruktionen oder Varianten - con (b) die vom Hörer verstandene propositionelle Bedeutung - prop (c) die vom Hörer verstandene illokutive (modale) Bedeutung - ill In der folgenden Darstellung wird die Prädikation P (behandel-, operier-, ...) mit den dazugehörigen semantischen Rollen Agens (Arzt, ...) und Patiens (Patient,...) als Input in die Bewertungstabelle eingesetzt. Im Output der Tabelle erscheinen verschiedene theoretisch mögliche Konstruktionen, deren Kodiervorteile und -nachteile im Folgenden auf verschiedenen Natürlichkeitsskalen bewertet werden sollen. Die Darstellung bezieht sich auf die deutsche Sprache, kann aber zumindest teilweise auch für andere Sprachen, insbesondere Nominativ-Akkusativ-Sprachen, Gültigkeit haben. Input Output Assertion: P (Agens, Patiens) (A) Hauptsatz (Aussagesatz, Aktivsatz), z.B. Der Arzt behandelt den Patienten (B) Hauptsatz (Aussagesatz, Passivsatz), z.B. Der Patient wird (vom Arzt) behandelt (C) Nebensatz (Aktivsatz), z.B. dass der Arzt den Patienten behandelt (D) Nebensatz (Passivsatz), z.B. dass der Patient (vom Arzt) behandelt wird (E) Infinitivsatz, z.B. den Patienten zu behandeln (F) Nominalisierung, z.B. die Behandlung (des Patienten) (durch den Arzt) 1. Ausgehend vom konstruktioneilen Ikonismus (k. Ikon.: ein komplexerer Inhalt korreliert mit einer komplexeren Form) sind (B), (D) und (F) bei fehlendem Agens (bzw. fehlendem Agens und Patiens) weniger ikonisch als (A) und (C), da die zuerst genannten drei Konstruktionen das Prädikat zweiteilig realisieren, die Konstruktion (E) hingegen immer kontraikonisch (aufgrund obligatorisch nicht realisiertem Agens). Daher erhalten (B), (D) und (F) Rang 2, (A) und (C) Rang 1, die Konstruktion (E) dagegen Rang 3. 2. Eine Sprechhandlung (hier eine Behauptung (Assertion)) wird durch einen Satz eindeutiger ausgedrückt als durch eine Nominalphrase, d.h. die illokutive Transparenz (ill. Tra.) von (A) und (B) ist größer als von (C), (D), (E) und (F). Daher erhalten (A) und (B) auf dieser Skala Rang 1 (den höchsten Rang), (C) und (D) Rang 2, (E) und (F) Rang 3. 3. Ausgehend vom Kodieraufwand (Ökon) ist eine Konstruktion für den Textproduzenten einfacher, wenn sie fakultative Bestandteile enthält bzw. wenn sie bestimmte Bestandteile einer Proposition nicht enthält. (A) und (C) enthalten nur obligatorische Bestandteile, (B), (D) und (E) enthalten eine fakultative Konstituente bzw. erlauben die Nicht-Realisierung einer Konstituente (nämlich des Agens), die Konstruktion (F) ermöglicht sogar die Aussparung von zwei Bestandteilen (nämlich des Agens und des Patiens). Daher erhält (F) Rang 1, (B), (D) und (E) Rang 2, die Konstruktionen (A) und (C) dagegen Rang 3. 4. Rektionsrichtung (Rekt.): Im Deutschen kann die Konstituentenstellung SOV als zugrunde liegend angesehen werden, d.h. Links-Rektion. In den Konstruktionen (B), (C), (D) und (E) regiert das Hauptverb nach links, daher erhalten sie Rang 1. In (A) kann das Hauptverb in Zweitstellung auftreten und dann nur indirekt (über die Spur) nach links regieren, daher erhält (A) Rang 2. In (F) wird nach rechts regiert, daher erhält (F) nur Rang 3. 5. Ausgehend von der anthropozentrischen und egozentrischen Perspektive sind in Nominativ-Akkusativ-Sprachen die Konstruktionen (A) und (C) typangemessener (Typ) als (B), (D), (E) und (F), denn in (A) und (C) wird die Erwartung „Agens vor Patiens" erfüllt, in (B), (D), (E) und (F) dagegen nicht. Daher erhalten (A) und (C) auf dieser Skala Rang 1, (B), (D) und (F) Rang 2, (E) hingegen wegen obligatorisch fehlendem Agens nur Rang 3. 6. Paradigmengröße (Paradig.): Alle Prädikate können in (A), (C) und (E) eingesetzt werden, nicht jedoch in (B), (D) und (F), denn es gibt einerseits nicht passivfähige Verben, andererseits aber nicht-nominalisierbare Verben. Die Paradigmen (A), (C) und (E) sind größer als die Paradigmen (B), (D) und (F). Daher erhalten (A), (C) und (E) Rang 1, (B), (D) und (F) dagegen nur Rang 2. 7. Prototypische Funktion: Sätze sind der prototypische Ausdruck für einen propositionellen Gehalt. Prototypische Sätze enthalten ein finites Verb (Vfin). Daher erhalten die satzartigen Konstruktionen (A), (B), (C) und (D) Rang 1. (E) enthält immerhin noch eine infinite Verbform und erhält daher Rang 2. Nominalisierungen sind dagegen als Nominalphrasen der prototypische Ausdruck für Teile von Sätzen, d.h. Satzkonstituenten oder Valenzpartner von Verben. Sie enthalten keine Verbform. Daher erhalten Sie nur Rang 3. 8. Hauptsätze sind funktionell weniger eingeschränkte Konstruktionen als Nebensätze, Infinitivsätze oder Nominalisierungen. Letztere sind gewöhnlich einem anderen sprachlichen Element untergeordnet (Subord.): (C) und (D) einem Subjunktor, (E) einem Matrixverb im übergeordneten Satz, von dem aus ersichtlich ist, ob Subjekt- oder Objektkontrolle vorliegt (in manchen Fällen, insbesondere bei adverbialen Infinitivsätzen, kann auch eine Abhängigkeit von einem Subjunktor vorliegen), und (F) einem Matrixverb im Trägersatz. (A) und (B) erhalten daher Rang 1, die übrigen Konstruktionen Rang 2. Input Output Subord. Vfin Paradig. Typ Rekt. Ökon III. Tra. K.Ikon. Sum. Ass: (A) HS-A 1 1 1 1 2 3 1 1 11 P (B) HS-P 1 1 2 2 1 2 1 2 12 (AR, (C) NS-A 2 1 1 1 1 3 2 1 12 Pat) (D) NS-P 2 1 2 2 1 2 2 2 14 (E) INF 2 2 1 3 1 2 3 2 16 (F) NOM 2 3 2 2 3 1 3 2 18 Mit Hilfe dieser (hierarchisch noch nicht geordneten) acht Parameter (d.h. acht Ordinalskalen) können die sechs Konstruktionstypen hinsichtlich ihrer Natürlichkeit eingeordnet werden. Auf der linken Seite der unten abgebildeten Skala werden die Konstruktionen abgebildet, die hinsichtlich der oben verwendeten Parameter höher eingeordnet worden (d.h. die Rangsumme ist niedriger). Auf der rechten Seite der Skala sind Konstituenten mit einer höheren Rangsumme abgebildet. Je weiter links eine Konstruktion auf der unten folgenden Skala eingeordnet ist, umso natürlicher ist sie für den Ausdruck eines propositionellen Gehalts bzw. umso präferenter ist sie in Texten. <+nat> <-nat> -- (A) (B) (D) (E) (F) (C) Die Reihenfolge der oben verwendeten acht Parameter ist noch nicht hierarchisch (d.h. nach ihrer Stärke) geordnet. Gemäß den Vertretern der NT 1987 ist jedoch vorgesehen, dass einzelsprachliche Parameter (wie Systemangemessenheit oder Paradigmengröße) einen höheren Rang einnehmen als typologische oder universelle Parameter, d.h. dass einzelsprachliche Parameter auf der Grundlage universeller und typologischer Prinzipien, aber auch im Konflikt mit diesen, die Herausbildung des Sprachsysterns entscheidend in bestimmte Bahnen lenken. Die Reihenfolge der acht Parameter in der obigen Tabelle richtet sich nach der Hierarchisierung der NT 1987 nur insofern, als die universellen (möglicherweise weniger einflussreichen) Prinzipien weiter rechts in der Tabelle erscheinen als typologische oder einzelsprachliche. Aufgrund der oben verwendeten Rangsummenmethode wird die Stärke der einzelnen oben verwendeten Prinzipien allerdings noch nicht berücksichtigt. Allerdings ist leicht einzusehen, dass bei hoher Prominenz des Parameters der illokutiven Transparenz (d.h. auch bekannt als Präferenz für direkte Sprechakte gegenüber indirekten) Konstruktionen wie (E) und (F) kaum noch in Frage kommen und Konstruktionen wie (A) und (B) an Präferenz gewinnen. Weitere Parameter können die oben dargestellte Rangfolge der Konstruktionen (wie sie sich aus der Rangsumme ergibt) auf jeden Fall verändern (vgl. auch Petrič 1995b). Eine Konstruktion, die nach den oben verwendeten acht Parametern als weniger natürlich eingestuft wurde (z.B. ein Passivsatz), kann unter besonderen Umständen an Akzeptabilität gewinnen und in der obigen Präferenzskala weiter nach links rücken: 9. Bei Definitheitsunterschieden (Def.): Das Definitheitsprinzip ist einflussreicher als die Belebtheitshierarchie oder das Agensprinzip, das Agensprinzip dominanter als das Belebtheitsprinzip. Sind sowohl Agens als auch Patiens defi-nit, dominiert das Prinzip »Agens vor Patiens". Ist jedoch das Agens indefinit, das Patiens dagegen definit, kommt das Prinzip „Definites vor Indefinitem" zur Geltung. Im Falle eines definiten Patiens und eines indefiniten Agens sind (B), (D) und (F) angemessener (Rang 1), (A) und (C) weniger angemessen (Rang 2). Auf (E) ist die Definitheitsskala (zumindest in dieser Ausprägung) nicht anwendbar. Entsprechende Definitheitsunterschiede bewirken im Deutschen, dass Passivkonstraktionen oder Nominalisierangen eine präferentere Stellung einnehmen können (in der unten verwendeten Skala durch einen Pfeil nach links abgebildet). Ein Passivsatz (B) mit definitem Patiens kann bei Verwendung der Rangsummenmethode zumindest als ebenso präferent wie ein Aktivsatz (A) eingeordnet werden. Ob dann im Text tatsächlich ein Passiv verwendet wird, hängt aber auch von anderen Faktoren ab. Mit Bezug auf eine Einzelsprache kann man die Stärke der einzelnen Parameter festlegen. Man könnte sich beispielsweise vorstellen, dass in einer flexionsarmen Sprache wie dem Englischen das Defi- nitheitsprinzip eine größere Stärke aufweist als in einer flexionsreicheren Sprache wie dem Deutschen. <+nat> <-nat> -*-► (A) (C) (D) <-<-<- (E) (F) (B) «-«-<- 10. Im Text ist auch die Thema-Rhema-Gliederung (TRG; thematische Progression) bei der Reihenfolge der Satzkonstituenten zu berücksichtigen: Nach dem Themaprinzip ist die Reihenfolge Thema vor Rhema vorherrschend. Ist beispielsweise das Patiens (trotz Indefinitheit) thematisch und das Agens (trotz Definitheit) rhematisch, wird das Agensprinzip zugunsten des Themaprinzips zurückgestellt, d.h. der Einfluss des Agensprinzips auf die Reihenfolge der Satzkonstituenten wird zugunsten des Themaprinzips vermindert. Da im Deutschen das Subjektprinzip (Subjekt vor Objekt) einflussreicher ist als das Themaprinzip, kann beispielsweise ein Aktivsatz (A) seinen hohen Präferenzrang zugunsten eines Passivsatzes (B) verlieren. Denn im Passivsatz (B) wird sowohl das Themaprinzip (Thema vor Rhema) als auch das Subjektprinzip (Subjekt vor Objekt) berücksichtigt. Entsprechendes gilt für den Vergleich von (C) und (D). Auf (E) und (F) ist das Themaprinzip in dieser Ausprägung aus bekannten Gründen nicht anwendbar. <+nat> <-nat> -- ->->-* (A) (D) <-<-<- (E) (F) (C) (B) (A) Der ARZT operiert einen Patienten (nicht die Krankenschwester). (B) Ein Patient wird vom ARZT operiert (nicht von der Krankenschwester). 11. Die Textsorte, die sich aus Globalziel, kommunikativen Umständen und Teilnehmern ergibt, beeinflusst ebenfalls den Gebrauch von sprachlichen Konstruktionen (vgl. Petrič 1998a). Anders ausgedrückt: sprachliche Konstruktionen zeigen nicht nur eine Affinität zu bestimmten sprachlichen Umgebungen, sondern auch zu bestimmten außersprachlichen Umgebungen. Nehmen wir die Grundtextsorten (GTS) aus Diewald 1991 (vgl. auch Petrič 1998a) als Grundlage zur Bewertung der Distribution sprachlicher Konstruktionen, stellt sich heraus, dass (B), (C), (D), (E) und (F) in markierteren Grundtextsorten (insbesondere in schriftlichen Fernmonologen) häufiger auftreten als in weniger markierten Grundtextsorten (wie mündlicher Nahdialog oder mündlicher Ferndialog). <+nat> <-nat> -*.- (A) ->-►-> (B) -►->-> (D) (E) -»-»-»(F) (C) 12. Letztendlich spielen überhaupt Textualitätskriterien (Beaugrande & Dressler 1992, Dressler 1989) wie Kohärenz, Kohäsion, Informativität, Akzeptabilität, Situationalität, Intertextualität eine Rolle, welche sprachliche Konstruktion sich unter welchen Umständen in realen Textzusammenhängen durchsetzt bzw. an Präferenz gewinnt. Welche Konstruktion im Text tatsächlich gewählt wird, hängt davon ab, welche Rangfolge (welches Gewicht) die einzelnen (universellen, typologischen, sprachsy-stemspezifischen, pragmatischen) Prinzipien einnehmen (vgl. auch das Ranking in der Optimalitätstheorie). 7. Einzelsprachliche Unterscheidung durch rangierte SNT-Prinzipien Die zweite Fragestellung betrifft die Rangfolge der SNT-Prinzipien (Parameter), um einzelsprachliche Unterschiede zu beschreiben und zu erklären. Ausgangspunkt meiner Darstellung Orešnik behandelt in seinem Artikel (Orešnik 2000) die Natürlichkeitsskalen >sem (+/-A, -A) und >sem (+/-A, +A). Die Natürlichkeitsskalen folgen wohl der allgemeinen Annahme der Markiertheitstheorie, dass das Spezielle (+A) oder (-A) markierter (weniger natürlich) ist als das Allgemeine (+/-A). Im Aufsatz möchte ich mich eingangs mit der unten angeführten Annahme und Vorhersage über englische abhängige Sätze mit fini-ter und infiniter Verbform aus Orešnik (2000) beschäftigen. Dabei soll der Blick danach vor allem auf meine Beurteilung der deutschen Infinitivsätze gelenkt werden. »English. With non-finite clauses,« [-fin] »the lack of a clause link« [-CLL] »is normal« (Orešnik 2000: 239, Nr. 2). Orešniks Fortsetzung (Orešnik 2000: 239-240, Nr. 2): Wenn (a) l>scm (+fin' -fin)1 clausel (vgl. Mayerthaler & Fliedl & Winkler 1998: 325) und wenn (b) |>sem (+/-CLL, -CLL) / clause type| und wenn (c) [>sem] [>sem] und (d) [sem-Parameter die Tendenz, sich mit zumindest einem anderen >sem-Parameter zu verbinden. Nach (d) zeigt ein Nullmorphem genug so darauf dass Die Einordnung auf der oben angeführten Skala richtet sich nach den folgenden Kriterien: (i) der Ausdruck leitet den abhängigen Satz [+depend] ein, also [+intro]; (ii) der Ausdruck ist bedeutungsentleert [+empty]; z.B. das Präpositionaladverb darauf hat die Eigenschaften [-intra, +empty]. Es bieten sich zwei extreme Alternativen an: (A) Wenn adverbiale Ausdrücke wie genug, so, darauf oder die Infinitivpartikel zu als [+CLL] gewertet werden, dann haben alle oben angeführten Infinitivsätze die Eigenschaft [+CLL]. (B) Wenn lediglich der Subjunktor daß / dass als [+CLL] eingeordnet werden, dann hat nur ein Infinitivsatz in den oben angeführten Beispielen die Eigenschaft [+CLL], nämlich der in (5). Beispiel (2) zeigt (so wie (1)) noch eine Unterscheidungsmöglichkeit: die Bedeutung des Subjunktors dass ist zu allgemein, als dass er ein konsekutives Verhältnis [+consec] im Satzgefüge anzeigen könnte. Dies wird erst durch das Adverb genug ermöglicht. Demnach wäre es angemessen, auch das Adverb genug als »clause link« einzuordnen. Der Unterschied zwischen den Sätzen [+fin] und [-fin] in (2) läge demnach darin, dass das „Clause Link" im ersten Satzyp [+fin] zweiteilig wäre (d.h. genug + dass), im zweiten Satztyp [-fin] dagegen nur einteilig (nämlich nur aus genug bestehend). Die Kombination des Präpositionaladverbs darauf mit dem Subjunktor dass könnte demnach als prototypischeres zweiteiliges „Clause Link" bezeichnet werden, die Kombination des quantitaven Adverbs genug mit dem Subjunktor dass dagegen als weniger prototypisches „Clause Link", so trägt doch die erst genannte Kombination die Eigenschaften [-intro, +empty und +intro, +empty], die zuletzt genannte Kombination dagegen die Eigenschaften [-intro, -empty und +intro, +empty]. Deshalb liegt die m.E. angemessenste Grenze zwischen den Kategorienwerten [+CLL] und [-CLL] gerade zwischen dem Adverb so [-empty] und dem Präpositionaladverb darauf [+empty]. Im Weiteren werde ich angesichts meiner Fragestellung, ob deutsche Infinitivsatztypen mehrheitlich die Eigenschaft [+CLL] aufweisen, dennoch die ungünstigere Definition des Begriffs „Clause Link" wählen, d.h. (B), wonach dem syntaktischen Kriterium Genüge geleistet wird, dass der Ausdruck in der Rolle eines „Clause Link" den abhängigen Satz einleitet und damit nicht Bestandteil des übergeordneten Satzes ist. Deutsche Neben- und Infinitivsätze mit und ohne „Clause Link" Wenn wir nun die Beschreibung der Infinitivsätze [+inf] in einer deutschen Standardgrammatik berücksichtigen, z.B. die in Heibig & Buscha (101987: 656-661), dann sind im Deutschen sowohl abhängige Sätze mit den Merkmalen [+fin, +/-CLL] als auch mit den Merkmalen [-fin, +/-CLL] zu finden, also im Gegensatz zum Englischen, wie aus der Schlussfolgerung (f) in Orešnik (2000) folgt, nicht nur bzw. nicht vorwiegend Infinitivsätze mit den Merkmalen [-fin, -CLL], Wenn ein deutscher Infinitivsatz in der Funktion einer Adverbialbestimmung [+adv] auftritt, weist er meist das Merkmal [+CLL] auf, erscheint ein deutscher Infinitivsatz dagegen als Subjekt oder Objekt [-adv], hat er gewöhnlich die Eigenschaft [-CLL]. Die meisten Adverbialbestimmungen sind nicht valenzgebunden [-val], während das Subjekt und alle Objekte valenzgebunden [+val] sind. Als Ausnahmen unter den Adverbialbestimmungen dürfen z.B. Richtungsbestimmungen, sogenannte Direktivergänzungen, gelten, denn sie weisen die Merkmale [+CLL] und [+val] auf. Allerdings erscheinen die Direktivergänzungen meines Wissens nach nicht in der Form von Infinitivsätzen. Hier folgen einige Beispiele aus Heibig & Buscha (101987: 656ff.) zur Illustration, und zwar zunächst Infinitivsätze mit dem Merkmal [+val], wobei ein Infinitivsatz in (7) als Präpositionalobjekt, in (8) als Subjekt, in (9) als Akkusativobjekt, in (10) als Genitivobjekt und in (11) als Präpositionalobjekt zum Substantiv Recht auftritt: (7) Wir freuen uns, ihn zu sehen, [-fin, +inf, -i-val, +prp, -CLL] (8) Ihn zu treffen ist mir peinlich, [-fin, +inf, +val, +sub, -CLL] (9) Sie erlaubte ihm, früher nach Hause zu gehen, [-fin, +inf, +val, +akk, -CLL] (10) Peter ist sich bewußt, seinen Freund verletzt zu haben, [-fin, +inf, +val, +gen, -CLL] (11) Er hat das Recht, diese Papiere zu lesen, [-fin, +inf, +val, -i-prp N, -CLL] Noch einige Beispiele aus Heibig & Buscha (101987: 656ff.) zur Illustration, und zwar mit nicht-valenzgebundenen adverbialen Infinitivsätzen mit den Merkmalen [+adv] [-val]: (12) Er geht in das Bad, fan)statt zu arbeiten, [-fin, +inf, -val, +subsütutiv, +CLL] (13) Er geht in die Bibliothek, um dort zu arbeiten, [-fin, +inf, -val, +final, +CLL] (14) Er geht in den Betrieb, ohne dort zu arbeiten, [-fin, +inf, -val, +kommitativ, +CLL] Die valenzgebundenen Satzglieder betrachte ich als natürlicher, die nicht-valenzgebundenen Satzglieder hingegen als weniger natürlich. Dies möchte ich damit begründen, dass der Textrezipient valenzgebundene Satzglieder eher erwartet als nicht-valenzgebundene Satzglieder, d.h. die zuerst genannten Satzglieder redundanter sind als die zuletzt genannten: (g) |>sym (+erwartet, -erwartet) / Redundanz] (h) |>sym ( +val, -val) / Erwartbarkeit der Satzglieder] Ähnlich (wenn auch nicht gleich) wird von Orešnik (2000: 246) zwischen »subject or object gap« einerseits und »adverbial or prepositional object gap« andererseits unterschieden, indem erläutert wird, dass »subject and objects are easier for the hearer to process than adverbial or prepositional object gaps.«: (i) |>sym (subject/object gap, adverbial/prepositional object gap) / inf. clause in English.| Aus den Natürlichkeitsbewertungen (a) - (d), (g) in (h) folgt, dass wir unter den deutschen Satztypen mit den Merkmalen [+fin, +val] häufiger solche mit dem Merkmal [+/-CLL] finden können als unter den Satztypen mit den Merkmalen [-fin, -val]. Hier folgen einige Beispiele aus der deskriptiven Grammatik Heibig & Buscha (101987: 653, 660) zur Illustration, und zwar zuerst Satztypen mit den Merkmalen [+depend, +fin, +val, -adv]: (15) Ich dachte, er hätte seine Prüfung abgelegt / daß er seine Prüfung abgelegt hätte. (16) Es ist besser, du kommst pünktlich / daß du pünktlich kommst. Die valenzgebundenen finiten Nebensätze [+fin, +val], die als Subjekt [+subj oder Objekt [+akk, +prp, ...] auftreten, erlauben also die Eigenschaft [+/-CLL], die wenigen valenzgebundenen finiten Nebensatztypen in Adverbialfunktion [+fin, +val, +adv] erlauben dagegen nur das Merkmal [+CLL], z.B. die Richtungsbestimmungen (Direktivergänzungen). Das Kriterium [+/-adv] ist demnach im Deutschen einflussreicher als das Merkmal [+/-val]. Sowohl finite abhängige Sätze in Subjekt- oder Objektfunktion, d.h. solche mit den Merkmalen [+depend, +fin, -i-val, -adv], als auch finite abhängige Sätze in Adverbialfunktion, d.h. solche mit den Merkmalen t+depend, -i-fin, +/-val, +adv], können entweder das Merkmal [+CLL] oder (in wenigeren Fällen) das Merkmal [-CLL] aufweisen (siehe Beispiele (17) - (19) unten). Hier folgen wiederum einige Beispiele aus der deskriptiven Grammatik Heibig & Buscha (101987: 653,660) mit den Satztypen [+depend, +fin, -val, +adv], die uns zeigen, dass es sich lediglich um drei Fälle von valenzungebundenen finiten Nebensätzen [+fin, -val] handelt, und zwar um Konzessivsätze [+conces], um Konditionalsätze [+cond] und Konsekutivsätze [+consec], die das Merkmal [+/-CLL] erlauben, während die meisten anderen Nebensatztypen mit den Merkmalen [+depend, +fin, -val, +adv], die im Deutschen wesentlich zahlreicher sind, lediglich das Merkmal [+CLL] zulassen: (17) Mag er auch noch wenig Erfahrung haben / Obwohl er noch wenig Erfahrung hat, so weiß er doch sehr viel. (18) Solltest du sie treffen / Falls du sie treffen solltest, sage ihr bitte Bescheid. (19) Er war klug genug, seinen Fehler einzusehen / daß er seinen Fehler einsah. Im Deutschen bestimmt die grundlegende Abgrenzung zwischen infiniten Satztypen mit den Merkmalen [+adv] und [-adv] auch im Wesentlichen die Grenze zwischen den inifiniten Satztypen mit den Merkmalen [+CLL] und [-CLL]. Das Valenzmerkmal [+/-val] leistet dazu im Deutschen keinen entscheidenden Beitrag, so dass man in der Natürlichkeitsskala (h) das Merkmal [+/-val] durch [+/-adv] ersetzen sollte. Im Deutschen folgen die finiten und infiniten abhängigen Sätze der Natürlichkeitsskala >sem (+/-A, +A) (vgl. dazu Orešnik 2000: 253ff.), denn mehrere deutsche finite Nebensätze [+fin] erlauben das Merkmal [+/-CLL], deutsche Infinitivsätze [+inf] dagegen relativ häufiger [+CLL], Deutsche Infinitivsätze mit dem Merkmal [+adv] gehören fast ausschließlich zum Typ [+CLL], was darauf deutet, dass sie im Rahmen der Natürlichkeitsskala >sem (+/-A, +A) und nicht gemäß der Natürlichkeitsskala >sem (+/- A, -A) betrachtet werden können - im Unterschied zu den englischen Infinitivsätzen. Zusammenfassung des 7. Abschnitts Zurück zu der Grundannahme in Orešnik (2000: 239, Nr. 2): In englischen Satzkonstruktionen mit infiniter Verbform [-fin] fehlt gewöhnlich ein „Clause Link" („Konjunktionswort"), also [-fin] [-CLL], Im Deutschen kann diese Schlussfolgerung höchstens auf Infinitivsätze in Subjekt-und Objektfunktion (also [+val, -adv]) übertragen werden, denn diese weisen tatsächlich die Eigenschaft [-CLL] auf, was bedeutet, dass sie als Fall der Natürlichkeitsskala >sem (+/-A, -A) betrachtet werden könnten. Infinitivsätze mit den Merkmalen [+/-val, +adv] gehören in den meisten Fällen zum Typ [+CLL], was bedeutet, dass sie als Fall der Natürlichkeitsskala >sem (+/-A, +A) aufgefasst werden könnten. Wenn man davon ausgeht, dass wesentlich weniger Objektsatztypen als verschiedene Arten von Adverbialsätzen unterschieden werden können, kann die Annahme, dass im Deutschen (so wie im Englischen) mehr Infinitivsatztypen ohne „Clause Link" [-CLL] auftreten, nicht bestätigt werden, denn die meisten Infinitivsatztypen [-fin, +inf] weisen im Deutschen das Merkmal [+CLL] auf. Im Deutschen kann man die Stärke der Parameter [+/-fin], [+/-val] und [+/-adv] folgendermaßen festlegen: [+/-fin] > [+/-adv] > [+/-val], d.h. der einflussreichste Parameter auf das Auftreten eines „Clause Link" in einem abhängigen Satz ist der Finitheitsparameter, gefolgt vom Adverbialparameter. Der Valenzparameter spielt im Vergleich zu diesen beiden eine untergeordnete Rolle. In deutschen Infinitivsätzen spielt somit der Adverbialparameter eine entscheidende Rolle für das Auftreten eines „Clause Link" („Konjunktionswortes"). In einer anderen Sprache, z.B. dem Englischen, scheint der Adverbialparameter diesbezüglich eine geringere Rolle zu spielen als im Deutschen. Literatur de Beaugrande, Robert Alain & Dressler, Wolfgang Ulrich (1992): Uvod v besediloslovje. Založba Park Ljubljana. Boretzky, Norbert & Dressler, Wolfgang & Orešnik, Janez & Teržan, Karmen & Wurzel, Wolfgang (1995) (Hgg.): Natürlichkeitstheorie und Sprachwandel / Teorija naravnosti in jezikovno spreminjanje. Beiträge zum internationalen Symposium über "Natürlichkeitstheorie und Sprachwandel" an der Universität Maribor vom 13.5.-15.5.1993. Bochum: Brockmeyer. Diewald, Gabriele (1991): Deixis und Textsorten im Deutschen. Niemeyer: Tübingen. Dobrovoljc, Helena (2005): Slovenska teorija jezikovne naravnosti. Ljubljana: Založba ZRC. Dotter, Franz (1990): Nichtarbitrarität und Ikonizität in der Syntax. Hamburg: Buske. Dressler, Wolfgang Ulrich (1989): Semiotische Parameter einer textlinguistischen Natürlichkeitstheorie. Wien: Verlag der Österreichischen Akademie der Wissenschaften. DRESSLER, Wolfgang Ulrich & Mayerthaler, Willi & Panagl, Oswald & Wurzel, Wolfgang Ullrich, (eds.) (1987): Leitmotifs in Natural Morphology. Amsterdam: Benjamins. Helbig & Buscha (101987): Deutsche Grammatik. Ein Handbuch für den Ausländerunterricht. Leipzig: Verlag Enzyklopädie. Marinčić, Branka (1990): Zapletenost oziralnih odvisnikov v delih Thomasa Elyota. «Komplexität der Relativsätze in den Werken von Thomas Elyot>. Magistrsko delo . Filozofska fakulteta «Philosophische FakultäO. Ljubljana. Mayerthaler, Willi (1981): Morphologische Natürlichkeit. Wiesbaden: Athenaion. Mayerthaler, Willi (1987): System-independent morphological naturalness. In: Dressler, W.U., Mayerthaler, W„ Panagl, O., Wurzel, W.U. (eds.) 1987, 25-58. Mayerthaler, Willi & Fliedl, Günther & Winkler, Christian (1998): Lexikon der natürlichkeitstheoretischen Syntax und Morphosyntax. Tübingen: Stauffenburg. Orešnik, Janez (1986): The Obligatorium of unemphatic pronoun subjects in Germanic languages. In: Filologija 14, 261-270. Zagreb. Orešnik, Janez & Snedec, Andrej & Teržan, Karmen & Trobevšek-Drobnak, Frančiška (1990): Introduction to the Subsequent Three Papers in the Present Volume. In: Linguistica XXX, 5-12. Ljubljana. Orešnik, Janez (1999): Krepke in šibke dvojnice v skladnji. Strong and weak variants in syntax. Ljubljana: SAZU. Orešnik, Janez (2000): The scale formats >sem (+/-A, -A) and >sem (+/-A, +A). In: Linguistica XL/2, 237-262. Ljubljana. Orešnik, Janez (2001): A predictable aspect of (morpho)syntactic variants. Predvidljiv vidik (obliko)skladenjskih dvojnic. Ljubljana: SAZU. Orešnik, Janez (2003): Naturalness. Some Norwegian (morpho)syntactic examples. In: Norsk Lingvistik Tidsskrift 21, 57 -70. Petrič, Teodor (1990): Posamostaljenja v knjižni nemščini . Filozofska fakulteta «Philosophische Fakultät>. Ljubljana. Petrič, Teodor (1995a): Naklonsi členki v nemščini. «Modalpartikeln im Deutschen». Doktorska disertacija «Dissertation». Filozofska fakulteta «Philosophische Fakultät». Ljubljana. Petrič (1995b): Modalpartikeln und Natürlichkeitstheorie. In: Boretzky, N. & Dressler, W. & Orešnik, J.& Teržan, K. & Wurzel, W. (1995) (Hgg.). 277 - 292. Petrič, Teodor (1996): Odgovori na pripombe prof. Orešnika. Workshop-Manuskript. Petrič, Teodor (1998a): Korrelationen zwischen merkmalhaften oder merkmallosen syntaktischen Varianten und verschiedenen Graden der Textsortenkomplexität. In: Linguistica XXXVIII, 1. 201 - 224. Petrič (1998b): Zu geschwächten Konstruktionen am Beispiel deutscher Nominalisierungen mit abstraktem Verbalnomen als Kopf. In: Teržan-Kopecky (1998) (ur., Hg.), 133 - 158. Schmid, Stephan. 1997. The Naturalness Differential Hypothesis: Cross-linguistic Influence and Universal Preferences in Interlanguage Phonology and Morphology. Folia Linguistica XXXI/3-4, 331-348. Snedec, Andrej (1988): Potek skladenjske spremembe. Angleški glagol do «Syntaktischer Wandel. Das englische Verb do>. Magistrsko delo «Magisterarbeit». Filozofska fakulteta «Philosophische Fakultät». Ljubljana. Teržan (1988): Trpnik kot skladenjski pojav v nemščini . Magistrsko delo «Magisterarbeit». Filozofska fakulteta «Philosophische Fakultät». Ljubljana. Teržan-Kopecky (1991): Prihodnjik kot skladenjska okrepitev v nemščini - diahroni in sinhroni vidiki . Doktorska disertacija «Dissertation». Filozofska fakulteta «Philosophische Fakultät». Ljubljana. Teržan-Kopecky (1998) (Hg.): Zbornik referatov II. mednarodnega simpozija o teoriji naravnosti 23. do 25. maja 1996. Sammelband des II. internationalen Symposions zur Natürlichkeitstheorie. Maribor: Pedagoška fakulteta, Pädagogische Fakultät. Trobevšek Drobnak, Frančiška (1987): Potek skladenjske spremembe. Angleški razširjeni časi . Magistrsko delo «Magisterarbeit». Filozofska fakulteta «Philosophische Fakultät». Ljubljana. Trobevšek-Drobnak, Frančiška (1991): Skladnja staroangleških glagolov s predpono ge- v primerjavi s skladnjo staroangleških glagolov brez predpone . Doktorska disertacija . Filozofska fakulteta . Ljubljana. Povzetek PREDLOGI ZA RAZŠIRITEV SLOVENSKE RAZLIČICE TEORIJE NARAVNOSTI Jezikovne zgradbe so rezultat optimalizacije v procesu sporočanja: tvorec besedila želi s čim manj kognitivnega napora doseči čim večji učinek (načelo minimax). Med jezikovnimi težnjami tvorca besedila in sprejemnika besedila prihaja do nasprotij: varčno izražanje otežuje razumevanje in obratno. Jezikovne zgradbe, ki nastopajo v besedilih, predstavljajo več ali manj uspešne kompromise za rešitev komunikativnih nasprotij. Sprejemnik besedila lažje razume pomen povedi (tj. stavčno vsebino in/ali namen povedi), če jo tvorec besedila posreduje z manj zaznamovanimi ali naravnejšimi (razvidnejšimi, ikoničnejšimi, tipsko ustreznejšimi, prototipičnejšimi, ...) jezikovnimi sredstvi. V sestavku avtor razglablja o možnostih za razširitev slovenske teorije naravnosti: kateri dejavniki povzročajo, da se npr. tvorec besedila odloči za potratnejšo, vendar razvidnejšo jezikovno zgradbo, in kako bi z razporejanjem načel naravnosti mogli prikazovati naravnostne razlike jezikovnih zgradb znotraj enega jezika in naravnostne razlike jezikovnih zgradb različnih jezikov. Varja Cvetko-Orešnik University of Ljubljana UDK 811.2Г02 : 81'367.625 THE AORIST AND THE PERFECT OF THE OLD INDIAN CAUSATIVES IN THE LIGHT OF NATURAL MORPHOSYNTAX Abstract The Old Indian causative conjugation features a reduplicated aorist and a periphrastic perfect. Within Natural Syntax, an attempt is made at predicting this state of affairs. Morphologically, the causative conjugation, the reduplicated aorist, and the periphrastic perfect are the most complicated structures within the verb system. The rules of alignment valid in Natural Syntax therefore predict the existing pairing of the two tenses within the causative conjugation. In Slovenia, we study the behaviour of (near-)synonymous syntactic expressions, here called syntactic variants. Whenever two syntactic variants are included in the same naturalness scale, and consequently one variant can be asserted to be more natural than the other, something can be said about some grammatical properties of the two variants. Naturalness values will be stated in naturalness scales. The basic format is >nat (A, B), i.e. with respect to cognitive complexity, A is more natural than B. In the present paper, the language examples are dealt with in »deductions.« Each deduction is valid only within the limits set by the language data stated at the beginning of any deduction. Each deduction contains at least two naturalness scales. The naturalness values of paired scales will be aligned by the principle of markedness alignment as stated in Andersen 1968 (repeated in Andersen 2001), and adapted to naturalness in the following way: what is more natural tends to align with another instance of more natural; what is less natural tends to align with another instance of less natural. A basic contribution of the »Slovenian School« to Naturalness Theory is the foll-wing list of instructions (criteria) for determining naturalness in (morpho)syntax: (A) The principle of least effort (Havers 1931:171). What conforms better to this principle is more natural. What is cognitively simple (for the speaker) is easy to produce, easy to retrieve from memory, etc. (B) Phylogenetic age. What is older phylogenetically is more natural. What is cognitively simpler (for the speaker) is acquired earlier by the language. (C) Prototypicality. What is nearer to the prototype is more natural. (D) Degree of integration into the clause. What is better integrated into its clause is more natural. This partially exploits (C): the prototypical syntactic situation is for a syntactic element to be well integrated into its syntactic construction. (E) Frequency (in the spirit of Fenk-Oczlon 1991). What is more frequent in a language tokenwise is more natural. What is cognitively simpler (for the speaker) is used more. (However, the inverse situation does not obtain: what is natural is not necessarily frequent.) (F) Small v. large class. The use of a unit pertaining to a small class is more natural than the use of a unit pertaining to a large class. During speech small classes are easier for the speaker to choose from than are large classes. (G) Specialised v. non-specialised use. If there exists a specialised way of expressing a category, that specialised way is very natural as expression of the category in question. Suppose a language has reflexive personal pronouns. In that case, they are specialised to express reflexivity (whereas other personal pronouns are not specialised to express reflexivity, even if they may express it under certain circumstances), and their use to express reflexivity is very natural: >nat (+reflexive, -reflexive) / personal pronoun as expression of reflexivity. (H) Use v. non-use. The use of a category or process is more natural than its non-use. This generalisation is based on the following consideration. All kinds of categories occur in the most natural lexical items, paradigms and constructions of the language, and ebb on the way out of that core. Take for example a language whose noun phrases distinguish singular, plural and dual. Although singular, plural and dual are not equally natural with respect to one another, each of them is highly natural in its own field. For instance, the dual is highly natural as an expression of duality: >nat (+, -) / dual in expressions of duality. This is correlated with the circumstance that the dual (in fact all three numbers) is present in personal pronouns, i.e. in the most natural noun phrases, while it may be present to different degrees in the remaining noun phrases of the language. The situation is similar with processes (such as raising); all processes are natural. (I) Acceptable v. non-acceptable use. What is acceptable is more natural than what is not acceptable. The very reason for the acceptability of a syntactic unit is its greater naturalness with respect to any corresponding non-acceptable unit. (J) What is more widespread in the languages of the world is more natural (the typological criterion). What is cognitively simpler (for the speaker) is realized in more languages. The current version of criteria (A-J) has resulted from the cooperation of Janez Orešnik with Helena Dobrovoljc. Each of criteria (A-J) is assumed to indirectly inform us about what is easy or difficult for the cognitive activity in the human brain. Criterion (B) has recently been removed from this list. Notice that the framework just outlined contains no generative component, and that it operates ex post facto. (The above is from work by Janez Orešnik and is here published with permission. Janez Orešnik has also translated the present paper into English.) The Naturalness theory will here be applied to the Old Indian causative conjugation, which is different from all other conjugations in that it is paired (a) with a periphrastic perfect (involving the auxiliary 'do, make' or, more frequently, the auxiliary 'be'), and (b) with the reduplicated aorist. For the (philological) detail, see Whitney 1889:378 ff. The question to be answered is, why the causative conjugation, of all conjugations, is combined with the two exceptional formations, the periphrastic perfect and the reduplicated aorist. More precisely, we shall enquire which presuppositions have to obtain so that this state of affairs becomes predictable. The answer will be couched in a deduction, as follows: *** Old Indian. Of all the conjugations, only the causative one is combined with a periphrastic perfect. The variants: the causative and the remaining conjugations. 1. The assumptions of Naturalness Theory: 1.1. >nat (other conjugations, causative conjugation) I.e. the causative conjugation is less natural than other conjugations. - The internal structure of the causative conjugation is richer than the internal structure of the other conjugations. The speaker generates and pronounces the forms of the causative conjugation using more effort than handling the forms of other conjugations. Therefore the causative conjugation is the least natural conjugation by the criterion of least effort, item (a) in the list of criteria. 1.2. >nat (other perfects, periphrastic perfect) 1.e. the periphrastic perfect is less natural than the other perfects. - The same argumentation as sub 1.1 applies. 2. Rules of alignment: 2.1. >nat tends to associate with another >nat, 2.2. u (with nm as a probable scribal inversion of a difficilior mri) = Ritamnel KaturiepuMor *Reitiamnel Catu-ritos / -retos = N (dedicatory Etruscoid Genitive II) + N (Rhaeticization of a Celtic nominative: Catu-ritos / -retos)\ cf. ritamne in FID 221 and see the research concordance in Table 1 and our discussion below. The twenty-two Magre votives conserved at Este were inscribed in a North Etruscan alphabet and were presumably dedicated, so the majority of scholars now believe, primarily to Reitia (probably a Veneticization of рцвегоа [dialectal peiGeioa] nominative singular feminine of the aorist passive participle of eeipco 'to say, to tell, typically through a divine intermediary', medio-passive 'to ask'; derivationally and functionally akin to Ilithyia = Juno Lucina, who, like Reitia, presided over the travails of women, from elrjXvdm to ep/o/uai 'to come'). For discussions of Reitia, the Venetic Artemis (Orthia) or Athena who was possibly represented in plastic art only on Venetic coins (cf. similarly the role of coins in the veneration of Juno Lucina), as the primary recipient of the Magre votives and on Magre as a possible sister foundation of the women's Baratella sanctuary-scriptorium at Venetic Este, see Kretschmer (1949), Mastrocinque (1987:97-128), Pascal (1964:113-114) and Whatmough (1922). The first publication of the Magre find was prepared by Giuseppe Pellegrini as a survey that appeared posthumously in 1918, a work that many authorities, such as Alberto Mancini (1973:365), still consider a remarkable piece of analysis. On March 31st, 1922, just two months after the death of Este's Alfonso Alfonsi and nearly a decade after the initial find, Joshua Whatmough (1897-1964), then barely twenty-five, read the Magre inscriptions at Este; see Whatmough (1921, 1923, 1933:33-48). Whatmough autopsied the Magre inscriptions at the request of his mentor, the University of Manchester's Robert Seymour Conway (1864-1933), with whom he the autumn of 1998, for kindly supplying biographical information on Pellegrini, an outstanding student of the justly celebrated Eduardo Brizio at Bologna. Incidentally, as the DAfs Horst Blanck kindly informs us, Alessandro Prosdocimi was director of the Este museum from 1887 until 1909, when he retired, whereupon administration of the museum was entrusted to Alfonso Alfonsi, but who was formally appointed director as late as 1921, just one year before his death. would collaborate in publishing the monumental three volume The Prae-Italic Dialects of Italy ( = PID) eleven years later. Since Whatmough's survey — and unfortunately many of his readings have by now been deemed both imprecise and questionable, the Magre votives have been addressed numerous times, most notably by Alberto Mancini (1975), Maria Grazia Tibiletti Bruno (1978), Stefan Schumacher (1992) and Alessandro Morandi (1999); see the concordance in Table 1. In our discussion below, the Magre and other Rhaetic inscriptions are referenced by their PID numbers, where available, optionally followed by the corresponding siglum in Schumacher's (1992) CIRCE (= Corpus inscriptionum raeticarum completum et emendatum) and a reference to Morandi (1999) = MOR + number. Venetic inscriptions are referenced by the sigla (e.g. Es 1 = Este 1) in Pellegrini and Prosdocimi (1967). The writer examined the Magre votives at the Museo Nazionale in Este on October 19th, 1998.2 Relative, much less absolute, chronologies for the Magre staghorns remain elusive. As an initial working hypothesis, however, it seems only reasonable to assume a relatively broad time span for a collection of votives that probably represents but a fraction of the ex voto remains that presumably accumulated in favissae over some three or four centuries during the sanctuary's putative existence. Even from the scant, regrettably largely decontextualized, archaeological evidence that has come down to us, it appears likely that votives were still being inscribed at Magre during the final decades (ca. 70-50 BC) prior to its destruction as a sanctuary -scriptorium, a result of the Roman conquest, and ensuing imposition of Roman construction on its surface. In his contribution to Italia omnium terrarum alumna (1988:121), Raffaele Carlo De Marinis, the authoritative and widely published Milan archaeologist, dated the 2 The following conventions have been adopted here: so-called "Etruscan f with an angled bar at its head = t, St. Andrew's Cross t = x, arrowhead tau - so-called "Magre thorn" = Јз, presumably the ultimate source of runic thorn. San is indicated by š. Gk. ф is retained for phi, while so-called "red" chi is represented by %, rather than v|f (psi) which, with regional variants with and without the handle (so-called "crowfoot chi", Lejeune's [1974] jr), was the actual morphology of chi in Etruscan, Rhaetic, Messapic, Venetic, Golaseccan / Lepontic and at Castaneda (crowfoot chi in two attestations to date); see Markey and Mees (2004). In passing, we note that phi is lacking in Golaseccan / Lepontic and that both phi and chi are absent at Tortora, at Novilara, in Oscan, Umbrian, Faliscan, Sicel and South Picene. In our transliterations here, syllabic punctuation is indicated by periods. Where indicated by puncts (typically three or six) in the inscriptions themselves, such "word" division is here indicated by a colon. We have inserted spaces to indicate "word" divisions, and these divisions often result from "etymological" interpretation of a scriptio continua text, while forward slashes (J) indicate actual line breaks in the originals. The smooth undersides — the horns were sawed in halves, of PID 221-233, 236-240, 242-243 bear graffiti that sometimes resemble Roman numerals (e.g. PID 227 V//, PID 238 11 X I, PID 239 X I); that is, X, D, V, I, , N, W, HI, but also crowfoot chi and decorative arrowhead zeta-like x's. In fact, of the twenty-two votives, only four (PID 220-230 = a composite, 234, 235 and 241) lack such graffiti; see Whatmough (1933: 33-48). These graffiti are similar and sometimes identical to those found at Padua, Monte Pore, on the Vadena stele, the votive alphabet pins from the Baratella sanctuary-scnpformm at Venetic Este, the Negau B helmet and the various Sanzeno-Casalini votives and castings; see fn. 7. Five of the horns have "partials": PID 226 (]emanis[), PID 235 (]šu), PID 240 Qixinatvixa), PID 241 (axt[) and PID 242 (atv[). These inscriptional fragments have been omitted by Morandi (1999) and Tibiletti Bruno (1978), and, in so far as "readings" are possible at all, have so far never been provided with "explanations"; see Table 1. related Vadena stele inscription (PID 196 = BZ-10) to the 5th century BC (ca. 450400?). He mistransliterates the stele's inscription, however, as pnake vitamu laze, apparently confusing Rhaetic arrowhead tau [{ф, t] with Camunic arrowhead zeta [s]. He offers no further interpretation, but then abruptly proceeds to assign the Magre inscriptions, which he also neglects to interpret, to the same century (ca. 450-400?). Pauli (1885:17 [Nr. 36], 107), Markey (2000) and numerous others have repeatedly interpreted Vadena's pnake as a Rhaeticized (Etruscoid-syncopated) Celtic Ben(n)acos, a reading with which Whatmough (1934:31) eventually agreed, although he abstained from interpretive translation at PID 196; see also Pellegrini's (1985:108109, fn. 50) abortive attempt. For De Marinis (1988:121), the chronologically diagnostic benchmark for dating Magre's votives is the ousting of closed alpha (= Latin a) by so-called "open alpha" (an a on which the central crossbar parallels, rather than horizontally joins, one of its sides, thereby forming a letter shape that approximates and can sometimes be confused with that of ^-shaped digamma). For De Marinis, the benchmark rule is: closed alpha until the end of the 5th century, but open alpha thereafter; see our discussion below and Markey (2001a:99-102). Conversely, in view of its two versions of closed alpha, the second putatively earlier than the other, Morandi (1999:78) dates Magre PID 221 = MA-1 to the Illrd century (ca. 250-200?). He then (1999:78) proceeds to date Magre PID 222 = MA-2, the final portion of which both he and Giuseppe Pellegrini convincingly reconstruct as a Celtic ka[turiei>u] (= Caturito for Caturitos), to the Illrd century as well; see Schmidt (1957:259) on -rito. In view, however, of the Celticity of its onomastics and the fact that its closed alpha is matched by Negau B from about 55 BC, Morandi's 3rd century date seems far too early. Morandi (1999:79) concludes his Magre chronologies by dating PID 231 = MA-11 to the 4th century (ca. 350-300?), though this is the only Magre inscription with two contrasting alpha's, the first of which is closed, and the second of which is open. Finally, in the PID, for reasons that are rather imperiously evasive (undefined "epigraphical evidence"), Whatmough (1933:34) categorically assigned the Magre votives to about the middle of the 3rd century; that is, within Este IV as he knew it, a period now generally considered roughly contemporaneous with La Tene B2 through CI (ca. 300-175 BC in Padania); see De Marinis (1986), Riemer (1998). Earlier, Giuseppe Pellegrini (1918:176) had dated the stratum in which the Magre votives were found to the closing phase of the pre-Roman Venetic period; that is, sometime just prior to about 150-100 BC. The fact is that, while contrasting alpha's such as those on PID 231 = MA-11 afford little or no uncontroversial evidence for dating the Magre inscriptions, they certainly do evidence the contemporaneous or near contemporaneous performance of different epigraphic disciplines at Magre. In reality, introduction of open alpha in the various North Italian epichoric systems must have varied widely, certainly earlier in Lepontic than Rhaetic, prospects De Marinis (1988:121) fails to entertain. In fact, the demonstrably late (ca. 55 BC or shortly before) Negau B inscription, epigraphically influenced as it clearly is by Magre, and the late Venetic Isonzo inscriptions (Is 1 and Is 2, ca. 100-50 BC) all have closed alpha\ see Markey (2001a:124, 130). Subsequently, De Marinis (1990-1991:216) suggested that the appearance of open alpha was contingent upon the disappearance of digamma, a contingency that seems indicated by the history of Golaseccan / Lepontic alphabetism with its closed alpha and digamma in inscriptions from the 6th through the 5th century vs. its open alpha and conspicuous absence of digamma in inscriptions from some two centuries later. Indeed, we find closed alpha beside upsilon but no digamma at Lepontic Banco (PID 273), Tesserete (PID 268/11) and Vergiate (PID 300), but this is presumably indicative of digamma'% marginalization as a semivowel. Cf. Magre's (PID 237 = MA 17) asua, possibly for *asva, but valt- and not *ualt- in the same inscription, and on upsilon vs. a marginalized digamma at Prestino (ca. 480-450 BC), see Markey and Mees (2003). Nevertheless, as we (2001:99-101) have demonstrated elsewhere, the universality of De Marinis's contingency hypothesis is vitiated by, inter alia, Camunic which deftly avoided the prospect of pernicious homographic confusion between open alpha, digamma and possibly even epsilon, as seems to have been the case on the late 6th century Providence Situla (RISD Museum, Providence, Rhode Island, Inv. No. 32.245); see our discussion below and fn. 6. In contradistinction to its two-pronged open alpha, Camunic introduced a four-pronged digamma that is unique among North Italian alphabets, whereupon Camunic continued both digamma and open alpha. Then, too, despite their virtual homography, open alpha and canonical digamma simply co-existed at, for example, Monte Pore and Serso and in the discipline that produced the remarkable Sanzeno-Casalini votives (SZ 1-15, ca. 420-380 BC). Open alpha and digamma also coexist on the late Vlth century (ca. 525 BC) Providence Situla (MOR 27), reputedly from Certosa, but here a two-pronged epsilon has been aligned with a two-pronged digamma. Interestingly enough, at Etruscan Balone (Rovigo) about 450 BC, we find closed alpha, while digamma has but one downward prong, though it cannot be confused with L-shaped lambda nor with F-shaped upsilon, while epsilon has three horizontal, rather than angled, prongs; see Marinetti (1994). Within the allegedly talismanic alphabetic disciplinae of northern Italy (with evident echoes throughout the eastern Mediterranean), alpha and digamma must have been open to persistent reciprocal morphological influence. In this tradition's Phase Two abecedaria, they would have been equipollent initials (a-e ... [full, unabbreviated consonant and vowel alphabet] / v-z = /v-d/... [abbreviated consonant only alphabet of Reitia veneration at Este]); see below for further definition of Phase Two alphabets. Identification of the language and ethnicity of the Magre inscriptions long remained as controversial as determination of their chronologies and epigraphic elucidation. Whatmough repeatedly (1923:69-72, 1933:4, 34) categorized their dialect as "Kelt(o)-Illyrian", a vacuously hypothetical branch of Indo-European that embraced an equally fictive and now generally discredited Illyrian. In comparison with Whatmough's categorization, we may cite Massimo Pallottino's (1963:77-78) rather evasive, prejudicial and hardly more informative assessment some forty years later, an evaluation embraced by the venerable Giovan Battista Pellegrini (1985:122), that the "Liguri e Reti" as "nuclei tribali" were "cuturalmente arretrati e privi di autentiche tradizioni storiche. La penetrazione etrusca da sud e le invasioni celtiche da nord debbono aver alterato profondamente i caratteri di queste genti ... ." Despite Whatmough's (1934) resistance to Thurneysen's (1932) stunning equation of Magre pinake with Etruscan zinace 'made, made for' (as an artistic signature) and thereby the insight that Rhaetic was necessarily Etruscan (or genetically Etruscoid or at the very least heavily influenced by Etruscan) and Kretschmer's (1932) impression that Rhaetic was a mixture of Indo-European and Etruscan, we now consider Rhaetic typologically Etruscoid, although it sometimes incorporated Indo-European elements, preeminently so with respect to (Celtic) onomastic items; see the relevant discussion and bibliographical references in Markey (200la: 135, 137). Further to the detriment of Whatmough's position, more than fifty years earlier, despite initial reservations (1885:101, 109), Carl Eugen Pauli (1839-1901) had been essentially correct in his contention that Rhaetic was probably Etruscan. Albeit clearly influenced by both Venetic and North Etruscan, the East Rhaetic alphabet that emerged at Magre was a distinctly independent epichoric accomplishment. It descended from a so-called (Etruscoid) Phase Two aev-Alphabet, that is, from an ancestral North Etruscan alphabet in which the first three letters were aev, not abg (abc); see Pandolfmi and Prosdocimi (1990:13). Such alphabets evolved after implementation of an Etruscan (Vulci?) reform that dictated elimination of beta, gamma (c), delta, s+, omicron and s'. The precise date of this early reform is unknown, but a best guess estimate is about 600 BC, though certainly prior to a later reform (ca. 570-560 BC) that ushered in so-called "Phase Three Alphabets" signaled by introduction, typically in alphabet final position (after chi), of an hour glass figure (8) to express /f/. Datable Phase Two Alphabets are evidenced as follows in northern Italy: Spina (from ca. 580-480), San Martino in Gattara (ca. 400), San Polo d'Enza near Reggio Emilia (Golasecca III A 3, ca. 400-375) and at Roncoferraro near Mantua (ca. 390); see Pandolfini's convenient catalogue in Pandolfmi and Prosdocimi (1990:1994). Note that both the San Martino in Gattara and San Polo d'Enza sites are in Celtic horizons; see Bermond Montanari (1969:218-224, 226-228), Colonna (1974:4-5, 1119) and de Marinis (1977:27b and Figure 3:1). Here we should underscore the fact that the role of Etruscan Mantua as a center for the spread of Etruscoid alphabetism in northern Italy, particularly to Golaseccan / Lepontic horizons, but probably also to Rhaetic sectors as well, should not be underestimated. Contrary to earlier impressions, the Mantua area must have hosted imposing and long-standing Etruscan enclaves, points of convergence for Greek, Celtic, Venetic and, of course, Etruscan trade and exchange; see the comprehensive state-of-the-art report and copious bibliographical documentation by De Marinis (1986), a survey now somewhat dated by continuing (at Ca' dei Monti in 2002) work at Etrusco-Celtic Monte Bibele (at Monterenzio south of Bologna, ca. 400-200 BC) and by De Marinis's own ongoing excavations at the Etruscan settlement at Forcello di Bagnolo S. Vito (southeast of Mantua on the triangle between the Mincio and the Po). In the absence, as demonstrated below, of an indigenous zeta (= /d/) and in contrast to Atestinian v-z = /v-d/ (the abbreviated consonant only alphabet of Reitia veneration at Este), the Magre alphabet's initial consonants would have been v-h. Like all North Etruscan systems and in contrast to the five vowel systems (with o) of Venetic and Golaseccan / Lepontic, Magre had but four vowels: a, e, i, u. Vocalic length was left unindicated (not by V+ h, gemination, macrons, or any other diacritic device), nor, in the same vein, was consonant gemination indicated. Magre's upsilon assumed both inverted (Л) and non-inverted V-shapes. With but one dramatic exception, discussed below, its epsilon was the cardinal angled three (never four!) pronged epsilon (E) found in virtually all North Italian systems. It had the conventional single-stroke vertical iota. As pointed out above, it had both open and closed alpha, though the latter shape was dominant. Explicit diagnostic arguments and evidential details need not detain us here, but it appears that, as was certainly the case in contemporaneous Etruscan, Rhaetic vowel systems, including that at Magre, were asymmetrical with a non velar [a]: i u e a and therefore /u/, but not /a/, could approach backward c; see Agostiniani (2000:495) and note Magre's (PID 239 = MA-19) ) below. On the other hand, the vocalic nucleus was eliminated after the glide in Negau B's harixasti, the Rhaeticization of a Germanic nom. sg. /'-stem *Harjagastiz; see infra. As for its consonants, Magre presents surd shapes and only surd shapes. For tau, Magre used both the cross-barred, so-called "Etruscan ?" and the St. Andrew's Cross t = , historically the x extruded from crossed theta, an innovation apparently spawned at Chiusi that spread northward. Indeed, Central Etruscan zila0 (with crossed theta) appears instead as zilax, that is, with St. Andrew's Cross tau, on the so-called "Second Rubiera Stele" (from about 640 BC), which was found as recently as 1984 in a riparian setting near Rubiera (Emilia); see De Simone (1992:10).3 Magre obviously eschewed scrittura fonetica (cf. fn. 4) and utilized both k and its aspirated counterpart % (chi) to the exclusion of c and/or qoppa, but, as pointed out by Whatmough (1933:36) and numerous others, while Magre attests cp, it lacked p; see our discussion below of ear-shaped vs. handled rho (not pi). Magre apparently merged p and cp as <Ф>, which seems to have been a general tendency in Rhaetic systems. As evidence of this merger, compare, for example, upiku (Siebeneich / Settequerce, PID 191 = BZ-3, MOR 10; Castelgrifo PID 192 = BZ-4, MOR 11) and ucpiku (Verona, PID 247 = VR-3, MOR 51) vs. uiiku (lx: Meclo, PID 210 = N0-3, MOR 13) as a facilior rendering of upiku with "unmarked" iota for pi; see fn. 6. Further to Rhaetic upiku ~ ucpiku 'to place, position (an object, institutionally, not physically)?' with the locative orpertinentivo, compare fragmentarily attested South Picene upeke[— at Crecchio (CH 1), on which see Marinetti (1985:113-117), who was apparently oblivious of the possible Rhaetic analogue, and Untermann's (2000:801) laconic entry. As a recap, then, while Magre merged p and

), while in others the reverse is true ( > p). Moreover, in some cases, Etr. t- has been assimilated to в-, and h- to f- and vice versa and b>ptof; cf. Lat. Fausta vs. Etr. Fasti, Fausti, Hasti; Lat. Faber : Etr. Hapre; Lat. Tiberius : Etr. 0efarie to Tiber : Etr. @epri, Qefri. In this context, it is significant to ask why some anticipated configurations do not occur: given Pupluna and Fufluna, then why not *Hupluna or *Hufluna1 Attempts to nativize non-Indo-European loans also altered "predictable" outcomes, construed as regularities, particularly so in the case of Phoenician where the underlying vocalism remains largely unknown and the consonantism is known to have lacked equivalent graphemes in the target orthography: Etr. Kardazie 'Carthage': QRTHDŠT'aew city', qutum 'pitcher' : KDM 'id.', španti 'basin' : §PN'id.'. See, further, the commentary to Table 2. is, Roman Z3-shaped pi plus triangular, ear-shaped, delta-like rho. Although ear-shaped rho (D) and handled (T-shaped) rho certainly co-existed in numerous Greek and Etruscan (e.g. Cerveteri, Vulci) systems, as well as in some (e.g. Sicel) Italic systems, ear-shaped rho and only ear-shaped rho is found in Oscan, Umbrian, Venetic, Lepontic, Camunic, at Novilara, at Monte Pore, in Rhaetic at Sanzeno and in North Etruscan at Mantua, Spina, Adria, Marzabotto and Villa Coviolo (Reggio Emilia), while handled rho and/or R (and only these shapes) are found in South Picene (only handled rho, both angular and rounded), Faliscan (also R), Messapic (also R) and at Tortora (only handled rho, 7x); see Guarducci (1995:99). A Magre type handled rho also occurs (lx) on the Castelrotto bone votive (VR-4) discussed below. Moreover, the Rhaetic inscription (PID 208 = NO-10) at Tavon near Sanzeno apparently also attests handled rho, probably ireke, although this has been read by some as ileke, cf. Schumacher's (1992:152) l]rileke:sa[l. Paradoxically, Tibilleti Bruno (1978:222) reads rho at Tavon, presumably because the "knob" of its handled rho is rounded rather than angular like the Magre handled rho's, which, however, she seemingly prefers to read as pi's. Were she consistent, then she would surely have read Tavon as ipeke with pi. Such speculation aside, the fact is that, at Magre, ear-shaped rho and handled rho were presumably in free (allographic) variation, not only initially, but also medially and finally: e.g. #ri- in PID 225 vs. #Pi- in PID 221, 222, 228, 229), cf. #Pi-vs. -uri- in PID 222, and note keP = [ker] in PID 229; see McLaughlin (1963:56-133) on the phonological significance of such graphemic orchestrations. Finally, we note that, with the exception of Camunic and its innovative paper-clip pi (a probable precursor of which is the North Etruscan humped or camelback pi on Stele 2 at Rubiera), presumably the progenitor of runic

, pi was one of the more immutable letter shapes of early northern Italian alphabetism.5 5 Basic pi shapes on the Italian peninsula fall into two major typological categories: (1) humped/semicircular, sometimes postulated as Proto-Italic (Sicel, Messapic, South Picene, Oscan), but presumably an outgrowth of (2), the shepherd's crook pi, a perpendicular stave with an oblique descender from its top, either to the left or the right (Etruscan, Umbrian, Golaseccan / Lepontic, Rhaetic at Sanzeno, and in the Novilara inscription). The shepherd's crook pi was sometimes hooked at the end, as in early Venetic, Messapic and at Tortora. Naturally, the unhooked shepherd's crook pi was sometimes easily confused with lambda's that had their oblique descender attached to the head, rather than the foot, of the perpendicular stave; see Mancini (1975:303-305) for an enlightening discussion of this problem. In Rhaetic, and to some extent in Venetic as well, maintenance of a clear distinction between lambda, pi and even asymmetrical upsilon's must have been problematic. This problem was solved in Camunic by introduction of the M-shaped so-called "paper-clip" pi, probably a fusion of a right-facing shepherd's crook pi (or lambda) and a left-facing shepherd's crook pi (or lambda), a combination that may well have served as the model for runic

; see Marstrander (1928:104). WTiile alignment of the paper-clip /f/ of Tortora (vs. arrowhead tau as /f/ in Faliscan and hour glass 8, simplified as a colon in South Picene, as /f/ in Oscan, Umbrian and South Picene) with the paper-clip digamma of South Picene may be a productive line of inquiry for mapping the transmission of alphabetism on the Italian peninsula, alignment of paper-clip digamma with runic

while omitting discussion of the Camunic paper-clip pi leads but to genetic confusion; pace the under-informed discussion in Letizia Lazzarini and Poccetti (2001:41-42), cf. Markey (2001a:95, 103). It was, of course, the humped/semicircular "Asiatic" pi that engendered "modern" Greek pi. It appears that one Rhaetic solution for maintaining a distinction between upsilon, lambda and pi was a cop-out, namely selection of iota as an "unmarked" neutral substitution for lambda or pi, particularly between upsilons. This was presumably a serviceable solution given the redundancy of votive "texts". Note ишш for upiku at Meclo (PID 210 = NO-3, MOR 13), as pointed out above. Consider, moreover, Magre's PID 227 = MA-8 (right-reading scriptio continua) reixemuiu ina e = Reite muiu ina e (with muiu ina e inverted, upside-down) for *Reitie muiu zina e 'dedicated, offered to (and) made for Reitia', with passive verbs in asyndetic Magre lacked the Sanzeno-Bolzano/Vadena and Negau (A [la]) arrowhead tau (or sampi) = t- Like Camunic, however, Magre presents a saw-toothed thorn (p) with either two or three teeth, probably historically a digraph, t + s. Indeed, in view of Magre (PID 228 = MA-9) pinaxe : Etr. zinace [tsinake] 'made, elaboravif, one necessarily infers that Magre p represented [ts]. However, Magre also records xinaxe (PID 231 = MA-11), with St. Andrew's Cross tau, which patterns with jina/e (SZ-1) at Sanzeno-Casalini, a site in proximity to Revo's \erisna = Vače's ] er is na. We thus retrieve <ф : x : | Т> as an etymologically braided chain of divergent graphemes that could represent surficial [ts]; see Markey (2001a:142-143). In contrast to Magre (and Rhaetic alphabets generally), we recall that Prestino rendered [ts] with zeta: kozis = Celtic gotsis < IE *ghosti-s (JEW 453).6 At Padua tandem, indeed the same 'dedicated' + 'made' formulaic sequencing as in Greek votives (civsOtike + л8ло(г|К8); see Rouse (1902) for examples. If this interpretation is correct, then this is the only attestation of mulu in Rhaetic, but note North Etruscan mvlvainice on the Providence Situla (ca. 525 BC). Alternatives are bleak: there is no comparable Etruscan muiu or Celtic moio, mojo, muio, etc. Moreover, selection of iota (mvtv) instead of L-shaped lambda (mvlv) avoided possible misinterpretation as (numerical?) mvw. 6 Zeta charts a checkered history in the early alphabetism of the Italian peninsula. In most Italic settings, as noted below, it denoted [ts], its value in Etruscan, which is the source of, for example, Faliscan's I-shaped zeta (at Narce). In other epigraphic settings, however, zeta is absent, lacking at Italic Tortora and in the cryptic Novilara inscription. While zeta makes single appearances at Prestino and Casate, it is otherwise absent in Lepontic. In the 2nd BC century Casate inscription, we have: zv ošoris = dv Ordorix 'dv prince of maces', in which dv plausibly represents a reversal of vd, the initials of a consonant only Phase Two abecedarium; see Schmidt (1957:252) on ordo- and Solinas (1995:341) on Casate. On deployment of san for dentals (t / d) and clusters with dentals in the Celtic alphabetism of northern Italy, see Markey and Mees (2004). In South Picene, which has ear-shaped delta, I-shaped zeta was apparently relegated to the tau slot to represent /t/. The top and bottom crossbars of I-shaped Picene zeta for tau were retooled as globes (dots), yielding a dumbbell-shaped letter that represents /d/ in the Celtic Ic-inscription on Negau A: Iubni gen. sg. of Dubnos; see Markey (2001a:105, 113-116). In Rhaetic, in addition to Magre and Castelrotto (VR-4) as detailed below, we find zeta shapes only at Lothen, Steinberg and Sanzeno-Casalini (lx). On the Lothen (PU-1, MOR 7) belt buckle (later half of the 5th century) from St. Lorenzo di Sebato, Valle Pusteria (Pustertal), eelzume (with a zeta often misread as sigma) presumably corresponds to Etruscan VELem; that is, (early) Rhaetic z : Etr. theta in an "inverse spelling"; for a classic restatement of the orthographic / phonological mechanics, see Hoenigswald (1960:9-10). In passing, we should point out that the buckle's deer motif is foreshadowed by comparable arrangements in the late Vllth century style of the "Maestro delle Spine di Pesce" at Cerveteri (cf., for example, Inv. No. 50364, Collezione Augusto Castellani, Villa Giulia) and in heavily Etruscan-influenced Faliscan venues (e.g. Vllth century Etrusco-geometric amphorae, particularly from Narce's necropolis of Contrada Morgi). Moreover, execution of the Lothen graphics finds a near match on the celebrated and contemporaneous Celtic terracotta from Matzhausen (Oberpfalz, Bavaria; Inv. No. IIcl244, Museum für Vor- und Frühgeschichte, Berlin). As Mitja Güstin has repeatedly informed us (p.c.), art forms that emerged in Etruria were quickly picked up in Celtic settings, either as imports or as skillful copies. Cross of Lorraine shaped zeta's (T), conventionally transliterated as f s, have crossbars in immediate parallel proximity to each other across the middle of a central perpendicular stave. In Rhaetic, they are found only at Steinberg (ST-2, ST-3: kasTri- = kastri- : Etr. Castr-: Kaaaävdpa ?); see Schumacher (1992:223). Finally, concluding our inventory of Rhaetic zeta's, we note that a "Camunic-style" (Castaneda, Montagna, Tresivio, Valcamonica - Cimbergo, Voltino) arrowhead zeta has so far been attested just once in a Rhaetic inscription, namely, as an isolated graffito on the "belly" of a votive from Sanzeno-Casalini (SZ-11, MOR 18; Inv. No. 7668, 'Castello del Buonconsiglio' [Monumenti e collezioni provenciali], Trent); see Pellegrini (1951:313, Fig. 11). In Camunic, however, arrowhead zeta represents /s/. On arrowhead zeta in Castaneda's demonstrably Celtic inscription; see Markey and Mees (2004). However, the Sanzeno-Casalini arrowhead zeta may well be a smith's mark, as it also occurs elsewhere at Sanzeno on castings; see Nothdurfter (1979:101, 103, Taf. 24-25) and de Marinis's (1986:107) enlightened and well documented commentary on this early practice for talismanic or apotropaic purposes and recall the graffiti on Negau B, Vadena and the undersides of the Magre votives. (PID 244 = PA-1, MOR 52), however, we find Etsuale (a possible Celticity?) vs. Magre's Estuale (PID 232 = MA-12), [Ej.stuva (PID 231 = MA-11) and Es.stuatel (PID 233 = MA-13). As a further possible, albeit remote, scribal Celticity in the Paduan inscription, note, if etymologically related, suspension of the nasal in Padua's utiku past passive participle 'brought' vs. utince 'led forth' in the Liber Linteus; see Facchetti (2002:11-12), who justifiably segments utince as ut-in- like alpn-in-, rep-in-. Suspension of nasals is, however, also periodically noted in Etruria, e.g. Atiuce = Avrioxog : Lat. Antiochus. Given the morphosyntactic comparability of Padua's kaian (ca. 500-480) and kaianin (with deictic, pronominal acc. -in 'this') on the Providence Situla (Certosa, ca. 525 BC), the Paduan ember palette would seem to share not only the same lexical, but also the same dialectal tradition. By the early fifth century, except in the dialects of the northwest, intervocalic -y- had been lost (cf. Kae < *Kaie : Lat. Caius), particularly between equal vowels: kaian > *kan, whence Etrurian carta, can(-l) 'gift, image, stele' vs. Xanu at Sanzeno-Casalini (SZ-1, ca. 420-380 BC) and Lothen (PU-1, ca. 450 BC). On the problematic semantic range and definitional complexity of cana; see the ample documentation and discussions by Agostiniani (1982:179, 189-190). While I-shaped zeta rendered [ts] at Prestino (and in Oscan, Umbrian, Faliscan, Sicel and Messapic), it represented /d/ at Este, where Etruscoid I-shaped zeta assumed a Cross of Lorraine format with two slanted bars across the midpoint of its perpendicular stave. Significantly, Magre attests this latter-day Este-shaped zeta just twice in a single inscription (PID 243 = MA-23: usiiJju / zezeve) that we now consider East Celtic rather than Rhaetic. In that inscription, detailed below, zeta represents /d/: zeze- = dede 'gave, dedif. In addition to zeta as /d/ at Magre, there is, in the entire Rhaetic sector, but one other probable occurrence of I-shaped zeta representing /d/; namely in VR-4 from Castelrotto (Comune di San Pietro di Cariano, Valpolicella, Verona). This epigraphically difficult left-reading inscription is in a Magre-derived font on a pig bone (votive?). Castelrotto's M-shaped s an, inverted upsilon (A), handled rho (P) and handled "red" chi (VP) are all uniquely matched at Magre. As is the case with PID 243 = MA-23, the Castelrotto inscription is presumably also Celtic. It possibly reads as follows: ruzineš^an = rudinešgan, in which at least ruz is epigraphically unmistakable, probably rudi- for Celtic roudi- 'red'; cf. Schumacher (1992:207 and Tafel 2.3), see Schmidt 1957:262) and revisit fn. 5 below. Although scribal practice and disciplinae scriptori at Magre apparently recognized graphemic as /t/ - /ts/ - /d/ respectively, and were obviously the indigenous graphemes most actively utilized to represent dentals to the general exclusion of zeta. Nonetheless, we contend that, in addition to its recognition of Atestine use of zeta as /d/, Magre alphabetism deployed its thorn in the same discretionary manner as theta was deployed in comparable North Italian Etruscoid systems, namely, as a "wild card" to render /dJ unambiguously in Indo-European, particularly Celtic, dialects. Just as discretionary theta was marked, so too was discretionary thorn. It was these marked, sometimes virtually moribund, graphemes that served as discretionary disambiguators in polyvalent systems in which, for example, St. Andrew's Cross tau could represent both /t/ and /d/. Consequently, arrowhead tau, a regional innovation, was deployed in the Vadena Stele inscription as a discretionary marked tau to represent /t/ unambiguously, while bivalent St. Andrew's Cross tau (t or d) was deployed to represent /d/: vixamu / i.л 1e = Celtic windamolatos 'having most conspicuous warriors'. This is obviously an archaic (bipartite) superlative bahuvrlhi immediately comparable to Prestino's uvAMOKOZis 'having supreme guests'. Vadena's final, after la|e, highly suspect sanlike figure is most credibly interpreted as a decorative graffito akin to the terminal decorative graffiti on Negau B and on the undersides of many of the Magre votives; see Markey (2000) and fn. 2. In line with these interpretations, we derive Magre's ustijiU (PID 243 = MA-23) from an underlying Celtic anthroponym *Ostedö(n). Cf. Osti-names (Vi 2, Es 6, Es 133, Ts 1, Tr 1, Tr 3) in Venetic settings at both Este and Lägole (Calalzo); see Morandi (1999:84), Pellegrini and Prosdocimi (1967:2.148-150), and RIG 4.4, 102, 103, 314. As a product of Thurneysen's Law (1946:Art. 75), Magre's cognomalizing -i})u = -idü necessarily derives from -edü < *-edö(n). We may therefore compare a Pre-Celtic *-edö(ri), segmentable as *-ed-5(ri), with the Oscan and Umbrian patronymic formans -idio- (~ -edio-), a thematic composite of -id- + -yo-, Messapic patronymic -id-, and Greek -idä, which formed anthroponyms, gentilicia and patronyms, e.g. ПршџгбгЈд-, cf. the abbreviated account by Thurneysen (1946:58-59) and the uninformed discussion by Prosdocimi in Pandolfini and Prosdocimi (1990:292-295). Having identified ustij>u (PID 243 = MA-23) as Ostidü < *Ostedü, we may now proceed to relate its composition to Magre's kuniijiu (PID 224 = MA-5) < *Cuniedü, Prestino's pliale8u (ca. 480-450 BC) = Plialedü, Salassian (Lepontic) numismatic sexe8u (ca. 400-377 BC) = Segedü, Verona's koliti.etu or, better, kaliti.etu (PID 248 = VR-5) = Caletiedü : Lingones numismatic ku/xtj:öoi) (RIG 4.91, 92, ca. 148 BC) = Caletedü, Ornavasso's oletu (PID 307, corrected reading, ca. 190-75 BC) = Olledü, Cenomanian kelešu (ca. 100-50 BC) = Celedü at Isola Rizza in Casalandri (Verona), built to *kel- 'to strike' (JEW 545-546); see Colbert de Beaulieu (1966), Solinas (1995:376, 1998:148), Schmidt (1957:160, 170, 186, 250-251, 265-266). On deployment of san for dentals (t /' d) and dental clusters in early Celtic alphabetism; see Markey and Mees (2004) and fn. 6. With respect to -idü / -edü, then, Magre lines up securely with Lepontic and Gaulish (including Cenomanian) onomastic practice recorded both long before and long after the Celtic onslaught of 388 BC. Moreover, as Magre's -idü evidences Thurneysen's Law and contrasts with -edü elsewhere in northern Italy, one might infer a comparatively late (ca. 70-50 BC?) date for PID 243 = MA-23; see our account further below. For now, however, compare Magre's -idü < -edü and Cembra's (PID 215 = CE-1) vino- < *veno- in vinutalina < *Venö-tal-inä, a situla inscription traditionally dated to between 390 and 350 BC whose OSV-syntax and dedicatory intent are highly reminiscent of the celebrated second century BC Briona (Novara) inscription; see Solinas (1995:379-381) for a convenient survey of the latter. Cf., further, Cembra's vinutalina < *Veno-tal-inä, nom. sg. fem., as a hypocoristic diminutive, lit. 'the little one having the clan's brow, forehead, face,' and Briona's patronymic tanotaliknoi, nom. pi. masc. of a sg. *tanotaliknos = Dan(n)o-tal-ikno-s 'son of one having a judge's brow, forehead, face' : (Briona) tanotalos = Danno-tal-o-s. We recall Latin Fronto, -önis, Celticized as Frontu, Frontunis, a caique on Celtic talo- 'forehead, brow, face' (Olr. taul < *talu-), along the peripheries of Romania, particularly in Romanized Noricum, but also in post-conquest (after ca. 15 BC) Valcamonica; see Falkner (1948), Markey (2001a:103), Schmidt (1957:91, 274, 289).7 Magre's sigma consistently has but three strokes, as opposed to excessively multi-stroked serpentine sigma's in early Lepontic systems vs. three or four-stroked sigma's (consistently) in later Lepontic systems. Magre apparently never evolved or adopted a butterfly or a Lägole san, and here one recalls that san = /Š/ designated palatal s, while sigma = /s/ designated apical s. Palatalization of 5 before /i/, /у/ or consonants was putatively a northern, rather than southern (Etrurian), phenomenon: so North Etruscan spur- 'city' vs. Etrurian spur-, like German šp- vs. English sp-. Magre inscriptions attest h, indeed a 3-ranged ladder h, but /h/ is generally absent in Rhaetic, and it is distributionally rare (regularly only word initially) in Etruscan. In fact, A is a beacon for word boundaries in deciphering Etruscan scriptio continua. At Magre, it is found in but two inscriptions: PID 221 = MA-1 (helanu) and PID 225 = MA-6 (triahi.s). Otherwise in Rhaetic, h is found twice (also 3-runged) at Lothen (PU-1), twice (three-runged) in the Ca' dei Cavri inscription (PID 247 = VR-3) and, finally, possibly three times (SR-1,2, 7) at Serso (Pergine, Trent) as follows: -vitahur (SR-1), -vitahu.r (SR-7) and, possibly, as a two-ranged variant in hetinu or hetina (SR-2). We approach the matter of h's representational status indirectly. PID 229 = MA-10 (right reading, scriptio continua) ri^iekerrinake = Ritie ker rinake for *Rietie(i) ker tinaxe. Having emended the putatively erroneous scribal dittography of rr by substituting rt and having etymologized tinake as tinaxe to fit the etymologically inferred semantactic requirements, this inscription seems transparent enough except for ker as an apparently problematic hapax. There is no comparable onomastic Ker in any probable donor language. We compare Ritie ker tinaxe '(it) was made as a ker for Reitia' and mi titasi ever menaxe (TLE 282, Bomarzo, a 3rd-2nd century mirror inscription) 'I was made as a kver for Tita'. We have seemingly comparable common denominator configurations and therefore equatable morpho-syntactic constructions: (it) /1+ N(dative /pertinentivo) + ever + V/e; see Agostiniani (1982:106, 219-220). We therefore equate Magre ker and Etr. ever, whatever its precise semantic value, but presumably 'gift, offering,' its long-standing traditional gloss supported in part by combinatory equation of ever turce and alpan turce. Moreover, in the majority of its occurrences, ever follows a theonym in the datival / pertinentivo case; see Zavaroni's (1996:324-325, fns. 97 & 98) thorough research history. On ES 398, ever is personified as Cvera 'Grace, Benevolence'. Finally, we n Segmentation of Cembra's vinutalina as vinu talina and interpretation of vinu as an abbreviated (why?) vinum and talina as an otherwise unattested Etruscoid gentilicium in -lina (cf. tau mlina on SZ-9 = MOR 20 at Sanzeno-Casalini, Etr. puplina, etc.) is counterproductive, raising, rather than solving, a multitude of epigraphic, syntactic and etymological problems. note that gradual elimination of labial closure and regression of aspiration are clearly observable in Etruscan from the early 4th century onward: */vestna- > * yes t na- > cestna- (cestnal; San Manno, Pe 5.2, TLE 619, 3rd/2nd century). If construed as a datival recipient, then Ritie in PID 229 = MA-10 presumably caiques Ven. re.i.ti(i)a .i (dat. sg. (2-stem). A Neo-Etruscoid (after ca. 375) Ritie could be legitimately derived from Venetic *Reitiai, albeit without remarking with -i. As argued by Agostiniani (2000:495), *Riti-e-i would derive from *Riti-ai + -i to yield a late Neo-Etruscan feminine "absolute" singular in -ei. Etruscan "absolutes", we recall, functioned as morphologically undifferentiated nominatives or accusatives; see Facchetti (2002:fn. 38). In fact, such late 4th century Neo-Etruscan remarking is apparently recorded by Magre's ritiei (PID 224 = MA-6). The argument here, then, is that Ven. dat. Reitiai was rephonologized (nativized) as a Rhaetic Ritie(i), a feminine "absolute", but that the thus nativized Ritie(i) retained its original Venetic datival (pertinentivo) morpho-syntactic functions within Magre's highly proscriptive votive setting. This argument seems validated by what may be presumed to be later replacement of Ritie(i) as an early nativized approximation of Venetic datival grammaticality by a fully Rhaeticized Ritale in PID 228 = MA-9 with the (later) datival (Pertinentivo II) desinence -ale common to both Rhaetic and Etruscan. PID 228 = MA-9 (1 line: right-reading scriptio continua) reads as follows: ritalelemais{>inake = Ritale Lemais pinake (N-ale = Pertinentivo II (recipient) + N-is = Ablativo I (agent) + Уке) 'Made for Reitia by Lema', where Lema is perhaps akin to Ven. Lemetor-; see Pellegrini and Prosdocimi (1967:2.129-130). This interpretation avoids positing dittography (-le-le-) and then presuming Mais to be an (otherwise unknown) idionym. Returning to PID 229 = MA-10, we contend that, if Magre ker = Etr. ever, then kw> k- in Rhaetic (or at least at Magre). By extension, one could assume that, in a structurally parallel rephonemicization and / or regraphemicization strategy upon general loss of labialization, hv- > h-. Recall that both k- and kw- (qu-) in Latin loans were represented by in Etruria. In both Etruscan and Atestine Venetic orthography, however, graphemic vh / hv = /f/, the digamma of which yielded Lat./8 Alternatively, Venetic at Lägole (Cadore) and Würmlach opted for н, rather than vh / hv, to represent /f/: ha.r.to (Gt 14) = Harto. In Etruria, h- = Lat. / e.g. Etr. Herme = Lat. Firmus, Etr. Hapre = Lat. F aber and revisit fn. 4. Reiterating and hopefully refining arguments presented in Markey (2001a:95), we accept the convention that archaic Etruscan (digamma / upsilon) represented semivocalic w, but suggest that w later emerged as a labiodental [v] upon acquisition of stridency. Phonologically, then, Proto-Etrascan *w later emerged as v as a result of o Incidentally, the late Arthur Gordon (1975:52-57) presented the fullest research history to date on recognition of the digraphs vh / hv as /f/, first by Wilhelm Deecke (1831-1897) in print in 1888 (if not in conversation before) and again three years later by Carl Pauli (1839-1901), then 52; cf. Hamp (1981:152-153). Unfortunately, both Deecke and Pauli found confirmation in Wolfgang Helbig's (1839-1915) "Manios hoax" of 1887, a masterful collaborative forgery that took epigraphers and classicists nearly a century to discredit. The list of worthies, in addition to Deecke and Pauli — Holger Pedersen, Carl Darling Buck, Ernst Pulgram and numerous others, even yet today, "sucked in" by Heibig et Cie. is no less than stunning. articulatory simplification (bilabial being arguably more marked than labiodental), while v subsequently gave rise to / upon acquisition of tenseness; that is: *w (idigamma) > v (digamma) > /(vh/hv and later 8). Just so, as Cortsen (1935:178) pointed out, while p> p^ (/in Etruscan,/never becomes p; cf. Breyer (1993:21). Note, further, Nevtlane on ES 231 for ШотпоХеџо^, nefts : OLat. neptis, NeOuns : Neptunus with cluster simplification in Etruscan. We may therefore infer that, upon acquisition of stridency, an affricated Rhaetic tp was variously (dialectally) identified with /v/ or /f/. This inference accounts for (pEL- ((pELZURiE-) at Lothen (PU-1) = Etruscan vel- (velöür) < *wel-. Note, too, affrication of 6 [t] > z [ts] in the non-etymological orthography of Lothen's (pELZURiE- vs. Etr. velOur. Moreover, фммд (SZ-2) at Sanzeno-Casalini, presumably a nom sg. fem. idionym or epithet, the putatively syncopated counterpart of Sanzeno-Casalini's uprniMA (SZ-1), is traditionally compared with Venetic Frema (vhrema.i.-). We conclude as follows. On the basis of etymological correlations, as selectively demonstrated below, we contend that, when confronted with possible representational ambiguities between digamma / upsilon and affricated cp, Rhaetic, specifically at Magre, optionally deployed = /f/ (reminiscent of Etr. h- = Lat. /-) to represent aspirated Celtic w = [wh]; on which see now De Bernardo Stempel (1990). In PIL) 221 = MA-1, as detailed below, h- in Helanu was presumably selected to represent Celtic Vellaunos (in which v- = [w^j ), thereby avoiding confusion with, for example, an Etruscoid Velna (in which v- = [w]). One is reminded of ulatucia ~flatucias at Larzac. We may categorize Magre's consonant inventory as follows where graphemic realizations are presented in angled brackets beside presumed phonetic values: Consonants Labials Apicals Gutturals Sibilants Laterals Resonants Voiceless Stops Full expiration (p «p> t (?) (6 <©>) i It (ts ) 1 Restricted expiration (p <

) t k< K> ts' > i Fricatives / S š Nasals m n Liquids w / / r Table 2. The Rhaeto-Etruscoid Consonant System at Magre Commentary. There appears to have been a general tendency toward a regression of aspiration just as there was in Central Etruscan. However, only the k vs. % contrast is diagnostic, as p is lacking and all instances of q> are etymologically /b/ at Magre. Nevertheless, a combination push and drag-chain movement in the direction of more restricted expiration is clear in the case of apicals and sibilants, whereby © and z were eliminated in favor of т and Jj respectively. The same progression is seemingly evidenced by р [ts'] > м (san) [š]; we have -ipu at Magre vs. comparable -ešu at Isola Rizza, perhaps a window on linguistic competencies in a contact situation: ts (z) > ts' (t>) > š (m). There was also an apparent tendency toward lenition whereby plosives were weakened to fricatives (p > f) along a strength hierarchy. Lack of a voiced / voiceless contrast may have mattered less to speakers of Continental Celtic who may have been more concerned with contrasts in fortition than voice, than to speakers of other languages, such as Venetic, with a critical need for this contrast; see Watkins (1955), Evans (1967:403), Jackson (1953:545-548). Observe what are necessarily fortition distinctions between the two 's of Rheto-Celtic at Sanzeno / Vadenä: fortis arrowhead tau for Celtic /t/ vs. lenis St. Andrew's Cross tan (x) for Celtic /d/. The paroxytonality of Celtic onomastic items embedded in Rhaetic inscriptions is reflected by accent-conditioned reduction (Etruscoid syncope) / assimilation; see De Bernardo Stempel (1994) and recall Celtic ben(n)äkos > pnäke at Vadena. Celtic onomastic desinences were optionally subject to Rhaeticization, e.g. Celtic nom. sg. mase. -os could be Rhaeticized as -e, that is, as an Etruscan and Rhaetic absolute desinence (vs. Celtic and Italic voc. sg. -e), perhaps as part of a facultative interlingual discourse strategy, a quasi-pidginization process seeking common ground. Alternatively, it seems, Celtic -os could also be Rhaeticized as -u, rather than -us, thereby avoiding confusion with Rhaetic's sigmatic genitive (Genitive I). The flexible rule-of-thumb for Rhaeticization of non-native onomastic desinences seems to have been as follows: forms with Italic (particularly Latin) analogues were Rhaeticized in the same manner as Italic (particularly Latin) forms were Etruscanized: e.g. Lat. -us (os) -ius (-ios) > Etruscan and Rhaetic -e, -ie respectively. Forms without Italic analogues were, however, optionally grammaticalized / nativized (Celtic -os > Rhaetic -u) as if Rhaetic (Etruscoid). With the above survey in mind, we now proceed to comment further on Celticity in a selection of Magre votives. (1). PID 243 = MA-23: ustij>u / zezeve = Ostidü dedeve < *Ostedö(n) dedhe(t)-we 'Ostido dedicated (me, this) emphatically'. Cf. Gaulish бебе 'posuit, placed' (with the sense of aveOrjKs) < *de-dhe-t, and for the enclitic emphatic -we ~ -we (> -wT), hereby corroborated as the initial morpheme of the Castaneda inscription, see, further, Markey and Mees (2004). Cf. Latin ve-cors 'mad' = Old Welsh gui-cr 'brave', lit. '(too) much heart.' Bernard Mees (p.c.) points out that *co(m)we 'emphatically, utterly' underlies the coui of secoui at Chamalieres in the figura etymologica of li. 7-8 (etic secoui toncnaman / toncsiiontio 'as well as they who may prosper that prosperity utterly') with etic reflecting formulaic utique and tone- 'to prosper' : Olr. tocad 'luck, fortune, prosperity' < *tonketo- vs. e-grade Olr. techt(a)e 'proper' < *tenk-tyo-; pace Lambert (1997:156). For the mechanics, cf. Olr. co(a)ir 'proper, fitting' < *com-wari-; see Thurneysen (1946:503). Magre's syntagmatic dedeve is, presumably, a regional caique on formulaic avEdrjKs-spi, exemplified by SOANE0EKEEDI = oco(g)-aviet]KS-spi on a votive at Adria (Cabinet 7, Museo archeologico di Adria, autopsied by the author on October 25th, 1998) ' in good cause (salvus) - dedicated - emphatically' with formulaic suffixation of the intensive prefix epi- to the perennial Greek votive operator avsdtjKE; pace Dubois (1995:177-178). A photograph of PID 243 = MA-23 is displayed on Adolfo Zavaroni's admirable website, and the inscription has been accurately reproduced by Morandi (1999:Fig. 29, No. 48). The orthography of ustiJju (AsxiJjv) with both inverted and uninverted upsilons, St. Andrew's Cross tau and a thorn with three teeth is consistent with Magre fonts. While the zeta's of zezeve are generic Venetoid zeta's, the epsilon's are uniquely diagnostic. They find their closest match in the epsilon of the late Negau A (II) Celtic inscription kerup = Kerub[ogios] 'he who harvests deer'; see Markey (2001 a: 116-118). This was probably the helmet's final inscription, plausibly engraved in Noricum, if not on the Magdalensberg, toward the middle of the first century BC; that is, just prior to its deposit at Ženjak (Negova / Negau). (2). PID 237 = MA-17: (right-reading scriptio continua) klevieva.l.xikinuasua = Klewie Valtikino ASUA < *Klewios Waletikenos ASUA (ASVA) 'Clevios Valeticnos LIBENS MERITO'; that is, a binomial (Celtic or Venetoid) idionym + patronym configuration naming a (votive) dedicator followed by what we presume to be the abbreviation of an Etruscoid formula for dedicatory motivation, so dedicator + (recipient) + dedication motivation formula. This presumed mixture of Celtic + a non-native votive formula, here Etruscoid, is echoed by Latin VSLM appended to the Gaulish votive from Sazeirat (Arrenes), on which see Lambert (1994:95). Compare the inferred dedication motivation-positioning of ASUA with the positioning of Etr. alpan turce 'gave the gift' in TLE 64 and the positioning of frequently abbreviated Latin motivation formulas such as v(otum) siplvit) liibens) m(erito) or d(onum) d(edit) liibens); e.g. C. Aemilius / Felix / Aecur(nae) / v(otum) siplvit) liibens) mierito), CIL 3.3831 (Emona) = dedicator + recipient + dedication motivation formula. Etr. alpan is, moreover, probably best glossed 'offering, gift' and not libens or libenter as traditionally proposed: alpan turce equates with donum dedit, not libens dedit. A plausible resolution of asua (= asva ?) might well be: ailpan) siacni-) viacl) ia)ra 'perform(s) the sacred offering in praise (of X)'; see Facchetti (2002:34, 61-63). The -ie outcome of *-ios in an underlying *Klewios (IEIV 605-607) 'the renowned one' (uel sim.) is a Rheticization: cf. Etr. -ie for Lat. -ius i-ios) beside Gaul, -ios : Olr.-e < *-e(s) < *-yos vs. Ven., Ose. and South Picene -is (-es), Goth, -jis (after a short stem syllable as in harjis) and, finally, Lith. -ysl-is. Cf. unrelated Etr. cleva 'offering', conceivably as a collective in -va and thus possibly related to the -va- in the non-human (inanimate) plural marker -ix!c)va-\ see our discussion of estua- / estva- / etsua- below. The patronym (or gentilicium) Valtikino has traditionally been paired with Ven. volti/enei (Es 1) = Voltigenei, vo.l.ti/Jn]o.s. (Es 8) = Voltignos and similar Venetic names with Volti-, see the history of early research in Pellegrini and Prosdocimi (1967:2.205-206). Untermann (1961:1.17, 89-90, 92-93, 170) has devoted the fullest discussion to date to this matter, but then waffled in conclusion by asserting that, if the names are related at all, then it is only their initial components, volti- resp. valti-, either by remote genetic relationship or by borrowing from Venetic. He presumed (1961:1.92) that the suffixes -genei / -gnos and -kinu were but coincidental resemblances; cf. Lejeune (1974:42, 44, 47-48, 52-53), who derives Ven. volti- 'of desired (birth)' from IE *wl-ti- : Goth wulpus, but fails to address Magre's Valtikinu. Finally, we note that, in his recent research history of the Valtikinu problem, Schumacher (1992:66) chides Untermann's conclusions, but then offers nothing memorable in return. For a dress rehearsal of Untermann (1961) with particular emphasis on the onomastic stock of the Rhaetic sector, mainly as reflected in post-Romanization Latin, see Untermann (1959-1961). As pointed out above, if Rhaetic had borrowed volt- from Venetic, then we would expect *vult-, not valt-, at Magre; cf. Lat. Clovatius and Ose. Kluvatiis as examples of formations recorded in systems with and without omicron respectively. Conversely, if Venetic volt- were ultimately Etruscoid, then one might well expect the underlying form to have been *velt-\ cf., for example, Etr. VelOurna vs. (pELZURiE (= Etr. *Vel6urie) at Lothen : Lat. Volturnus; see fn. 7. Having identified Rhaetic as genetically Etruscoid, one would then expect *velt-, not valt, at Magre; We thus infer that valt- is neither Etruscan nor Etruscoid and neither Venetic nor Italic, and, by a further process of elimination, having also discounted Illyrian, Ligurian, or some other nebulous non-Indo-European (Mediterranean) speech community as entirely specious possible sources, we necessarily infer that valt- is ultimately Celtic: valti- < valeti-, cf. Olr. failte 'gladness', Caesar's Valetiacus (BG.V 11.32) : Lat. Laetus. If, however, Valtikino = Valtikina (Etr. *Valtici-na) were from Italic, then we clearly lack an underlying Italic *Valtikio~; cf. Etr. Melacina < Ital. *Melakio- as conjectured by Prosdocimi (1986:88-89). On the syncopation (valeti- > valti-) assumed here, cf. Breton Verlamio < *Verolamion and see De Bernardo Stempel (1994). We conclude that Valtikinu represents Rheticization of a Celtic *Valetikenos, a patronym in *ken- (-keno---kno-) 'son of the prosperous, propitious one' (uel s im.); see Lejeune (1974:Art. 38), Schmidt (1957:100, 216-218) and note derivationally similar Celtic Enicenius (CIL 5.2620) from the Roman period at Este. (3). PID 236 = MA-16: (right-reading) VAL.TEtpNU = Valtebnu < *Valetibenou 'of the two Valeti Women (suppliants)' < Pre-Celtic *wH2eleti-gwn-ou 'of the two joyous, pleasing, propitious women' (uel sim.) gen. dual : *wH2eleti-gwen-eH2 nom. sg. (IEW473-474, 1111-1112); cf. Mandubenos and Olr. gen. dual ban < *gwHin-ou and see Thurneysen's (1946:Art. 287) prophetic hypothesis on a possible fem. gen. dual in -ou and Schmidt (1957:147-148) on names with benos / bena, which are never fronted as initial components. Cf. *Valetibena and Etr. Alpan : Lat. Laetitia, and consider possible reference to members of a college of sibyls. This is a "possession inscription" comprised of the genitive of a syncopated compound feminine idionym with lowering of its composition vowel: -i- > -e- before b as in Prestino's ARiuoNEPos = Ariwonebos < *Ariwonibhos, on which see Markey and Mees (2004). Syncopation of a gen. dual *-benou (< *gwHn-ou) > -bnu = - pnake, but here we presumably have elimination of a weakly articulated vocoid before a nasal. Except for its lowered composition vowel, the first member of our compound here is identical with the Valti- in PID 237 = MA-17. (4). PID 239 = MA-19: (left-reading scriptio continuä) LASXEtpuxixiNU = Laste Boudigino for * Lastos Boudiginos] once again a binomial (Celtic, Venetoid) idionym + patronym configuration naming a (votive) dedicator. Boudiginos contains boudi-(Olr. buaid, W budd) 'victory', as in British Boudicca, while -ginu corresponds to -kinu in PID 237 = MA-17 (Valtikinu); see Schmidt (1957:100, 152, 154). Cf. фuxiiako• s• (Ca 17) = Boudiakos at Venetic Lägole within the context of a well documented Celticity; e.g. Lägole's -mmn- > -nmn- in Venetic; cf. Lejeune (1971:420-412, n. 220; 1974:128-129, 138), Schumacher (1998:95-96). This is possibly a local hypocoristic form of Ven. Reitia, but, more likely, an interlingual (Rhaetic / Celtic) abstraction for epiclesis. We note Celtic abstracts with -mn- such as Olr. flaithemna < *vlati-mn-ad-tu-s 'lordship' and the epiclesis implied by Rhaetic theonyms: Magre's esium.ninušu.r (PID 234 = MA-14) with Esiumni '(for) Esia', Steinberg's esimnesi (ST-3) = Esi(a)mnesi < *Esiamnai-si 'for or on behalf of Esia' and Sanzeno's esiunne = Esiunne < *Esiamnai\ see Morandi (1999:48, 63-64, 81). On Etruscoid alternation of a and u, particularly before velar nasals; cf., for example, Velianas [Pyrgi A] ~ Veliiunus [Pyrgi B], Priumne [ES V 118] : Priamus, Jlpia/ioc, and see Morandi (1991:119-120). Pace Helmut Rix's myopic insistence (p.c. Sept. 21st, 2001, at the Gubbio convegno) that the Sanzeno-Casalini votives fail to attest theonyms. Our suggestion, then, is that Rhetic regularly suffixed -mne(i) < -mna-i to names of (ethnonymic) goddesses, pre-eminently Reitia at Magre, but also Esia at Magre, Steinberg and Sanzeno-Casalini, as deities venerated by state cults. With respect to this suggestion, cf. the analogous deployment of -äli- in Latin: Saturnus (theonym) vs. Saturnalia (veneration). Note, further, Central Etruscan VelOa (Veltune) : Lat. Voltumna :: VelOurna : Lat. Volturnälia; see Varro (L.L. 6.21, 7.45) on Volturnälia and Bader (1991) on -äli-, a provocative evaluation, whether one ultimately subscribes to her conclusions or not. We note, moreover, what appear to be reflexes of Esia- in Valtellina and Valcamonica: Montagna (Sondrio) PID 252 esiavau, Tresivio (Sondrio) PID 253 esiau (for Esiol) and Piancogno (Valcamonica) Rock 25 h 37 esi(ul)msz. On the other hand, Valcamonica presents items that appear comparable to es tu- at Magre and etsu- at Padua: Camunic (Berzo Demo 3 c) eztiau and the recently (2001) discovered Camunic espiau or estiau (Campanine di Cimbergo); see Tibiletti Bruno (1990:38-40, 92-93), Gavaldo and Solano (2001), Morandi (1998:109, 124) and our discussion below of Magre ast-forms. Interestingly enough, Esia is otherwise known only from the South. She was apparently an (archaic) Etruscan Ariadne / Persephone / Core and, as such, quite possibly the counterpart of an Italic (Oscan, Umbrian, Volscian, Marsian) Ves una.9 An Etrurian Esia is, in fact, identified but twice, on two Praenestine mirrors that are reverses of each other; Brussels (Inv. No. R 1260 [818]) and Bologna (Inv. No. It. 746). For accounts of these important mirrors, see Colonna (1975:216), Lambrechts (1978: No. 10, pp. 67-73), Sassatelli (1981:No. 10, pp. 27-30, Figs. lOa-f) and Van der Meer (1995:51-54).10 Dating the mirrors is complicated, but is discussed by all four authors. On both mirrors, Esia is depicted as a shade brought by Artemis to Fufluns (functionally Bacchus / Dionysus / Liber) in the company of Minerva. The legends on the mirrors, inscribed by what are clearly different hands at widely different times, read as follows (letters within square brackets are missing; those in italics barely legible): 9 The Esia-Vesuna constellation is the primary focus of an ongoing study being conducted together with Fausto Orioli, a portion of which was presented by the author on September 21st, 2001, at the Gubbio conference: "L'umbro e le altre lingue dell'Italia mediana antica", hosted by IRDAU (Istituto di recerche e documen-tazione sugli antichi umbri). Vesuna is apparently attested but once in an Etruscan context, namely on the so-called "Vesuna Mirror" (Walters Art Gallery, Baltimore, Maryland, Inv. No. WAG 54.85). The mirror portrays the goddess in an Ariadne-like setting, thereby permitting an equation: Ariadne (Etr. Агеава, Ara a) = Esia = Vesuna; see the discussion and documentation by Zavaroni (1996:331-332). While our study discusses a possible Rhaetic Vesuna at Sanzeno-Casalini, it also presents evidence to show that the Vesuna Mirror, which mysteriously vanished to scholarship in 1884 only to resurface on a somewhat clandestine private art market some forty-two years later in 1926, is implicated as yet another Helbig hoax. The mirror's integrity as a probative integer of Vesuna veneration appears to have been compromised as part of a rather elaborate 19t'1 century gallery scam. Details are by now elusive, but the scam seems to have involved the avaricious Count Biagio Bucciosanti of Orvieto (d. 1889) as a rather reprobate property owner of the discovery site: one of eight Etruscan tombs said to have been located about three meters from the ancient Via Cassia at a place known locally as Fattoraccio ('the evil bailiff') in Comune di Castel Giorgio, sometime between October 14^ and 20'k, 1877. The archaeologist who made the discovery was Riccardo Mancini (1843-1915), who emerges as a rather venal Helbig sycophant, while the overly ambitious Helbig probably orchestrated the scam. Alessandro Castellani (1824-1883) seems to have been the prospective fencer, possibly aided by the notorious Francesco "Checco" Martinetti (1833-1895) as the forger, and these two surely played major roles in the subsequent "Manios hoax" (fabricated sometime between 1880 and 1887); see the tantalizing hints of foul play in the confiscated (upon Helbig's dismissal) Mancini - Helbig dossier (Archiv Nr. M 571) at the Deutsches Archäologisches Institut (DAI) in Rome. Torquato Castellani, Alessandro's son, is known to have been active ca. 1877-1879 as an unusually skillful ceramic ware (maiolica) copyist in Orvieto; see Savage (1963:135). 10 We are grateful to Richard Daniel De Puma (p.c.) for these initial references. Dating the mirrors is difficult, but an Etruscan £s/'a-mythology was probably in place well before production of the Sanzeno-Casalini votives (ca. 420-380 BC). Recently (September 5th, 2001), Roger Lambrechts kindly informed us (p.c.) as follows: "Le miroir Esia de Bologne est certainement le plus ancien. Sassatelli 1'a date du deuxieme quart du Ve siecle. II a certainement raison. Celui de Bruxelles est mieux conserve et plus beau. Dans Miroirs etrusques etprenestins des Musees Royaux, je l'ai date de la seconde moitie du Ve siecle. C'est certainement Minerva Fufluns Artemis Esia Bologna (ca. 475-450 BC): menarva fu[flu]nuv artame[s] esia Brussels (ca. 330-300 BC): menarea fuflunus artc[me]ns esia At Magre, then, we seem to glimpse a triad of Rhaetic lararian guardians of communal munificence and fertility: Estia - Esia (Ariadne / Vesuna) - Reitia (Artemis I Athena), perhaps in this order, cf. Pausanias 5.14.4. One recalls the mid-5th century (Eleusinian?) choral prayer in The Suppliant Maidens (li. 674-677): "And that other guardians be always renewed (тгкхеаваг), we pray; and that Artemis-Hecate watch over the child-bed (Aoyovg) of their women." (Loeb Library, Smyth translation). The ordering, ziktco 'to bring forth (as a mother)' followed by Xo/axn 'to give birth to', is surely intentional, probably formulaic and quite possibly an expression of wish and fulfillment excerpted from a cult song. (6). PID 234 = MA-14 (1 line: right-reading scriptio continua) esiu.mninušu.r = Esiumni Inušur for *Esiumnei Inosur[ix] 'Inosurix for Esia (at the Esialia)' with *Esiumnei ultimately from *Esiamnai as discussed above. Pace Morandi's (1999:81) reading Esiu.mni Nušur and contention that Nušur is an otherwise unknown idionym. The final iota of the first term (Esiu.mni) is "re-used" as the initial iota of the second term (Inusu.r). Note the syllabic punctuation. The major epigraphic advantages of syllabic punctuation, made scrupulously regular to the point of becoming otiose in Venetic, are as follows. It aids "word / morph" identification in scriptio continua by: (1) revealing syllable boundaries and (2) blocking "the re-use option". The re-use option was exercised in scriptio continua whenever the final vowel or consonant of a foregoing word was "re-used" as the initial vowel or consonant of a following word. This practice was wide-spread, and it is also found in (early) Runic inscriptions. For example, the heroic Möjbro Stone (ca. 480-500 AD, Hagby Parish, just southwest of Uppsala in the heart of the ancient realm of the Svear), presents islaginar for is#slaginar 'was slain'. In the Inosurix retrieved here, we posit ino < eno; cf. Cembra's (PID 215 = CE-1) vino- < *veno- in vinutalina as discussed above. We thus have Inosurix from Enosurix 'the grand (good) king of the Inn'; see Schmidt (1957:91, 206, 226) on eno-and note that the majority of eno- / шо-names are concentrated in neighboring Noricum. Even from this rather cursory presentation we can firmly posit an East Celtic presence at Rhaetic Magre and point to Celtic as an essential component of the diverse mixture that has enriched Slovenia's cultural heritage. une erreur de ma part. Je me suis base uniquement sur des criteres stylistiques, qui effectivement nous repor-tent au Ve siecle. Mais j'aurais du aceorder plus d'attention a la typologie (prenestine) de l'objet, qui n'appa-rait qu'ä partir du dernier tiers du IVe siecle. Toutefois ce desaccord entre le style de la gravure et le type du miroir cree un gros probleme. La seule explication (provisoire) de ce mystere semble etre Celle proposee par Sassatelli: un artiste du IVe s. aurait repris une composition plus vieille d'un siecle. Beaucoup d'auteurs elimi -nent le probleme en declarant (toujours sans l'avoir vu!) que le miroir de Bruxelles est un faux. C'est une solution de facilite. J'ai eu ce miroir assez d'heures, et de jours, sous les yeux et dans les mains pour affirmer avec force qu'il est authentique. Je maintiens aujourd'hui encore cette position, meme si l'explication de Sassatelli n'est pas entierement convaincante." References Adiego Lajara, Ignacio-Javier. 2001. Observaciones sobre el plural etrusco. Unpublished paper delivered at the 1° Convegno Internazionale sugli Antichi Umbri, L'umbro e le altre lingue dell Italia mediana antica. Gubbio, September 21st, 2001. Agostiniani, Luciano. 1982. Le "iscrizioniparlanti" dell'Italia antica. (Lingue e iscrizioni dell'Italia antica, 3). Firenze: Leo S. Olschki Editore. _. 2000. La lingua. In: gli Etruschi, ed. Mario Torelli. Pp. 485-499. Milan: Bompiani. Alfieri, Nereo and Paolo Enrico Arias. 1958. Spina. Die neuentdeckte Etruskerstadt und die griechischen Vasen ihrer Gräber. Aufnahmen von Max Hirmer. Munich: Hirmer, 1958. Bader, Fran^oise. 1991. Comparaison typologiques de l'etrusque et les langues indo-europeennes: structures agglutinantes. Studi Etruschi 56.155-173. Bermond Montanari, Giovanna. 1969. La necropoli protostorica di S. Martino in Gattara (Ravenna). Studi Etruschi 37.213-228. Breyer, Gertraud. 1993. Etruskisches Sprachgut in Lateinischen unter Ausschluß des spezifisch onomastischen Bereiches. (Orientalia Lovaniensia Analecta, 53). Louvain: Editions Peeters. CIE = Corpus inscriptionum Etruscarum. Carl Pauli et al./Academia litterarum regiae Borussica et Societatis litterarum regiae Saxonica/Istituto per l'archeologia etrusco-italica del Consiglio nazionale delle ricerche, eds. 1893-[1996], Vol. 1-3. Leipzig: A. Barth/A. Meiner; and Rome: Istituto per l'archeologia etrusco-italica del Consiglio nazionale delle ricerche. CIL = Corpus inscriptionum Latinarum. Theodore Mommsen et al. / Academia litterarum regiae Borussica (and successor bodies), eds. 1863-. Vol. 1-17. Berlin: Georg Reimer / Walter de Gruyter & Co. Colbert de Beaulieu, Jean-Baptise. 1966. La monnaie de Caletedu et les zones du statere et du denier en Gaule. Revue archeologique du Centre 18.101-129. Colbert de Beaulieu, Jean-Baptiste and Brigitte Fischer. 1998. Recueil des inscriptions gauloises (R.I.G.). Volume IV: Les legendes monetaires. (XLVe supplement ä "Gallia"). Paris: CNRS Editions. Colonna, Giovanni. 1974. Ricerche sugli Etruschi e sugli Umbri a nord degli Appennini. Studi Etruschi 42.3-24. _____. 1975. Praeneste. 19. Studi Etruschi 43.216-218. Cortsen, Saren Peter. 1935. Literaturbericht 1928-1934: Etruskisch. Glotta 23.145-187. De Bernardo Stempel, Patrizia. 1990. Einige Beobachtungen zu indogermanischem /w/ im Keltischen. In: Celtic Language, Celtic Culture: A Festschrift for Eric P. Hamp, ed. A. T. E. Matonis and Daniel F. Melia. Pp. 26-46. Van Nuys, CA: Ford & Bailie, Publishers. _. 1994. Zum gallischen Akzent: eine sprachinteme Betrachtung. Zeitschrift für celtische Philologie 46.14-35. De Marinis, Raffaele Carlo. 1977. The La Tene Culture of the Cisalpine Gauls. (Keltske študije) Posavski muzej Brežice 4.23-50. _. 1986. L'etä gallica in Lombardia (IV-I secolo a. C.): Risultati delle ultime ricerche e problemi aperti. In: Atti. 20 Convegno archaeologico regionale. La Lombardia tra protostoria e romanitä. Como villa Olmo. 13.14.15. aprile 1984. (Societä Archeologica Comense). Pp. 93-152. Como: New Press. _. 1988. Le popolazioni alpine di Stirpe retica. In: Italia omnium terrarum alumna. La civiltä dei Veneti, Reti, Liguri, Celti, Pičeni, Umbri, Latini, Campani e Iapigi, ed. G. Pugliese Carratelli. Pp. 99-155. Milano: Libri Scheiwiller. 2nd ed. 1990. _ . 1990-1991. Una nuova iscrizione lepontica su pietra da Mezzovico (Lugano). Aspetti epigrafici e inquadramento cronologico. Sibrium 21.201-218. De Simone, Carlo. 1992. Le iscrizione etrusche dei Cippi di Rubiera. (Archeologica Regiensia, 6). Reggio Emilia: Comune di Reggio Emilia, Assessorato alle istituzioni culturali, Civici musei. DIL = Dictionary of the Irish Language. Dublin: Royal Irish Academy, 1983. Dubois, Laurent. 1995. Inscriptions grecques dialectales de Grande Grece. I. Colonies eubeennes. Colonies ioniennes. Emporia. (Ecole pratique des hautes Etudes, IVe Section, Sciences Historiques et Philologiques, III. Hautes etudes du monde greco-romain, 21). Geneva: Librairie Droz S.A. ES = Etruskische Spiegel, see Körte, Körte, Gustav and (Johann) Adolf Klügmann. 1834-1897. Evans, D. Ellis. 1967. Gaulish Personal Names. A Study of Some Continental Celtic Formations. Oxford: Clarendon Press. Facchetti, Giulio M. 2002. Elementi di morfologia etrusca. Milano: Arcipelago Edizioni. Falkner, Margit. 1948. Die norischen Personennamen auf -u und ihre kulturgeschichtliche Bedeutung. In: Frühgeschichte und Sprachwissenschaft. (Arbeiten aus dem Institut für allgemeine und vergleichende Sprachwissenschaft herausgegeben von W. Brandenstein, Graz, 1. Heft). Vienna: Gerold & Co. Gavaldo, Silvana and Serena Solano. 2001. Le nuove inscrizioni in caratteri nordetruschi in Valcamonica. Paper presented at: Etrusche nelle Alpi (Capodiponte, November 10-11, 2001). Capodiponte: Centro Camuno di Studi Preistorici. (In press). Gordon, Arthur E(rnest). 1975. The Inscribed Fibula Praenestina. Problems of Authenticity. (University of California Publications: Classical Studies, Vol. 16). Berkeley and Los Angeles: University of California Press. Guarducci, Margherita. 1995. Epigrafia greca. I. Caratteri e storia della disciplina. La scrittura greca dalle origini all'etä imperiale. Roma: Istituto poligrafico e zecca dello stato. 2n^ printing with corrections. Ist ed. = 1967. Hamp, Eric P. 1976. On the Celtic Names of Ig. Acta Neophilologica 9.3-8. _. 1978. Further Remarks of the Celtic Names of Ig. Acta Neophilologica 11.57-63. _. 1981. Is the Fibula a Fake? American Journal of Philology 102.151-153. _. 1993. Venetic ke 'and'. Glotta 71.1-2.65. _. 1995.perendi andperenddn. Linguistica Baltica (Kurylowicz Memorial Volume, Part Two) 4.277-278. Helbig, Wolfgang. 1880. Due specchi etruschi. Annali dell 'instituta di corrispondenza archeologica (= Jahrbuch des Deutschen Archäologischen Instituts). 52.256-264. Roma: Coi tipi del Salviucci. Hoenigswald, Henry M. 1960. Language Change and Linguistic Reconstruction. Chicago: The University of Chicago Press. IEW = Pokomy, Julius. 1959. Indogermanisches etymologisches Wörterbuch. 1. Band. Bern-Munich: Francke Verlag. Jackson, Kenneth Huristone. 1953. Language and History in Early Britain. A Chronological Survey of the Brittonic Languages 1st to 12th c. A.D. Edinburgh: Edinburgh University Press. Kaenel, Gilbert. 2000. Les relations transalpines ä l'Age de Fer: territoire "lepontien" - Suisse occidentale - Jura. In: ILeponti tra mito e realtä. Raccolta di saggi in occasione della mostra, Locarno 20 maggio - 3 dicembre 2000, ed. Raffaele Carlo de Marinis and Simonetta Biaggio Simona. Vol. 2.151-158. Locarno: Gruppo Archeologico Ticino / Armando Dado editore (distribuzione). Körte, Gustav and (Johann) Adolf Klügmann. 1834-1897. Etruskische Spiegel. Founded by Eduard Gerhard. Vols. 1-5 (I-V). Berlin: Georg Reimer. Kos, Marjeta Šašel. 1999. Pre-Roman Divinities of the Eastern Alps and Adriatic. (Situla, 38). Ljubljana: Narodni muzej Slovenije. Kretschmer, Paul. 1932. Die Etruskerfrage und die Inschriften von Magre. In: Symbolae philologicae O. A. Danielsson Octogenario dicatae, ed. Axel Nelson. Pp. 134-142. Uppsala (Upsaliae): A.-B. Lundequistska bokhandeln. _ . 1949. Wem waren die raetischen Täfelchen von Magre geweiht? Die Sprache (Zeitschrift ßr Sprachwissenschaft) 1.30-36. Lambert, Pierre-Yves. 1994. La langue gauloise: description linguistique, commentaire d'inscriptions choisies. (Collection des Hesperides). Paris: Editions Errance. Rpt. 1997. Lambrechts, Roger. 1978. Les miroirs etrusques et prenestins des Musees Royaux d'Art et d'Histoire ä Bruxelles. Brussels: Musees Royaux d'Art et d'Histoire. Laroche, Emmanuel. 1966. Les noms des hittites. (Etudes linguistiques, IV). Paris: Librairie C. Klincksieck. Lejeune, Michel. 1971. Lepontica. (Monographies linguistiques, 1). Paris: Societe d'edition "les Belles Lettres" [= Documents gaulois et para-gaulois de Cisalpine. Etudes celtiques 12.2.337-500], _ . 1974. Manuel de la langue venete. (Indogermanische Bibliothek, I. Reihe: Lehr- und Handbücher). Heidelberg: Carl Winter Universitätsverlag. Letizia Lazzarini, Maria and Paolo Poccetti. 2001. L 'iscrizione paleoitalica da Tortora. (II mondo enotrio tra VI e V secolo a.C., Atti dei seminari napoletani (1996-1998)). Napoli: Loffredo Editore. Mancini, Alberto. 1973. Retico. Revista di epigrafia italica. Studi Etruschi 41.364-371. _. 1975. Iscrizioni retiche. Revista di epigrafia italica. Studi Etruschi 43.249-306. Marinetti, Anna. 1985. Le iscrizioni sudpicene. I. Testi. (Lingue e iscrizioni dell'Italia antica, 5). Firenze: Leo S. Olschki Editore. _. 1994. La documentazione epigrafica. In: Balone. Insediamento etrusco presso un ramo del Po. Rp. 8187. Rovigo: Comune di Rovigo. Markey, T. L. 2000 (1997). Rheto-Celtic pnake-vitamulate = *ben(n)acos windamolatos. General Linguistics 37.37-40. _. 2001a. A Tale of Two Helmets: The Negau A and B Inscriptions. The Journal of Indo-European Studies 29.69-172. _. 2001b. 'Ingvaeonic' *ster(i)r- 'star' and Astral Priests. NOWELE (North-Western European Language Evolution) 39.85-113. MARkey, T. L., and Bernard Mees. 2003. Prestino, Patrimony and the Plinys. Zeitschrift für celtische Philologie 53.116-167 _. 2004. A Celtic Orphan from Castaneda. Zeitschrift fur celtische Philologie 54.54-120. Marstrander, Carl J. S. 1928. Om runene og runenavenes oprindelse. Norsk Tidsskriftfor Sprogvidenskap 1.85188. Mastrocinque, Attilio. 1987. Santuari e divinitä dei Paleoveneti. Padova: La Linea Editrice. McLaughlin, John C. 1963. Graphemic-Phonemic Study of A Middle English Manuscript. The Hague: Mouton & Co. Minsky, Marvin. 1980. A Framework for Representing Knowledge. In: Frame Conceptions and Text Understanding. (Research in Text Theory / Untersuchung zur Texttheorie), ed. Dieter Metzing. Pp. Berlin -New York: Walter de Gruyter. Morandi, Alessandro. 1991. Nuovi lineamenti di lingua etrusca. Roma: Erre emme edizioni. _. 1998a. Epigrafia camuna. Osservazioni su alcuni aspetti della documentazione. Revue beige de philologie etd'histoire 76. 99-124. _ . 1999. II cippo di Castelcies nell'epigrafia retica. (Studia archaeologica, 103). Roma: "L'Erma" di Bretschneider. Nagy, Gregory. 1974. Perkunas and Perimu. In: Antiquitates Indogermanica. Gedenkschrift fur Hermann Güntert zur 25. Wiederkehr seines Todestages am 23. April 1973, ed. Manfred Mayrhofer, Wolfgang Meid, Bernfried Schlerath, Rüdiger Schmitt. Pp. 113-131. Innsbruck: Innsbrucker Beiträge zur Spachwissenschaft. _. 1990. Greek Mythology and Poetics. (Myth and Poetics). Ithaca and London: Cornell University Press. Pallottino, Massimo. 1963. Etruscologia. Milano: Editore Ulrico Hoepli. 5^ edition. Pandolhni, Maristella and Aldo Luigi Prosdocimi. 1990. Alfabetari e insegnamento della scrittura in Etruria e nell'Italia antica. (Istituto nazionale di studi etruschi ed italici, biblioteca di "Studi Etruschi," 20.). Firenze: Leo S. Olschki Editore. Pascal, Cecil Bennett. 1964. The Cults of Cisalpine Gaul. (Collection Latomus, Vol. LXXV). Brussels: Latomus, Revue d'etudes latines. Pauli, Carl Eugen. 1885. Die Inschriften nordetruskischen Alphabets. Mit sieben lithographischen Tafeln. (Altitalische Forschungen, 1. Bd.). Leipzig: Johann Ambrosius Barth. Pellegrini, Giovan Battista. 1951. Osservazioni sulle nuove iscrizioni nordetrusche di Sanzeno. Archivio per l 'Alto Adige 45.303-329. _. 1985. Reti e Retico. In: L'etrusco e le lingue dell'Italia antica. Atti del Convegno della Societä Italiana di Glottologia. Pisa, 8 e 9 dicembre 1984, ed. Adriana Quattordio Moreschini. Pp. 95-128. Pisa: Giradini Editori e Stampatori. Rpt. as Chapter 7 (= Reti e Retico) in: Dal venetico al veneto. Studi linguistici preromani e romanzi. (Filologia veneta, Testi e studi 2). Pp. 101-132. Padova: Editoriale Programma, 1991. Pellegrini, Giovan Battista, and Aldo Luigi Prosdocimi. 1967. La lingua venetica. 1. Le iscrizioni. 2. Studi. Padova: Istituto di glottologia dell'Universitä di Padova. Pellegrini, Giuseppe, ed. 1914. Guida descrittiva del R. Museo Archeologico di Venezia. Sezione classica. Con speciale riguardo ai marmi scolpiti. I-П. Padova: Fratelli Drucker. _. 1918. Magre (Vicenza). Tracce di un abitato e di un santuario, corna di cervo iscritte ed altre reliquie di una stipe votiva preromana, scoperta sul colle del Castello. Notizie degli Scavi di Antichitä 16.169-207. PID = The Prae-Italic Dialects of Italy. Conway, Robert Seymour, Joshua Whatmough and Sarah Elizabeth (Jackson) Johnson. 1933. The Prae Italic Dialects of Italy. Vols. 1-3. London: Milford, and Cambridge, MA: Harvard University Press. Rpt. Hildesheim: Olms, 1968. See Whatmough (1933). PID numbers in the text refer to inscriptions, not pages. The Rhaetic Corpus = Vol. II, Part III. Pp. 7-62, Nos. 188-253. Prosdocimi, Aldo Luigi. 1986.1 piü antichi documenti del celtico in Italia. In: Atti. 2° Convegno archaeologico regionale. La Lombardia tra protostoria e romanitä. Como villa Olmo. 13.14.15. aprile 1984. (Societä Archeologica Comense). Pp. 67-92. Como: New Press. REE = Rivista di epigrafia etrusca, self-contained supplement of Studi etruschi. Riemer, Heiko. 1998. Eisenzeitliche Aschenaltäre aus dem Reitia-Heiltigtum von Este-Baratella / Italien. Ein Vorbericht. Archäologisches Korrespondenzblatt 28.3.423-430. RIG (4) = Receueil des inscriptions gauloises, Vol. 4. See Colbert de Beaulieu and Fischer (1998). Rix, Helmut. 1985. Descrizioni di rituali in etrusco e in italico. In: L 'etrusco e le lingue dell'Italia antica. Atti del Convegno della Societä Italiana di Glottologia. Pisa, 8 e 9 dicembre 1984, ed. Adriana Quattordio Moreschini. Pp. 21-37. Pisa: Giradini Editori e Stampatori. Rouse, William Henry Denham. 1902. Greek Votive Offerings. An Essay in the History of Greek Religion. Cambridge: Cambridge University Press. Rpt. Chicago: Ares Publishers, Inc. 1998. Sasel, Jaroslav. 1953. Emona. Rpt. in Opera Selecta. (Situla 30). Ljubljana: Narodni muzej, 1992. P. 559. Sassatelli, Giuseppe. 1981. Corpus Speculorum Etruscorum: Italia 1. (Bologna, Museo Civico). Roma: "L'Erma" di Bretschneider. Savage, George. 1963. Forgeries, Fakes and Reproductions. A Handbook for the Collector. London: Barrie and Rockliff. Schmidt, Karl Horst. 1957. Die Komposition in gallischen Personennamen. Zeitschrift für celtische Philologie 26.33-301. Also separately: Tübingen: Max Niemeyer, 1957. Schumacher, Stefan. 1992. Die rätischen Inschriften. Geschichte und heutiger Stand der Forschung. (Innsbrucker Beiträge zur Kulturwissenschaft. Sonderheft 79). Innsbruck: Verlag des Instituts für Sprachwissenschaft der Universität Innsbruck. _. 1998. Sprachliche Gemeinsamkeiten zwischen Rätisch und Etruskisch. Der Schiern 72.2.90-114. Solmas, Patrizia. 1992-1994. Sulla celticitä linguistica nell'Italia antica: il leponzio. Da Biondelli e Mommsen ai nostri giorni. 1-2. Atti dell'Istituto veneto di scienze, lettere ed arti. Tomo CLI (1992-1993).1237-1335 and Tomo CLII (1993-1994).873-935. _. 1995 (1994). II celtico in Italia. Revista di epigrafia italica. Sezione monografica. Studi Etruschi (Serie III). 60.311-408. __. 1998. Le iscrizioni in alfabeto leponzio dalla necropoli di Casalandri (Isola Rizza -VR-). In: La necropoli gallica di Casalandri a Isola Rizza (Verona), ed. Luciano Salzani. Pp. 143-148. Mantova: Societä Archeologica Padana / Editrice S.A.P. Thurneysen, Rudolf. 1933. Italisches. I. Die etruskischen Raeter. Glotta 21.1-7. _. 1946. A Grammar of Old Irish. Dublin: The Dublin Institute for Advanced Studies. Rev. ed. 1966 Tibiletti Bruno, Maria Grazia. 1978. Camuno, retico e pararetico. In: Lingue e dialetti dell'Italia antica. (Popoli e civiltä dell'Italia antica, VI), ed. Aldo Luigi Prosdocimi. Pp. 211-255. Roma: Biblioteca di storia patria. _. 1990. Nuove iscrizioni camune. Quaderni Camuni 49-50.29-171. TLE = Testimonia linguae Etruscae. Selegit recognovit et indice verborum instruxit Maximus Pallottino. (Biblioteca di studi superiori. Storia antica ed epigrafia, 24). 1954. Firenze: «La Nuova Italia» Editrice. Untermann, Jürgen. 1959-1961. Namenlandschaften im alten Oberitalien. Beiträge zur Namenforschung 10.1.74108 (1959), 10.2.121-159 (1959), 11.3.273-318 (1960), 12.1.1-30 (1961). _. 1961. Die venetischen Personennamen. 1-2. Wiesbaden: Otto Harrassowitz. _. 2000. Wörterbuch des Oskisch-Umbrischen. (Handbuch der italischen Dialekte. Begründet von Emil Vetter. III. Band). Heidelberg: Carl Winter Universitätsverlag. Van der Meer, L(ammert) Bouke. 1995. Interpretatio etrusca. Greek Myths on Etruscan Mirrors. Amsterdam: J. C. Gieben. _ . 1999. Erus, Zinthrepus and Zimuthe: A Greek Myth in Etruscan Dress. In: Proceedings of the XVth International Congress of Classical Archaeology, Amsterdam, July 12-17, 1998. Classical Archaeology Towards the Third Millennium: Reflections and Perspectives (Allard Pierson Series, 12), ed. Roald F. Docter and Eric M. Moormann. Pp. 430-432. Amsterdam: Allard Pierson Museum. Vitali, Daniele. 1998.1 Celti e Spina. In: Spina e il delta padano: Riflessioni sul catalogo e sulla mostra ferrarese: Atti del convegno internazionale di studi Spina: due civiltä a confronto, Ferrara, Aula Magna dell 'Universitä, 21 gennaio 1994, ed. Fernando Rebecchi. Pp. 253-273. Roma: "L'Erma" di Bretschneider, 1998. Watkins, Calvert. 1955. The Phonemics of Gaulish: The Dialect of Narbonensis. Language 31.1.9-19. Whatmough, Joshua. 1921. Inscribed Fragments of Staghoms from North Italy. The Journal of Roman Studies 11.245-253. _. 1922. Rehtia, the Venetic Goddess of Haeling. Journal of the Royal Anthropological Institute 52.212-229. _. 1923. Inscriptions from Magre and the Raetic Dialect. The Classical Quarterly 17.61-72. _. 1933. The Prae-Italic Dialects of Italy. Vol. II. Part III. The Rhaetic, Lepontic, Gallic, East Italic, Messapic and Sicel Inscriptions. Cambridge, MA: Harvard University Press. Rpt. Hildesheim: Georg Olms Verlagsbuchhandlung, 1968. _. 1934. The Raeti and their Language. Glotta 22.27-31. Zavaroni, Adolfo. 1996.1 documenti etruschi. Reggio Emilia: Sherpa. Table 1: Concordance of Presentations of the Magre Votives PID CIRCE Inv. No. Este Museum Tibiletti Bruno (1978) Morandi (1999 )No. 221 MA-1 p. 237 38 222 MA-2 58811 p. 237 39 223 MA-4 p. 237 45 224 MA-5 58813 p. 237 36 225 MA-6 58812 p. 237 35 226 MA-7 227 MA-8 58808 p. 237 34 228 MA-9 58810 p. 237 40 229 MA-10 58809 p. 237 37 230 MA-3 58811 p. 237 39 231 MA-11 58816 p. 237 41 232 MA-12 P. 232 42 233 MA-13 58815 p. 234 43 234 MA-14 p. 237 44 235 MA-15 236 MA-16 58819 p. 237 47 237 MA-17 p. 237 46 238 MA-18 58814 p. 237 50 239 MA-19 58814 p. 237 49 240 MA-20 241 MA-21 242 MA-22 243 MA-23 58817 p. 238 48 Povzetek ZGODNJA KELTŠČINA V SLOVENIJI IN V RETSKEM KRAJU MAGRE (SCHIO) Zgodnje oblike celinske keltščine (okrog 550-50 pr. Kr.) so bile razširjene v severnem alpskem pasu italijanskega polotoka od jezer Maggiore in Como na zahodu do Slovenije na vzhodu. Fragmenti keltskih napisov s tega območja so se ohranili v različnih lokalnih abecedah, na katere je vplivala prvenstveno etruščina (retščina). V teh sicer skopih besedilih lahko spoznavamo, kako se je v železni dobi pismenost razširjala z mediteranskega področja proti severu v keltsko Evropo. Zgodnje jezikovne in kulturne stike med etruščino (retščino), keltščino in venetščino pa lahko še posebej opazujemo na osnovi votivnih napisov, povezanih s čaščenjem boginje Reitije v svetišču v kraju Magre (Schio) severno od Vicenze. Zdi se, da imamo tu pred seboj tri larijske čuvajke blaginje in plodnosti skupaj: to so Estija - Esija (Ariadna / Vesuna) - Reitija (Artemis / Ateno). V članku avtor z jezikoslovnega stališča analizira nekaj magrejskih napisov, predvsem v odnosu do keltščine, pa tudi z vidika njihovih epigrafskih značilnosti. Vezi med otoško keltščino in vzhodno keltščino, kot se odraža v italijanskem Magreju in na slovenskem Igu, so presenetljivo in obenem enigmatično močne. Tako npr. v enem izmed magrejskih votivnih napisov najdemo onomastično pojavitev keltske osnove boudi- ' zmaga', prisotno tudi v britanskem imenu Boudicca (sic, Tacitus, Agr. 16,1) za slavno kraljico Icenov, kije v letih 60 - 61 pr. Kr. vodila tragično neuspešen upor proti rimskim okupatorjem. Irena Orel University of Ljubljana UDK 22.043=163.6 : 8Г25 PREPOSITIONAL PHRASES WITH VERBA DICENDI FROM DALMATIN'S TRANSLATION OF THE BIBLE (1584) IN RELATION TO FOREIGN LANGUAGE TRANSLATIONS 0 INTRODUCTION In a diachronic perspective from the 16th century to the present, this article investigates translated interlinguistic agreement and difference in the use of the temporally marked Slovenian prepositional phrases that appeared in the semantic group of verba dicendi in the first two books of the Old Testament and the New Testament of the oldest Slovenian translation of the Bible, from 1584, and that were replaced in the modern literary language in the 19th century by the introduction of prepositionless or other prepositional patterns. A comparison is made on the basis of Internet publications of parallel sections of six foreign language translations (Latin, German, two English [17th century and modem], French and Russian), and the extent to which these prepositional phrases are covered by older or modern literary Slovenian syntactic patterns is determined. 1 PREPOSITIONAL PHRASES WITH VERBA DICENDI IN A DIACHRONIC PERSPECTIVE 1.0 For the older period of literary Slovenian, from the diachronic perspective the differential use of prepositional phrases is significant, especially in certain syntactic senses with verb phrases with free prepositional morphemes, known as prepositional verbs, in verb phrases and in certain figurative meanings that were adopted on the model of foreign language literary patterns (in this case, on the basis of Biblical pattern), and therefore a certain interlinguistic agreement, and even interlinguistic universality of equivalent prepositional phrases in the valency of obligatory and non-obligatory syntactic senses in a particular time are expected, and on the other hand deviation in the direction of differential linguistic specifics as well. 1.1.0 The sample group of prepositional phrases that was selected with regard to a developmentally disproportional distribution of use in Slovenian and great lexical frequency is limited to a comparison of right valence prepositional complements as part of the semantic field of structurally similar verba dicendi such as govoriti, odgovoriti, reči, praviti, dejati, povedati, klicati, vpiti, pridigati, etc. that from the 16th century up to the 19th century partially corresponded to the model govoriti/reči ...h komu kaj od koga/česa (skozi koga/kaj), for which the omission of the use of prepositional free morphemes and especially the substitution of non-Slovenian calqued prepositional phrases with suitable native elements (e.g., govoriti, reči... komu kaj o kom/čem (po kom/v čem, s čim) is typical in historical development. To show which prepositional structures agree with those in foreign languages, interlinguistic comparison will cover a number of such prepositional phrases from the first two books of the Old Testament and New Testament in six different languages and eight translations. These are: Dalmatin's Bible from 1584 (DAL), the Slovenian standard translation from 1996 (SSP), Luther's German Bible from 1545 (L45), the Latin Vulgate from the 4th to 5th century (VUL), the English King James translation from 1611 (KJV) and modern English (the Good News Bible, GNB), a French translation from 1910 (Louis Segond, SEG) and the Russian synodal translation from 1876 (RUS). The results were verified with the same Latin and German foreign language phrases in Hutter's six-language Biblia sacra (1599), which did not significantly differ from the samples in the two editions mentioned above. 1.1.1 All of the prepositional phrases mentioned above primarily express spatial relations. Their figurative use, which in Slovenian linguistics this is defined as calqued, semantically also derives from them and, with a link to the abstraction of the spatial component, expands its usage to verbs that express this semantic relationship at the figurative, metaphorical level, which would agree with the prototype theory of prepositional meaning and the radial interpretation of the cognitive linguistics approach (Sicherl 2000, 416-420). Prepositions - as polysemously complex relators or a subtype of junctors that prototypically connect a given base with a nominal or pronominal adjunct and distinguish three determining types or syntactic functions: attributive, predicative and applicative, in which the base is a noun, a subject and a copula as well as a verb (Weinrich 1993, 610) - show a varying degree of preservation of the original sense, which is partially lost (made meaningless), and take on the grammatical function of what are known as free morphemes, the semantic trace of which is nonetheless marked (with selected verbs, the orientation of the verbal action toward the participant in the communication as opposed to the goal with the free dative), while their exchangeability with non-prepositional verbs defines their non-obligatory nature or non-necessity in synthetic languages as opposed to analytical languages. 1.2 The sample studied is limited to the semantic group of verba dicendi, which František Danes (1999,105-106) defined as verbs of action whose agentive participant is the agent of an action based on the use of language in its spoken (auditory), graphic or figurative shape and express only a spoken activity or its content or thought (goal, function) and have the role of performatives (e.g., opravičiti se, ukazati, prisiliti, prositi, etc.) with a declarative function (termed illocutionary power by Austin) in which the use of language is only one of the possible means for attaining the same goal. A similar basic scheme of obligatory and non-obligatory participant complements is characteristic for all of them. 1.2.1 In Dalmatin the right argument role that names the person responded to with verba dicendi is the most numerous, for example: the most general unidirectional or bidirectional communicative verb govoriti, the typical unidirectional reči, less often poreči, the grammatically limited verb dejati, which, like today, appeared only with a temporal-functionally expressed limitation on the past (the type je djal 'said'), izgo -voriti, exceptionally praviti, DAL: GOSPVD Bug je djal k'Sheni (Gen. 3.13), k'Sheni je on rekäl (Gen. 3.16); SSP: GOSPOD Bog je rekel ženi; ženi je rekel). In the four books of the Bible mentioned, in Dalmatin the verb dejati is most frequently used, with nearly one thousand attestations, with a formal-temporal limitation on the reporting of past verbal actions that primarily introduce a literal quotation or report it, and with it the free morpheme k/h consistently appears. It follows the verb reči with nearly 450 attestations - which, however, in cases with a stated receiver of the message expressed with a dative participant does not always realize the morpheme к with itself (14 times)), and a content complement does not appear with it. In third place is the verb govoriti (over 200), which appears in two distinct forms of the past participle, govoril and govuril, which surprisingly appear in a two-to-one relationship at the expense of the non-Lower Carniolan phonetic variant (86 to 46), whereas the entirety of examples from all available books exhibits a diametrically opposed picture (145 to 272). Next is povedati (approximately 180), which does not require a free morpheme, e.g. DAL: de fie timu Moshu povedali (Gen. 46.31). The verb praviti is more rarely represented (approximately 70 examples), and is usually followed by a literal quotation, and otherwise the morpheme may be given or not: DAL: od kateriga je Bug njemu pravil (Gen. 22.3). Usage without the prepositional morpheme sporadically appears (e.g., Gen. 19.12, 34.11, 41.55, Mt 22.13, Mk 3. 9, etc.). In analytical languages it is expressed with a complement for the indirect participant in the action, which also implies the orientation of the verbal action (KJV: the LORD God said unto the woman; GNB: and he said to the woman; KJV: Unto the woman he said; SEG: 1'Eternel Dieux dit ä safemme; L45: da sprach GOtt der HErr zum Weibe, whereas in Russian a bare dative is used (I skazal GospodbBog žene; žene skazal), and in Latin both usages are attested (et dixit Dominus Deus ad mulierem, mulieri quoque dixit ei - dixit ad eum, dixit mihi, dixit Sarai uxori suae,etc.), which also occurs in French with a personal pronoun complement (lis lui dirent), and in German only with the verb sagen, and not sprechen (Und sie sagten ihm versus und sprach zu ihm). 1.2.2.1 The basic bivalent grammatical pattern of the verb govoriti, which alone anticipates bidirectional communication (SN (+hum.) V (govoriti) Sj (+hum.) 'govoriti s kom' 'to speak with someone'), was also dominant in the 16th century, and in the modern translation is substituted for with unidirectional verbal action (s'fvojim fynom Esavom govoril; SSP: govoril sinu Ezavu (Gen. 27.5)), whereas unidirectional verbal action of the subject is designated (directed) toward the listener or interlocutor; i.e., the type Sj^ (+hum.) V (govoriti) Sd(A-//j) (+hum) 5д (+inan.) 'govoriti komu kaj' 'to say something to someone': (od)govoriti, reči, povedati, dejati, praviti, poročati, komu kaj in the 16th century was very often expressed by a prepositional dative with the free morpheme k (e.g., govoriti h komu kaj), and one also finds complements without it, always with the verb povedati. The unidirectional verb, the perfective pair of the upper (odgovoriti), which appears rarely (16) and generally has no complement (e.g., Moses je pak timu Folku odgovoril (Ex. 20.20)), which also agrees with the government of the German verb (e.g., Mose antwortete ihm), also does not appear with a prepositional morpheme in the 16th century (the type (odgovoriti komu kaj), which corresponds to the German translation, in which the verb odgovoriti (antworten) does not have the free morpheme zu, and therefore it also does not exist in Slovenian. Russian is similar, where the response to the participant in the message is rarely stated, whereas in French and Latin the free morpheme for the indirect object is also found with this verb (e.g., et repondit ä Abraham). A prepositional morpheme is not always found with the verb reči (reči komu kaj), e.g. DAL: Jeft hozhem gori pojti inu Pharaonu povedati, inu hozhem k'njemu rezhi; SSP: Pojdem poročat faraonu in mu rečem (Gen. 46.31), where it corresponds with the German text: und Pharao ansagen und zu ihm sprechen-, GNB: said to him, tell so.-, KJV: say unto so., the Russian and Latin translation also have a bare dative (nunciabo Pharaoni dicamque ei). The verbs zapovedati and poročiti have only a bare dative (poročiti komu kaj 'to announce'), which agrees with German. 1.2.2.2 According to the theory of strong and weak syntactic variants, the use of a prepositional verb as opposed to a prepositionless verb could be defined with regard to the complexity of syntactic structures as more or less complex, which may be diachron-ically interpreted as a change from more complex structures to simpler structures. 1.23 With verbs that include semantic content of the manner (strength, intensity) of expression, such as vpiti, klicati h komu 'to cry, to call' the prepositional morpheme of directionality in Biblical texts is also preserved in all newer translations up to the most recent, and is also known in foreign language translations; e.g., DAL: vpili h'GOSPVDV (Ex. 17.4); SSP: je vpil h GOSPODU, although today's normative rules do not allow this valency possibility (only vpiti na koga/kaj, emphatic vpiti po kom/čem and prepositionless vpiti komu/kaj (e.g., orders to workers)). 1.2.4 With govoriti the prepositions proti and pred also sporadically appear in the role of orientation toward the interlocutor or location (DAL: Moses je pak pred GOSPVDOM govoril (Ex. 7.2), whereas the semantic opposition is expressed by zoper, which takes the accusative (super: DAL: govory super Synu (Mt 12.32)), which can also be preserved today (SSP: če kdo reče besedo zoper sina),pruti: DAL: je Jesus sazhel pruti folku govoriti ad Joannefa: SSP: je Jezus začel množicam govoriti o Janezu (Mt 11.7). In foreign translations there is no change in the prepositional morpheme - k remains, whereas the phrase with od for the most part corresponds with an older translation (German von, Latin de, except English about in newer (in older concerning (KJV)) and Russian (o, ob). 1.2.5.0 The object complement of content, substituted for by the preposition o (SL0+/-hum.) ('about someone/something'), the type govoriti, reči o kom/čem was expressed in the 16th century with the preposition od (govoriti od koga/česa), a free valency phrase of circumstance (a circumstant) that defines the verbal action with regard to the manner in which it is carried out or the mediator of the verbal action that is not also the agent of the action in the role of the subject was introduced in the 16th century with the preposition skozi, but in the standard Slovenian translation has been replaced by po for human mediators and by the prepositions v and zJs with nouns lexically and verbally defined for manner or communication. 1.2.5.1.1 With verbs of reporting, speaking, knowing and feeling, the typical prepositional phrase for a content object complement in the 16th century was expressed with the free morpheme od (the type govoriti od česa), which appears with the verbs brati, disputirati, govoriti, misliti, pisati, povedati, praviti, prerokovati, pridigovati, (za)slišati, (z)vedeti, vprašati, zamerkati, and zdeti ...od and with an appropriate noun that implies the meaning of the verbal communication (e.g., DAL: befsedo od krajleft-va (Mt 13.19); SSP: besedo kraljestva), which was used in the literary language until the middle of the 19th century and is still found in Wolf's edition of the Bible of 1856 (except for vprašati od, where it is replaced by vprašati za), and is still alive in the spoken language today, where the meaning of the spatial starting point or separation, or in a metaphorical sense the origin of the verbal action or deduction, has already lost the sense of a semantic connection to spatial separation. Since the end of the 19th century it has been replaced by the morpheme o (čem), because there was a desire to root out the same surface structure as there was in German, upon which the Slovenian use is apparently based. A similar change is shown in English translations from various times expressing of - about (e.g., DAL: je on od nyh govuril; SSP: da govori o njih; KJV: that he spake of them; GNB: that he was talking about them (Mt 21.45). The Russian translation always contains the morpheme o, and the others are the equivalent of Slovenian od: German von, Latin and French de, except in the noun phrase, where the right non-agreeing attribute in the modern translation is a bare genitive, as in Latin (SSP: besedo kraljestva poslušati, VUL: verbum regni). 1.2.5.1.2 With the performative verb prositi the right human participant is expressed in the 16th century with an original od (e.g., prositi kaj od koga), while today this is a grammatically accusative object complement (prositi koga kaj): DAL: je eno rezh od njega prosila; SSP: da bi ga nekaj prosila; KJV: to want something from someone: desiring a certain thing of him; GNB: to ask for a favour: asked him for a favor, French differently, VUL: petens aliquid ab eo (Mt 20.20). 1.2.5.2 Non-obligatory prepositional phrases define the mediator of the verbal action (the type govoriti skozi koga) or define the manner of the utterance (e.g., go -variti skozi kaj), which proceeds from the meaning of transition within something from one end to the other, and according to the semantic shift expresses mediation (today more suitably with po), spoken medium or means (the preposition zJs čim), manner (the preposition v čem, s čim), purpose (to speak in memory [of]): govoriti v podobah, z besedami, etc.; e.g., DAL; Satu jeft k'nym govorim fkusi priglihe (Mt 13,13), SSP: Zato jim govorim v prilikah, which is expressed with the preposition v [in] all languages except in German, which has durch, corresponding to the Slovenian preposition. Passive structures also occur with verba dicendi when the agent (left participant) is unexpressed and, along with the active, it is often replaced in translations with Latin per, German durch: DAL: kar je govorjenu fkusi Preroke (Mt 2.23); SSP: kar je bilo rečeno po prerokih). 2 CONCLUSION From an interlinguistic comparison of Biblical translations in the expression of the participant role with verba dicendi with regard to the use of bound and free (with regard to valency) prepositional verb phrases that do not express continuity in Slovenian, but were substituted for by a bare complement (the type govoriti, reči, etc. h komu, prositi od koga), a semantically similar metaphorically used preposition (skozi - po) or an apparently semantically different prepositional morpheme (od-o), it is possible determine connections to or dependencies on foreign language translations that were often limited only to a German (or broader Germanic) grammatical linguistic example as an imitation of this, or to infer a mutual linguistic type or agreement at the level of prepositional phrase use in general, which probably derives from deep-structure semantic argumentation, as, for example, in the type govoriti od koga/česa the diachronic tendency is apparent to omit semantically undetermined prepositional use with verbs that express direct speech toward an interlocutor in the speech interaction as well as certain other verbal senses (with the sememe of association, approaching), as well as a tendency for semantic simplification or other semantic motivation of certain prepositions (od - o) in certain languages (Slovenian, English of - about) and with continuity in spoken linguistic practice. The tendency toward prepositional verbal forms in the older periods was, like other syntactic parallels, part of the heritage of Western European civilization and culture and was fostered by formal linguistic variation and similar semantic motivation. BIBLIOGRAPHY Hutter, Elias. 1599. Biblia sacra. Nuremberg. : The Bible on the Internet: - Dalmatin, Jurij. 1584 (2004). Biblia. Društvo svetopisemska družba Slovenije. (DAL) - Slovenski standardni prevod (Slovenian standard translation). Second, revised edition, 1997. Copyright 1996, 2000. Društvo Svetopisemska dražba Slovenije. (SSP) - Good News Bible (Today's English Version). 1992. American Bible Society. (GNB) - The Holy Bible (King James Version). 1611. Crown copyright, UK. (KJV) - Segond, Louis. 1910. La Bible (French translation). Public domain. (SEG) - Luther, Martin. Die Heilige Schrift (German translation). 1545. Public domain. (L45) - Biblija. Russkij sinodal'nyj perevod (Russian translation). 1876. Public domain. Distribution: Rossijskoe Biblejskoe obščestvo. (RUS) - Vulgata. (Latin translation), 4th-5th century (textual critical edition, 1969). Deutsche Bibelgesellschaft. (VUL) REFERENCES Cervoni, Jean, 1991: La preposition - Etude semantique et pragmatique. Duculot, Paris - Louvain - la Neuve, Belgique. Daneš, František, 1999: Verba dicendi a vypovednf funkce. Jazyk a text I. Vyybor z lingvistickeho dfla Františka Daneše. Univerzita Karlova v Praze, Filozofickä fakulta. Praha, 105-114. Dirven, Rene (ed.), Verspoor, Marjolijn, (ed.) 20042: Cognitive Exploration of Language and Linguistics. J. Benjamins. Amsterdam, Philadelphia. Langacker, Ronald W., 1991: Concept, Image, and Symbol. The Cognitive Basis of Grammar. Mouton de Gruyter. Berlin, New York. Rizzi, Luigi, 1991 : II sintagma prepositionale. In: RENZI, Lorenzo (ed.), Grande grammatica italiana di consultatione. La frase. I sintagmi nominale e preposizionale. Vol. I. II Mulino, Bologna. Schröder, Jochen, 1986: Lexikon deutscher Präpositionen. VEB Verlag Enzyklopädie. Leipzig. Sicherl, Eva, 2000: Sodobne jezikoslovne raziskave predložnih pomenov. Vestnik, Društvo za tuje jezike in književnosti Slovenije, 34/1-2, 413-424. - — 2001: O predložni vezljivosti v angleščini in slovenščini. Vestnik, Društvo za tuje jezike in književnosti Slovenije, 35/1-2, 299-312. Vidovič Muha, Ada, 2000: Slovensko leksikalno pomenoslovje - Govorica slovarja. Filozofska fakulteta. Ljubljana. Weinrich, Harald, 1993: Textgrammatik der deutschen Sprache. Dudenverlag. Mannheim - Leipzig - Wien -Zürich. Žele, Andreja, 2001a: Vezljivost v slovenskem jeziku (s poudarkom na glagolu). ZRC SAZU. -- 2001b: Vezljivost kot pomensko- in strukturno skladenjski pojav. Slavistična revija 49/3, 137-161. Povzetek PREDLOŽNE ZVEZE OB GLAGOLIH GOVORJENJA IZ DALMATINOVEGA PREVODA BIBLIJE (1584) V RAZMERJU S TUJEJEZIČNIMI PREVODI V prispevku je v diahroni perspektivi od 16. stoletja do sodobnosti raziskana prevodna medjezikovna ujemalnost in različnost v rabi tistih časovno zaznamovanih slovenskih predložnih zvez, ki se pojavljajo v pomenski skupini glagolov govorjenja v prvih dveh starozaveznih in novozaveznih knjigah najstarejšega slovenskega prevoda Biblije iz 1. 1584 in ki so jih v 19. stoletju zamenjali v sodobnem knjižnem jeziku uveljavljeni nepredložni ali drugi predložni vzorci. S primerjavo na podlagi internetne objave vzporedno izpisanih mest v šestih tujejezičnih prevodih (latinskem, nemškem, dveh angleških (iz 17. stoletja in sodobnem), francoskem in ruskem) je bilo ugotovljeno, v kolikšni meri se te predložne zveze prekrivajo s starejšimi ali sodobnimi slovenskimi knjižnimi skladenjskimi vzorci. Iz medjezikovne primerjave svetopisemskih prevedkov pri izražanju udeleženskih vlog ob glagolih govorjenja glede rabe vezljivostno obveznih in prostih predložnih glagolskih zvez, ki v slovenskem jeziku ne izkazujejo kontinuitete, ampak so bile nadomeščene z brezpredložnim dopolnilom (tip govoriti, reči ipd. h komu, prositi od koga), blizupomenskim metaforično rabljenim predlogom (skozi -po) ali navidezno pomensko različnim predložnim morfemom (od - o) lahko ugotovimo tujejezično prevodno navezavo oz. odvisnost, ki se večkrat omejuje le na nemški (ali širše germanski) slovnični jezikovni zgled kot posnemanje le-tega, ali pa lahko predvidevamo medsebojno jezikovno tipiko oz. skladnost na ravni predložnozvezne splošnosti rabe, ki izhaja verjetno iz globinske pomenske argumentacije, kot je razvidna npr. v tipu govoriti od koga/česa jezikovnorazvojna težnja po opuščanju pomensko neodločevalne predložne rabe pri glagolih usmerjenosti govorjenja na soudeleženca v govorni interakciji in tudi nekaterih drugih glagolskih pomenih (s sememom pridruženja, približevanja), ter težnja po pomenski razbremenjenosti oz. drugačni pomenski motivaciji nekaterih predlogov (od - o) v nekaterih jezikih (slovensko, angleško of - about) in s kontinuiteto v govorjeni jezikovni praksi. Pogostejša predložnosklonska glagolska raba v starejših obdobjih je bila kot druge skladenjske vzporednosti del zahodnoevropske civilizacijske in kulturne dediščine, za katero je bila značilna formalnojezikovna različnost in podobna pomenska motivacija. Alja Ferme University of Ljubljana UDK 811.163.6'282.3 Celje : 8Г34 FINAL SONORANT SEQUENCES IN THE CELIE DIALECT1 1 Introduction In this paper I will analyse final sonorant sequences2 in the Celje variety of Slovene. In §2 various definitions of a consonant cluster will be discussed and the definition needed for further development of the article will be provided. In §3 I will present pretheoretical arguments against treating all final sonorant sequences as consonant clusters. In addition, a seemingly special behaviour of a small group of sequences will be pointed out. The government phonology framework will be introduced in §4. In §5 the data from the Celje dialect will be analysed within the given theoretical framework. 2 Consonant clusters Let us have a look at definitions of the term consonant cluster. Toporišič (1959, 203) claims that consonant clusters are sequences of consonants in words, word units and on word boundaries. They fall into three categories: initial, intervocalic and final clusters. The order of cluster members is not arbitrary; vowels are flanked by sonorants, in front of sonorants and after them appear obstruents. Toporišič observes the "increase" of sonorants when preceding a vowel, which can be interpreted as a sequence of sonorants ordered according to the sonority sequencing principle. The sonority sequencing principle requires the more sonorous segments to appear closer to the vowel (the most sonorous part of the syllable) (Kenstowicz, 1995). The relative sonority of segments is read off the sonority scale (as proposed in Srebot Rejec (1992, 228) for sonorants). Although Toporišič (1959) is not specific about the "decrease" of sonorants in positions following a vowel we can assume it exists as a logical opposition to the "increase" in initial positions. Unuk (2003, 48) uses the term consonant sequences instead of consonant clusters to underline the importance of the ordering of consonants. The ordering of members in a sequence is determined by phonotactic principles, the most prominent among them being the sonority sequencing principle, which relies on the sonority scale (1). (1) stops, fricatives, affricates < nasals < sonorants < r < vowels (taken from Unuk (2003)) 11 would like to thank dr. Erika Rržišnik for her helpful comments on the draft of this paper. The remaining errors are mine. This paper focuses merely on final sonorant sequences. However, the same analysis I propose for final sequences would also apply to sonorant sequences in preconsonantal position. In addition, Unuk distinguishes true consonant sequences (consonant clusters in our terminology, see (2)) from bogus sequences. The former belong to the same syllable while the members of the latter belong to different syllables. The distinction is guided by the sonority sequencing principle: the former obey it, the latter do not. Srebot Rejec (1992, 228) similarly observes that any arbitrary sequence of consonants does not necessarily stand for a consonant cluster, since the latter is conditioned by the sonority sequencing principle. She argues that final consonant sequences in butelj [butl]'an idiot'3 and siten [sitq j'fussy, masc.' do not represent consonant clusters but sequences of two consonants with the second consonant having the value of a syllable.4 Although the definition of a consonant cluster does not seem very unified, a common tendency can be recognized in all the given works. For the purpose of this paper let us assume the following definition of a consonant cluster: (2) Final consonant clusters are those final consonant sequences that belong to the same syllable as the vowel preceding them. It will be assumed that consonant clusters are subject to the sonority sequencing principle as in (3). (3) Sonority must decrease from the vowel to the margins of a syllable. The ranking of consonants according to their sonority is given in a sonority scale. When not referring to a particular work which uses the term cluster the term sequence will be used from now on in the most general way to avoid any implication as to whether a series of consonants in question are a part of the same syllable as the preceding vowel or not. 3 Sonorant sequences and syllabic consonants 3.1 The sonority sequencing principle in final sonorant sequences Srebot Rejec (1992) proposes the following sonority scale for sonorants (4). (4) v < m,n < 1 < r < j She argues that sonorant clusters in final positions must obey the sonority sequencing principle, which means that they are of falling sonority (as in the word gostiln 'a pub, gen.pl.'). The Celje pronunciation of the word kalen 'muddy, masc.' shows vocalic reduction in the last syllable, thus we get [kalp]. This kind of a final sonorant 3 The data are in the nominative singular, unless indicated otherwise. 4 The phenomenon of sonorants being the most sonorous parts of the syllables is also found in languages such as Slovak, Czech, English, German and others. sequence obeys the sonority sequencing principle, which suggests that it could belong to the same syllable as the vowel preceding it and could thus be a consonant cluster as defined in (2). Alongside the falling sonority final sequences the rising sonority final sequences (again as a result of vocalic reduction) can be found in the Celje dialect, as in the word kanilo [kan]]'to drip, past part., neuter'. According to (3) a final sequence of /n/ and /1/ can never be a consonant cluster. This is also confirmed by the intuition of the speakers that such a sonorant sequence is a syllable on its own.5 However, the same intuition applies for the final sequences in [kalr^] and [gostil^], which could, according to (4), represent true consonant clusters but apparently never do. Recall that Srebot Rejec (1992, 228) proposes that the second member of the final consonant sequence as in [butl] has a value of a syllable, and is thus called syllabic. Due to the apparent similarity, let us propose the same for the second member of a final sonorant sequence in the Celje dialect. The Celje dialect permits not only the given pair of sonorant sequences [nl]/[ln] but also most other sonorant sequences6 regardless of the sonority sequencing principle. In each of the cases the second member of a sequence is always syllabic. We can conclude this section by claiming that final sonorant sequences in the Celje dialect are not subject to the sonority sequencing principle. The aim of the sonority sequencing principle is to predict distribution of segments within the scope of a syllable. We have seen that in the Celje dialect a final sonorant sequence does not belong to the same syllable as the preceding vowel,7 but is a syllable on its own. Therefore any attempt to analyse the given sequences by referring to the sonority sequencing principle would be misleading. 3.2 Toporišič's categorisation of words with final sonorant sequences Toporišič (1978) observes that words with final sonorant sequences fall into four categories. The first one is represented by words that insert a schwa between the two sonorants in the nominative singular (jarem 'a yoke'). The second group consists of words in which a schwa immediately precedes the sonorant cluster, therefore the cluster is not broken by an additional schwa (trn 'a thorn', obrv 'an eyebrow'). The same is true for loan words (film 'a film'). However, Toporišič himself observes that words from this group have a tendency to syllabify the sonorant sequence by means of inserting an additional schwa (trn is realised as [taran]) or by vocalising the final sonorant (obrv is realised as [obsru]). The third group is represented by words in the genitive plural that tend to be as 5 The data concerning the number of syllables in this paper are acquired by using simple experiments on rhythm and verse. 6 Sequences with /r/ as the second member are exceptions to this. The reason for the inability of /r/ to become syllabic lies most likely in its melodic structure. (A discussion on this is beyond the scope of the present paper.) Whenever /r/ is to appear as the second member of a final consonant sequence, a vowel (usually a schwa) is phonetically realised immediately before it. n With the exception of sequences with a glide /j/ as the first member of the sequence (see §3.2). faithful as possible to the the nominative singular form, therefore they avoid the insertion of a schwa (tovarn 'a factory, gen.pl.', trm 'stubbornness, gen.pl.'). The fourth group consists of words in the genitive plural (morij 'a sea, gen.pl.', slavij 'a celebration, gen.pl.'), which have undergone the "the hardening of j into -i, which is then followed by an additional j due to the analogy with other cases" (Toporišič, 1978, 46). Let us now go back to the Celje dialect. What we observe is that words that Toporišič painstakingly devides into the first three groups behave exactly the same: the second member of the sonorant sequence becomes syllabic (5). (5) kalen [kaln], trn [tarn], film [film] The only exception we observe here are sequences with /j/ as the first member, as in fajn 'good, colloquial' and mejl 'e-mail, colloquial', which are according to the speakers' intuition monosyllabic words. Srebot Rejec (1992, 229) independently proposes that these sequences are not really clusters since [j] forms a diphthong with a previous vowel. Another explanation might be that these sequences are actually the only final sonorant clusters that can be found in the Celje dialect. We observe that the behaviour of words that belong to Toporišič's fourth group seems different from other words in the Celje dialect as well. Toporišič's explanation seems to be rather complicated, since it involves two processes: the "hardening" of a segment into another segment and the analogy (the relevant example is morje/morij 'a sea, nom.sg./gen.pl.'). It is disputable whether it is necessary to make use of two unrelated processes that turn one segment into another one and then call back the identity of the first one. If the behaviour of final sonorant sequences with /j/ as the last member is truly different from the behaviour of other final sonorant sequences, then a phonological theory that attempts to explain not only describe phonological phenomena should find reasons for this discrepancy. We will come back to this issue once we are armed with a good theory. For the time being let us pretheoretically propose that since other sonorants (except /r/) can become syllabic in the final position of a final sonorant sequence, /j/ also becomes syllabic. 4 Theoretical background We shall be analysing final sonorant sequences within the government phonology framework (Kaye et al., 1985, 1990), more precisely in the strict CV approach (Lowenstamm, 1996; Scheer, 2004). In government phonology a phonological string consists of a series of positions. There are two types of positions: C and V. The former type roughly corresponds to consonants and the latter to vowels. These two types of positions have to alternate in a phonological string. A C starts a phonological string and a V finishes it. A set of items that determine pronunciation and perception is attached to the positions in autosegmental fashion (see Goldsmith, 1976). These items are called elements and represent melodic material attached to a particular position. The theory uses only a small number of elements, the most radical versions recognize only six of them (for discussion cf. Neubarth and Rennison, 1998, 2005). The same elements can appear in a C or a V position—the nature of the position will determine the exact pronunciation. For instance, the phonetic realization of the element I will be a glide /j/ when attached to a C position, and a vowel /i/ when attached to a V position. Adjacent positions can also share melodic material. The melody sharing can either be lexical (already in our mental lexicon) or the result of a phonological process called spreading that allows elements to spread from a position to its adjacent positions of the same or different type. To use a skeleton of alternating Cs and Vs to represent the structure of words in a language that only permits syllables consisting of one consonant followed by one vowel hardly poses any problems. What about a language with consonant clusters? In such cases the strict CV approach allows a position to remain inaudible if it is empty, i.e.without melodic material. Also, the distribution of inaudible V positions is not arbitrary. It is constrained by the empty category principle (ECP), which can be satisfed in different ways. For the purpose of this paper let us discuss only two of them. Let us take a Slovene word like kanta 'a bin.' Although it lacks an audible vowel between n and 11 claim that its structure consists of three CV pairs as shown in (6)8. The theory allows positions to become involved in two different types of relations. The relations are strictly local and always involve exactly two adjacent positions, the direction of a relation is from right to left. The source of the relation is a phonetically audible position. Since position V2 in (6) is inaudible, it has to satisfy the empty category principle. A way of satisfying the ECP is to be the target in the governing relation (or government) with the following V. We can say that position V3 governs position V2 and thus allows it to remain inaudible (the governing relation is marked by a single arrow). If position V2 was not governed, it would be phonetically interpreted. The interpretation of an empty unsilenced V varies across languages: it usually appears as some sort of a schwa (Neubarth and Rennison, 1998). An empty V position can also remain inaudible by virtue of being the last position in a phonological domain. The theory proposes the (parametrized) final empty nucleus (FEN) principle, which allows domain-final V positions to remain unexpressed as in kot 'an angle' (depicted in (7)). Q Orthographic symbols stand for an informal summary of the melodic elements attached to a particular skele- Ci V] C2 V2 c3 v3 (6) kan t a tal position. FEN (7) Vi C2 V2 kot 5 Syllabic sonorants in government phonology and the analysis of the Slovene data We have pretheoretically proposed that (almost) any final sonorant sequence in the Celje dialect has a value of a syllable and that it is the second member of a sequence that functions as the most sonorous part of a syllable. In other words, the second member in a sequence becomes syllabic. Let us now see whether these observations are the expected outcomes in the given theoretical framework. Despite a more or less unified idea about the building blocks of government phonology, the representation of syllabic consonants due to their vocalic and consonantal characteristics remains open to dispute. One way to represent them is to propose that melodic material indicating the phonetic interpretation of a sonorant is attached to a V position (Toft, 2002; Harris, 1994). The shortcomings of this proposal are the following: (i) whether the phonetic realisation is that of a consonant or a vowel cannot be determined on the basis of melodic material itself; the type of the position elements attach to plays a decisive role, and (ii) alternations between syllabic and non-syllabic realisations of the same consonant in different forms of a word would imply resyllabification (switching of melodic material from one position to another), which is not recognized as a phonological process in government phonology. Another proposal is that the same melodic material is attached to at least two adjacent positions (Scheer, 2004; Blaho, 2001). These positions are of different types, which implies the twofold character of a syllabic consonant. However, this proposal does not entirely overcome the shortcoming of the previous one concerning the interpretation of melodic material. For the purpose of this paper we need to make the following assumptions: (i) A CV pair represents a unit, therefore the interpretation of the melodic material shared within a unit is specific, in this case a segment with a vocalic function and consonantal phonetic realisation, (ii) Melodic material is shared by precisely two adjacent positions: a C and a following V position.9 (iii) In Slovene there are no syllabic consonants present lexically; whenever they appear they are the result of the application of phonology on a phonological string. If a final empty V receives melodic material from the adjacent C position by means of spreading, we get a syllabic consonant, which is the case in the Celje dialect (depicted in (9)). This V position is thus phonetically realised and can therefore silence (by governing) the preceding V position. If there is no spreading into the last V position and consequently no sharing of the melodic material, the last empty V position is not phonetically realised and thus incapable of silencing the preceding V position. In this case a schwa is realised between the two sonorants (depicted in (8)). 9 Other possibilities are discussed in Scheer (2004); Blaho (2001); Ferme (2004). C V C V C v kana 1 C V C V C V (9) J k a ii Let us take the alternation obrv/obrvi [obaru]/[ob9rvi] 'an eyebrow, nom. sg./gen.(=nom.)pl.' In contrast with the example above, phonetic realisations of the segment under inspection differ: [u] in the position of an expected syllabic consonant and [v] in the position where a non-syllabic consonant is expected. Toporišič (1978) argues that [u] in the given example is the vocalised variant of the final sonorant. If 'vocalised' is translated to government phonology vocabulary as 'syllabic,' then Toporišič (1978) and our theory agree in explaining a phenomenon which at first glance seems very different from what we have observed in (9), but is in fact very similar depicted in (10) The structure (not taking into account the melodic material) of a syllabic /v/ is the same as in the case of other syllabic consonants, the unexpected part is the phonetic interpretation.10 This is not surprising since glides are phonetically very close to vowels, therefore the distinction, if there is one, can be easily blurred. Further elaboration of this stipulation, however, is beyond the scope of this paper. Let us recall that Toporišič (1978) proposes the hardening of /j/ to [i]11 when it is a second member of a final sonorant sequence. In addition to this process another /j/ is added to the sequence due to the analogy with other cases. This way of dealing with sonorant sequencing with /j/ as a second member is radically different from previous observations made for other sonorant sequences. Let us see if a different analysis is really necessary. 10 Despite the difference in pronunciation, I claim that the melodic representation of a segment does not change (it is always the element U), the only difference being the type and number of relations that target the given position. A discussion on the real character of the alternation is beyond the scope of this paper and will be dealt with elsewhere. 11 It is interesting to note that Toporišič (1978, 46-47) actually uses two explanations concerning this phenomenon interchangeably. The first one is the one mentioned above, that is, the hardening of/j/ to [i] and adding another /j/, the second one is the insertion of the [i]. The two proposals provide the same surface output, however, there is a fundamental difference between them: the first one involves the alophonic explanation (/j/ turns into [i]) while the second one calls for the insertion of a new phoneme. cvcvcvcv v We encounter two problems if we want to translate this proposal directly into our theory. The first one concerns the existence of a process that changes a consonant into a vowel, provided that each segment is represented by one position. The theory of government phonology does not recognize such a process. The second problem concerns the source of the element I, which represents the additional /j/. We could stipulate that I first breaks the bond with its original position, attaches to the adjacent position and in the end spreads into its original position. Not only that this process is not recognized by the theory, it is also intuitively hardly plausable that an element that is frequently lost in weak positions (e.g. unaccented [i] is very often lost) would have enough strength to delink, link again and finally spread. In addition, from this point of view the phenomenon does not seem to be motivated by analogy in any way. Having shown that Toporišič's proposal cannot be treated in our theory, let us go back to the previous analysis of final sonorant sequences and propose that sequences with /j/ behave exactly the same. We find that the surface result is very similar to the /v/ case: syllabic /j/ is phonetically identical to the vowel /i/. The phonological structure of the word morij is depicted in (11). c v c v^TTV III (11) mor j As for the additional /j/, the intuition of speakers shows that there is no such object.12 To sum up, final sonorant sequences with /j/ as the second member are phonologically the same as other final sonorant sequences. In all cases alike the theory predicts that the second member becomes syllabic, i.e. that its melodic material spreads and attaches to the following V position. "The hardening of/j/ in to [i]," "the insertion of /i/" or "adding a /j/ due to analogy" turn out to be merely illusions fed by orthography. 6 Conclusion In this paper I examined final sonorant sequences in the Celje dialect. After having provided the definition of a consonant ciuster, I pointed out that given sequences cannot be treated as sonorant clusters, since they do not need to obey the sonority sequencing principle and always form a syllable on their own. These pretheoretical proposals were strongly supported by the analysis in the government phonology framework. Furthermore, the analysis in the given theoretical framework dispensed with a belief in the idiosyncratic behaviour of sequences with /j/ as the second member, and showed that these sequences behave phonologically exactly the same as other final sonorant sequences. 19 1 An experiment shows that speakers cannot hear the difference between ladij 'a ship, gen.pl' and Ladi (a proper name). References BLAHO, Sylvia (2001). The representation of Slovak syllabic consonants in strict CV. In The Odd Yearbook 6, 3-24. Department of English Linguistics, School of English and American Studies, Eötvös Loränd University (ELTE). FERME, Alja (2004). Syllabic consonants in Slovene. Ms. GOLDSMITH, John A. (1976). Autosegmental Phonology. Ph. D. thesis, MIT. HARRIS, John (1994). English Sound Structure. Oxford: Blackwell Publishers. KAYE, Jonathan D., Jean Lowenstamm, and Jean-Roger Vergnaud (1985). The internal structure of phonological elements: a theory of charm and government. Phonology Yearbook 2, 305-328. KAYE, Jonathan D., Jean Lowenstamm, and Jean-Roger Vergnaud (1990). Constituent structure and government in phonology. Phonology 7, 193-231. KENSTOWICZ, Michael (1995). Phonology in generative grammar. Oxford, Cambridge: Blackwell. LOWENSTAMM, Jean (1996). CV as the only syllable type. In J. Durand and B. Laks (Ed.), Current Trends in Phonology: Models and Methods. Salford, Manchester: European Studies Research Institute (ESRI), University of Salford. NEUBARTH, Friedrich and John R. Rennison (1998). An x-bar theory of Government Phonology. In S. Ploch (Ed.), Living on the Edge: 28 Papers in Honour of Jonathan Kaye, Volume 62 of Studies in Generative Grammar, 95-130. Mouton de Gruyter. NEUBARTH, Friedrich and John R. Rennison (2005). Structure in Melody, and vice versa. Leiden Working Papers in Linguistics 2(4), 95-124. SCHEER, Tobias (2004). A lateral theory of phonology. Vol 1: What is CVCV, and why should it be? Mouton de Gruyter, Berlin. SREBOT Rejec, Tatjana (1992). Initial and final sonorant clusters in Slovene. Linguistica 32(2), 227-230. TOFT, Zoe (2002). The Phonetics and Phonology of some Syllabic Consonants in Southern British English. ZAS Papers in Linguistics 28, 111-144. TOPORIŠIČ, Jože (1959). Alternativni soglasniški sklopi slovenskega knjižnega jezika. JiS 4(7), 203-207. TOPORIŠIČ, Jože (1978). Glasovna in naglasna podoba slovenskega jezika. Založba Obzorja. UNUK, Drago (2003). Zlog v slovenskem jeziku. Ljubljana: Rokus: Slavistično društvo Slovenije (Slavistična knjižnica 7). Povzetek ZVOČNIŠKA IZGLASJA V CELJSKEM GOVORU Članek obravnava izglasne zvočniške nize v celjskem govoru. Na podlagi definiranega termina izglasni soglasniški sklop ugotavljam, da celjski govor pravih izglasnih zvočniških sklopov ne pozna. Prav zato se v njem lahko uresničujejo skoraj vsi končni zvočniški nizi, pri čemer se drugi zvočniški člen skoraj brezizjemno ozloži. Nazadnje so podrobneje preučeni primeri z izglasnim /j/, ki je edini izmed zvočnikov, pred katerega naj bi se vrival samoglasnik /i/. V okviru teoretičnega aparata vezalne fonologije pokažem, da v tovrstnih primerih ne gre za izjemno obnašanje, temveč za popolnoma pravilen proces ozloženja drsnika /j/, katerega fonetična realizacija (ne pa tudi fonološka zgradba!) je enaka samoglasniku /i/. Chikako Shigemori Bučar University of Ljubljana UDK 811.521 : 811.163.6 CAUSATIVE CONSTRUCTIONS IN JAPANESE AND SLOVENE 0. Introduction Whereas Japanese has a productive causative morpheme -(s)ase- that is added to a non-causative verb to derive a causative verb, Slovene has no such single morpheme that may be considered to directly correspond with the Japanese morpheme. This article considers causative constructions first from a typological perspective, and then in present-day Japanese and Slovene by examining a small corpus of literary texts. The use of the Japanese causative morpheme is analyzed according to the characteristic of agency. For the most typical causative situation (with agentive causer and causee), corresponding Slovene constructions are analyzed. 1. Causative situation, causative constructions and causative semantics A causative situation is such that an event is brought about by some outside entity. For example, Comrie (1985) states: If we take a sentence containing a non-causative verb (or other predicate) to be describing a certain situation [s], then a sentence containing the corresponding causative verb will describe a situation [s/cause/] where some entity (person, thing, abstract force) either brings about situation [s] or, at the very least, fails to prevent [s]. Syntactically, one of the main differences between non-causative and causative constructions will be the increased valency (or potential valency) of the latter, since in addition to the participants in [s] there will also be the actant that brings about [s], (Comrie 1985: 330-331) Comrie (1985) continues with English and Turkish examples to illustrate analytic (or syntactic), morphological (or synthetic), and lexical causative constructions. (1) Sam slid off the roof. (2) Mary brought it about that Sam slid off the roof. (3) Mary caused Sam to slide off the roof. (4) Mary made Sam slide off the roof. If one takes what is expressed in sentence (1) as the certain situation fs], then corresponding analytic causatives in English would be sentences (2), (3), and (4), where the idea of causation is expressed by the verbs bring about, cause, or make (Comrie 1985: 331). A morphological causative means that the predicate of [s] undergoes some derivational process in order to express causativity, as in (6) below from Turkish, where (5) represents [s]; the derivational process is the addition of the suffix -diir: (Comrie 1985: 331) (5) Hasan öl-dii. Hasan die-PAST 'Hasan died.' (6) Ali Hasan-i öl-dür-dü. Ali Hasan-DO die-CAUSE-PAST 'Ali caused Hasan to die. / Ali killed Hasan.' The alternative English gloss to sentence (6), according to Comrie, exemplifies lexical causatives. The verb used in describing [s] in English here is die, but to indicate that Ali brings about [s] one may use a completely different lexical item, kill, which has no morphological resemblance to die; kill is thus the lexical causative of die (Comrie 1985: 323,331). According to Comrie, formation of analytic causatives is completely productive. Formation of lexical causatives is by definition non-productive because there is no regular pattern to be extended to new items. The degree of productivity of morphological causatives varies from language to language, from languages such as English with only a few fossilized pairs (lie/lay, sit/seat) to languages such as Turkish, in which virtually any verb (including a causative verb) can form a causative (Comrie 1985: 332). As for the semantics of causatives, the degree of closeness between the cause (i.e., the causer's actions) and the effect (resultant situation) comes into question (Comrie 1985: 332—333). Although an absolute distinction between mediated and immediate causation is difficult to draw, one often finds that, if a language has both analytic and morphological or lexical constructions, the former imply less direct causation than the latter, as in the Russian examples below (Comrie 1985: 333): (7) Anton slotnal paločku. Anton(NOM) break(TRANS)-PAST stick(ACC) 'Anton broke the stick.' (8) Anton sdelal tak, ctoby paločka slomala-s'. Anton(NOM) do-PAST so that stick(NOM) break(INT)-PAST 'Anton brought it about that the stick broke.' (For example, Anton pushed against the lever that released the weight that fell on the stick, Comrie 1985: 333) A similar distinction is sometimes found between non-productive (lexicalized or virtually lexicalized) and productive morphological causatives, the former implying greater closeness between cause and effect; for example, in Nivkh, the productive morphological causative in -gu (Comrie 1985: 333): (9) Lep ce-d'. bread dry(INT)-PAST 'The bread dried.' (10) If lep seu-d'. he bread dry(TRANS)-PAST 'He dried the bread.' (11) If lep ce-gu-d'. he bread dry(INT)-CAUSE-PAST 'He caused the bread to get dry (e.g., by forgetting to cover it).' Another aspect of the relation between causer and effect is the difference between causative proper and permissive meaning. For example (Comrie 1985: 333): (12) I made John climb the tree. (13) / let John climb the tree. Analytic causative constructions generally have distinct predicates for causation proper and permission; for example, English cause, make verses allow, let. Lexical causatives typically express causation proper, but morphological causatives often cover the range of both causation proper and permission. Consider the following example in Nivkh (Comrie 1985: 334): (14) Ni atik-ax ph-emek vi-gu-d'. I younger brother-CAUSEE own-mother go-CAUSE-PAST 'I made/let my younger brother follow my mother.' 2. Causative expressions in Japanese The syntactic, semantic, and morphological characteristics of the Japanese causative 'shieki' are defined by Teramura (1982) as follows: In the typical sentence structure: W ga Xwo(Yni)/ni W-(s)ase-ru, W NOM X ACC (Y DAT) /DAT V-CAUSE-PRESENT W brings about, or cannot prevent, a situation in which X does something (to Y); or W feels responsible for the situation brought about (Teramura 1982: 290). The distinctive Japanese morpheme -(s)ase-, called shieki-ji ('causative suffix'), is a productive means to derive causative verbs from non-causative verbs. For example: (15) Tarö g a gakkö e itta. Tarö NOM school DIR go-PAST 'Tarö went to school.' (16) H aha ga Tarö wo gakkö e ikaseta. mother NOM Tarö ACC school DIR go-CAUSE-PAST 'Mother made Tarö go to school.' (17) H aha ga Tarö ni gakkö e ikaseta. mother NOM Tarö DAT school DIR go-CAUSE-PAST 'Mother let Tarö go to school.' Taking what is expressed in sentence (15) as the certain situation [s], the corresponding morphological causative to express a situation [s/cause/] in Japanese is sentence (16). The lexical verb iku 'to go' combines with the derivational morpheme -(s)ase- to form the causative verb ikaseru 'to make (someone) go'. Sentence (17) is another variation of the causative expression: whereas sentence (16) expresses immediate causation with the case particle wo, sentence (17) conveys permissive meaning with the case particle ni. The semantics of sentences (16) and (17) may be compared somewhat parallel to the example sentences (12) and (13) in the previous section, below as (18a) and (18b): (18a) I made John climb the tree. (18b) I let John climb the tree. If the certain situation [s] is a transitive action with a direct object as in sentence (19) below, the case particle wo is already taken, so the corresponding causative sentence expressing [s/cause/] has no variation. Sentence (20) below may convey the causative proper as well as permissive meaning. (19) Tar6 ga ring o wo tabeta. Taro NOM apple ACC eat-PAST 'Taro ate an apple.' (20) H aha ga Tar6 ni ringo wo tabesaseta. mother NOM Taro DAT apple ACC eat-CAUSE-PAST 'Mother made Taro eat an apple. (= Mother fed Taro with an apple.)/ Mother let Taro eat an apple.' Teramura (1982) states that the causer of the Japanese causative construction is usually an animate being. If not, the expression is felt as a personification (1982: 293). Most typically, the Japanese causative is used when both the causer and the causee are human beings, and usually when the causer is superior to the causee (Teramura 1982: 301). However, there are cases when the morpheme -(s)ase- appears even when the causee is inanimate. This happens when there is no transitive counterpart of a certain intransitive verb: in this case, the derived causative verb with the causative morpheme takes over the role of the transitive verb that happens to be absent in the Japanese lexicon (Teramura 1982: 293—296). For example, although there is a pair of intransitive and transitive verbs with the common root ak-, aku 'to open [intr.]' and akeru 'to open [trans.]', there is no transitive counterpart to the intransitive verbs hikaru 'to shine, glisten' and sumu 'to be settled, finished'. Observe: (21a) Mado ga aku. (21b) Mado wo akeru. window NOM open(INT) window ACC open(TRANS) 'The window opens.' '(I) open the window.' (22a) Gar asu ga hikaru. (22b) Gar asu wo hikaraseru. glass NOM shine(INT) glass ACC shine(INT)-CAUS 'The glass shines.' '(I) make the glass shine.' (23a) Kanjo ga sumu. (23b) Kanjo wo sumaseru. account NOM settle(INT) account ACC settle(INT)-CAUS 'The account is settled.' '(I) settle the account.' For expressions that may stand parallel to example (21b) with the transitive verb akeru, the derivational causative morpheme -(s)ase- appears for the transitive expressions (22b) and (23b). The morpheme -(s)ase-, which has just been seen above, is an example of grammatical means to create the voice opposition of non-causative vs. causative. What is called the 'semi-lexical' opposition of voice in Japanese (Nöda 1991) are morpholo -gical forms such as -as-, -s-, -ar-, and -r-, which appear only in certain verb pairs with a common verbal root. In the example pairs below, the morphological form -,v-, which seems historically related to the derivational causative -(s)ase-, is found in the transitive counterpart of each pair: Intransitive Transitive kowareru 'to break(INT)' kowasu 'to break(TRANS)' okiru 'to get up, rise' o kosu 'to wake, raise' The third type of opposition, so-called 'lexical' by Nöda (1991), is realized by verbs of no common morphology. A typical example is the verb pair shinu 'to die' vs. korosu 'to kill'. (24) Tar o ga shinda. Tarö NOM die-PAST 'Tarö died.' (25) Hanako ga Taro wo shinaseta. Hanako NOM Taro ACC die-CAUS-PAST 'Hanako let/made Taro die.' (26) Hanako ga Taro wo koroshita. Hanako NOM Taro ACC kill-PAST 'Hanako killed Taro.' Taking the expression in sentence (24) as the certain situation [s], the corresponding morphological causative to express a situation [s/cause/] is (25). Here the derived causative verb conveys mediated causation, something like 'Hanako let Tarö die by not caring for him enough, etc.' The transitive counterpart of the lexical opposition, the verb korosu 'to kill' in sentence (26), on the other hand, expresses direct causation; i.e., 'Hanako killed Tarö by stabbing with a knife, etc.', though of course the degree of closeness between the cause and the effect depends on the individual context in which each expression appears. 3. Causative expressions in Slovene Toporišič's encyclopedia of the Slovene language (1992) carries an entry »causative verb« (vzročniški glagol) and it is explained that causative verbs express the achievement of someone's purpose that some event comes true. Examples are (Toporišič 1992: 362): pojiti/napajati = povzročiti/doseči, da kdo/kaj pije 'to supply (with water, fuel etc.) = to cause/achieve that s.o./s.th. drinks' točiti = povzročiti/doseči, da kaj teče 'to pour = to cause/achieve that s.th. flows' uspavati = povzročiti/doseči, da kdo spi 'to lull = to cause/achieve that s.o. sleeps' In Toporišič's grammar under »morphological means of expressing transitivity« (2000: 353—354), the following three derivational means are listed: i) Pairs of verbs with vowel gradation: Intransitive Transitive rumeneti rumeniti 'to turn yellow' vs. 'to make yellow' oživeti oživiti 'to become animated' vs. 'to revive' ii) Preposition as a free verbal morpheme: hoditi 'to walk' hoditi za 'to try to obtain s.o.'s affection' iii) Taking the morpheme se from the intransitive verb: Steklenica se je razbila. bottle(NOM) break(INT)-PAST 'The bottle broke.' Razbil je steklenico. break(TRANS)-PAST bottle(ACC) 'He broke the bottle.' The first means (i) is very similar to the »causative verb« above. Although the systematic vowel alternation -e-l-i- is not seen in the former examples of the causative verb, the shared etymology is observed in each pair. If the pairs are examined seman-tically, the intransitive/transitive opposition is seen in the same manner with the following verb pairs: piti/pojiti, teči/točiti, and spati/uspavati. It is in fact the causativiza-tion process of gaining a new argument, an agent, whose purpose is to realize the event expessed by the intransitive counterpart (Shigemori Bučar 2003: 100). The second (ii) is mentioned in Dular (1982: 147) as one of the word formation processes to gain a place for a new argument. The example cited in Toporišič (hoditi vs. hoditi za) is not a simple semantic pair of intransitive and transitive verbs by way of derivation, but the transitive counterpart has an additional semantic value, as pointed out by Dular (1982: 147). Another such example is found in the verb pair poslati 'to send' vs. poslati po 'to send for'. The preposition po in this case combined with the lexical verb poslati derives a new verb with a causative meaning (Shigemori Bučar 2003: 242): (27) Janeza smo poslali po zdravnika. Janez(ACC) send(l.p.pl.)-PAST for doctor(ACC) 'We sent Janez to get the doctor./We sent Janez for the doctor.' The third means mentioned by Toporišič (iii) is one of the main means of intransitive/transitive opposition in Slovene. These are all derivational processes of verbs between 1-place (intransitive) and 2-place (transitive) variations of the same semantic base of events, but none of them are productive in modern Slovene. The vowel alternation -e-l-i- is observed in a few fossilized pairs of verbs and, together with intransitive/transitive opposition of the type pojiti/napajati, most of these pairs have become semantically quite distant and are actually used in quite different contexts in modern Slovene (Shigemori Bučar 2003: 214). These and the lexical oppositions of the type razbiti se/razbiti may be called "semi-lexical" oppositions of voice in Slovene. "Lexical" opposition would be, for example, ostati 'to stay' /pustiti 'to leave', or pasti 'to faW/spustiti 'to let fall, drop', etc. Orešnik (1992) expected that morphological causative verbs would display 3-role semantic valency. However, his analysis of all verbs under the letter j shows that most of the verbs of 3-role semantic valency are verbs of saying. Instead, many causative verbs are present among the verbs of 2-role semantic valency (Orešnik 1992: 184—185). An example of a rare 3-place causative verb in Slovene is učiti 'to teach'. Sentence (29) is an analytical interpretation of sentence (28), also by Orešnik (1982: 185): (28) Janez poučuje Micko matematiko. Janez(NOM) teach-PRES Micka(ACC) mathematics(ACC) 'Janez teaches mathematics to Micka.' =(29) Janez dela neposredno, da se Micka uči matematiko. Janez(NOM) do-PRES directly that Micka(NOM) learn-PRES mathematics(ACC) 'Janez does it directly that Micka learns mathematics.' 4. Hypothesis for corpus analysis Both Japanese and Slovene have lexical and semi-lexical means of causative expressions, but they are limited to certain verbs and are not productive today. The productive derivational morpheme for the causative, on the other hand, is found only in Japanese. Based on the survey of Slovene grammar (in the previous section), one may expect the causative expressions in Slovene to be found in one of the following constructions: A. With a lexical or semi-lexical transitive (causative), 2-place, or 3-place verb B. Analytical expression: B-l. With 2 verbs of which the latter is in a subordinate clause e.g., (30) Janez mu je dovolil, da vstopi. Janez(NOM) he(DAT) permit(PAST) that enter(3 .p .sg .PRES) 'Janez permitted him to enter.' (31) Janez mu je rekel, naj vstopi. Janez(NOM) he(DAT) say(PAST) should enter(3.p.sg.PRES) 'Janez told him that he should enter.' B-2. With 2 verbs, of which the latter is in the infinitive e.g., (32) Janez mu je dal piti. Janez(NOM) he(DAT) give(3.p.sg.PAST) to drink 'Janez gave him to drink.' B-3. With a verb and a verbal noun e.g., (33) Janez mu je omogočil vstop. Janez(NOM) he(DAT) enable(3.p.sg.PAST) entrance(ACC) 'Janez enabled him entrance ./Janez made it possible for him to enter.' C. With a phrasal verb of the type hoditi za 'to try to obtain s.o.'s affection', poslati po 'to send for'. 5. Corpus analysis Literary texts in modern Japanese literature with direct translations in Slovene were chosen (altogether 197,730 Japanese characters, the list is found at the end of the article). The analysis is in 2 parts: First, passages with the Japanese causative morpheme are searched throughout the chosen corpus and classified according to the agency of the causer and causee (Section 5.1). Second, for the most typical causative situation with the Japanese causative morpheme, corresponding Slovene expressions are analyzed according to their grammatical constructions (Section 5.2). 5.1. Japanese causative morpheme and agency There are 140 passages in which Japanese causative morphemes (the productive suffix -(s)ase- and the semi-lexical morpheme -s-) appear. These Japanese causative expressions were classified according to the distinction of agency. In accordance with the assertion by Teramura (1982: 293—301; see Section 2 above) and other Japanese grammarians, the productive causative morpheme is usually used and is also taught (e.g., in Japanese language teaching) within the context of both the causer and the causee being human beings; therefore, with agency or self-control of these two. Languages vary with respect to how often agency is lexicalized in verbs (e.g., Van Valin & Wilkins 1996: 309—340), and it has been observed that Japanese verbs are more sensitive to the distinction of agency than are verbs in English (Shigemori Bučar 2003:15—16) and Slovene (Shigemori Bučar 2003: 199). However, one must also bear in mind that in some cases the derived causative verb in Japanese assumes the role of a transitive verb. (Teramura 1982: 293—296; see Section 2 above) Therefore, it may be expected that some passages with the derived causative may be of an inani-mate/non-agentive causee. Thus, the 140 cases of Japanese causative expressions are classified below (Table 1) according to the agency of the causer (the one that brings about the event [+agency] or a thing/force causing the event [-agency]), and the causee (the one that is acting in the event [+agency] or a thing with which some change is brought about in the event [-agency]). Table 1: Agency of the causer and causee in Japanese literary texts (total 140 cases) Agency of causer Agency of causee Number of cases and % + + 63 (45%) + - 51 (36.5%) + 10 (7%) 16(11.5%) It is interesting, when considering the definition of the Japanese causative and the sensitivity of Japanese to agency, that inanimate causers are still found in the Japanese corpus (18.5% of the corpus). In Hasegawa (1992) it is argued that Japanese transitive verbs such as korosu 'to kill' and kowasu 'to break' take true agent arguments, not simply effectors. The Japanese situation is compared to the quite different situation in English. Van Valin & Wilkins (1996: 309—340) cite the following example sentences: (34) Larry killed the deer. (35) Larry accidentally killed the deer. (36) The explosion killed the deer. and argue that sentence (34) is neutral and leaves room for our interpretation whether or not the subject Larry is an agent, that in sentence (35) the adverb accidentally is »agent-cancelling« and thus the subject Larry is interpreted as an effector but not the direct agent, and that in sentence (36) the subject is non-human and thus cannot be interpreted as the agent, but as the »force« that effected the killing. When the English sentence (36) is translated into Japanese, the following is observed (Shigemori Bučar 2003: 15): (37) *Bakuhatsu ga shika wo koroshita. explosion NOM deer ACC kill-PAST 'The explosion killed the deer.' (38) Bakuhatsu de shika ga shinda. explosion INST deer NOM die-PAST 'The deer died due to the explosion.' Whereas sentence (37) is difficult to accept, a translation possibility is found in sentence (38) in which the noun phrase 'explosion' is accompanied by the case particle de to express the cause of the event and 'deer' is in the nominative/agentive case as the central entity for which the event 'to die' occurs. In the same manner, the passages found in the corpus with an inanimate/non-agen-tive causer (effector) may be formulated otherwise. Examples (39a) and (40a) are the original passages in the Japanese corpus 'Kagiri naku...', and (39b) and (40b) are my reformulations: (39a) Toppü ga boku wo fukitobasu. sudden wind NOM I ACC blow away(TRANS) 'A gust of wind blows me away.' (39b) Toppü de boku ga fukitobu. sudden wind INST I NOM be blown away(INT) 'I am blown away by a gust of wind.' (40a) Pinku no keroido ga ue no ho made hifu wo hikitsurasete iru. pink keloid scar NOM upwards skin ACC crinkle(INTR)-CAUS-STATE 'The pink keloid scar drew the skin upwards.' (40b) Pinku no keroido de ue no ho made hifu g a hikitsutte iru. pink keloid scar INST upwards skin NOM crinkle(INTR)-STATE 'Due to the pink keloid scar the skin is drawn upwards.' The original passage (39a) contains an inanimate causer (toppu. 'sudden wind'), a human causee (boku 'me'), and a transitive/causative verb with the morpheme -s- (fukitobasu 'to blow away'). In the reformulated sentence (39b), the animate participant of the event is in the nominative case with the particle ga (boku ga T), the sudden wind is accompanied by the particle de to express the cause of the event, and the verb is intransitive (fukitobu 'to be blown away'). In (40a), which is also an original literary expression, the scar is the inanimate causer (keroido ga) and the skin is the causee, which is also inanimate (hifu wo). The verb in this passage is affixed with -ase-, which causitivizes the originally intransitive verb hikitsuru 'to crinkle'. My reformulation puts the skin in the center of the expression with the nominative case particle ga, and the inanimate causer keroido is accompanied by de as the cause of the event. These are all grammatically acceptable sentences in Japanese, but (39a) and (40a), especially (39a), are felt to be stylistically characteristic. It may be assumed that these cases of a non-animate/non-volitional causer are characteristic for the literary genre. In many cases in these passages, the inanimate causer is a natural force (wind, the sound of rain, clouds, the sun, etc.), and in other cases traces of personification are seen; for example, another passage from 'Kagiri naku...': (41) ...kage ga, kaibutsu ga ha wo mukidasu yöni boku-tachi wo odorokaseru. shadow NOM moster NOM teeth ACC bare as-if we ACC surprise-CAUSE-PRESENT 'The shadow frightens us, like a monster baring its teeth.' It may be concluded that, in the modern literary genre, Japanese causative expressions with the productive causative morpheme -(s)ase- or causative verbs with -s- are mainly used with the causer being agentive (more than 80% in the present corpus), and most often when both the causer and the causee are agentive (45% in the corpus). There are also cases in which the causer is not an animate agent but simply an effector/cause of the event (a little less than 20% in the corpus). 5.2. Slovene constructions for the causative situation with animate participants The 63 passages of the present corpus with both the causer and the causee being animate/agentive were further analysed in order to see which Slovene constructions correspond to the most typical Japanese causatives with the distinctive morphemes -(s)ase- and -s-. The results are listed below in the same order as the hypothesis in section 4 above. The classification of verbs as to whether they are 2-place or 3-place and the English translations of each expression are given according to the specific contexts of individual verbs. Some prepositional cases (z 'with' and v 'in, into') are also counted as the third argument in clauses with a 3-place verb. A. With a lexical or semi-lexical transitive (causative) verb: 34 cases 2-place (14 cases): dvigniti 'to lift, raise', narediti konec 'to put an end to', oploditi 'to impregnate', peljati 'to drive, lead', pobrati 'to pick up', posaditi 'to plant', posesti 'to seat', presenetiti 'to surprise', priganjati 'to drive, spur, hurry', poslati nazaj 'to send back', udariti 'to beat, strike', ustaviti 'to stop', utrujati 'to tire', uvoziti 'to import'; 3-place (20 cases): govoriti 'to speak', izpostaviti 'to expose', naliti z 'to pour, fill with' (2), narediti 'to do', naznaniti 'to notify, inform', pokazati 'to show',poslati 'to send', povedati 'to tell' (3), predstaviti 'to present, introduce', reči 'to say, tell', vključiti v 'to include in' (3), stisniti v 'to press, squeeze into', vpisati 'to register, enroll in', vtakniti v 'to put into', zadeti 'to hit'; B. Analytical expression: 10 cases B-l. With 2 verbs, of which the latter is in a subordinate clause: (5 cases) narediti, da počuti 'to make s.o. feel', pusititi, da opravi 'to let s.o. settle', pustiti, da počne 'to let s.o. do' (2), ukazati, da sede 'to order s.o. to sit'; B-2. With 2 verbs, of which the latter is in the infinitive: (4 cases) dati piti 'to give to drink', dati stkati 'to have weave', dovoliti iti 'to permit to go', pustiti opazovati 'to let s.o. watch'; B-3. With a verb and a verbal noun (1 case): ponuditi izkušnje 'to offer experience' C. With a phrasal verb: none 34 out of 63 cases are expressions with a transitive verb (either 2-place or 3-place); altogether, 54% of 63 cases. Analytical expressions are found in 10 cases (16%), much less than the cases with a transitive verb. The only disappointment is that no case of word formation (phrasal verb) of the type poslati po, hoditi za is found in the corpus. For the analytical expression, it must be noted that the distinctive predicates show the semantic subcategories of the original Japanese morpheme in one form. Whereas the semantic distinction between direct or mediated causation must be made according to the context in Japanese, the Slovene analytical expressions make this distinction by the choice of lexical verbs: ukazati 'to order' denotes direct or forceful causation, whereas pustiti 'to leave, let' and dovoliti 'to permit' denote mediated causation or more distance between the causer and the causee. One case is found in which the Slovene corresponding expression is a set phrase without any verb: po njegovih besedah 'in his words' for the original Japanese expression kare ni iwasereba 'if (I/we) let him say'. The remaining passages are difficult to analyse in the given framework and need further discussion mainly due to the change in the viewpoint of narration. The following two tendencies are observed: 1. The causer-part of the expression is dropped in the Slovene corresponding passage, probably because the analytical expression is felt to be too complicated and stylistically awkward. The semantics of the causer are often included in a wider range of the narrative. 2. The original Japanese causative expression is a part of a more complicated pre -dicate construction; i.e., causative-passive, causative-imperative, and links with giving and receiving verbs to denote benefit or favor. These constructions trigger a switchover in the viewpoint. 6. Conclusion This paper is an attempt to search for common ground for contrastive analysis of the causative in Japanese and Slovene. Within the limits of a small corpus analysis in this paper (i.e., the most typical causative semantics with agentive causer and causee, and in the modern literary genre), the majority of passages with the Japanese causative morpheme found their equivalent in a transitive verb or in an analytical construction of two verbs in Slovene. Further research with a broader view of the semantics of the causative, using more corpora in various genres, may lead to more detailed accounts of this subject. References Comrie, Bernard (1985) Causative verb formation and other verb-deriving morphology, Language Typology and Syntactic Description Vol III: Grammatical categories and the lexicon, ed. Timothy Shopen, Cambridge Univ. Press, 309-348. Dular, Janez (1982) Priglagolska vezava v slovenskem knjižnem jeziku (20. Stoletja) (Verbal goverment in literary Slovene (20^ century)), doctoral dissertation, Faculty of Arts, University of Ljubljana. Hasegawa, Y. (1992) Syntax, Semantics and Pragmatics of te-linkage in Japanese, doctoral dissertation, University of California at Berkeley. Nöda, Hisashi (1991) 'Bunpö-teki-na Voisu to Goi-teki-na Voisu no Kankei' ('The Relationship between Grammatical Voice and Lexical Voice') Nihongo no Voisu to Tadösei (Japanese Voice and Transitivity) Kurosio, Tokyo, 211—232. Orešnik, Janez (1992) Udeleženske vloge v slovenščini (Semantic Roles in Slovene), Classis II: Philologia et Litterae 37, SAZU, Ljubljana. Shigemori Bučar, Chikako (2003) Voice in Contrast—Japanese and Slovene (Protistava glagolskega načina v japonščini in slovenščini), doctoral dissertation, Faculty of Arts, University of Ljubljana. Teramura, Hideo (1982) Nihon-go no sintakusu to imi I (Japanese Syntax and Semantics I), Kurosio, Tokyo. Toporišič, Jože (1992) Enciklopedija slovenskega jezika (Encyclopedia of the Slovene Language), Cankarjeva založba, Ljubljana. Toporišič, Jože (2000) Slovenska slovnica (Slovene Grammar), Založba obzorja, Maribor. Van Valin, Robert D. Jr. & David P. Wilkins (1996) The Case for 'Effector': Case Roles, Agents, and Agency Revisited, Grammatical Constructions: Their Form and Meaning, ed. Masayoshi Shibatani & Sandra A. Thompson, Clarendon Press, Oxford, 289—322. Slovene translations of the first three texts are parts of undergraduate theses presented to the Dept. for Asian and African Studies, Faculty of Arts, University of Ljubljana; the following three are in Literatura, 16 (2004), Nos. 161 — 162, LUD Ljubljana; the last one was independently published as Skoraj prosojna modrina (2005), Cankarjeva založba, Ljubljana. Medtem ko v japonščini obstaja produktivni vzročniški mortem -(s)ase-, ki se doda nevzročniške-mu glagolu za izpeljavo vzročniškega, slovenščina ne razpolaga s čim podobnim, ki bi ustrezalo japonskemu morfemu. Članek najprej tipološko povzema vzročniške zgradbe, nato na osnovi korpusa modemih literarnih besedil primerja vzročniške izraze v japonščini in slovenščini. Japonščina dobro označuje in ločuje živost aktantov, kar je delno potrdil tukaj obravnavani korpus. Več kot 80 % vzročniških izrazov v japonskem korpusu (z vzročniškim morfemom -(s)ase- ali z vzročniškim glagolom s -i-) spremlja človeški aktant (oz. povzročitelj) in pogosto sta oba člena, tj. povzročitelj kot tudi vzročno vplivano, živo bitje oz. človek. Našlo se je tudi nekaj primerov, pri katerih je aktant neživ (manj kot 20 %). Večina japonskih izrazov z vzročniškim morfemom in s človeškim aktantom ima v slovenščini svojo ustreznico v zgradbah s prehodnim glagolom (54 %) ali z analitično zgradbo dveh glagolov (16 %). Primerov predložnomorfemskih glagolov po hipotetičnem pričakovanju v obravnavanem korpusu ni bilo. Nadaljnje delo na področju protistave japonskih in slovenskih vzročniških zgradb zahteva širši oz. obsežnejši pregled semantike vzročniškosti in predvsem z večjimi korpusi. Text corpus Chinmoku, MURAKAMI Haruki, 1998 Gotai Fumanzoku, Ibarin-bo, OTOTAKE Hirotada, 1998 Toshishun, AKUTAGAWA Ryünosuke, 1920 Danmari-ichi, TSUSHIMA Yoko, 1983 Yasashii Sayoku no tame no Kiyükyoku, SHIMADA Masahiko, 1983 Hokago no Ki-noto, lay-Walk, YAMADA Eimi,1989 Kagiri-naku Tomei ni Chikai Buru, MURAKAMI Ryü, 1976 (18,400 characters) (19,244 characters) (10,741 characters) (10,660 characters) (31,434 characters) (12,751 characters) (94,500 characters) Povzetek VZROČNIŠKE ZGRADBE V JAPONŠČINI IN SLOVENŠČINI Aleksandra Derganc Univerza v Ljubljani UDK 811.163.6 : 811.161.1 81*25 TRIJE ZGLEDI NEPOSREČENEGA PREVODA ZARADI »LAŽNIH PRIJATELJEV« Znano je, da posebej v sorodnih jezikih, genetsko zares ali navidezno sorodne besede, ki imajo enako ali podobno zunanjo podobo, pogosto nosijo različen pomen. Takšnim besedam pravimo »lažni prijatelji«. Če jih ne prepoznamo oz. ne zaznamo, da je pomen, ki ga pripisujemo kaki besedi, v nekakšni opreki s kontekstom, postanemo njihova žrtev. Ob prebiranju prevoda knjige Zakladi Slovenije M. Kmecla v ruščino sem naletela na mesto, ki mi je vzbudilo nevero: ali je mogoče, da je Kmecl napisal kaj takega? Mesto se je glasilo (tu in v naslednjih zgledih so okrepljeno tiskana mesta moj dodatek): B замечателБнои красотм Шалецкои долине (красоту ее в многом приумножила современнаи индустриализацил) пока нашли один-единствешљш римскии каменв, что длл нас более чем скромно. (Кмецл 1990: 70) Prevod tega mesta nazaj v slovenščino se glasi: V prelepi Šaleški dolini (njeno lepoto je v mnogočem povečala sodobna industrializacija) so doslej našli en samcat rimski kamen, kar je za nas nadvse skromno. Seveda se mi je zazdelo skrajno neverjetno, da bi Kmecl napisal, da je lepoto Šaleške doline povečala sodobna industrializacija, zato sem brž poiskala ustrezno mesto v slovenskem izvirniku. Le-to se glasi: V vsej nenavadno lepi (do neke mere jo je razlepotila šele zadnja industrializacija) Šaleški dolini so menda doslej našli en sam rimski kamen, kar je za slovenske razmere pravcato uboštvo. (Kmecl 1979: 70) Prevajalka slovenskega besedila v ruščino je kompetentna ruska slovenistka, ki je delo kot celoto dobro prevedla. Niti najmanj ni moj namen kako posmehovanje, saj velja, da naj prvi vrže kamen tisti, ki ni nikoli grešil, in verjamem, da se je skoraj vsakemu, ki je kdaj prevajal, kdaj zgodila kaka nerodnost, le odkrije se bolj poredko. Ta zgled bi rada opisala le zato, ker gre za zelo nazoren zgled o pomenu poznavanja vrednot in prepričanj kake družbe, da lahko pravilno interpretiramo besedila, ki so nastala v njej. Tu je nadvse ustrezna misel Radojke Vrančič: Za prevajalca ni dovolj, da je dober lingvist, ampak mora biti tudi kos sociologa, etnologa, antropologa in še kakega drugega strokovnjaka za tisto družbo, tisto okolje, tisto dobo, o kateri govori njegov avtor. Če ta aspekt zanemarja, pride lahko do prav ljubkih pomot, nad katerimi se, če jih odkrijemo pri svojem bližnjem, seveda prav prisrčno in hudobno zabavamo. (Vrančič 1996: 77-78) Kako je torej lahko prišlo do te »ljubke pomote«? Pojdimo po vrsti: Jezikovni vir nesporazuma je a) redkost glagola razlepotiti, ki ga ni v Slovarju slovenskega knjižnega jezika, b) večpomenskost slovanskih glagolskih predpon in c) obstoj »lažnega prijatelja«, ruskega glagolapasyicpacumb, ki pomeni okrasiti, olepšati z mnogim. Razlepotiti je tako redka beseda, da sem sprva menila, da je to Kmeclov neologizem, vendar kaže, da le ni popoln hapaks, saj v korpusu Nova beseda najdemo še dva zgleda etnologinje Marjetke Kavčič Golež. Ta dva zgleda sta naslednja (kot ju najdemo v korpusu Nova beseda): ... zanika mitološkost Lepe Vide, jo razlepoti ... ... ta svet razlepoti, podre realizem ljudskega ... V SSKJ je pomen predpone raz- opisan takole: v glagolskih sestavljenkah izraža a) premikanje ali usmerjenost v več krajev, smeri', razgnati, raznesti, razvoziti ... b) dejstvo, da kaj ni več skupaj, v prvotnem položaju: ragrebsti, razriniti, razsuti ... c) delitev, ločitev na več delov, razcepiti, razdeliti, razrezati ... č) nastop stanja, navadno v veliki meri: razbesneti, razjeziti, razžariti ... d) doseg zaželenega namena, cilja: razgreti, razmehčati ... e) nasprotnost tega, kar pomeni glagol z drugo predpono: razbarvati, razčlovečiti, razelektriti, razorožiti, raztovoriti ... V okviru gornjih razlag bi lahko razlepotiti uvrstili pod pomen e). Nekoliko drugače opredeljuje skupine glagolov, tvorjenih z raz- A. Vidovič-Muha (1993: 179). Glagol razlepotiti bi po tej tipologiji verjetno sodil v skupino a) (ob levi in desni tožilniški vezljivosti), v kateri tvorjeni glagol izraža odvzem, prekinitev dejanja podstavnega glagola, npr. razbarvati les, razbremeniti avto, razočarati koga, razdediniti sina, razdružiti podjetje. V ruščini je opis pomena predpone раз- v Slovarju Ožegova in Svedove naslednji: 1. delitev na dele, razdelitev po površini, po različnih krajih: paздeлитb(razdeliti), разбросатБ (razmetati), разложитв (razložiti), размазатБ (razmazati); 2) okrepitev, intenzivnost dejanja: разукраситБ (okrasiti), разгуливатБ (razgrajati); ... 5) prekinitev dejanja, stanja: разлгобитБ (prenehati ljubiti); 6) nasprotno dejanje: разбинтоватБ (razvezati), разминироватБ (razminirati), размагнити (razmagnetiti) ... Vidimo torej, da so pomeni predpone раз- (oz. pomeni skupin tvorjenih glagolov) v ruščini precej podobni pomenom predpone raz- v slovenščini. Tudi tu smo priča precej običajnemu pojavu: glagolske predpone imajo v slovanskih jezikih več po -menov, te skupine pomenov so si precej podobne, vendar nikakor ni nujno, da nastopajo v dveh slovanskih jezikih predpone v enakih oz. podobnih pomenih pred genetsko ali pomensko ustreznimi glagoli. V ruščini obstaja glagol разукраситЂ, vendar nima pomena 'nasprotno dejanje od podstavnega glagola', ampak 'okrepljeno dejanje pod- stavnega glagola'. Približno tak pomen imajo v slovenščini glagoli razmehčati, razgreti ali razjeziti, razžariti. Pomen nasprotnosti tega, kar pomeni podstavni glagol, pa imajo glagoli разбинтоватћ, разминироватћ, размагнититћ. Разукрасит0 ima v slovarju Ožegova in Svedove oznako pogovorno in pomeni okrasiti obilno ali na mnogih krajih. Zgleda sta pa3yKpacumb новогодншго елку (okrasiti novoletno jelko) in разукраситћ рассказ n o dp 06 u o c m hm u (okrasiti pripoved s podrobnostmi). Ruska prevajalka je skoraj gotovo povezala glagol razlepotiti z ruskim pa3yKpacumb, čeprav potem v prevodu ni uporabila tega glagola. Ko je prevajalka pripisala glagolu razlepotiti nasproten pomen, tj. olepšati, je to gotovo storila pod vplivom na videz enako tvorjene glagolske sestavljenke v ruščini, pri kateri pa se je realiziral drag pomen predpone kot pri slovenski. Vendar mislim, da bi do te napake pri razumevanju ne moglo priti, če ne bi bilo v sovjetskem okolju tedaj še vedno vsaj deloma sprejemljivo oz. še ne čisto zastarelo prepričanje, da je industrializacija tako pozitivna vrednota, da lahko celo olepša kako pokrajino. Takega prepričanja nikakor ne pripisujem prevajalki osebno, o njenem mnenju ne vemo ničesar. Gre za to, daje prevajalka menila, daje tako prepričanje vsaj deloma še sprejemljivo. Možno je celo, da je menila, da je tako prepričanje možno v Sloveniji. Če tega ne bi menila, bi se ji zazdela ta interpretacija tako absurdna (kot se je zazdela meni kot bralki njenega prevoda), da bi nadalje raziskovala pomen glagola razlepotiti in bi se dokopala do pomena, ki ga je imel v mislih avtor izvirnika. Poglavitni razlog za ta resnično ponesrečeni prevod je torej poleg »lažnega prijatelja« premajhno poznavanje sistema vrednot in prepričanj v družbi izvornega besedila, še posebej seveda prepričanj in vrednot avtorja. Naslednji »lažni prijatelj«, ki bi ga rada opisala, se je pojavil v prevodu začetka znanega romana V. Jerofejeva Moskva-Petuški. Roman se v ruščini začne takole: Первое издание »Москва-Петушки«, благо бвшо в одном зкземшгаре, бБгстро разошлосв. (Ерофеев 1990: 14) Prevod pa je takle: Prva izdaja dela »Moskva-Petuški«, blago je bilo v enem primerku, je hitro pošla. (Jerofejev 1980: 5) V slovarju Ožegova in Svedove najdemo za благо dva slovarska članka, v prvem gre za samostalnik, v drugem pa za veznik: Благо 1. Blagor, sreča, dobrina (visoko). Благо 2. Veznik (pogovorno, pogosto ironično). Zahvaljujoč temu, da. Korekten prevod bi se torej moral glasiti: Prva izdaja dela »Moskva-Petuški« je zahvaljujoč temu, daje bila v enem primerku, hitro pošla. Prevajalec je torej prehitro menil, da ruska beseda благо po pomenu ustreza slovenski besedi blago in da je samostalnik. Po pravici rečeno njegov prevod ne moti kaj posebno in ni v kaki očitni opreki z nekakšnim »splošnim« smislom besedila, vendar bi bil pravilen prevod bolj tekoč, smiseln in bi bolje posredoval ironijo izjave, čeprav tudi pri dejanskem prevodu ni čisto izgubljena. Tu bi prevajalca pred napako morda obvaroval premislek o statusu besed z nepolnoglasnim korenom v ruščini. Besedo благо v ruščini lahko takoj prepoznamo kot cerkvenoslovanizem, saj ima v svojem korenu južnoslovansko cerkvenoslovansko nepolnoglasje. Fonetično avtohtona vzhodnoslo-vanska beseda bi imela polnoglasje in bi se glasila болого, vendar take besede v knjižni ruščini ni, obstajajo pa nekatere besede iz polnoglasnega korena v ruskih dialektih (Porohova 1988: 77). Načeloma velja, da imajo cerkvenoslovanizmi, ki so se v knjižni ruščini pojavili zaradi večstoletnega bivanja cerkvene slovanščine v funkciji kulturnega jezika na vzhodnoslovanskih tleh, v ruščini abstrakten, vzvišen, duhoven pomen in točno to vidimo pri благо, ki ima pomen blagor, dobrina sreča in ima kvalifikator visoko. Uporablja se v zvezah kot so благо народа (blagor ljudstva), длп блага Родинм (za blagor domovine) itd. Kadar obstajajo v ruščini besede s polnoglasnim in nepolnoglasnim korenom, imajo ponavadi prve konkretnejši, bolj vsakdanji pomen, druge pa abstraktnejši, vzvišenejši. Tako imamo v pomenu 'zdravje' besedo 30opoebe, 'zdravstveno varstvo' pa je здравоохранение, v pomenu 'glas' imamo голос, 'soglasnik' pa je согласнип, 'mlad' je молодоп, 'mladogramatiki' pa so младограмматики. Skratka, besede s koreni z nepolnoglasjem, ki so po zunanji podobi taki kot v slovenščini, imajo v ruščini ponavadi vzvišen, knjižen, poetičen pomen. Slovenščina je tako kot stara cerkvena slovanščina južnoslovanski jezik in ima seveda v vseh teh besedah avtohtono nepolnoglasni koren ter ne pozna take stilistične oz. semantične diferenciacije med besedami s polnoglasnim oz. nepolnoglasnim korenom. Zato bi prevajalec lahko postal nekoliko sumničav, ko je pripisal ruski besedi z nepolnoglasnim korenom povsem konkreten, trgovski pomen slovenske besede blago. Pa še tretji »lažni prijatelj«, tokrat iz znamenitega romana Mojster in Margareta M. Bulgakova. Na začetku romana Berlioz razlaga pesniku Ivanu Bezdomnemu, ki mu je bilo naročeno napisati antireligiozno pesnitev, naslednje: Нет ни однои восточнои религии, - говорил Берлиоз, - в которои, как правило, непорочнал дева, не произвела 6bi на свет бога. (Булгаков 1969: 14) Prevod se glasi: »Prav nobene vzhodne vere ni,« je govoril Berlioz, »da ne bi v njej kot po pravilu neporočena devica spravila na svet boga.« (Bulgakov 1977: 9-10) Tu je lažni prijatelj pridevnik иепорочш, ki ne pomeni neporočen, ampak brezmadežen, brezgrešen,uporablja se npr. v zvezi непор очт e зamrnue 'brezmadežno spočetje'. ГЂ p o к pomeni v ruščini greh, napaka, порочш pa grešen, nenraven, napačen. V ruščini so besede v zvezi s poroko: c ead m (poroka, svadba), вешлше (cerkvena poroka), etno d umb з ам/ ж (poročiti se, omožiti se), ж ш mu: n (poročiti se, oženiti se). 'Neporočena' je нгзалу жшл. Pravilen prevod bi bil: »Prav nobene vzhodne vere ni,« je govoril Berlioz, »da ne bi v njej praviloma brezmadežna devica rodila boga.« Skratka, prevajalec je bil prenagel v sklepanju o sorodnosti pomenov na podlagi zunanje podobnosti. Spet moram poudariti, da je prevod na splošno dober, pa tudi v navedenem citatu pravzaprav ni v opreki s splošnim smislom. Čeprav sem se tudi sama prav prisrčno in hudobno zabavala, ko sem odkrila omenjene napake, poudarjam, da so vsi trije prevodi kot celota dobri. Služijo pa kot opozorilo študentom, da naj pri prevajanju čimvečkrat preverjajo pomene besed v slovarjih in skušajo pridobiti čim več vedenja o jeziku, kulturi, kontekstu in okolju, iz katerega prevajajo. Viri in literatura M. Булгаков: Macmep u Маргарита. Frankfurt/M.: Посев, 1969. M. Bulgakov: Mojster in Margareta. Ljubljana: Cankarjeva založba, 1977. B. Ерофеев: Москва-Петушки. Москва: ИздателБСтво СП «Интербук», 1990. Venedikt Jerofejev: Moskva-Petuški. Ljubljana: Cankarjeva založba, 1980. Matjaž Kmecl: Zakladi Slovenije. Ljubljana: Cankarjeva založba, 1979. МатБНЖ Кмецл: Сокровигца Словении. Лгобллна: СанкарБева заложба, 1990. C И. Ожегов и Н. К>. Шведова: Толковшп cAoeapb русского лзика. Москва: Азђ, 1993. О. Г. Порохова: Полногласие и неполногласие в русском литературном лзшке и народнмх говорах. Ленинград: Наука, 1988. Radojka Vrančič: Iskanje slovenskih rešitev za Iskanje: hvalnica praksi. V: Prevod besedila/Prevajanje romana/20, prevajalski zbornik. Ljubljana: Društvo slovenskih prevajalcev, 1996, str. 77-80. (Citat iz diplomske naloge Milanka Ilić: Harmsova starka v prevodih. Stilistična analiza. Oddelek za slavistiko. FF. Ljubljana 2006, str. 6.) A. Vidovič-Muha, Glagolske sestavljenke - njihova skladenjska podstava in vezljivostne lastnosti. SR 41 (1993), 1,161-192. Nova beseda: hhtp://bos.zrc-sazu.si/s_beseda.html Abstract THREE INSTANCES OF INCORRECT TRANSLATIONS DUE TO FALSE FRIENDS Some examples of incorrect translation from Slovene to Russian resp. from Russian to Slovene are discussed in the paper. All mistakes are due to »false friends«, especially frequent in genetically related languages. An interesting mistake can bi found in a translation from Slovene to Russian. The author used a rare verb razlepotiti in the meaning of 'taking away the beauty' when speaking of what happened to a valley after industrialization. The Russian translator connected the verb razlepotiti because of the structural similarity with the Russian verb pa3ytcpacumb, which means exactly the opposite: to decorate, to adorn. It is suggested in the paper that this mistake wouldn't happen if the translator payed more attention to values prevailing in Slovenia in the period when the book had been written. In the 70ties (and of course later) the idea of industrialization embellishing a valley was inconceivable. Eva Sicherl, UDK 8Г367.633 : 8Г25 University of Ljubljana Andreja Žele, Scientific Research Centre of the Slovenian Academy of Sciences and Arts CQMBINABILITY WITH PREPOSITIONS IN SLOVENE AND ENGLISH: THEORETICAL AND CONTRASTIVE VIEWS The present article aims to shed some light on the combinability of verbs, nouns and adjectives with prepositions - a phenomenon which is common both in Slovene and in English, but has not been extensively commented on by Slovene linguists. In Anglo-American linguistics, such combinations with prepositions are basically divided into two groups, which tend to be named differently by authors of different linguistic "orientations". Keywords', semantic- and structural-syntactic aspect of valency, combinability with prepositions, prepositional free morpheme, prepositional collocation, free combination 0 Introduction Valency theory, originating in dependency grammar as developed by Tesniere (1959) and particularly developed in the 1960s and 1970s by German linguists (e.g. Heibig, Schenkel, Heringer, Schumacher, Engel), has also found its place in Slovene linguistics. A rather complex multi-level treatment of valency from the semantic- and structural-syntactic aspect (with an original orientation from form to meaning and vice versa and with account taken of transformational grammar linguistics) is found as early as the second half of the 1970s in Toporišič (1976). Approximately at the same time valency in Slovene was dealt with by the French author Vincenot in his grammar Essai de Grammaire Slovene (1975). In the following years, valency was treated in more detail and depth by Toporišič (1982), as well as by Dular (1982), Križaj (1981, 1982, 1989), and Vidovič Muha (1993), who all researched the transformational grammar aspect of valency more intensively. Kunst Gnamuš (1981) and Orešnik (1992) have contributed to shaping the theory of Slovene valency with a more exact treatment of the semantic composition of the utterance and by presenting semantic orientation. These two works rely theoretically and methodologically on Anglo-American studies1 and treat valency within the framework of the semantic level. Mainstream Anglo-American linguistics has largely neglected the findings of valency theory, but has often gone parallel ways to describe the same phenomena (cf. Herbst and Roe 1996: 179-180). Nevertheless, several attemps have been made to apply the concept of valency to English (e.g. Emons 1974,1978), Matthews (1981), Allerton (1982), Herbst (1983, 1988), Somers (1984), von Randow (1986)). One aspect of valency pertains to complementation of verbs (but also of nouns and adjectives) by prepositions. The present article aims to shed some light on this phenomenon, which is common both in Slovene and in English, but has not been extensively commented on by Slovene linguists. In Anglo-American linguistics, such combinations with prepositions are basically divided into two groups, which tend to be named differently by authors of different linguistic "orientations" (cf. Herbst and Roe 1996: 180), yet most of them seem to agree that the key issue lies in the obligatoriness of the preposition and its dependence on the lexical word. Most speakers of either of the languages thus tend to "feel" there exists a difference between sentences like: John lectures on || semantics. Janez predava o || semantiki. John lectures || on Fridays. Janez predava || ob petkih. In the first sentence, the link between the lexical word (verb) and the following preposition is felt to be stronger than is the case with the combination in the second sentence. In the latter sentence, the phrase boundary comes after the verb, and the preposition with the following noun forms a (prepositional) phrase. The latter combinations are usually referred to as free combinations of lexical words with prepositions; they are constructed in accordance with general syntactic rules and freely allow substitution (e.g. John lectures || in the classroom, John lectures || from five to six thirty / Janez predava || v razredu, Janez predava || od petih do pol sedmih). The former combinations may be referred to as collocations2 consisting of a lexical word (a verb/noun/adjective forming the base of the unit) and a preposition (functioning as the collocator); this type of word combination could also be named 'prepositional collocation' (cf. Sicherl 1999, 2004). Prepositional collocations can thus be defined as typical, recurrent combinations of verbal, nominal or adjectival bases with prepositional collocators.3 However, the distinction between the two groups is not always clear cut - while some cases may be immediately classified as either prepositional collocations or prepositional free combinations, others may prove to be extremely unclear and vague, and there are many borderline cases which may satisfy some of the syntactic criteria4 for inclusion among prepositional collocations but not others. Semantic criteria for dis- 2 Hausmann (1985: 178) defined collocations as "typical, specific and characteristic two-fold relationships between words" (typische, spezifische und charakteristische Zweierbeziehungen von Wörtern). Characteristically, collocations are not produced creatively by the speaker as are free combinations; rather, they are retrieved from the speaker's memory as combinations forming a unit, as some kind of language "semiproducts". Another typical trait is a relationship of affinity between the two words forming a collocation, so that they often appear together (see also Hausmann 1984: 398 and Bahns 1996: 24). In The BBI Dictionary of English Word Combinations such combinations are referred to as a subclass of grammatical collocations (see categories Gl, G5 and G8D in Benson, Benson, Ilson 1997: xvi, xviii, xxi). These criteria were mainly defined by British and German linguists (for further details see, for example, Quirk et al. 1985: 1163 ff, Palmer 1988: 229-231, Schröder 1986: 13-22, Helbig and Schenkel 1975: 40^19, Lerot 1982: 263-265), and include the choice of the question form, possibility of passive transformation, possibility of cutting the prepositon off from the lexical verb in relative clauses and wh-questions, substitutability of preposition, etc. tinguishing between free combinations and prepositional collocations seem to be even more elusive,5 yet they often prove decisive for inclusion in one group or the other, especially when combined with the syntactic ones (see also Quirk et al. 1985: 1163). The distinction between the two categories remains to a certain extent arbitrary and subject to gradience. The existence of different semantic-syntactic roles of prepositions in Slovene, however, was commented on already in the first Slovene grammar by Bohorič (1584); the author felt the preposition to be gravitating towards the verb in some cases, and towards the adverb in others.6 A survey of later Slovene grammars shows that the treatment of prepositions under either verbal or adverbial government is not accidental; rather, it suggests the authors' language feeling which made them treat governed prepositions as part of the verb, i.e. as free morphemes which help to form the meaning of the verb7, or, in other words, they perceived the entire combination as a prepositional collocation. However, the fact that Slovene as a synthetic language realizes cases by means of case inflections also has to be taken into account, and prepositional free morphemes can thus also be used to mark the case in the following free combination. 1 Valency and Combinability with Prepositions Combinability with prepositions is related to valency on the morpho-syntactic level. Thus, the prepositional free morpheme (as described by Helbig 1984, 1992) which combines with a content or lexical word (usually a verb, but also a (morphologically related) adjective or noun) can be said to primarily modify the meaning of the lexical word it co-occurs with. The preposition is used to "direct" the verbal (adjectival or nominal) meaning; however, the morphological extension (the addition of a preposition to a lexical word) can also introduce new lexemic meanings, which results in new lexemes. The following examples may illustrate this point: 5 Naturally, since the borders between the two groups remain fuzzy, no hundred-percent agreement can be expected between different analysts as to the inclusion of certain cases in either of the two groups. Thus, for example, the degree of metaphoricalness or idiomaticity may influence our decision. If "disagree with " as used in the sentence Orange juice seems to disagree with some babies, is perceived as a prepositional collocation, does the same hold true for "disagree with " in the sentence I respect the president but I disagree with his decision ... (illustrative sentences taken from Collins Cobuild English Dictionary 1995: 464), or is the latter a free combination? It seems that the verb "disagree " when used with an inanimate subject becomes semantically specialized and is therefore often treated differently from "disagree" with an animate subject. In such cases obviously the context, and not the word combination itself, causes semantic modification which may result in a different classification. Other factors which may influence our analysis, apart from the combination of strictly semantic and syntactic ones, are also usage and frequency. " Bohorič (see Toporišič ed. 1984: 258) used the terms "dostavljen" and "pristavljen" and classified prepositions according to the cases they govern based on Latin patterns, quoting examples such as od Buga, pres njega, s'Bugom, jkrivjhipredozhetom. The preposition, according to Bohorič, is either syntactico-semantically closely related to the verb ("dostavljen") or just added to it ("pristavljen"). Centuries later, Toporišič (1976) introduced the terms "vezavnovezljiv" and "vezavnodružljiv". 7 In his description of the Slovene language, Gutsman (1777: 110-111) makes a distinction between sentences like Je v'Zelovzi rojen, Mi ßno vu Velikouzi prebivali, Eni Je v'DoberleveSs, eni pak v'Tershizh popelajo and sentences such as On je sa frednika med nami biu pofiaulen, On bo sa mießnega Jbdnika svolen gratau. Počakati moram. I need to wait. Počakati moram na šolski avtobus. I need to wait for the school bus. Tone je prišel, ('prispeti') Tony has come, ('arrive') Tone je prišel do odločitve, ('odločiti se') Tony has come to a decision, ('decide') While the Slovene language is generally oriented into using the direct accusative, it tends to express itself more accurately and analytically by using combinations with prepositions (cf. Žele 2001). English as an analytic language seems to depend on prepositions to an even greater degree; there prepositions are, like other grammatical items, primarily used to mark the syntactic relation between lexical words forming a construction (cf. also Lyons 1968: 435-438 and Collinge 1990: 144). However, the semantic distinction between the illustrative sentences cited below seems to be identical in both languages: Streljali so zajce. They shot rabbits. Streljali so na zajce. They shot at rabbits. When the direct accusative is used with the verb shoot!streljati, the sentences imply that the rabbits were either killed or injured. When the verb is combined with the preposition atlna, the preposition introduces the meaning of 'orientation/direction' of the shooting, without implying the result of the action. When used in combinations, prepositions form part of the semantic content of verbs, nouns and adjectives (particularly those nouns and adjectives which are morphologically related to the corresponding verbs) they combine with. When a preposition enters into a combination with a lexical word, it either helps form the meaning of the entire combination (resulting in lexicalization8 of the whole) or merely adds some stress to the lexical word. Zivi skupaj s psom. ('deliti si življenjski prostor') He lives with his dog. ('share a living place with') Živel je z njo, še preden sta se poročila, ('živeti v partnerstvu') He lived with her before they got married. ('live together as if married') A practical and useful differentiation between free combinations with prepositions and grammatical, or, to be more precise, prepositional collocations taken from Anglo-American linguistics may be parallelled to the classification of combinations with prepositions as used by Slovene linguists. However, the distinction between the two Q The term lexicalization is used here to denote that the preposition becomes part of the meaning of the lexical word; a syntactic unit consisting of a lexical word and a preposition becomes a dictionary unit, a multi-word lexeme. categories is subject to gradience. Prepositional collocations can be regarded as results of natural semantic- and structural-syntactic combining of words, as mutual semantic and syntactic expectancy of words.9 When used in collocations or in free combinations, the prepositions usually express those semantic components which tend to participate in valency both semantically and syntactically. Prepositions thus help express intentional/valency-related semantic components of verbs, nouns or adjectives also in structural-syntactic terms. Prepositions express more clearly the mutual dependency between the semantic component of a verb/adjective/noun and its valency, or, in other words, they show the reverse effect valency has on the semantic structure of a verb/adjective/noun.10 It needs to be stressed, however, that in prepositional collocations the two-way semantic-syntactic affinity observed between the lexical word and the preposition (cf. Sicherl 2004: 44) is not equally strong: the one leading from the dominant lexical word to the preposition is felt to be much stronger than the one leading from the preposition to the lexical word. The mutual expectancy therefore seems to work more strongly from the lexical word to the preposition than vice versa. The following pairs of Slovene and English equivalents may illustrate this point: povabiti na (zabavo) invite to (a party) jezen na (mamo) angry at (Mum) prepir o (politiki) argument over (politics) Naturally, such mutual semantic-syntactic ties are to be found in other languages as well and are not characteristic of Slovene or English only. However, they make a particularly interesting subject of study when viewed contrastively. While free combinations with prepositions seem relatively straightforward and unproblematic when two languages such as English and Slovene are contrasted11, the mutual semantic-syntactic ties existing between the verbal/nominal/adjectival base and the prepositional collocator in collocations are more interesting to observe, particularly when the same base can be combined with several different prepositions with little or no change in meaning. The addition of the prepositional collocator stresses a certain semantic component, and, usually, also introduces valency. Thus, for example, the Slovene preposition na introduces the semantic component of 'expectation' in combinations such as pripraviti se na, čakati na, upati na, etc.; in English, there are prepositional 9 In linguistics, the mutual semantic and syntactic expectancy of words could only be subjected to adequate research in the last decades of the 20t'1 century when extensive computerized corpora of over ten million words enabled analysts to collect relevant data on the basis of which a certain word combination could be classified as a collocation with more certainty. However, statistical data cannot be the only criterion for the inclusion of a combination among collocations. 10 Prepositions combined with verbs, particularly those lexicalized, are treated in the Slovar slovenskega knjižnega jezika (1975-1991) in the form of notes on grammatical patterns: "s predložnim povedkovim določilom" [used with a prepositional subject complement]; "s predložno zvezo" [used with a prepositional phrase]; "s predlogom" [used with a preposition]; "v zvezi z/s" [followed by the preposition s/z]. For more on this, see Sicherl (1999). collocations with for carrying a similar meaning, as in prepare for, wait for, hope for, etc. And, similarly, German equivalents of those bases tend to combine with the preposition auf io introduce 'expectation' as in sich vorbereiten auf, warten auf, hoffen auf, etc. Pripravil se je na izpit. He prepared for the exam. Er bereitete sich auf das Examen vor. Čakal je na vlak. He waited for the train. Er wartete auf den Zug. Upal je na pozitiven rezultat. He hoped for a positive result. Er hoffte auf ein positives Resultat. Another example to illustrate the above point is provided by prepositional combinations in which the verbal/nominal base carries the meaning of 'intellectual activity' while the combining preposition is used to introduce the 'theme/content' of this intellectual activity. The preposition used in Slovene in these combinations is typically o, in English it is substituted by on (with alternative combining options possible in several cases, such as about, concerning, over, of) with bases of the same or a similar meaning, while German tends to use the preposition über with most equivalents: Prosiii so ga, naj spregovori o pomembnosti umetnosti v izobraževanju. He was asked to talk on the importance of arts in education. Er wurde gebeten, über die Bedeutung der Kunst im Unterrricht zu sprechen. Napišite spis o napredku civilizacije. Write an essay on the advance of civilization. Schreiben Sie einen Aufsatz über den Fortschritt der Zivilisation. 2 Classification of Prepositions Used in Combinations Prepositions used in combinations have developed out of spatial (and temporal) adverbs which define an action more accurately, so it is not surprising that their primary semantic-syntactic position is next to the verb. In Slovene linguistics, verbal-prepositional combinations can be classified into the following three types (according to Žele 2001):12 a) lexicalized verbal-prepositional combinations b) non-lexicalized verbal-prepositional combinations c) verbal-prepositional free combinations A preposition is lexicalized when it semantically co-forms the meaning of the verb. 12 1 Enciklopedija slovenskega jezika (Toporišič 1992: 230,351) distinguishes between "predložni vezavni morfem" [prepositional morpheme] and "prosti morfem" [free morpheme]. See also Švedova (1980: 156) for Russian. Žic Fuchs (1991: 123) mentions J.M. Anderson's classification into "inherent lexical content", which corresponds to the lexicalized verbal-prepositional combination in semantic-syntactic terms, and "derived lexical content", the latter corresponding to non-lexicalized verbal-prepositional combination in semantic-syntactic terms. The preposition in such a lexicalized combination thus behaves as if it were an "inherent" part of the verb (cf. also Dixon 1991: 268); it becomes part of the verbal lexeme as a dictionary unit and, simultaneously, part of this verb's valency. A preposition is said to be non-lexicalized when it is semantically related to (usually) only one sense of the verbal lexeme and expresses its semantic- and structural-syntactic valency. Delal je na projektu, ('ustvarjati') lexicalized verbal-prep. combination He worked on the project, ('have sth as the subject of thought') prep, collocation Delal je na polju, ('obdelovati, kmeto-vati') non-lexicalized verbal-prep. combination He worked in the fields, ('grow crops') free combination A lexicalized preposition is bound to the entire meaning of the multi-word unit (dictionary unit) it co-forms with the lexical word, a non-lexicalized preposition is bound only to one sense of the dictionary unit. The differentiation between lexicalized and non-lexicalized prepositions in Slovene can also be illustrated by delexical primary verbs. The preposition is lexicalized in the following combinations where the verb is delexical, or rather, does not carry a full lexical meaning. Semantically equivalent English sentences do not necessarily use a prepositional combination. Bil je ob hišo. ('izgubiti') He lost his house. Imel jo je za pametno, ('ceniti kot') He regarded her with respect. Dala je na njegov nasvet, ('upoštevati') She paid attention to his advice. Below, the preposition is non-lexicalized; it is combined with the verb with a full lexical meaning. Again, parallel semantically equivalent English combinations are given, yet they do not always use a combination with a preposition: Bil je ob hiši. ('nahajati se') He was next to the house, ('stand') Prihranke je imel za slabe čase. ('hraniti') He kept his savings for hard times, ('set aside') Krožnike je dal na mizo. ('položiti') He put the plates on the table, ('place') Analyzing the above combinations from the point of view of the collocation - free combination dichotomy, the non-lexicalized are mainly free combinations, while the lexicalized can pass for collocations. 2.1 The lexicalization of the preposition in combination with a full (specialized) non-primary verb can neutralize the meaning of that verb. Along with the neutralization of the original, specialized verbal meaning, the addition of a preposition results in transitivization of the lexicalized unit, and often gives the combination an additional tenor in terms of register, formality, etc. Tista kača se je že levila. That snake has already shed its skin. Janez se levi v pravega podjetnika. John is turning into a proper businessman. Pri izgovorjavi težkih besed seje zapletel. When pronouncing difficult words he stumbled. Zapletel se je v umazane posle. He got mixed up in some shady business. With some other verbs, the addition of a preposition merely changes the valency of the verb in semantic- and structural-syntactic terms. Kokoš vali jajca. The hen hatches her eggs. Vedno je valil krivdo na koga drugega. He always shifted the blame onto somebody else. Ponudba še velja. The offer still holds. Velja za poštenjaka. He is believed to be an honest man. Syntactic characteristics of lexicalized prepositional combinations: 1) The preposition introduces rightward valency a) when combined with originally intransitive verbs Prišlo je do vojne. It came to war. Spustil se je v dolgo razlago, zakaj me mora videti . He went into a long explanation of why he had to see me. b) when combined with primary verbs Se boš moral delati na šibkih točkah svoje slovenščine. You will have to work on the weak points in your Slovene. Nenadoma je bil ob vse svoje bogastvo. He suddenly lost all his fortune.13 c) when combined with verbs of motion; the preposition in these combinations retains the meaning of orientation, and gives a secondary, metaphorical meaning (of phaseness of action/characteristic/state or of relation of action) to the verbal base. Oddaljil se je od teme. He digressed from the subject. Planili so po sovražniku. They fell on the enemy. Dan se je prevesil v noč. The day turned into night. Nehaj dregati vame! Stop picking on me. Pes se je spravil nad lastnega gospodarja. The dog turned on his own master. 1 T J The English equivalents of the Slovene illustrative examples of lexicalized prepositional combinations given here may not be constructed on the same pattern and may lack a prepositional combination altogether. 2) Combinability with prepositions can also result in generalization of the verbal meaning, as in: a) purpose(fulness)/intensity of action/state b) conclusion of action c) excitement or diligence d) intrusiveness e) active/passive relationship or acquisition or use a) Moral se bo vreči na delo. Zakopal se je v študij. He will need to bury himself in his work. He dug into his studies. b) Vstala je od mize. She rose from the table. c) Zagorel je za novo idejo. He burned with excitement for the new idea. d) Tiščali so vanj, pa jim ni povedal. They plied him with questions, but he refused to tell them anything. e) Zivi od pisanja knjig. She manages to live on her writing. 3) The preposition can change the meaning of an originally transitive verb and thus also its rightward valency: Konj je vlekel voz. The horse pulled the cart. Barva je vlekla na zeleno. The colour was greenish. 4) The preposition causes semantic transition to a non-spatial evaluation of the state: Nagibal se je čez ograjo. He leaned across the fence. Nagibajo se k praktičnim rešitvam. They tend to lean towards practical solutions. 5) The preposition stresses a characteristic of an actant: Visel je med življenjem in smrtjo. He hovered between life and death. Fantje so se kar lepili nanjo. The boys were really stuck on her. 2.2 A non-lexicalized preposition in a combination retains its spatial meaning or has at least some orientational meaning. As part of verbal valency, a non-lexicalized preposition can be obligatory and introduce the so-called "partially obligatory verbal valency" (cf. also Samardžija 1986: 119). Od včeraj leži v bolnišnici, ('biti hospitaliziran') He's been in hospital since yesterday, ('be hospitalized') It is precisely within the group of non-lexicalized prepositional combinations that the boundary between prepositional collocations and free combinations with prepositions as known from Anglo-American linguistics can be drawn. However, this bound- ary is often hard to define and remains fuzzy. Nevertheless, it can be said that in most cases non-lexicalized combinations with prepositions either introduce or retain valency and form some kind of non-idiomatic prepositional collocations. Syntactic characteristics of non-lexicalized prepositional combinations: 1) The preposition introduces new participant roles, making what was originally a rather general meaning of the verb more precise and concrete: a) prepositions combined with verbs of oriented control, transfer, introduction/continuation, limitation, prevention tend to have an orientational meaning Poslal je sporočilo. He sent a message. Poslal je sporočilo po elektronski pošti. He sent a message by e-mail. Nadaljeval je. He continued. Nadaljeval je z razpravo. He continued with the discussion. b) prepositions combined with verbs of motion retain a spatial meaning of aim or target Zdrsnil je. He slipped. Drsel je v depresijo. He sank/slid into depression. c) prepositions may render the meaning of end of state Prebudil se je. He woke (up). Potrebuje nekaj časa, da se prebudi iz nočnega spanca. It takes him some time to wake from his night's sleep. 2) Prepositions combined with prefixed verbs of motion in Slovene tend to be homonymous with the prefix used in the verb; such prepositions introduce a certain "orientational valency": Vstopil je v stranko. He joined the party. Sonce zahaja za gorami. The sun sets behind the mountains. 3) The preposition can act as a permanent (stylistically neutral) free morpheme combining with a verb, the two making up a prepositional collocation. These combinations are particularly interesting to compare when the preposition is either used explicitly to express spatial meaning (the combination is non-lexicalized) or has a changed, figurative meaning (the combination is lexicalized). Njegova parcela meji na športno dvorano. ('biti na meji z' - non-lexicalized) His plot of land borders on the sports hall. Njeno obnašanje meji na blaznost, ('biti zelo podoben' - lexicalized) Her behaviour verges on madness. 4) The preposition is optional with some verbs whose meanings already include 'purpose' or 'aim'; in the following illustrative sentences the Slovene prepositions na and za can be omitted without a change in meaning: Čakam (na) šolski avtobus. I wait for the school bus. Igral je (na) flavto. He played the flute. Pazila je (na) otroke. She looked after the children. Pritisnil je (na) gumb. He pressed the button. Lovil je (za) njegov rokav. He grabbed at his sleeve. In the above combinations, Slovene allows for variant forms (prepositional and non-prepositional), which testifies to the relative weakness of the link between the verb and the preposition. Consequently, this makes an ellipsis possible. A similar phenomenon can be observed in English, but usually a (slight) change of meaning is observable there: He grabbed his sleeve, ('take something suddenly and roughly') He grabbed at his sleeve, ('try to take/grab something') He kicked the door, ('hit something forcefully with one's foot') He kicked at the door, ('aim one's foot at something') 3 The Influence of Prepositions on the Formation of Verbal Valency Groups in Slovene14 In the following classification, the entire spectrum of combinations with prepositions as they appear in Slovene can be found, ranging from prepositional collocations proper to free combinations with prepositions. Again, their semantic equivalents in English may not be constructed on the same pattern and may lack a prepositional combination altogether. 3.1 With specialized verbs denoting physical or mental states the preposition stresses the state itself and/or the resulting characteristics Norčujejo se iz starega učitelja. They make fun of the old teacher. Dvomi v njene sposobnosti. He doubts her abilities. 3.2 Combined with specialized verbs denoting action/operation/creation, the preposition can express the prevalent semantic component, which is also the intentional/valency-related semantic component. The preposition thus introduces: 14 For participant roles in the Slovene language see Orešnik (1992). a) enabling the originating/origin of something Organizirajo kmete za sodelovanje z zadrugo. They are helping to organize farmers to work for the co-operative. b) a stressed semantic component of motion Odstranil je navlako z dvorišča. He removed the junk from the yard. c) a stressed semantic component of co-occurrence/co-appearing/appurtenance Pacienti čakajo (na) zdravnika. The patients are waiting for the doctor. Čaka na ugodno priliko. He's waiting for his chance. Sodelujejo z različnimi organizacijami. They cooperate with various organizations. d) a stressed semantic component of a change of property Pretvarjal se je v čudaka. He was turning into a loony. 3.3 Combined with specialized verbs of communicating/thinking/understanding, the preposition introduces: a) emphasis on 'receiving and having information' Dokumentiral je podatke s tabelami. He tabulated the data b) emphasis on 'understanding and reacting to information' Projekt je argumentiral s številkami. He backed up the project with statistics. c) emphasis on 'giving information' Agitirali so za kandidata. They canvassed for the candidate. 3.4 Combined with specialized verbs with a general meaning of change, the preposition introduces emphasis on 'self-movement'. Navijači so se pulili za vstopnice. The fans scrambled for tickets. 3.5 Combined with specialized verbs of movement, the preposition stresses: a) the course of movement when used with a process verb; b) the goalness of movement or purpose when used with a goal-directed verb; c) the content of the event when used with an event verb. a) Vozi se v šolo. He drives to school. b) Redno hodi na obiske. He visits them regularly. c) Rad se vrača k domačim. He likes to go back to his family. 4 Contrastive View of Prepositions Used in Combinations A contrastive treatment of combinability with prepositions and valency patterns in two or even three different languages may often help shed some light on the characteristics of this phenomenon, outline potentially difficult areas in foreign-language learning as well as suggest techniques which may help sensitize users of a foreign language to divergences between their mother tongue and the foreign language. While native speakers of a language may find prepositional collocations quite unproblematic as word associations of the collocational type are relatively uniform within one language community (cf. Herbst 1996: 389-390), these have proven to be one of the most difficult areas of linguistic knowledge to master for non-native speakers. To native speakers, collocations allow a certain degree of syntagmatic predictability which they have gradually learnt to master; however, non-native speakers, even very advanced ones, often lack this type of knowledge and end up combining their words in an untypical and unusual way. On the whole, it seems that Slovene learners of English, after having reached a certain level of proficiency, tend to have less difficulty producing free combinations than prepositional collocations (cf. Sicherl 1999)15. The reason for this may lie in the (changed, usually abstract and hardly definable) content of the preposition when this is used as collocator. During recent decades linguists began to realize that even in the case of prepositional collocations we cannot talk of "semantic emptiness" (see Schröder 1986: 10), and that the choice of the preposition is semantically motivated (cf. also Dixon 1991: 271). It has become clear that in some collocations there is a possibility of variation in the preposition (e.g. talk of/about/on) without a radical change in meaning. That the preposition when used as a collocator acts, at least to some extent, as a carrier of meaning, can also be proven by valency patterns which the collocational bases enter. If these prepositions were entirely meaningless, we would probably not tend to combine meaning-related content words (collocational bases) with identical prepositions. However, in most cases, this is precisely what we do, and the emerging valency patterns turn out to be surprisingly uniform in the languages compared. To illustrate this point, let us compare instances of prepositional collocations consisting of a verb/noun/adjective denoting sensations perceived through the nose by the olfactory nerves, and the following preposition which introduces a substance emitting an odour16. We can see that in English practically all these bases regularly combine with the preposition of, similar combinability with a preposition seems to be present in ^ Typically, Slovene speakers of English opt for that English preposition to combine with a lexical word which is "nearest" to the prototypical meaning of the Slovene preposition used in the semantically equivalent Slovene combination. Since the English preposition for is felt to be the nearest semantic equivalent of the Slovene preposition za, they tend to "translate" the combination značilen za as * typical/characteristic for and not as typical!characteristic of. Even when there is a choice between two prepositions in English with little or no change in meaning, they usually select the one that is felt to be semantically nearer to the Slovene preposition: thus, an accidental sample of 33 advanced students of English at Ljubljana University preferred the combinations fair to (63%), suited for (91%) and rich with (47%) to the combinations fair on (6%), suited to (3%) and rich in (44%) as translation equivalents of the Slovene combinations pravičen do (otrok), primeren za (delo) and bogat z (vitamini). 16 The illustrative sentences below were taken from various sources found in the Internet and simplified by the authors. other languages as well. Below, comparison has been made with Slovene and German, where most verbal, nominal, and adjectival semantic equivalents of the English bases regularly combine with the prepositions po and nach respectively. His room smelled of his dog. V njegovi sobi je dišalo po psu. Das Zimmer roch nach seinem Hund. The pale blue flowers scented of almonds. Svetlomodri cvetovi so dehteli po mandljih. Die blauen Blüten dufteten nach Mandeln. My clothes were stinking of tobacco. Moja obleka je smrdela po tobaku. Meine Klamotten stanken nach Tabak. The cabin reeked of kerosene. V kabini je zaudarjalo po kerozinu. Im Cockpit roch es nach Kerosin. An odour of lemons was left in the room. V sobi je ostal vonj po limonah. Im Zimmer blieb der Duft nach Zitronen. The male was attracted by the natural scent of a female. Samca je privabil duh po samici. Der Geruch nach Weibchen zog das Männchen an. Try to remove the smell of vomit from your carpets. Poskusite odstraniti smrad po bruhanju iz preprog. Versuchen Sie, den Gestank nach Erbrochenem aus den Teppichen zu entfernen. The aroma of sizzling bacon filled the kitchen. Kuhinjo so napolnile vonjave po cvrčeči slanini. Das Aroma von brutzelndem Speck füllte die Küche. I wish I could bottle the fragrance of magnolia blossom as perfume. Rad bi ustekleničil vonjavo magnolijinih cvetov kot parfum. Ich wünsche mir, ich könnte den Duft nach Magnolienblüten als Parfüm in Häschchen abfüllen. The stench of burning rubber was in the air for days. Smrad po zažgani gumi je še več dni ostal v zraku. Der Gestank nach verbranntem Gummi lag noch tagelang in der Luft. His lungs were filled with the malodour of diesel fumes. Pljuča mu je napolnil smrad po dizelskem gorivu. Seine Lungen füllten sich mit dem Gestank von Dieseldämpfen. The old pharmacy was odorous of herbs and flowers. Stara lekarna, dišeča po zeliščih in cvetju ... Die alte Apotheke, wohlriechend nach Kräutern und Blüten ... Her clothes were redolent of moth balls and cedar closets. Njena oblačila, vonjajoča po naftalinu in omarah iz cedrovine ... Ihre Kleider, riechend nach Mottenkugeln und Zedernholzschränken ... She ran out of the dirty room reeky of stale cigarettes. Pobegnila je iz umazane sobe, zaudarjajoče po cigaretnem dimu. Sie rannte aus dem schmutzigen, nach Zigarettenrauch stinkendem Zimmer. Its blooms, deliciously scented of orange blossom ... Cvetje, čudovito dišeče po pomarančnih cvetovih ... Die Blumen, herrlich duftend nach Orangenblüten,... Naturally, one can find exceptions (e.g. the noun "aroma" in Slovene regularly combines with the non-prepositional genitival case, as in aroma sveže kave/the aroma of fresh coffee; the German nouns Aroma and Gestank in the above examples are combined with the preposition von if the source of the odour is immediately present), yet the pattern which emerges from the illustrative examples above seems distinct enough. The finding that prepositions in collocations are not devoid of content, but contribute a certain content to the meaning of the collocational base, even if their content is even more abstract and difficult to define than that of prepositions used in free combinations (which, per definition, tend to have more general and dispersed meanings than lexical words with their lexical meanings), is particularly helpful for language learners and non-native speakers: in the apparent chaos of word combinations, there seem to appear certain patterns which may make easier the memorizing of prepositional collocations, and thus in general improve the quality and naturalness of the texts they produce. References Allerton, D. J., Valency and the English Verb. London, Academic Press. 1982. Bahns, J., Kollokationen als lexicographisches Problem. Eine Analyse allgemeiner und spezieller Lernerwörterbücher des Englischen. In: Lexicographica, Series maior 74. Tübingen, Max Niemeyer. 1996. Bajec, A., Besedotvorje slovenskega jezika IV. In: Predlogi in predpone. Ljubljana, SAZU. 1959. Benson, M., The Structure of the Collocational Dictionary. In: International Journal of Lexicography II/l. 1989, 1-14. Benson, M., Benson, E., Ilson, R., The BBI Dictionary of English Word Combinations. Amsterdam/Philadelphia, John Benjamins. 1997. Bohorič, A., Arcticae horulae succisivae (Zimske urice proste). Wittenberg, 1584. ed. J. Toporišič. Maribor, Obzorja. 1987. COLLINGE, N. E. et. al. eds., An Encyclopaedia of Language. London/New York, Routledge. 1990. Courtney, R„ Longman Dictionary of Phrasal Verbs. Harlow, Essex, Longman. 1983. Dixon, R. M. W„ A New Approach to English Grammar, on Semantic Principles. Oxford, Clarendon Press. 1991. Dular, J., Priglagolska vezava v slovenskem knjižnem jeziku (20. stoletja): Ph. D. diss. Ljubljana, FF. 1982. Dular, J., Združena vezava v desni vezljivosti slovenskega glagola. In: Jezik in slovstvo, XXIX/8. 1983/84, 289-293. Emons, R., Valenzen englischer Prädikatsverben. Tübingen, Max Niemeyer. 1974. Emons, R., Valenzgrammatik für das Englische. Tübingen, Max Niemeyer. 1978. Gutsman, O., Windijche Sprachlehre verfajjet von Oswald Gutsman. Gratz. 1777, 81-133. Hausmann, F. J., Wortschatzlernen ist Kollokationslernen. In: Praxis des neusprachlichen Unterrichts 31: 1984, 395^106. Hausmann, F. J., Kollokationen im deutschen Wörterbuch: ein Beitrag zur Theorie des lexikographischen Beispiels. In: Lexikographie und Grammatik, eds. Н. Bergenholtz in J. Mugdan. Tübingen, Max Niemeyer. 1985,118-129. Helbig, G., Buscha, J., Deutsche Grammatik. Ein Handbuch für den Ausländerunterricht. Leipzig, VEB Verlag Enzyklopädie. 1984. Helbig, G., Probleme der Valenz- und Kasustheorie. Tübingen, Max Niemeyer Verlag. 1992. Helbig, G., Schenkel, W., Wörterbuch zur Valenz und Distribution deutscher Verben. Leipzig, VEB Bibliographisches Institut. 1975. Herbst, T., Untersuchungen zur Valenz englischer Adjektive und ihrer Nominalisierungen. Tübingen, Narr. 1983. Herbst, T. A Valency Model for Nouns in English. In: JL 24. 1988, 265-301. Herbst, T., What Are Collocations: Sandy Beaches or False Teeth? In: English Studies 4. 1996, 379-393. Herbst, T., Roe, I., How Obligatory Are Obligatory Complements? An Alternative Approach to the Categorization of Subjects and Other Complements in Valency Grammar. In: English Studies 2. 1996, 179-199. Križaj Ortar, M., Glagolska vezljivost (na podlagi korpusa črke b v SSKJ). B. A. diss. Ljubljana, FF. 1981. Križaj Ortar, M., Glagolska vezljivost. In: Slavistična revija ХХХ/2. 1982, 189-213. Križaj Ortar, M., Vezljivost: iz pomena v izraz. In: XXV. seminar slovenskega jezika, literature in kulture. Ljubljana. 1989, 129-140. kunst Gnamuš, O., Pomenska sestava povedi. Ph.D.diss. Ljubljana, FF. 1981. Lerot, J., Die verbregierten Präpositionen in Präpositionalobjekten. In: Satzglieder im Deutschen: Vorschläge zur syntaktischen, semantischen und pragmatischen Fundierung, ed. W. Abraham. Tübingen, Narr. 1982, 261-291. Lindstromberg, English Prepositions Explained. Amsterdam/Philadelphia, John Benjamins. 1997. Lyons, J., Introduction to Theoretical Linguistics. Cambridge, Cambridge University Press. 1968. Matthews, P., Syntax. Cambridge, Cambridge University Press. 1981. Orešnik, J., Udeleženske vloge v slovenščini. Ljubljana, SAZU. 1992. Palmer, F. R., The English Verb. 2nd ed. Harlow, Essex, Longman. 1988. Pawley, A., Syder, F. H., Two Puzzles of Linguistic Theory: Nativelike Selection and Nativelike Fluency. In: Language and Communication. Applied Linguistics and Language Study, Ur. eds J. C. Richards and R. W. Schmidt. Harlow, Essex, Longman. 1983, 191-226. Quirk, R. et al. eds., A University Grammar of English. Harlow, Essex, Longman. 1973. Quirk, R. et al. eds., A Comprehensive Grammar of the English Language. Harlow, Essex, Longman. 1985. von Randow, E., Valente Substantive des Englischen. Tübingen, Narr. 1986. Samardžija, M., Valentnost glagola u suvremenom hrvatskom književnem jeziku. Ph. D. diss. Zagreb, FF. 1986. Schröder, J., Lexikon deutscher Präpositionen. Leipzig, VEB Verlag Enzyklopädie. 1986. Sicherl, E., Predložne kolokacije v slovenščini in angleščini. Ph. D. diss. Ljubljana, FF. 1999. Sicherl, E., On the Content of Prepositions in Prepositional Collocations. In: English Language Overseas Perspectives and Enquiries: Studies in the English Language and Literature in Slovenia 1/1-2. 2004, 37-46. Sinclair, J., ed., Collins Cobuild English Guides 1: Prepositions. London, HarperCollins. 1991. Sinclair, J., ed., Collins Cobuild English Dictionary. London, HarperCollins. 1995. Slovar slovenskega knjižnega jezika (SSKJ) I-V. Ljubljana, DZS. 1970, 1975, 1979, 1985, 1991. S Omers, H., On the Validity of the Complement-Adjunct Distinction in Valency Grammar. In: Linguistics 22. 1984,507-530. Švedova, N. Ju., Russkaja grammatika II. (Sintaksis). Moskva, Nauka. 1980. Tesnićre, L., Elements de syntaxe structurale. Paris, Klincksieck. 1959. Toporišič, J., Slovenski knjižni jezik I-IV(SKJ). Maribor, Obzorja. 1965-1970. Toporišič, J., Slovenska slovnica. Maribor, Obzorja. 1976. Toporišič, J., Nova slovenska skladnja (NSS). Ljubljana, DZS. 1982. Toporišič, J., Prva slovenska skladnja. In: Slavistična revija XXXII/3. 1984, 159-181. Toporišič, J., Enciklopedija slovenskega jezika (ESJ). Ljubljana, Cankarjeva založba. 1992. Vidovič Muha, A., Nova slovenska skladnja J. Toporišiča. In: Slavistična revija ХХХП/2. 1984, 142-155. Vidovič Muha, A., Glagolske sestavljenke - njihova skladenjska podstava in vezljivostne lastnosti (Z normativnim slovensko-nemškim vidikom). In: Slavistična revija XLI/1. 1993, 161-192. Vincenot, C., Essai de Grammaire Slovene. Ljubljana, MK. 1975. Žele, A., Vezljivost v slovenskem knjižnem jeziku. Linguistica et philologica. Ljubljana, Založba ZRC, ZRC SAZU. 2001. Žic Fuchs, M., Znanje o jeziku i znanje o svijetu (Semantička analiza glagola kretanja u engleskom jeziku). Ph. D. diss. Zagreb, FF. 1991. Povzetek PREDLOŽNE ZVEZE OZ. PREDLOŽNOMORFEMSKOST V SLOVENŠČINI IN ANGLEŠČINI: TEORETIČNI IN KONTRASTIVNI VIDIK Predložnomorfemskost kot aktualna stalnica različnih jezikovnih sistemov vedno znova zahteva tudi aktualne teoretične opredelitve - zlasti povedne, širše informativne in uporabne so kontrastivne obravnave. Izhodiščno je z aktualno slovensko-angleško jezikoslovno terminologijo komentirana predložnomorfemskost v slovenščini. Potem je kontrastivno, z vidika prevajanja in žive vsakodnevne prakse, predstavljena problematika doseganja logične in hkrati normativno ustrezne (ne)idiomatične rabe predložnih morfemov v slovenščini, angleščini in nemščini. V slovenščini kljub prevladujoči usmeritvi v neposrednejše tožilniško izražanje težnja po natančnosti in analitičnosti sporočanega hkrati vodi tudi k pogostejši predložnomorfemski uporabi glagolov. Kako glagolski pomen odloča o izbiri predložnega morfema kažejo npr. zgledi delati na projektu (pov. dol.): govoriti za njega / govoriti o njem (obvezno dol.): klestiti s šibo (neobvezno dol.). Prosti glagolski morfem je leksikalizirani kot del pomenskosestavinskosti glagolov v smislu, da jo sotvori (leksikalizacija) in hkrati del glagolskega leksema in glagolske vezljivosti, ali neleksikalizirani. ko pomensko izhaja vsaj iz enega pomena glagolskega leksema in izraža njegovo pomensko- in strukturnoskladenjsko vezljivost ter je zato del povedkove vezljivosti; izven vezljivosti ostajajo še obglagolski vezavnodružljivi predložni morfemi. V anglo-ameriškem jezikoslovju se je v zadnjih desetletjih uveljavila delitev predložnih zvez na proste predložne zveze in slovnične/predložne kolokacije; kolokacije so opredeljene kot povsem naravno pomensko- in strukturnoskladenjsko povezovanje besed v smislu vzajemnega pomensko-skladenjskega pričakovanja besed (mutual expectancy of words). V obeh primerih pa so s prostimi morfemi navadno ubesedene tiste pomenske sestavine, ki so tako pomensko- kot strukturnoskladenjsko udeležene pri vezljivosti. S prostimi morfemi so torej tudi strukturnoskladenjsko izražene intenčne/vezljivostne pomenske sestavine glagolov. Tako npr. predložni morfem na poudari pomensko sestavino 'pričakovanja' v zvezah kot pripraviti se na, čakati na, upati na ipd.; v angleščini imamo podobnopomenske predložne kolokacije s for v primerih kot prepare for, wait for, hope for ipd. in še v nemščini z auf v primerih tipa sich vorbereiten auf, warten auf, hoffen auf ipd. Dejstvo je, da je ravno pri vezljivosti precej velika interferenca -govorci si vezljivosti v materinščini niti ne uzavestijo, zato vezljivostne vzorce prenesejo tudi v jezik prevoda. Marjeta Vrbinc and Alenka Vrbinc University of Ljubljana UDK 8Г374.81 A RESEARCH-BASED STUDY OF FOREIGN STUDENTS' USE OF GRAMMATICAL CODES IN FIVE LEADING BRITISH LEARNERS' DICTIONARIES Abstract Grammatical codes are one of several ways of including grammar in learners' dictionaries. In our research we focussed on the usability and user-friendliness of learners' dictionaries as regards grammatical information. The results presented and discussed in this article are based on answers obtained by a questionnaire that tested the understanding of codes found in five leading British monolingual learners' dictionaries and the success of the explanations of the same codes provided in the front matter of each dictionary. The results are presented by dictionaries and by codes. The most important finding of this research is that the understanding of the code and thus its usefulness depends on the code itself rather than on the dictionary. 1. Introduction Dictionaries, especially monolingual dictionaries designed to meet the needs of a foreign language learner, contain a large amount of information of various types. Early studies conducted into dictionary use (Barnhart 1962, Quirk 1973, Bejoint 1981, Tomaszczyk 1979) show that dictionaries are most often used for checking the meaning of unknown words. Apart from meaning, other types of information are also included and used by dictionary users but to a lesser extent than meaning. Undoubtedly, one of the important elements of monolingual learners' dictionaries is the inclusion of grammar (McCorduck 1993). Grammar has been included in learners' dictionaries since the very beginning (Cowie 1999), but the amount of grammar in dictionaries has increased noticeably in the course of time (Bejoint 2000: 29). Typically, all monolingual learners' dictionaries include the part of speech label, which represents grammatical information about words, where they can operate in the syntax of sentences, and what their combinatorial possibilities are. Other grammatical information refers to various grammatical properties of each individual word class, such as countability in nouns, verb patterns, attributive, postpositive, and predicative use of adjectives, as well as gradability in adjectives and adverbs (Herbst 1989, Jackson 2002, Landau 2001, Cowie 1999). Grammatical information is provided in the form of codes or abbreviated phrases as well as more indirectly, i.e. in definitions and examples of use. The question of what should be taken into consideration when including grammatical information has been widely discussed (Jackson 1985, Lemmens and Wekker 1986, Sinclair 1987), but it is of the greatest importance to know what dictionary users understand and make use of when consulting their dictionaries in terms of grammar. The only way to find this out is to carry out research into different aspects of dictionary use, including grammar. Surveys should enquire into students' awareness of the existence of grammatical information in learners' dictionaries. Several aspects have been tested so far, such as syntax, complementation codes and labels (Nuccorini 1992), the obligatory use of the article, verb patterns, and countability (Tono 2001), and verb patterns, especially verb complementation (Bogaards, van der Kloot 2001). The results show that grammar is rarely looked up by the test subjects; they do not use all the available information, and they tend to choose the first definition in the entry rather than read the whole entry (this applies especially to long entries), even though useful grammatical information may be included later in the entry. Several comparisons of codes used in different learners' dictionaries have also been made, investigating the system of grammatical codes (Aarts 1991, Dalgish 1995, Strevens 1987), the thoroughness and consistency in the use of grammatical codes (Heath 1982), and the transparency of codes (Aarts 1991, Herbst 1989, Strevens 1987). The most commonly expressed criticism refers to the fact that codes are not sufficiently transparent, therefore not sufficiently user-friendly. The main proposal is that a coding system should be self-explanatory and reasonably easy to use. In order to test the transparency and usability of grammatical codes included in the latest editions of five leading British monolingual dictionaries, we carried out a survey, the results of which are presented and discussed in this article. Research was undertaken to prove our hypothesis that dictionary users do not regularly consult the front matter in order to check a code or an abbreviated phrase they do not understand. We, therefore, wanted to find out to what extent the codes or abbreviated phrases are understood without being explained (are they so clear that the explanation is redundant, or are they so difficult that they cannot be understood without an explanation?). 2. Methodology 2.1 Test Subjects Research was carried out among second- and third-year students of the Faculty of Economics (hereafter referred to as FE) and fourth-year students of the Faculty of Arts (hereafter referred to as FA), Department of English, University of Ljubljana, Slovenia. Together the groups comprised 162 respondents. One hundred and fifty-seven respondents (i.e. 96.9 %) speak Slovene as their mother tongue, and 5 (i.e. 3.1 %) speak some other language as their mother tongue. The test subjects were asked to give details about the number of years they had been studying English. Most of the respondents (i.e. 69.1 %) had been studying English for 10-12 years; 15.4 % had been studying it for less than 10 years, and 15.4 % for more than 12 years. Our test subjects were considered as belonging to a quite homogenous group: most of them attended schools in Slovenia; they had the same number of English lessons per year and covered the same syllabus. For these reasons we considered it unnecessary for them to undergo a placement test. However, we did enquire about the students' average grades in English in grammar school. The majority of students indicated that their average school grade in English was either A or B (37 % stated Grade A and 34.6 % stated Grade B); 24.1 % had a Grade of C and only 4.3 % a Grade of D. 2.2 Test Design The questionnaire consisted of two parts: a Dictionary User Profile Form and a Dictionary Research Test. The Dictionary User Profile Form was completed by all test subjects and was aimed at obtaining information about the dictionary users, i.e. their mother tongue, how long they have been studying English, and their grades in grammar school (cf. 2.1 Test Subjects). The Dictionary Research Test concentrated on the respondents' abilities to decipher the grammatical codes included in five leading British monolingual learners' dictionaries: COBUILD4, OALD6, CALD2, LDOCE4, and MED1. It consisted of two tasks. In both tasks, the respondents were subdivided into five groups depending on the dictionary tested. Thirty-two students were given ten grammatical codes or abbreviated phrases from MED1, 32 from OALD6, 31 from LDOCE4, 36 from COBUILD4, and 31 from CALD2. In Task 1, the students were supposed to explain the code or the abbreviated phrase taken from the above-mentioned dictionaries. It was aimed at testing students' understanding of the codes. The students did not have to use the appropriate terminology; they could also explain their understanding in their mother tongue or provide an example. Task 2, on the other hand, included the same codes with the explanations taken from the front matter of each dictionary. It was aimed at testing the understanding of the explanations of these codes or abbreviated phrases. 2.3 Procedure of data collection For the purpose of this study, data were collected by means of a questionnaire. Test subjects were given the questionnaire, and the researchers explained what they had to do. The respondents were supposed to complete the questionnaire in 45 minutes. The respondents' answers were appropriately coded and prepared for a statistical analysis using the Microsoft Excel program. Standard statistical methods were used for the data processing, which was carried out by the SPSS for Windows, version 11. 3. Results of the Dictionary Research Test In Task 1, the range of correct answers in all the dictionaries tested is quite broad (from 12.5 % to 87.5 % in OALD6, from 3.2 % to 90.3 % in LDOCE4, from 11.1 % to 88.9 % in COBUILD4, and from 0.0 % to 90.3 % in CALD2). The only dictionary where the percentage of correct answers is somehow more balanced is MED 1, where the difference between the best and the worst result is also quite large but still smaller than in other dictionaries (between 46.9 % and 96.9 %). In Task 2, where the understanding of the explanations of the codes or abbreviated phrases was tested, COBUILD4 has to be mentioned as the dictionary where the difference between the explanation that was understood by the highest percentage of the respondents (97.2 %) and that which was understood by the lowest percentage of the students (22.2 %) is most noticeable. In MED1, the difference between the explanation that was understood by the highest percentage of the students (100.0 %) and that which was understood by the lowest percentage of the respondents (62.5 %) is the smallest, whereas the results range from 58.1 % to 100.0 % in CALD2, from 43.8 % to 100.0 % in OALD6, and from 38.7 % to 100.0 % in LDOCE4. Let us consider the individual codes and abbreviated phrases included in all the above-mentioned dictionaries. The codes or abbreviated phrases in MED1 do not seem to pose great problems to dictionary users. Among the codes that were the most problematic we have to mention code 8 'in imperative', code 1 'linking verb', and code 7 'usually progressive'. Code 6 ('usually passive') with 96.9 %, code 4 ('never before noun') with 87.5 %, and code 10 ('in negatives or questions') with 87.5 % are the most successful ones. A comparison of results concerning the explanation of the codes found in the front matter shows that only the explanation for code 6 ('usually used in the passive') was understood by all the test subjects. The explanations for code 3 ('adjectives that can only be used before a noun'), code 4 ('adjectives that can never be used before a noun'), code 10 ('usually used in negative sentences or questions'), and code 9 ('usually used in the infinitive') also proved successful, since they were all understood by more than 90 % of the respondents. The respondents found the explanations for code 7 ('usually used in the progressive') and code 8 ('usually used in the imperative') the most difficult ones, since they were understood by 62.5 % and 65.6 % of the test subjects respectively. Here a parallel can be drawn with the results in Task 1, where the same codes were understood by a relatively small number of respondents. For more detailed results concerning the success in deciphering the codes and understanding their explanations in MED1, see Table 1. Table 1: Success in deciphering and understanding the grammatical codes in MED1 (in %) Stud, resp. Code 1 Code 2 Code 3 Code 4 Code 5 Code 6 Code 7 Code 8 Code 9 Code 10 Cor. 50.0 78.1 84.4 87.5 68.8 96.9 53.1 46.9 75.0 87.5 U 78.1 87.5 96.9 93.8 87.5 100.0 62.5 65.6 90.6 93.8 Stud. resp. - students' responses Cor. - the respondent managed to explain the code correctly U - the respondent understood the explanation of the code Among the grammatical codes taken from OALD6, code 10 ('VN speech') proved to be the most difficult one, since it was correctly explained by only 12.5 % of the respondents. The most difficult part of this code is most certainly the second element ('speech'), since 71.9 % of the respondents correctly explained code 3 ('VN'), which corresponded to the first element in code 10. Less than 50 % of the test subjects were able to explain the meaning of code 4 ('VN + adv./prep.') and code 9 ('VN -ing'). Code 2 ('pt') and code 5 ('V-ADJ') were correctly explained by 81.3 % and 87.5 % of the respondents respectively, a result which means that they are among the most successful codes in OALD6. As regards the explanations of the codes, code 2 ('past tense') was the most successfully explained, which is understandable, given that all language learners are familiar with the term 'past tense'. Apart from code 2, code 3 ('verb + noun phrase'), code 1 ('past participle'), code 5 ('verb + adjective'), and code 7 ('verb + that clause') are also explained in such a way that the majority of respondents are able to understand them. The most problematic code seems to be code 10 ('verb + noun phrase + direct speech'), whose explanation was understood by only 43.8 % of respondents. The same code also yielded the worst result in Task 1, and even the explanation provided in the front matter does not help as much as it should. Detailed results of student success in diciphering the codes and understanding of the explanations of the codes are presented in Table 2. Table 2: Success in deciphering and understanding the grammatical codes in OALD6 (in %) Stud, resp. Code 1 Code 2 Code 3 Code 4 Code 5 Code 6 Code 7 Code 8 Code 9 Code 10 Cor. 62.5 81.3 71.9 40.6 87.5 71.9 68.8 53.1 40.6 12.5 U 90.6 100.0 93.8 68.8 90.6 84.4 90.6 78.1 78.1 43.8 In comparison to MED1 and OALD6, the codes included in LDOCE4 are less transparent. Only code 5 ('only before noun') was understood by more than 90 % of the respondents. As many as five codes were understood by less than 50 % of the students, which is far from being satisfactory. Code 9 ('+ adj/adv') was deciphered by 3.2 % of the respondents only. The explanation of this code ('shows that an adverb of degree is used before adjectives and adverbs') seems to be more successful, since 51.6 % of the respondents were able to understand it. Apart from code 9, code 10 ('also + plural verb BrE') was also a problematic one, since less than 20 % of the test subjects managed to explain it correctly. Again, the explanation ('shows that a group noun can take a plural verb in British English') was understood by 77.4 % of the respondents, which is quite a good result in comparison to the ability of the test subjects to explain the code. Another code that gave students problems is code 8 ('sentence adverb'), which was appropriately explained by only 23.3 % of the respondents. The explanation of this code ('shows that an adverb modifies a whole sentence') provided in the front matter is not of great help to dictionary users, either, since it was understood by only 38.7 % of the respondents. Apart from code 5 (already mentioned) the most successful codes include code 7 ('only after noun'), and code 6 ('not before noun'), as they both were successfully deciphered by more than 80 % of the respondents. Similarly, the explanations of these three codes were clear enough to be understood by the majority of respondents (the explanations for codes 5 and 6 were understood by all the respondents, and the explanation for code 7 by 96.8 %). The greatest difference between correct answers that show students' ability to decipher the code and those that show that respondents understood the explanation of the code can be found with code 2. Only 41.9 % of students understood the code, as opposed to 100.0 % of students who understood the explanation of the code. Further details concerning students' ability to decipher the code and understand its explanation are presented in Table 3. Table 3: Success in deciphering and understanding the grammatical codes in LDOCE4 (in %) Stud, resp. Code 1 Code 2 Code 3 Code 4 Code 5 Code 6 Code 7 Code 8 Code 9 Code 10 Cor. 41.9 41.9 54.8 54.8 90.3 80.6 87.1 23.3 3.2 19.4 U 67.7 100.0 93.5 93.5 100.0 100.0 96.8 38.7 51.6 77.4 Generally speaking, codes in COBUILD4 give cause for concern. Only one code (code 7 - 'V-PASSIVE') was understood by 88.9 % of the students. The results show that code 4 ('N-VAR') with 11.1 %, code 1 ('PREP-PHRASE') with 13.9 %, and code 3 ('N-COUNT-COLL') with 27.8 % were the most difficult ones. The results showing how well the students understood the explanation of the codes indicate that the explanations for code 7 ('passive verb'), code 2 ('comparative form'), and code 9 ('imperative') were understood by the majority of respondents, whereas the explanation for code 4 ('variable noun') was understood by only 22.2 % of respondents. For more detailed results, see Table 4. Table 4: Success in deciphering and understanding the grammatical codes in COBUILD4 (in %) Stud, resp. Code 1 Code 2 Code 3 Code 4 Code 5 Code 6 Code 7 Code 8 Code 9 Code 10 Cor. 13.9 61.1 27.8 11.1 33.3 58.3 88.9 41.7 66.7 58.3 U 66 7 83.3 50.0 22.2 52.8 69.4 97.2 77.8 80.6 63.9 The questionnaires tested the success of 50 codes - i.e. 10 codes taken from each dictionary - but the only code that was not understood by a single respondent was code 5 ('M') from CALD2. Even the explanation itself (i.e. 'phrasal verb with a particle that can come before or after the object') is not really revealing, since only 58.1 % of the students managed to understand it. This result is not a complete failure, but if we compare it with other results for explanations of codes in this dictionary, it is the worst one. Besides code 5, we have to mention code 4 ('L') and code 8 ('R'), which were appropriately explained by less than 10 % of the respondents. Again, a single letter is not self-explanatory; therefore students did not manage to decipher 'L' as referring to linking verbs and 'R' as referring to reflexive verbs. Also worth mentioning is code 9 ('S'), which was correctly explained by 35.5 % of the students. All the users understood this code after having read the explanation. A comparison of the results showing how well the students understood the explanations of the codes reveals a completely different picture. As many as three explanations (code 3 - 'adjective that is placed only before a noun'; code 9 - 'singular'; code 10 - 'verb usually used in the passive') were understood by all respondents. Effective explanations also include those for code 6 ('verb with an object followed by a noun or an adjective') and code 7 ('verb with an object followed by the -ing form of another verb'), which were understood by more than 90 % of the test subjects. More detailed results are presented in Table 5. Table 5: Success in deciphering and understanding the grammatical codes in CALD2 (in %) Stud, res p. Code 1 Code 2 Code 3 Code 4 Code 5 Code 6 Code 7 Code 8 Code 9 Code 10 Cor. 25.8 67.7 77.4 6.5 0.0 58.1 58.1 9.7 35.5 90.3 U 83.9 71.0 100.0 64.5 58.1 93.5 93.5 61.3 100.0 100.0 4. Discussion These results clearly show that the test subjects find some codes or abbreviated phrases fairly transparent, whereas others prove to be quite difficult. It is surprising, for example, that respondents had problems with the abbreviated phrase 'in imperative', since they come across the term 'imperative' in the course of learning English, but it is understandable that they did not know the meaning of the abbreviated phrase 'linking verb', since this is not used when teaching English grammar at school. The term 'linking verb' is known only to the students of English who get acquainted with it in their grammar courses at the university, when discussing the classification of the verbs. The general dictionary user, on the other hand, cannot possibly be familiar with the meaning of this term. The term 'progressive' also poses problems because it is not used in textbooks, the result being that students are not familiar with it. At school, the term 'continuous' is used, and it can be assumed that this term would be much better understood. It has to be stressed that all the textbooks that are officially approved for use in primary and grammar schools in Slovenia are published by the same publishing houses as the monolingual learners' dictionaries (e.g. Cambridge English for Schools - Cambridge University Press; Project - Oxford University Press; New Headway English Course -Oxford University Press; Matrix - Oxford University Press; Natural English - Oxford University Press). The same terminology can justifiably be expected in textbooks as well as dictionaries intended for the same target audience. It is true that in primary school and at least at the beginning of grammar school the level of English is not sufficient to enable the use of the dictionaries tested, which are intended for advanced learners. Advanced learners, however, have experienced many years of learning English, and the basic learning aid they have used has been the school textbooks. One would, however, expect a better result with 'v-cont' (in COBUILD4), where the term 'continuous' is used, as opposed to the term 'progressive', which is used in other learners' dictionaries. It is possible that our respondents did not manage to understand the abbreviated form of the term (cont for continuous), but even the results showing how many students understood the explanation provided for this code are not as high as one would expect. The students seem to have been puzzled by the entire expression 'continuous verb', which is used in the front matter of this dictionary, as they are more familiar with the term 'continuous tense' rather than with the term 'continuous verb'. Another complicated abbreviated phrase is 'sentence adverb', since a general dictionary user does not understand the concept underlying this term. This is a term language learners do not come across when learning English; they become acquainted with it as late as their grammar classes at the university (if they study English). The terms 'variable noun', 'prepositional phrase', 'collective countable noun', and 'reciprocal verb' are also much too complicated for an average dictionary user. The results for these abbreviated phrases are unsatisfactory and the reason for such poor results is most certainly the terminology used. What about one-letter codes? The code 'S' used in CALD2 proved to be quite difficult, which is surprising if we know that this code refers to 'singular'. In this particular example, the difficulty lies in the code itself not in the terminology used. In other learners' dictionaries the code 'sing.' is used (OALD6 and COBUILD4), or the word 'singular' is written in full (MED1 and LDOCE4). We also cannot ignore the code 'M' (used in CALD2), which is the only code (out of all 50 codes tested) that was not understood by a single respondent. What is it that makes this code so complicated? As we learn from the explanation of this code, it refers to a 'phrasal verb with a particle that can come before or after the object'. The code 'M' suggests that the object can occupy the medial position between the verb and the particle, but the code itself is not self-explanatory. How can a dictionary user know that 'M' stands for medial position? In codes consisting of several elements, such as 'VN speech', 'VN + adv./prep.', ' VN -ing', we can presuppose that it was the second element that the respondents were unable to understand, taking into account that the code 'VN' was not problematic. Probably, the reason cannot be attributed to the fact that the students do not know what 'speech', 'adv./prep.' or '-ing' mean; it is rather that they do not know what the entire code refers to. The same holds true of the codes '+ adj/adv', 'always + adv/prep' found in LDOCE4, since they consist of elements that average dictionary users would know, but probably users are unable to link the elements and understand what the entire code means. Consequently, the verbalization of these two codes ('shows that an adverb of degree is used before adjectives and adverbs' and 'shows that a verb must be followed by an adverb or a preposition') is perfectly clear, and is thus understood by all respondents. However, the explanation does not always prove successful, which is confirmed by the explanation of the code 'VN speech', meaning 'verb + noun phrase + direct speech'. The test subjects understood all parts of the explanation even the last part, i.e. 'direct speech', since they get acquainted with this term in the course of learning English, but in all likelihood, they were not able to make a connection between all three elements of the explanation. With a high degree of certainty it can be claimed that understanding of the code or abbreviated phrase and thus its usefulness depends on the code or abbreviated phrase itself rather than on the dictionary used. The compilers of the dictionaries provide as much grammatical information as possible, much of which is included in the form of grammatical codes or abbreviated phrases. As is shown in our research, the transparency and simplicity of the codes or abbreviated phrases are of the utmost importance if we want users to understand them and to make use of them. These results also confirm the findings of Aarts (1991), Herbst (1989), and Strevens (1987). 5. Conclusion There is no doubt that dictionary reference skills involve highly complicated cognitive skills. Dictionary skills are problem-solving skills: users are trying to solve a particular linguistic conflict by consulting a dictionary. However, it should be emphasized that dictionary use is not as simple as one would imagine. A dictionary abounds with information of various types, and very often a dictionary user is not even aware of the wealth of information he/she can find when consulting a dictionary. We strongly believe that students should be made aware of the whole range of information included in monolingual learners' dictionaries, and the inclusion of codes and abbreviated phrases is no exception. The front matter of each dictionary where the types of information included in this particular dictionary are explained can be of great help. Unfortunately, many dictionary users never read it, the result being that they cannot be regarded as proficient dictionary users. The educational system can play an important role, since the teaching of dictionary skills could be integrated into the foreign language syllabus. An important tool in the training of dictionary users is the use of separate companion booklets that are meant to train users and offer them practice in using the special features, including grammatical codes. Systematic dictionary training is absolutely necessary to enable learners to become efficient dictionary users. References: Dictionaries Rundell, M. (ed.) 2002. Macmillan English Dictionary for Advanced Learners. (First edition.) Oxford: Macmillan Publishers. (MED1) Sinclair, J., Sinclair Knight, L., and Clari, M. (eds.) 2003. Collins Cobuild Advanced Learner's English Dictionary. (Fourth edition.) Glasgow: HarperCollins Publishers. (COBUILD4) Summers, D. (ed.) 2003. Longman Dictionary of Contemporary English. (Fourth edition.) Harlow, Essex: Pearson Education Limited. (LDOCE4) Wehmeier, S. (ed.) 2000. Oxford Advanced Learner's Dictionary. (Sixth edition.) Oxford: Oxford University Press. (OALD6) Woodford, K. and Jackson, G. (eds.) 2003. Cambridge Advanced Learner's Dictionary. (Second edition.) Cambridge: Cambridge University Press. (CALD2) Other literature Aarts, F. 1991. "Lexicography and Syntax: The State of the Art in Learners' Dictionaries of English" in J. E. Alatis (ed.), Linguistics and Language Pedagogy: The State of the Art. Georgetown University Roundtable on Languages and Linguistics, 1991. Washington, D.C.: Georgetown University Press, 567-582. Barnhart, C. 1962. "Problems in Editing Commercial Monolingual Dictionaries" in F. Householder and S. Saporta (eds.), Problems in Lexicography. International Journal of American Linguistics 28 (2) (April), Bloomington, Indiana: Indiana University Research Center in Anthropology, Folklore, and Linguistics, 161-181. Bejoint, H. 1981. "The Foreign Student's Use of Monolingual English Dictionaries: A Study of Language Needs and Reference Skills." Applied Linguistics 2 (3), 207-222. Bejoint, H. 2000. Modern Lexicography: An Introduction. Oxford: Oxford University Press. Bogaards, P. and van der Kloot, W. 2001. "The Use of Grammatical Information in Learners' Dictionaries." International Journal of Lexicography 14: 97-121. Cowie, A. 1999. English Dictionaries for Foreign Learners. A History. Oxford: Oxford University Press. Dalgish, G. 1995. "Learners' Dictionaries: Keeping the Learner in Mind?" in B. B. Kachru and H. Kahane (eds.), Cultures, Ideologies, and the Dictionary: Studies in Honor of Ladislav Zgusta. (Lexicographica Series Maior 64.) Tübingen: Max Niemeyer Verlag, 329-338. Heath, D. 1982. "The Treatment of Grammar and Syntax in Monolingual English Dictionaries for Advanced Learners." Linguistik und Didaktik 49/50: 95—107. Herbst, T. 1989. "Grammar in Dictionaries" in M. L. Tickoo (ed.), Learners' Dictionaries: State of the Art. (Anthology Series 23) Singapore: SEAMEO Regional Language Centre: 94-111. Jackson, H. 1985. "Grammar in the Dictionary" in R. Ilson (ed.), Dictionaries, Lexicography and Language Learning. Oxford: Pergamon Press, 53-59. Jackson, H. 2002. Lexicography: An Introduction. London, New York: Routledge. Landau, S. I. 2001. Dictionaries: The Art and Craft of Lexicography. Cambridge: Cambridge University Press. Lemmens, M. and Wekker, H. 1986. Grammar in English Learners' Dictionaries. (Lexicographica Series Maior 16.) Tübingen: Max Niemeyer Verlag. McCorduck, E. S. 1993. Grammatical Information in ESL Dictionaries. (Lexicographica Series Maior 48.) Tübingen: Max Niemeyer Verlag. Nuccorini, S. 1992. "Monitoring Dictionary Use" in W. H. Tommola, K. Varantola, T. Salmi-Tolonen, and J. Schopp (eds),Euralex '92 Proceedings I-II (Part I). Studia Translatologica, Ser. A, Vol. 2. Tampere, Finland: University of Tampere, 89-102. Quirk, R. 1973. "The Social Impact of Dictionaries in the U.K." in R. I. McDavid and A. R. Duckert (eds.), Lexicography in English. New York: Annals of the New York Academy of Sciences, 76-88. Sinclair, J. M. 1987. "Grammar in the dictionary" in J. M. Sinclair (ed.), Looking up: An Account of the COBUILD Project in Lexical Computing and the Development of the Collins COBUILD English Language Dictionary. London, Glasgow: Collins, 104-115. Strevens, P. 1987. "The effectiveness of learners' dictionaries" in R. Burchfield (ed.), Studies in Lexicography. Oxford: Oxford University Press, 76-93. Tomaszczyk, J. 1979. "Dictionaries: Users and uses." Glottodidactica 12: 103-119. Tono, Y. 2001. Research on Dictionary Use in the Context of Foreign Language Learning. (Lexicographica Series Maior 106.) Tübingen: Max Niemeyer Verlag. Povzetek RAZISKAVA O UPORABI SLOVNIČNIH KOD V PETIH VODILNIH BRITANSKIH ENOJEZIČNIH SLOVARJIH ZA TUJCE NA VZORCU TUJIH GOVORCEV Slovnične kode so eden od načinov vključevanja slovnice v enojezične slovarje za tujce. V raziskavi smo preučevali, v kakšni meri so ti slovarji uporabni in prijazni do uporabnika pri vključevanju slovničnih podatkov. Rezultati, predstavljeni v tem članku, temeljijo na odgovorih, ki smo jih dobili s pomočjo vprašalnika, s katerim smo testirali razumevanje kod, ki jih najdemo v petih vodilnih britanskih enojezičnih slovarjih za tujce. Poleg tega smo testirali razumljivost razlag istih kod, ki jih najdemo v uvodnem delu vsakega slovarja. Rezultati so predstavljeni po slovarjih in po kodah. Najpomembnejša ugotovitev te raziskave je, da sta razumevanje kode in posledično njena uporabnost odvisna od kode same in ne od slovarja. Appendix I The Dictionary User Profile Form was produced in Slovene, so that respondents could complete the form in their native language. The questionnaire here is a translation of the original form. DICTIONARY USER PROFILE FORM The aim of this questionnaire is to discover how dictionaries can be improved. You can help us by filling in this questionnaire. The questions are about monolingual (i.e. English-English) dictionaries. 1. What is your native language?_ 2. How many years have you been learning English for?_ 3. What was your average grade in English in grammar school?_ Appendix II The Dictionary Research Test was produced in Slovene to ensure that participants in the project would read the questions in their own language. The questionnaire here is a translation of the Slovene original. DICTIONARY RESEARCH TEST Task 1: The following grammatical codes can be found in Oxford Advanced Learner's Dictionary (6"1 edition). Please write down what each code means (you can explain it or provide an example). 1. PP 2. pt 3. VN 4. VN + adv./prep 5. V-ADJ 6. VN-ADJ 7. V(that) 8. V wh- 9. VN -ing 10. VN speech Task 2: The codes from the previous task are accompanied by the explanations that can be found in the introductory part of the dictionary. Please indicate whether the explanation is clear enough for you to make use of it when using English actively. 1. pp past participle a. understand b. don't understand 2. pt past tense a. understand b. don't understand 3. VN verb + noun phrase a. understand b. don't understand 4. VN + adv./prep. verb + noun phrase phrase a. understand b. don't understand 5. V-ADJ verb + adjective a. understand b. don't understand 6. VN-ADJ verb + noun phrase a. understand b. don't understand 7. V(that) verb + that clause a. understand b. don't understand 8. Vwh- verb + wh-clause a. understand b. don't understand 9. VN -ing verb + noun phrase a. understand b. don't understand 10. VN speech verb + noun phrase a. understand b. don't understand OR: Codes from Longman Dictionary of Contemporary English (4th edition). 1. linking verb 2. always + adv/prep 3. not in progressive 4. no comparative 5. only before noun 6. not before noun 7. only after noun 8. sentence adverb 9. + adj/adv 10. also + plural verb BrE Codes from LDOCE4 with explanations 1. linking verb a verb that is followed by a noun or adjective complement that describes the subjects of the verb a. understand b. don't understand 2. always + adv/prep shows that a verb must be followed by an adverb or a preposition a. understand 3. not in progressive a. understand 4. no comparative a. understand 5. only before noun a. understand 6. not before noun a. understand 7. only after noun a. understand 8. sentence adverb a. understand 9. + adj/adv a. understand 10. also + plural verb BrE a. understand b. don't understand shows that a verb is not used in the progressive form, that is, the -ing form after 'be' b. don't understand shows that an adjective is not used in the comparative or superlative form, that is, not with -er and -est, or 'more' and 'most' b. don't understand shows that an adjective can be used only before a noun b. don't understand shows that an adjective cannot be used before a noun b. don't understand shows that an adjective is used only immediately after a noun b. don't understand shows that an adverb modifies a whole sentence b. don't understand shows that an adverb of degree is used before adjectives and adverbs b. don't understand shows that a group noun can take a plural verb in British English b. don't understand OR: Codes from Macmillan English Dictionary (1st edition). 1. linking verb 2. auxiliary verb 3. only before noun 4. never before noun 5. + (that) 6. usually passive 7. usually progressive 8. in imperative 9. in infinitive 10. in negatives or questions Codes from MED1 with explanations 1. linking verb verbs that are followed by a noun or adjective complement describing the subject a. understand 2. auxiliary verb a. understand 3. only before noun a. understand 4. never before noun a. understand 5. + (that) a. understand 6. usually passive a. understand 7. usually progressive a. understand 8. in imperative a. understand 9. in infinitive a. understand 10. in negatives or questions a. understand b. don't understand verbs 'be', 'have', and 'do' when they are used with other verbs to form questions, show their tense, etc b. don't understand adjectives that can only be used before a noun b. don't understand adjectives that can never be used before a noun b. don't understand can be followed by a clause beginning with 'that' but you can miss out the 'that' b. don't understand usually used in the passive b. don't understand usually used in the progressive b. don't understand usually used in the imperative b. don't understand usually used in the infinitive b. don't understand usually used in negative sentences or questions b. don't understand OR: Codes from Collins Cobuild English Dictionary for Advanced Learners (4t'1 edition). 1. PREP-PHRASE 2. adj-compar 3. N-COUNT-COLL 4. N-VAR 5. V-RECIP 6. V-LINK 7. V-PASSIVE 8. v-cont 9. imper 10. pron-refl Codes from COBUILD4 with explanations 1. PREP-PHRASE a. understand 2. adj-compar a. understand phrasal preposition b. don't understand comparative form b. don't understand 3. N-COUNT-COLL collective count noun a. understand b. don't understand 4. N-VAR variable noun a. understand b. don't understand 5. V-RECIP reciprocal verb a. understand b. don't understand 6. V-LINK link verb a. understand b. don't understand 7. V-PASSIVE passive verb a. understand b. don't understand 8. v-cont continuous verb a. understand b. don't understand 9. imper imperative a. understand b. don't understand 10. pron-refl reflexive pronoun a. understand b. don't understand OR: Codes from Cambridge Advanced Learner's Dictionary (2nd edition). 1. always + adv/prep 2. + question word 3. before n 4. L 5. M 6. + obj + n/adj 7. + obj + v-ing 8. R 9. S 10. usually passive Codes from CALD2 with explanations 1. always + adv/prep a. understand 2. + question word a. understand 3. before n a. understand 4. L a. understand 5. M verb that must be followed by an adverb or preposition b. don't understand verb with a question word b. don't understand adjective that is placed only before a noun b. don't understand linking verb (followed by an adjective or noun) b. don't understand phrasal verb with a particle that can come before or after the object а. understand б. + obj + n/adj a. understand 7. + obj + v-ing a. understand 8. R a. understand 9. S a. understand 10. usually passive a. understand b. don't understand verb with an object followed by a noun or an adjective b. don't understand verb with an object followed by the -ing form of another verb b. don't understand reflexive b. don't understand singular b. don't understand verb usually used in the passive b. don't understand Jasmina Markič Universidad de Ljubljana UDK 8Г367.625.5 : 811.134.2 VALORES Y USOS DE LAS PERiFRASIS VERBALES DE GERUNDIO CON LOS AUXILIARES IR, ANDAR Y VENIR El presente articulo tiene como fin presentar las perffrasis verbales de gerundio con los verbos de movimiento andar, ir y venir como auxiliares. En la primera parte se pre-senta la problemätica de la perffrasis verbal en general y la definition del concepto y los valores que se expresan con estas estructuras verbales complejas. A continuation el articulo se centra en las perffrasis verbales aspectuales de gerundio y en particular en las que tienen como auxiliar los verbos de movimiento. Partiendo de algunos ejem-plos de obras literarias espanolas y latinoamericanas y sus traducciones al esloveno se comparan las dos lenguas y se comentan las "(im)posibilidades" de traducir adecua-damente estas estructuras verbales al esloveno. 1. El concepto de perffrasis verbal La perffrasis verbal es la combination de dos elementos verbales, uno flexionado con funcion de verbo auxiliar y otro en una forma no flexionada o no personal,1 unidos directamente sin ningün tipo de nexos (las perffrasis verbales de gerundio, de par-ticipio y algunas de infinitivo) o indirectamente con conjunciones y preposiciones (perffrasis verbales de infinitivo). La perffrasis verbal es una unidad sintäctico-semän-tica que actüa como un solo predicado y su significado propio proviene de la union del auxiliar y de la forma no personal. Se distingue de otras construcciones no perifrästi-cas parecidas y de las locuciones verbales compuestas de dos verbos2. El espanol, sobre todo algunas variantes latinoamericanas, es muy rico y creativo en la formaciön de nuevas perffrasis verbales y en la agrupaciön de varias perffrasis verbales especial-mente con fines expresivos, estilfsticos y/o pragmäticos3. 2. Las perffrasis verbales de gerundio En las perffrasis verbales de gerundio el verbo auxiliado en gerundio estä relacio-nado directamente con el verbo auxiliar sin ningün tipo de nexos. Se distinguen estas Las perffrasis verbales denominadas copulativas son una exception a esta definicion ya que los dos verbos unidos por la conjunciön copulativa y van en forma personal: el auxiliar ir (coger, tomar, agarrar) + verbo conju-gado. Expresan acciones puntuales, vistas como bruscas, räpidas, inesperadas. Ej.: -Y ahora va y se nos duerme, el tfo - farfullo Adrian... (HDC, 190) "No, zdaj pa gre in nam zaspi, " je zajecljal Adrian. (KH, 190) 2 La construction de dos verbos en el ejemplo "deseo viajar" no es perifrästica. Se trata de dos verbos plenos. Tampoco son perifrästicas las locuciones verbales formadas de dos verbos como "echar a perder". 3 " Che, Cristobal. Vamos a tener que ir yendo pa'casa que estä por Hover. No me gusta eso de tener que mez-clar el vino con el agua." (Markić 1990:204) perifrasis de estructuras semejantes no perifrästicas en las que hay un verbo no auxi-liar con una oraciön subordinada adverbial de gerundio (Van cantando a la escuela) o un gerundio como predicado de complemento directo de un verbo de percepciön (La vio charlando con su amiga) (Garcia Fernandez 2006:31). En este ultimo caso es posi-ble construir una oracion subordinada sustantiva, lo que es una prueba importante de que se trata de una construcciön no perifrästica (La vio que charlaba con su amiga). En los usos perifrästicos, es la base lexica del gerundio, y no la del auxiliar, la que determina la estructura argumental de la perifrasis, por lo que coinciden necesariamente el sujeto del auxiliar con el sujeto referencial del gerundio y los complementos inciden sobre el conjunto y no exclusivamente sobre el verbo en forma personal. (Yllera 1999:3393) El esloveno no conoce las estructuras perifrästicas con gerundio4, pero existen algunas estructuras correspondientes a las estructuras no perifrästicas en espanol de verbo no auxiliar seguido de un gerundio en funciön adverbial aunque en la lengua hablada son raras ya que son de uso literario u obsoleto (V šolo gredo pojoč). El esloveno emplea otros medios lingüisticos para expresar los valores de las perifrasis verbales espanolas de gerundio (formas verbales imperfectivas, complementos circun-stanciales, etc.) como se presenta en los apartados siguientes. Entre los valores que se expresan con las perifrasis verbales de gerundio predomi-nan los aspectuales. Algunas perifrasis de gerundio funcionan como marcadores de discurso: empezar/comenzar + gerundio como marcadores de apertura y acabar/ter-minar + gerundio como marcadores de cierre. Otras tienen valores temporales-aspec-tuales como estar + gerundio. La nociön de aspecto es esencial para estudiar las perifrasis verbales en las lenguas romänicas en general y en espanol en particular. En la opinion de Coseriu y Dietrich (1983: 12-13) las perifrasis verbales estän creando un sistema verbal paralelo especia-lizado para la expresion de valores aspectuales. Para los fines de este anälisis el con-cepto de aspecto es considerado mäs allä de la tradicional division del aspecto verbal en perfectivo e imperfectivo. Bajo el termino de valores aspectuales se tienen en cuen-ta los siguientes conceptos5: a) las caracterfsticas del desarrollo de la accion verbal en el tiempo (acciones momentäneas, reiterativas, habituales, resultativas, progresivas, durativas), es decir, el modo de la accion, lo que frecuentemente se denomina Aktionsart; 4 En esloveno hay algunas estructuras perifrästicas indicadoras de la fase inicial de la accion formadas con infinitivo parecidas a las perifrasis verbales espanolas: Začel je peti. / Empezö a cantar. Tambien existen en esloveno las perifrasis copulativas (vease la nota 1). ^ La definition general del concepto de valores aspectuales se desprende de las definiciones de aspecto de los autores Miklič, T. (1983), "L'opposizione italiana perfetto vs imperfetto e l'opposizione slovena dovršnost vs nedovršnost nella verbalizzazione delle azioni passate". Linguistica XXIII. Ljubljana: Filozofska fakulteta. 53123; Comrie, B. (1976), Aspect. Cambridge: Cambridge University Press; Coseriu, Е. (1980), Aspect verbal ou aspects verbaux? Quelques questions de theorie et de methode. La notion d'aspect, Colloque organise par le Centre d' Analyse Syntaxique de l'Universite de Metz, Paris: Klincksieck; y otros. b) el tipo de la acciön: acciones telicas (realizaciones y logros) y atelicas (estados y actividades) referentes a la clasificacion de Vendler (en Garcia Fernandez 2006: 41); c) las fases de la accion se refieren al grado de la realization de la acciön (comien-zo, mitad y fin): fase inminencial, fase incoativa o ingresiva, fase media, fase ter-minativa; d) el aspecto en el sentido estrecho, lo que normalmente se denomina aspecto verbal y expresa como el hablante ve la accion, desde que punto de vista o perspectiva la observa, es decir, como la focaliza: - abarca la accion globalmente, en su totalidad (con el inicio y el final) - per -spectiva global o perfectiva (en espanol se suele indicar con el preterito perfecta simple), - abarca una parte de la accion sin interesarle el inicio ni el final, la perspectiva cursiva o imperfectiva (en espanol se suele indicar con el preterito imperfecto), - abarca una parte de la accion senalando el inicio pero sin indicar el final; focaliza una accion desde su inicio hasta un momento central de su desarrollo (perspectiva imperfectiva continuativa, se suele senalar en espanol con el preterito perfecto compuesto). 2.1 Las perifrasis verbales de gerundio con los verbos de movimiento an dar, ir y venir como auxiliares Estas construcciones son perifrasis verbales aspectuales de contenidos complejos condicionados en parte por el significado original del auxiliar que anade un significa-do de dinamismo y un importante valor estilistico. Cuando los verbos andar, ir y venir funcionan como verbos plenos, verbos lexicos de movimiento, entonces la construction andar/ ir! venir + gerundio no constituye perifrasis verbal.6 Como lo senala Gomez Torrego (1988:127-134) no es siempre fäcil determinar si se trata o no de una construction perifrästica. Para determinar el grado de gramatica-lizaciön de la perifrasis verbal, es decir, para comprobar que se trata de una perifrasis verbal, son necesarias pruebas sintäcticas entre las que figura la imposibilidad de omi-tir el gerundio o sustituirlo por otros terminos equivalentes (el adverbio interrogativo /como?; las subordinadas adverbiales, un adverbio).7 6 Ana va andando a la oficina. ^Cömo va a la oficina? Andando. Los chicos han venido cantando a la escuela. ^Como han venido? Cantando. Josefa anda moviendo los brazos. / Como anda? Moviendo los brazos. 7 Ando dändole vueltas a este asunto. *^Como ando a este asunto? Este coche viene costando tres millones. *^Como viene este coche? *Este coche viene tres millones. La calle se iba despoblando. *La calle iba - porque/ cuando - se despoblo. *La calle se iba de ese modo/asi. Otra de las pruebas del grado de gramaticalidad de las perffrasis verbales es la posi-ciön de los clfticos delante del auxiliar o deträs del auxiliado.8 2.2 Andar + gerundio es una perffrasis de valor aspectual durativo con la idea del desarrollo de un movimiento sin direction fija con efectos de ironfa, ansiedad, irrefle-xividad. A menudo se le anaden ideas de movimiento interno y/o externo. Es una perffrasis con un alto significado expresivo. El esloveno no dispone de este tipo de construction perifrastica y tiene que emplear otros medios lingtifsticos para expresar los matices expresivos y aspectuales de la perffrasis. En el ejemplo la el valor aspectual durativo/imperfectivo de andaba refunfunando se transmite al esloveno mediante el uso del verbo imperfectivo godrnjati en preterito pero los matices expresivos de la perffrasis no se transmiten: ko je Fernanda že godrnjala po kotih significa literalmente cuando Fernanda ya refunfunaba por los rincones. En la traduction eslovena, se pierde la idea del desarrollo de un movimiento sin direccion fija y el valor expresivo con matices de ironfa de la perffrasis verbal con andar. la) No bien Remedios, la bella, habfa subido al cielo en cuerpo y alma, y ya la des-considerada Fernanda andaba refunfunando en los rincones porque se habfa lle-vado las säbanas (CAS, 199) lb) Remedios, prelepa, je z dušo in telesom komaj izginila v nebo, ko je Fernanda že godrnjala po kotih, da ji je odnesla rjuhe.(SLS ,220) Con andar pidiendo (ej. 2a) el hablante no expresa solamente una action durativa reiterativa sino que se burla del interlocutor; en esloveno (ej. 2b) la perffrasis no se transmite sino se usa simplemente el presente (ker zahtevate - por pedir) sin los matices expresivos de la perffrasis verbal (por andar pidiendo). El tono sarcästico, sin embargo, se transmite a traves del resto del texto. 2a) - A mf no me venga a hablar mierda - le dijo -. Vamos a lo que se trata. Usted tiene todo empantanado por andar pidiendo huevonadas y aquf no hay sino una vaina: simplemente, los tipos tienen que entregarse y confesar algün delito por el cual se les puede meter doce afios. (NS, 112) 2b) »Nehajte ga srat,« mu je rekel. »Gremo k stvari. Vi ste vse skupaj zamočvirili, ker zahtevate sama jajca, gre pa samo za to: tipi se morajo vdati in priznati kak prestopek, za katerega lahko dobijo dvanajst let.« (PU, 104) Si el verbo auxiliar en la perffrasis andar + gerundio se encuentra en un tiempo verbal imperfectivo (preterito imperfecta), se focaliza una action desde la perspectiva o Anda acusändome por ahf de haber robado su dinero. Me anda acusando por ahf ... Me lo han ido revelando poco a poco. Han ido reveländomelo poco a poco. cursiva, sin su inicio ni su final (ejemplo la). Si el tiempo verbal del auxiliar es per-fectivo (preterito perfecta simple), la focalizaciön de la accion es completa con el inicio y el final (ej.: Anduve buscando mis llaves toda la tarde). 2.3 Ir + gerundio es una perffrasis aspectual con la que el hablante acentüa el trans-currir progresivo y la duracion de la accion. Si el predicado es puntual telico, indica tambien la repetition. El valor progresivo de la accion se refuerza con complementos circunstanciales de tiempo o modo: poco a poco, paulatinamente, progresivamente. Tiene un fuerte valor expresivo y se usa a menudo para describir la lenta progresion de las acciones. En la perspectiva cursiva, cuando la perffrasis verbal con el auxiliar en preterito imperfecto (u otro verbo imperfectivo) focaliza una accion sin indicar el inicio ni el final, se acentüan aün mäs el lento progresar, la duracion de la accion y se crean imägenes prolongadas. La perffrasis tiene un fuerte efecto estilfstico. En la perspectiva global (el auxiliar se encuentra en preterito perfecto simple), cuando se indi-can el inicio y el fin de la accion, la perffrasis verbal expresa la progresion de la accion vista globalmente, es decir, senala una accion progresiva pero limitada temporalmen-te. El ejemplo 3a muestra el uso estilistico de la perffrasis, donde un juego de focali-zaciones produce efectos especiales. En esloveno (3b) la diferencia entre la perspectiva global y la cursiva se expresa con ayuda de complementos circunstanciales (zmeraj bolj /čada vez mäs; polagoma/ gradualmente, poco a poco) y con los verbos en forma imperfectiva. 3a ) Poco a poco, sin embargo, y a medida que la guerra se iba intensificando v extendiendo. su imagen se fue borrando en un universo de irrealidad. Los pun-tos y ray as de su voz eran čada vez mäs remotos e inciertos, y se unian y com-binaban para formar palabras que paulatinamente fueron perdiendo todo senti-do. (CAS, 132) 3b) Kljub temu pa se je sčasoma in v enaki meri kot vojna, ki je bila zmeraj bolj razvlečena, njegova podoba razblinjala v vesolju neresničnosti. Pike in črte njegovega glasu so bile zmeraj bolj oddaljene in negotove, družile in prepletale so se, da bi tvorile besede, ki so polagoma izgubljale vsakršen smisel. (SLS, 145) Teniendo en cuenta la naturaleza del predicado la perffrasis en cuestion se combi-na tanto con los predicados durativos telicos, orientados a alcanzar una meta (ej. 7a) -lo que se explica con el propio verbo ir como verbo de movimiento de direction orientada (Garcia Fernandez 2006:175) -, como con predicados durativos atelicos que van acompanados de complementos indicadores de progresion gradual (ej. 8a). Con predicados puntuales telicos la perffrasis indica iteraciön (ej. 5a) y con predicados estativos se indica un proceso (Voy entendiendo el problema cada vez mejor - Garcia Fernändez 2006:174). Una caracterfstica de la perffrasis verbal ir + gerundio es tambien la de que el auxiliar y el auxiliado pueden ser el mismo verbo ir (La gente se iba yendo poco a poco - Gömez Torrego 1988:160). En los textos eslovenos analizados hay diferentes soluciones al problema de tra-ducciön de la perffrasis ir + gerundio. En muchos casos la version eslovena se empo-brece por falta de una traducciön adecuada de los valores aspectuales y estilisticos de la perffrasis. En 4b la traducciön eslovena no transmite el valor progresivo de la perffrasis. En esloveno ni bilo več no corresponde a se iban acabando, simplemente quie-re decir »cuando no habfa mäs (pastas)«. En 5b y 6b la traducciön de la perffrasis logra transmitir la progresividad mediante el uso de un verbo imperfectivo (se je opotekal, se razkraja), pero la traducciön eslovena serfa la mišma si en 5a y 6a tuvieramos un preterito imperfecta (ahora daba tumbos, se desmoronaba), es decir, que en la traducciön eslovena falta un realce de la progresividad. En 7b la traductora anade el adver-bio počasi (lentamente) al verbo en forma perfectiva (naberemo) que logra transmitir en parte el significado de la perffrasis del original espanol (vamos reuniendo) posible-mente por ser un predicado durativo telico. En 8b la progresividad se transmite con el adverbio sčasoma (con el tiempo), mientras que el verbo estä en presente. 4a) Devoraba guinda tras guinda y la masa la tiraba a una papelera. A medida que se iban acabando las pastas de ese tipo, los Camareros-Gorilas trafan mäs. (HDC 190) 4b) Goltal je eno višnjo za drugo, testo pa je metal v koš. Ko teh kolačkov ni bilo več, sta natakarja-gorili prinesla nove. (KH, 190) 5a) Ahora iba dando tumbos por la calle, con los ojos desenfocados y brillantes. (HDC, 191-192) 5b) Zdaj se je opotekal po cesti, z medlimi, sijočimi očmi. (KH, 192) 6a) El protagonista, no recuerdo ahora cömo se llamaba, habla participado de muchacho en la revoluciön de 1848, y de mayor mostraba el mismo desencan-to ante sus suenos juveniles que el que pude sentir yo al ver cömo se iba des-moronando el ideal libertario. (HDC, 327) 6b) Protagonist, zdaj se ne spomnim, kako mu je bilo ime, je kot fant sodeloval v revoluciji leta 1848, in ko je bil starejši, je bil nad svojimi mladostnimi sanjami enako razočaran kot jaz, ko sem videl, kako se razkraja anarhistični ideal. (KH; 324) 7a) Asi es que todos vamos reuniendo nuestro montoncito de porquerfas y lo lle-vamos rodando delante de nosotros como escarabajos peloteros (...) (HDC 328) 7b) Tako da počasi vsi naberemo svoj kupček gnoja, ki ga potem kot skarabeji kotalimo pred sabo. (KH, 325) 8a) - Pero hay algo que compensa todo eso, y es la sabidurfa. Al crecer ganas cono-cimiento. Es en el ünico registra de la vida en el que vas mejorando con el tiempo, pero es importante. (HDC, 329) 8b) »A nekaj vendar vse skupaj poplača, in to je modrost. Ko rasteš, pridobiš vedenje. To je edini register v življenju, v katerem si sčasoma vse boljši, in to je pomembno.« (KH, 325) 2.4 Venir + gerundio es una perffrasis verbal aspectual que indica acciones durativas y progresivas. Focaliza la accion desde su inicio hasta un momento central sin definir el final (perspectiva imperfectiva continuativa). Es incompatible con verbos que expre-san alejamiento, rasgo que la diferencia de la perffrasis con ir. Ello se debe al valor defctico del auxiliar venir que indica acercamiento al lugar del hablante desde un punto determinado: La tormenta se viene acercando !La tormenta se va acercando. Es agra-matical *la tormenta se viene alejando, pero correcto la tormenta se va alejando. La perffrasis tiene restricciones en cuanto a los tiempos verbales: predomina el pre-sente y el preterito imperfecto, tambien son posibles las formas compuestas (preterito perfecto compuesto y preterito pluscuamperfecto), pero no el preterito perfecto simple que aspectualmente focaliza la accion en su inicio y en su final. El esloveno no conoce este tipo de estructura perifrästica. En los textos revisados hay pocas perffrasis de gerundio con el auxiliar venir. En las traducciones las traduc-toras encuentran diferentes soluciones. En el ejemplo 10a la idea del desarrollo expre-sada por la perffrasis indicando el acercamiento hacia el presente, desde un punto ini-cial hacia el momento del habla sin indicar el final, estä traducida por el sustantivo potek boja (el desarrollo de la lucha). El preterito perfecto compuesto acentüa el valor aspectual de una accion iniciada en el pasado que se desarrolla hacia el momento del habla pero queda abierta, sin indication del final.9 En el ejemplo lib, en la traduction eslovena, el valor aspectual imperfectivo continuativo se traduce con un verbo imper-fectivo (razmišljati). El complemento desde hacia tiempo / že precej časa le da el valor de acercamiento de un pasado al presente del hablante. 9a) »Recuerde, doctor Villamizar, que la extradition ha cobrado muchas victimas, y sumarle dos nuevas no alterarä mucho el proceso ni la lucha que se ha veni-do desarrollando». (NS, 215) 9b) »Ne pozabite, dr. Villamizar, daje ekstradicija pokosila že veliko žrtev; dve več ne bosta kaj prida vplivali na ta proces in na potek boja.« (PU, 199-200) 10a) Este capto la idea de inmediato, pues el tambien venfa pensando desde hac/a tiempo en una manera de judicializar el problema del narcoträfico. (NS, 84) 10b) Ta je pri priči dojel zamisel, saj je tudi sam že precej časa razmišljal, kako bi se dalo pravno urediti problem preprodaje mamil. (PU, 78) 9 Este uso aspectual del preterito perfecto compuesto es frecuente en el espanol de America, ya que el uso temporal de presente reciente frecuente en el espanol peninsular no es usual en la variedad americana del espanol. 3. Conclusion El breve anälisis de las perffrasis verbales de gerundio con los auxiliares andar, ir y venir y de los valores aspectuales que se expresan con estas perffrasis demuestra que son unos potentes recursos lingüisticos para la expresiön de la aspectualidad, poseen un fuerte valor expresivo y se usan para lograr efectos estilfsticos especiales. Las tra-ducciones al esloveno plantean el problema de la traductibilidad de los matices expre-sados por estas estructuras perifrästicas y la necesidad de analizarlas en profundidad para entenderlas y encontrar los medios correspondientes en la lengua eslovena que puedan transmitir toda la riqueza de significados tempo-aspectuales y expresivos de las perffrasis verbales espanolas. Referencias bibiiogräficas Bello, A. (1847/1988), Gramätica de la lengua castellana destinada al uso de los americanos, I, II con las notas de Rufino Jose Cuervo. Estudio y ediciön de Raraön Trujillo. Madrid: Arco Libros. Dietrich, W. (1983), El aspecto verbal perifrästico en las lenguas romdnicas. Madrid: Gredos. Fernändez de Castro, F. (1990), Las perifrasis verbales en espanol. Comportamientos sintäcticos e historia de su caracterizaciön. Oviedo: Series Minor. Garci'a Fernändez, L. (2006), Diccionario de perifrasis verbales. Madrid: Gredos Gömez Torrego, L. (1988), Perifrasis verbales. Sintaxis, semantica y estilistica. Madrid: Arco/Libros. Markič, J. (1990), "Sobre las perifrasis verbales en espannol", Linguistica, XXX. 169-206 Toporišič, J. (1984), Slovenska slovnica. Maribor: Založba Obzorja Yllera, A. (1999), "Las perifrasis verbales de gerundio y participio", en I. Bosque y V. Demonte (eds.), Gramätica descriptiva de la lengua espanola. Madrid: Espasa-Calpe. 3391-3441. Textos citados GarcIa Märquez, G. (1986), den anos de soledad, Bogota: Editorial La Oveja Negra. (CAS) GarcIa Märquez, G. (1997), Sto let samote. Ljubljana: MK. Prevod Alenka Bole Vrabec. (SAS) Garcia Märquez, G. (1996), Noticia de un secuestro. Barcelona: Grijalbo Mondadori. (NS) Garcia Märquez, G. (1996), Poročilo o ugrabitvi. Ljubljana: MK. Prevod Vesna Velkovrh Bukilica. (PU) Montero, R. (1997), La hija del Cambal. Madrid: 1997 (HDC) Montero, R. (2000), Kanibalova hči. Ljubljana: MK. Prevod Barbara Vuga (KH) Povzetek VREDNOSTI IN RABE GLAGOLSKIH PERIFRAZ S POMOŽNIKI IR, ANDAR IN VENIR V GERUNDIJU Uvodni del prispevka predstavlja problematiko perifrastičnih struktur, ki jih špansko jezikoslovje opredeljuje kot glagolske strukture, sestavljene iz dveh glagolov, pomožnika in glavnega glagola v neosebni obliki (nedoločniku, gerundiju ali participu), s katerimi govorci izražajo aspektualne, časovne, modalne in pragmatične pomene. Glagola sta povezana neposredno ali posredno s predlogi in vezniki, nastopata pa kot enovit povedek. V osrednjem delu je govora o glagolskih perifrazah z gerundijem, predvsem o tistih, ki imajo za pomožnike glagole gibanja andar, ir in venir. Opredeljene so njihove aspektualne vrednosti in slogovne rabe. Predstavljeni so tudi nekateri primeri rabe teh struktur iz pripovednih del Rose Montero in Gabriela Garcfe Märqueza ter analiza prevodov v slovenščino, ki prikazuje (ne)možnosti prevajanja teh perifrastičnih struktur. Vladimir Pogačnik Universite de Ljubljana UDK 81*373.21 Ljubljana NQUVELLE CONTRIBUTION A L'ETYMOLOGIE DU TOPONYME LJUBLJANA Ce n'est sans doute pas un hasard si les toponymes Ljubljana, Lubiana et Laibach se sont vu attribuer une serie d'explications etymologiques et ceci par d'eminents lin-guistes slovenes tels que Miklošič (1859), Kelemina (1950), Ramovš (1953), Bezlaj (1956) auxquels on ajoutera egalement le grand croate Petar Skok (1934) et surtout ä la fin des annees soixante-dix, Anton Grad, professeur titulaire ä la chaire de fran9ais de 1'Universite de Ljubljana (1978, 1980). Pour ce qui conceme Ljubljana les slavisants auraient choisi une racine anthropo-morphique slave L'ub ä laquelle s'ajouteraient des suffixes ethniques du type -jan-ine au pluriel; Kelemina a prefere le choix d'une base adjectivale l'ub suivie des suffixes -ja-na. Notons cependant que Bezlaj reste inexplicablement indecis et qu'il laisse ouvertes toutes hypotheses possibles et imaginables, entre autres une origine preslave de l'hydronyme initial Ljubija. Quant ä Grad, il bätit son explication sur une origine romane (6 combien acceptable de nos jours et combien mal vue, il y a un bon quart de siecle !) - *alluviana, forme elliptique de (aqua) *alluviana »eau qui deborde, inonde, submerge le terrain riverain, y depose de la boue etc., cree des marais, ce qui, en effet, a toujours ete (et est encore) un trait caracteristique de la riviere en question«. Ces interpretations semantiques nous semblent effectivement fort acceptables. II est evident, cependant, que les deve-loppements phonetiques vers les formes Lubiana, Ljubljana etLaibach ont pu paraitre, ä l'epoque, quelque peu simplificateurs, pourtant le fait qu'il n'y ait pas eu d'oppo -sition definitive, laisse l'espoir qu'elles peuvent paraitre plausibles encore aujourd'hui, bien que parfois non tres ortodoxes. Mis ä part le respect que nous devons ä notre eminent predecesseur ä la chaire, mais peut-etre mus par le desir de renouveler le defi en matiere diachrono-socio-linguis-tique, plus particulierement ä l'occasion de ce volume de Linguistica que son meritant redacteur en chef, Mitja Skubic, offre en hommage ä l'academicien Janez Orešnik (lequel aurait fourni des donnees enrichissantes au dernier »interprete« du toponyme de Ljubljana), nous proposons deux solutions possibles et de par ce fait critiquables : 1) *luppa: etymon celtique signifiant » argile, terre grasse, glaise« ou encore preceltique *lappa, signifiant »schlämm« + i-ana (suffixe hydronimique latin): Wartburg, FEW, V, J-L(p.457) 2) *lupia (luoghi sabbionosi ed incolti; campi dove il fiume si spande (onde il verbo lubia 'scoscendere') + -ana (suffixe hydronymique latin): D. Olivieri: Toponomastica Veneta (p. 105) Nous avons une petite preference pour la premiere proposition: elle reconcilie les deux variantes ulterieures: la latino-slave Lubiana-Lubljana et la germanique Laibach selon les gradations phonetiques proposes par A. Grad(1980), et correspond ä la fois aux exigences semantiques et logistiques de la sociolinguistique moderne. Notons que le toponyme d'une grande metropole culturelle europeenne Louvain s'alignera peut-etre un jour, si ce n'est pas tres prochainement, dans ce contexte precis, ä celui de la capitale Slovene. Et pourquoi pas le Louvre - greve inondee? Et le village Slovene de Ljubno dans la valee de la Savinja, ayant depuis son existence, des problemes avec l'eau »roide« qui l'inonde? II nous faut avouer cependant que, voulant rester fideles aux pistes romanes parcourues par le professeur Grad, nous preconisons aussi l'etymologie »louparde« de notre capitale, bien que contestee par A. Dauzat (et al.: Diet. etym. des noms des rivieres...): »les formes medievales lupa sont de mauvaises latinisations (par attraction de lupus, loup); des rivieres dormantes ne peuvent avoir re£u le loup pour embleme et le -p- serait devenu -v-; il s'agit d'une rac. hydron. low-, sans doute preceltique, cf. irlandais lo, eau (p.61).« Dans un contexte plus leger, culturel cependant, il faudrait citer les voisins europeens: Louviers, Louveciennes, La Louviere. Et pourquoi ne pas leur associer Luwigana (?), ancienne appelation de Ljubljana, Ljubno en Haute-Carniole (loin de toute eau inondatrice), et Ljubelj, col de montagne se pretant parfaitement ä la denture du loup - Lupillum, et rejoignant tant d'exemples sur le territoire de la Gallia »romanica«? Sans parier de Ljubogojna, Ljubečna, Ljubela, Ljubinski potok, Libeliče; Voše, Vovše, Volčji potok, Volčja draga etc. La confusion est parfaite. La phonetique historique impose ses regies, parfois quelque peu trop rigides... A qui d'instituer l'ordre? Aux scientifiques, bien evidemment. La course au premier prix est ouverte... Bibliographie: Bezlaj, F.- Slovenska vodna imena, I., Ljubljana, SAZU, 1956 Dauzat, A. - La Toponymie frangaise, Paris, Payot 1946 Dauzat, A. et al.- Dictionnaire etymologique des noms de rivieres et de montagnes en France, Paris, Ed. Klincksieck, 1978 Grad, A.- K etimologiji toponima Ljubljana; Onomastica jugoslavica, 7, Zagreb 1978 Grad, A,- Prispevek k etimologiji toponimov Ljubljana, Lubiana, Laibach; Slavistična revija, 1, Ljubljana, 1980 Olivieri, D,- Toponomastica veneta, Istituto per la collaborazione culturale, Venezia-Roma, Schlamberger M., Pogačnik V., Perko G.- Parlons Slovene, Paris, THarmattan, 2005 Wartburg, W. von.- Französisches Etymologisches Wörterbuch, V., Basel.Helbing&Lichtenhahn, 1950 Povzetek NOV PRISPEVEK K ETIMOLOGIJI TOPONIMA LJUBLJANA Prispevek k obnovljenemu razmišljanju o etimologiji toponima Ljubljana nakazuje več možnih, a vsekakor neslovanskih oz. neslovenističnih rešitev: keltski etimon luppa, predkeltski lappa (oba: blato, močvirje, morost)- slednji se bo verjetno bolje prilegal germanskemu postreliktu Laibach - ter romanski lupia »volčji vrelec, ponikovalne narave«, hidronimi, ki se zdijo naravnejši in zlasti manj moteče »tehnični« od Gradove »alluviane«, poplavne reke, in ki se več ali manj vsi kar dobro prekrivajo s prvim vrelcem Ljubljanice, Ljubijo. Kar zadeva vodotočno pripono -ana in vse različne fonetične špekulacije, se avtor ne spušča v polemiko s prof. Gradom niti z njegovimi morebitnimi oponenti, misleč skromno, da ni nič bolj fonetično od nefonetičnega, seveda v strukturiranem jezikovnem kontekstu. VSEBINA - SOMMAIRE Hans Goebl Sur le changement macrolinguistique survenu entre 1300 et 1900 dans le domaine d'oi'l. une etude diachronique d'lnspiration dialectometrique Jezikovne inovacije med 14. in 20. stoletjem na ozemlju francoščine. Diahronična študija z vidika dialektometrije..............................................3 Marek Gawelko Sur un aspect des etudes historiques Razvojni vidik v indoevropskih jezikih......................................................45 Zuzanna Topolinjska Toward a definition of the generic noun phrase K definiciji t.i. generične imenske sintagme............................................55 Lara Burazer Referring and the context of situation in view of the Human Linguistics approach Pojmovanje reference (nanašanja) in situacijskega konteksta s stališča jezikoslovja človeka.....................................................................................61 Martina Ožbot On the possibilities and impossibilities of cross-cultural communication O možnostih in nemožnostih medkulturne komunikacije............................71 Karmen Teržan Кореску Translate als kognitive Entitäten oder die Natürlichkeitstheorie im Kontext der sprachlichen und psychosozialen Paradigmata der mehrsprachlichen Kommunikation Translati kot spoznavne entitete ali teorija naravnosti v kontekstu jezikovnih in psihosocialnih paradigem večjezikovnega sporazumevanja.................... 81 Teodor Petrič Vorschläge zur Erweiterung der slowenischen Variante der Natürlichkeitstheorie Predlogi za razširitve slovenske različice teorije naravnosti......................99 Varja Cvetko-Orešnik The aorist and the perfect of the Old Indian causatives in the light of Natural Morphosyntax Aorist in perfekt staroindijskih kavzativov v luči naravne oblikoskladnje.....117 Zorica Vučetić il linguaggio artistico. il lessico Jezik umetnosti. Besedje..........................................................................121 T. L. Markey Early Celticity in slovenia and at Rhaetic magre (schio) Zgodnja keltščina v Sloveniji in v retskem kraju Magre (Schio).............145 Irena Orel Prepositional phrases with verba dicendi from Dalmatin's translation of the Bible (1584) in relation to foreign language translations Predložne zveze ob glagolih govorjenja iz Dalmatinovega prevoda Biblije (1584) v razmerju s tujejezičnimi prevodi................................................173 Alja Ferme Final sonorant sequences in the Celje dialect Zvočniška izglasja v celjskem govoru........................................................ 181 Chikako Shigemori Bučar Causative constructions in Japanese and Slovene Vzročniške zgradbe v japonščini in slovenščini.......................................191 Aleksandra Derganc Trije zgledi neposrečenega prevoda zaradi »lažnih prijateljev« Three instances of incorrect translations due to false friends................203 Eva Sicherl, Andreja Žele Combinability with prepositions in Slovene and English: theoretical and contrastive views Predložne zveze oz. predložnomorfemskost v slovenščini in angleščini: teoretični in kontrastivni vidik.................................................................209 Marjeta Vrbinc, Alenka Vrbinc a research-based study of foreign students' use of grammatical codes in five leading british learners' dictionaries Raziskava o uporabi slovničnih kod v petih vodilnih britanskih enojezičnih slovarjih za tujce na vzorcu tujih govorcev...............................................227 Jasmina Markič Valores y usos de las perifrasis verbales de gerundio con los auxiliares ir, andar y venir Vrednosti in rabe glagolskih perifraz s pomožniki ir, andar in venir v gerundiju...................................................................................................243 Vladimir Pogačnik nouvelle contribution a l'etymologie du toponyme ljubljana Nov prispevek k etimologiji toponima Ljubljana....................................251 LINGUISTICA XLVI Izdala in založila Filozofska fakulteta Univerze v Ljubljani Revue publiee et editee par la Faculte des Lettres et Philosophie de l'Universite de Ljubljana Glavni in odgovorni urednik - Redacteur en chef Mitja Skubic Tajnica redakcije - Secretaire de la redaction Jožica Pire Nasloviti vse dopise na naslov Priere d'adresser toute correspondance ä Mitja Skubic Filozofska fakulteta Aškerčeva 2 SI-1000 Ljubljana linguistica@ff.uni-lj .si Tel.: +386 1 241 14 06 Fax: +386 1 425 93 37 Računalniški prelom - Mise en page Žiga Hrček Tisk - Imprimerie Tiskarna Littera pieta, d.o.o. Rožna dolina c. IV/32, SI-1000 Ljubljana