L I N G U I S TI C A
UREJUJETA MILAN GROŠELJ IN STANKO SKERLJ
LETO I	19 5 5	ŠT. 1
OB IZIDU PRVE ŠTEVILKE
Slovenski lingvisti — razen slavistov — doslej niso imeli periodičnega glasila, v katerem bi lahko objavljali kratke razprave in članke, ocene in beležke. (V glavnem velja to tudi za lingviste v ostalih naših republikah.) Kolikor so obravnavali predmete, ki so se tikali tudi slavistike, so vedno našli gostoljubno streho v Slavistični reviji. Ni pa bilo mogoče obremenjevati slavističnega glasila s prispevki, popolnoma posvečenimi na primer klasični filologiji ali romanistiki ali splošni lingvistiki; vrh tega bi bilo nesmotrno, take članke objavljati tam, kjer jih po pravici nihče ne pričakuje in ne išče. In še eno vprašanje je važno: ali je umestno, take specialistične razprave objavljati v slovenščini — kar je pri slavističnih v slavistični reviji samo po sebi umevno —, ko so v glavnem namenjeni bralcem (klasičnim filologom, romanistom itd.) v raznih deželah, ki našega jezika ne razumejo dovolj, iz posnetka pa ne morejo temeljito povzeti vse vsebine člankov.
Iz takih misli se je v Lingvističnem krožku pri filozofski fakulteti v Ljubljani rodil načrt, da osnujemo skromno glasilo, kjer naj imajo svoje mesto neslavistični prispevki — taki, ki zadevajo tudi slavistično področje, pa le izjemoma, kadar bo namreč težišče obravnavanega problema izven tega področja — in kjer se bodo lahko v celoti objavljali v tujih jezikih (s slovenskim posnetkom).
Uresničitev našega načrta je bila v sedanjem trenutku omogočena s tem, da je vodstvo Slavistične revije pristalo na to, da se temu že uglednemu časopisu priključijo, kot apendiks, naša Linguistica, in da je SPK LRS dal na razpolago potrebna sredstva. Za naklonjeno razumevanje naših pobud izrekamo Svetu za prosveto in kulturo in izdajateljem Slavistične revije iskreno zahvalo.
^	Uredništvo
A l'exception des slavistes, les linguistes slovenes 11'avaient pas jusqu'ici d'organe periodique pour y publier des etudes et des articles de moindre etendue, des coinptes-rendus, des notes. (D'une fa?on generale, c'est aussi le cas pour les linguistes dans les autres Republiques yougoslaves.) Dans la mesure oil ils traitaient ties sujets ayant aussi rapport aux etudes slaves, ils pouvadent tou-jours etre surs de l'hospitalite de la Slavistična revija (Revue de Slavistique) de Ljubljana. Mais il etait impossible de charger l'organe de slavistique de contributions entiercmcnt consucrees a la philologie classique, par exemple, ou a la philologie romane ou a la linguistique generale. De plus, il ne semblait pas raisonnable de faire paraitre de tels articles dans un periodique ou, a dire vrai,
t Linguistica
i
nul ne les attendrait et par consequent ne les chercherait. Enfin, comme il s'agit d'eiudes specialisees qu'on desircrait porter entierement — et non pas seulement en resume — a la connaissance et a la discussion d'un public plus etendu que celui de notre pays, nous nous demandions s'il n'etait pas utile de les publier dans une langue etrangere.
C'est de ces reflexions qu'est nee, au sein du Cercle linguistique a la Faculte des Lettres de Ljubljana, l'idee de fonder un modeste organe ou auraient leur place, redigees, si l'auteur le desire, en langue etrangere, des contributions de contenu non-slavistique ou d'autres qui, tout en touchant a des questions de philologie slave, traiteront des sujets appartenant essentiellement a d'autres dornaines de linguistique.
La realisation de notre projet est devenue possible du moment ou la Direction de la Slavistična revija a accepte que nos Linguistica se joignent, en appen-dice, a ce periodique deja connu dans le monde des linguistes, et ou le C-onseil pour l'lnstruction publique et la Culture de la Republique Populaire de Slovenie a mis a la disposition de la Direction les moyens necessaires. Nous tenons a exprimer au Conseil pour l'lnstruction publique et la Culture ainsi qu'aux editeurs de la Reoue de Slaoistique notre vive reconnaissance pour leur bien-veillante comprehension de nos buts et de nos efforts.	j DirecfeUTS
Milan Grošelj
O POSESIVNEM ADJEKTIVU V SLOVANŠČINI IN TOHARŠČINI
V sla. jezikih je la posebnost, da se ne reče hiša očeta, klobuk brata, ampak očetova hiša, bratov klobuk, ne klobuk sestre, ampak sestrin klobuk. Še starejši je sufiks -ji: bil je neindividualno svojilen, na njegovo mesto je očitno z uvedbo zasebne lastnine stopil -oo, -in, ki se rabi, če gre za enega lastnika. Gen. stoji le, če ima poleg sebe atribut: hiša mojega očeta. Tako je z izrazi za posest že v stcsl. (Vondrak2 II 231 s.). Kakor nas pouči Nahtigal, Slovanski jeziki2 223 s., je v ruščini in poljščini raba po«, adjektiva »dandanes že precej omejena«, nasprotno SO' pa ti adjektivi v češčini »še povsem v rabi« in v južni slovanščini so v »polni moči«. O starosti tega načina izražanja v sla. jezikih torej ne more biti dvoma.
Pos. adjektiv pozna izmed ievr. jezikov le še toharščina. Ta jezik so govorili v prvem tisočletju po n. št. v Turkestanu, prve ostanke tega jezika so objavili leta 1908. Sieg-Siegling-Schulze, Toch. Gramm. § 41 ugotavljajo: toh. »rada izraža, podobno kot sla. jeziki, pripadnost s pos. adjektivi tudi tedaj, ko bi drugi jeziki imeli rajši genetiv ali sestav-
Milan Grošelj: O p osesivnem adjektivu
Ijenko«. W. Schulze, eden izmed avtorjev, je gotovo imel pregled čez ievr. jezike. Vprašanje je potemtakem, ali je kakšna zveza med rabo pos. adjektiva v sla. jezikih in v toh. in, če je, od kod ta raba izvira.
J. Wackernagel je leta 1908 postal pozoren na pos. a d jektiv v grščini in latinščini (Melanges Saussure 137 ss.: Genitiv und Adjektiv). Trdil je, da je tudi v teh dveh jezikih bil prvotno adjektiv namesto pos. genetiva kakor v sla. jezikih. Toda znameniti švedski latinist Lofstedt (Syntac-tica I1 83 ss.) je 20 let nato to trditev zavrnil in pokazal, da je v grščini pos. adjektiv omejen na homerski jezik, ki ima to posebnost od eolskega narečja, in da so ti adjektivi v glavnem patronimika, n. pr. TtXa^wvios Aiag. Dobijo se pa pri Homerju tudi v pravem posesivnem pomenu, n. pr. vi]i>s 'Ayauefivover]. Posebno je treba omeniti, da se pri Homerju dobijo tudi pos. adjektivi, ki so napravljeni od občih imen, n. pr. xaAxfiiov i; 6ofiov »v kovačevo hišo«. Tudi v latinščini je ta adjektiv omejen v glavnem na patronimika, ki so pa v zgodovinski dobi že dedna gentilna imena: (Marcus) Tullius (Cicero) je bil najprej »Tulov (sin)«. Sicer je v rabi le patrius, ki mu ustreza skr. pitriyah, gr. naiQios (za mater takšnega enotnega adjektiva ni, ker mati v patriarhalni družbi, kakršna je postala indoevropska iz matriarhalne, ni imela lastnine); v stari latinščini (Pit.) je v rabi erilis »gospodarjev«, pozneje dominicus ozir. -a (dies), iz česar je fr. le dimanche, ital. /a domenica »nedelja«, in še nekaj drugih, ki so osamljeni v raznih nazivih. Po tem Lofstedtovem popravku je stanje takšno, da je pos. adjektiv živ le v sla. jezikih in toharščini, medtem ko so patronimika v rabi tudi še v sanskrtu, iranščini, ilirščini, venetščini in germanščini.
Jezikoslovje se je razvilo pred 150 leti, ko so začeli primerjati jezike med seboj. Temeljno načelo je, da ne verujemo v spontano nastajanje podobnih oblik, besed in sintaktičnih zvez, ampak da vedno najprej poskušamo najti neko zvezo med podobnimi pojavi, tudi če so zemljepisno tako daleč narazen, kakor so sla. jeziki in toharščina. Skupine ljudi so se namreč razselile z mnogo ožjega prostora. Kje je bila prvotna ievr. domovina, o tem so bila že razna mnenja: iskali so jo od Mezopotamije do Nemčije. H. Krahe je v predavanju Sprachnermandtschaft im alten Europa (1951) prišel do zaključka, da so bili zametki ali pra-celice ievr. jezikov okoli leta 2000 med Skandinavijo in Alpami ter Atlantikom in Vislo (ali še malo čez njo proti vzhodu). Do tega naziranja je prišel takole. Na omenjenem področju se dobijo številna rečna imena, ki so izvedena iz korenov *au-, *al-, *ar-, *ak'J- itd. s pomenom »voda« ali pod., prim. Ooa (pritok Njemena; O- namesto A- je pod sla. vplivom),
i'
3
Ala (v Skandinaviji), Oka (pri Moskvi). Te korene najdemo podaljšane s sufiksi -antia, -entia, -ara ipd., prim. Aoantia, Aoentia, Aoara; Alantia, Alara; Arantia itd. Ista imena najdemo tudi južno od Alp, toda tam so bolj na redko posejana, in sicer med imena sredozemskega izvora, ki so nam nerazumljiva. Indoevropci so jih na jug zanesli šele po letu 2000. Ker se rečna imena od vseh najbolje ohranijo, so po Kraheju najzanesljivejša priča o nekdanjih jezikih; kažejo, da so na omenjenem ozemlju severno od Alp bili okoli leta 2000 sosedi predniki naslednjih jezikov ozir. narodov: na severu so bili Germani, južno od njih od zahoda proti vzhodu pa Latino-Faliski in Italiki (Osko-Umbri), Iliri in Veneti ter Balti; južno od Italikov si je treba misliti Kelte. Germani imajo namreč z Latini skupno ime za bron (lat. aes, got. aiz, stnord. eir, stvn. er), s Kelti pa skupno za železo (gall, isarno-, stir, lam, germ, isarnan, stvn. isam »Eisen«). Torej so Germani prišli s Kelti v stik šele, ko so se odselili Latini. Za Grke in Hetite pravi Krahe, da so morali biti izven te skupine: njihovo bivališče za ta čas da ni znano. Najverjetneje se mu zdi, da so Grki prebivali vzhodno od tega področja, manj verjetno, da so se odselili, še preden so nastala omenjena rečna imena. Vsekakor so pa Grki odšli proti jugu pred Italiki, Iliri in Veneti.
Zgoraj navedenih rečnih imen res ni v Grčiji. Vzrok je morda tudi v tem, ker so Grki prišli v kraje, ki so bili gosto naseljeni in kjer so rekam bila že dana imena, ki so jih Grki kar prevzeli. Zaključki ex si-lentio niso zanesljivi. Krahe, Beitrage z ur Namenforschung 4 (1953), 47 s. pogreša obča imena, iz katerih so omenjena rečna imena šele nastala. Opozoril bi na to, da se ravno v grščini dobijo. To so &a~ auaxruia vdatog (lies., Phot.) »vsedlina vode«, č/Us »motna tekočina« in dpdj »vodeni del mleka, sirotka«: aa je moglo nastati iz *aua; 6A.6s in dpig sta v prevoju s kor. *al- in *ar- (prim. &y<o : Syftos). Reka Ooa ima torej natančno isto ime *aua, ki je ohranjeno v grščini, Al a ustreza maskulinu dAog, grškega dpčg ni treba več vezati z latinskim serum (kjer je spiritus lenis namesto pričakovanega ostrega pridiha delal težave: pomagali so si s tem, da so v 6g6g videli jonsko obliko).
Ker so torej v grščini ohranjeni omenjeni koreni v občih imenih, ki so vendar prvotnejša od lastnih imen, sledi, da so se Grki ali ločili od omenjenih skupnosti jezikov, preden je nastal omenjeni sistem rečnih imen iz kor. *au-, *al-, *ar-, ali pa da so Germani i. dr. pustili Grke za seboj in odšli proti zahodu ter se naselili na ozemlju, kjer jih najdemo po Kraheju okoli leta 2000. Vse je odvisno od smeri naseljevanja, toda ta nam do sedaj še ni zanesljivo znana.
Področje severno od Alp seveda ni bilo prazno, arheologija ima zato nešteto dokazov v ostankih starih kultur. Za zahodno Nemčijo, Francijo, Britanijo in Irsko pa tudi za Španijo je arheologija in antropologija dognala za predzgodovinsko dobo vsaj že štiri selitve: ljudi stare in srednje kamene dobe, sredozemskega človeka nizke rasti, ki se je širil iz Afrike, človeka visoke rasti, ki je — zopet iz Afrike — prinesel s seboj megalitsko kulturo, in končno, okoli leta 1850, človeka, ki je zanj značilna čaša zvončaste oblike. Indoevropci so torej dali neke skupne poteze raznim rodovom, ki so pa v izgovoru in besedišču, v načinu izražanja ohranili nekaj svojih posebnosti. Tudi je verjetno, da se je proces indo-evropeizacije starejšega prebivalstva vršil v več fazah, med katerimi so bili krajši ali daljši presledki. Nove oblike in sintaktične zveze so se širile iz raznih 6redišč v obliki koncentričnih krogov (»valovna teorija«). Okoli pracelic germanščine, latinščine itd. so se jeziki šele polagoma izkristalizirali, ko so nastale večje družbene celote.
Izven omenjenega jezikovnega področja so ostali poleg Hetitov, Indo-irancev i. dr. tudi Slovani in Toharci. Vprašanje je, ali so bili Slovani in Toharci kdaj bližnji sosedje. Če so bili, potem bi mogli rabo pos. adjektiva v teh dveh jezikih povezati med seboj. Na soseščino dveh ali več jezikov je pa mogoče sklepati iz skupnih posebnosti, ki jih drugi jeziki nimajo. W. Porzig, Die Gliederung des idg. Sprachgebiets (Heidelberg 1954), res ugotavlja na str. 182 ss., da sta si bila ta dva jezika nekoč soseda. Na to kažejo sledeča dejstva: sla. sufiksu -ynja (sle. blag-inja, drag-) ustreza le lit. -une in toh. -une. Drugje tega sufiksa ni. Adjektiv na -lo-, ki ga poznajo razni ievr. jeziki, so vključili v glagolski sistem le sla. jeziki (bi-l, dela-l), armenščina in toharščina. Isti trije jeziki so spojili sufiksa -sko- in -io- v -skio-: od tod sla. -i-šte (sle. -i-šče; druge razlage glej pri Bajcu, Besedotvorje I § 206), arm. -(a)fi in toh. A -si, B -sse. Od inovacij v pomenu besed naj omenim le, da je kor. *rek~, ki v indoir. in germ, jezikih pomeni le »določiti« (ta pomen je ohranjen tudi v sle. rok »termin«), le v sla. jezikih in toharščini dobil pomen »reči«. To ne more biti slučaj. Važno je sledeče: neki jezik, n. pr. tohar-ski, more imeti skupno posebnost z enim jezikom ali s skupino jezikov. Če se oziramo na en jezik, pravi Porzig 183 s., »moramo po tem, kar sedaj vemo, domnevati, da je toharščina nastala v bližini grščine, trako-frigijščine in slovanščine«. Če pa upoštevamo skupine jezikov, potem kažejo inovacije na to, da je toharščina nastala v bližini baltščine, germanščine in slovanščine (str. 187). V obeh primerih je torej bila soseda
slovanščine toharščina. Na podlagi vsega tega smemo sklepati, da je tudi raba pos. adjektiva v teli dveh jezikih enotnega izvora.
Zdaj se moremo vprašati, na katerem kraju sta si bili slovanščina in toharščina sosedi. Spomnimo se zopet Krahejeve lokalizacije ievr. jezikov za konec tretjega tisočletja! Domnevati moramo, da je bila pra-celica slovanščine vzhodno od Visle in severno od Črnega morja. Seveda so se predniki Slovanov morali najkasneje okoli leta 750 umakniti od Črnega morja pred invazijo Skitov. Grki so ob Črnem morju poznali le Skite.
Toharci so v prvem tisočletju po n. št. živeli v Turkestanu. Tja so morali priti od zahoda neznano kdaj. Sosedje Slovanov so mogli biti edino v bližini Črnega morja. Na tem področju so morali skupaj s Slovani razviti ali od kod drugod prevzeti rabo pos. adjektiva.
Vzrok za jezikovne inovacije so lahko nove razmere v prirodi ali družbi. Nastati pa morejo tudi, če jezikovna skupina pride v stik z drugim jezikom. Tudi od jezika zasedene dežele še vedno nekaj ostane kot plast pod jezikom prišlecev: jezikovni substrat.
Raba pos. adjektiva je omejena na področje Črnega morja. Od grških narečij jo pozna edino eolščina, ta je pa tista, ki leži Črnemu morju najbliže. Tu je tudi prastaro jezikovno in kulturno področje: Mala Azija s Kavkazom. Kavkaz je s svojimi težko pristopnimi pobočji in dolinami zadnje pribežališče za plemena, ki so' nekoč stanovala daleč proti zahodu, čez Malo Azijo do Grčije. Severno od Kavkaza so kav-kaška plemena segala do Kubana še do velike selitve leta 1864. Še danes, ko so ta plemena stisnjena na Kavkazu, govore tam čez 30 jezikov; Strabon pa pripoveduje (VII p. 498), da je bilo na sejmu v Dioskuriadi (kjer se Kavkaz spušča v morje) slišati 70 jezikov. Koliko jezikovnih plasti je bilo v Mali Aziji in na Kavkazu, ni še niti približno dognano. Saj niti ne vedo, koliko so severni kavkaški jeziki sorodni z južnimi. Za vpliv teh jezikov, ki so v starem veku segali daleč proti zahodu, naj navedem le tri primere.
Po Grčiji in Mali Aziji so razširjena številna imena mest, rek in gora na -ssos (at. -ttos): Halikarnessos, Ilissos, Parnassos. Sufiks je neievr., ravno tako tudi koreni. Torej ta in druga imena izvirajo od predievr. prebivalcev. To staro prebivalstvo so bili Lelegi ali Legi Afjyeg, Ar\yai): Le- v Ac'Xeyeg je pluralni prefiks, sg. tega imena nam je ohranil gramatik Herodian: A<?|. Legi so bili v prvem stoletju pred n. št. le še malo kav-kaško pleme na severni strani Kavkaza ob Kaspiškem morju. To so današnji Lahki: imeni sta identični. V štirinajstem stoletju pred n. št.
se je pa z utrjenim mestom Lakkus vojskoval hetitski kralj Mursilis II., dokler ga ni zavzel. — V jeziku današnjih Lakov je še sedaj živ sufiks -ssa, ki služi za tvorbo adjektivov od poljubnih substantivov.
Drugi primer: v hetitščini se ojevska debla končujejo v nom. sg. na -as < *-os, gen. se pa tudi končuje na -as, kar je nenavadno. Edina sprejemljiva razlaga za to dejstvo je, da je gen. na -as v resnici pos. adjektiv (Kretschmer, Glolta 21, 94 s.). Seveda ta adjektiv ni indoevropski, ampak hatski. Ko so Hetiti vdrli od vzhoda v Malo Azijo, so tam našli neievr. Hate; da ne pride do mešanja s Hetiti, ki so njihovo ime podedovali pri sosedih (tudi Armenci se imenujejo sami Hajkli), imenujemo Hate rajši Protohate.
Tretji primer: v ievr. jezikih je ta posebnost, da predmetom (stolu, mizi) pripisujemo (gramatični) spol: moški ali ženski, ko< bi vendar pričakovali, da bodo predmeti neutra, t. j. ne eno ne drugo. Ta pojav, ki je nam domač, je pa med jeziki na svetu nekaj osamljenega: dobi se le v ozkem pasu v Severni Afriki, v Prednji Aziji in v ievr. jezikih v Evropi. Pojav spada med tako imenovane »nominalne kategorije«, kakor so še »živo — neživo« (v sla. jezikih!), »razumno — nerazumno bitje«, »moško — žensko (po prirodnem spolu)«. Največ teh kategorij je pa še danes živih v kavkaških jezikih in upravičena je domneva Wilh. Schmidta, da se je gramatični spol razširil s kavkaškega področja proti Afriki in Evropi. Med jeziki, ki imajo to posebnost, je torej »tipološko sorodstvo«, čeprav si sicer med seboj niso v sorodstvu.
Ce potemtakem najdemo v Mali Aziji ali na Kavkazu jezike ali celo jezikovno plast, ki rabi pos. adjektiv namesto pos. genetiva, smemo rabo pos. adjektiva v sla. jezikih, toharščini in eolščini pripisati vplivu takšnega jezika ali plasti. Videli smo, da het. genetiv razlagajo kot po«, adjektiv in da ga pripisujejo protohatskemu jeziku. Za ta stari jezik je pa ugotovil W. Brandenstein, da mu »je manjkal genetiv« in da so namesto njega rabili adjektiv na -il. Genetiva pa tudi ni imela lidij-ščina. Ohranjen je dvojezični nagrobni napis z imenom pokojnika: Ndvvag Aiovvain/leoc; ('.) — lid. nannaš bakioalis (-lis je sufiks za adjektiv). K moškemu imenu Manes je pos. adjektiv manelid, k Zevsovemu imenu: civvalis. Lidijščina odraža tu protohatski vpliv. Številne pos. adjektive je pa končno imela tudi etruščina in Etruski so prišli v Italijo pred letom 800 iz Lidije. Tako je za Malo Azijo izpričana raba pos. adjektiva v protohatščini, pod njenim vplivom jo imata lidijščina in etruščina, živa je pa še danes na Kavkazu v lakiškem jeziku. Laki so pa zadnji ostanek Lelegov, ki so segali tja do Makedonije.
Sedaj bi bilo ireba dognati, kje so predniki Slovanov in Toharcev prišli v stik s protohatsko plastjo, da so od nje prevzeli rabo pos. adjektiva. — Eolci so jo mogli prevzeti od Lelegov, ko so zasedli severno Grčijo, Lezbos in severno Malo Azijo. Za Slovane in Toharce je pri današnjem stanju jezikoslovja in arheologije nemogoče to točno dognati. V poštev pride ali zahodna obala Črnega morja in zaledje severno od tod ali pa področje severno od Kavkaza proti Kubanu in onstran tega. Na zahodu kažejo na protohatski vpliv krajevna imena -ssos ob Črnem morju: pokrajina in kraj Salmydessos, Odessos (dan. Varna) in kraj istega imena pri dan. Odesi, ki ga pa. Plinij star. in Ptolemej imenujeta Ordessos. Zadnje ime ima pa tudi levi pritok Donave (dan. Ardžiš) pri Herodotu 4, 48. To rečno ime je posebno pomembno, kajti rekam dajejo imena le dalj časa naseljeni prebivalci, ne pa trgovci, na kar bi pri mestnem imenu mogli misliti. Tu na zahodu ali na vzhodu je torej bilo mogoče prevzeti ob daljšem sožitju rabo pos. adjektiva od proto-hatske plasti.
Jezikoslovju bo lahko mnogo pomagala arheologija, če se kdaj posreči vsaj največje arheološke kulture pripisati določenim ljudstvom. Vsi se zavedajo, da kultura ni vezana na jezikovne meje, vendar je na drugi strani neki paralelizem mogoč. O. Menghin je leta 1928 v Zborniku na čast Hruševskemu (po Glotta 28, 259) zapisal, da je z arheološkega stališča mogoče ugotoviti na zahodu črnomorskega področja dva pred-njeazijska kulturna tokova, nadalje tok, ki je prihajal iz zahodne Male Azije in ki ga z vso rezervo imenuje — »protohatskega«, in še dva poznejša. Še zanimivejše mnenje pa beremo v Menghinovi knjigi IVelt-geschichte der Steinzeit (Wien 1931), 557. Med arheološkimi ostanki na ljubljanskem barju iz prve polovice drugega tisočletja pred n. št., po katerih je dobila ime »ljubljanska kultura«, je posebno značilna skledica 6 podstavkom (nogami) v obliki križa, poleg tega štirioglato šilo ipd. Te oblike ljubljanske kulture imajo najboljšo paralelo v kubanski kulturi. Ljubljanska kultura se nadaljuje v Vučedolu in v Bosni (proti severu sega do Dunaja) in v grški zgodiljeheladski kulturi. Na drugi strani pa peljejo iz Ljubljane tokovi v Italijo. Torej gredo iz Ljubljane tokovi proti Italiji, Grčiji in Kubanu. Zato sklepa Menghin takole: če naj pripišemo te tri smeri trem ievr. narodom, potem morejo to biti samo ltaliki, Grki in — Toharci. Ti namreč spadajo k tako imenovani skupini kentum ievr. jezikov, ker so ievr. palatale ohranili kot zapornike. Toharci, nadaljuje, pridejo tembolj v poštev, ker ima toharščina — »kavkaške« elemente v sebi. — Poudarjam, da je Menghinov poskus, identificirati
arheološko kulturo z narodom, samo domneva, podana z izrecno rezervo. Ce bi se pa kdaj res posrečilo to domnevo dokazati, bi iz nje sledilo, da so predniki Slovanov in Toharcev bili v stiku s protohatsko jezikovno plastjo severno od Kavkaza.
Naj na koncu povzamem glavne točke. Raba pos. adjektiva je med ievr. jeziki živa le v sla. jezikih in toharščini. Ta dva jezika sta bila nekoč soseda. To sta mogla biti le ob Črnem morju. Grščina pozna to rabo le v eolščini, torej na severu, v bližini Črnega morja. Na tem področju pa pozna rabo pos. adjektiva protohatščina, od katere jo imata tudi lidijščina in etruščina. Protohatščina je bila okoli leta 2000 že star jezik, ko so bili bodoči ievr. jeziki šele v zametkih. Protohatska plast je segala tja v Grčijo in ob Črnem morju do Odese, severno od Kavkaza pa do Kubana do leta 1864. Upravičena je torej domneva, da je mogoče rabo pos. adjektiva pripisati protohatski jezikovni plasti.
Resume
Sur Vadjectif possessif en slave et en tokliarien
L'empkw de 1'adj. poss. rempla<;ant le gen. poss. est un des traits caracte-riistiques des langues slaves, cf. pour le detail Nahtigal, Slov. jezikP 223 s. Le suffixe -bjh qui se pretait a designer indifferemment un ou plusieurs posses-seurs, a ete supplante par -odi>, -irvb, s'il s'agissait d'un possesseur individuel. C'etait la consequence d'un cliangement dans la structure sociale.
L'adj. poss. est aussi bien represente en tokli., voy. Sieg-Siegling-Schulze, Toch. Gramm. § 41: «Das Tocharische liebt es, ahnlich wie die slavischen Spra-elien, die Zugeliorigkeit durch abgeleitete Adjektiva zu bezeichnen.» Dans les autres langues i.-e., l'emploi de l'adj. poss. est limite aux patronymiques et a quelques inots isoles comme scr. pitriyah, gr. nainiog, lat. patrius.
J. Wackernagel, Melanges Pedersen (1908), 145 ss.; Vorl. II2 68 ss. etait d'avis que, dans une periode reculee, l'adj. poss. avait ete aussi plus frequent en grec et en latin: le gen. poss. ne serai t que le prod uit d'un developpement recent. Maiis E. Liifstedt, Syntactica I1 (1928), 83 ss. s'est senti oblige d'apporter des corrections essentielles aux vues de W. Selon Liifstedt, l'emploi de l'adj. poss. est, en grec et en latin, assez restraint. II etait d'un usage limite dans la langue lio-nierique, influencee en ceci par l'eolien, dans des cas comme vtjvg 'Ayaftefivovetj, Kanavfjiog vlig; on trouve aussi des adjectifs tires de noms communs, p. ex. Od. 18, 328 /aAtifjiov ig Adfiov. Ce n'etait que dans le diolecte eolien que les patronymiques en -tog, c. a d. les adj. poss., etaient d'un usage courant et offi-ciel, cf. čol. 2&cvclag 6 Ntxiaiog, thess. 'AAevag Aaitoo&ivetog, beot. Mvaatiixog 'A&avoStopiog ou meine IJovidXa TToviaAeia x/ipa Tnvpet'a yvva (Fraenkel PWRE XVI 1656, 8ss.: Schwyzer, Griecli. Gramm. II 177). En latin, ce sont les genti-
lioes qui font pendant aux patronymiques eoliens: (Quintus) Marcius etait d'abord «(Q.) fils de Marcus* avant de devenir un nom.
II est done certain que l'emploi de l'adj. poss. n'est de regie qu'en slave et en tokh. Une question se pose, a savoir, si cette concordance est plus que fortuite.
Des traits commons unissent, on le sait, le tokh. au grec, au thraeo-phrygien et au slave. Pour en rester au slave, on citera le suffixe v. si. -ynja, Lit. -une en regard de tokh. -une; en face de v. si. rekg, rešti «dire», on a tokh. A rake, B reki «mot» (Porzig, Die Gliederung des idg. Spracligebiets, 1954, 183 s.). Si on oonsi-dere la parente du tokh. avec des groupes de langues, e'est le groupe constitue par le balte, le germanique et le slave, qui offre des coincidences: l'adj. i.-e. en -lo- est devenu en v. si. le participe passe act. et, en tokh., l'adj. verbal; le suffixe -sko- a ete elargi dans les deux langues par -io- pour former des adj. poss.: v. si. -iste (selon quelques auteurs) en regard de tokh. A -si, B -s.se (ibid. 186). 11 est done probable que le tokh. s'est assimile l'emploi de l'adj. poss., tel qu'on l'y connait, a une epoque ou le tokh. etait encore dans le voisinage du slave. Puisque les Slaves out etendu leur territoire en partant de la region de la Mer Noire, et que les Tokhariens ont emigre vers l'Est, le point de contact des deux langues ne peut avoir ete que la region entre les Karpatlies Orientales et la Mer Noire.
Or on sait que ladite region est riche en couches de langues non i.-e. Ces langues ont laisse des traces nombreuses dans le vocabulaire et dans la structure des langues i.-e. P. Kretschmer, Glotta 32 (1953), 16 ss. a relie le suffixe -oodg, frequent dans l'onomastique grecque, an suffixe lakke -ssa qui sert a former des adjectifs et des participes. — Le gen. sg. hittite en -aš remonte, selon le meme savant, a un adj. poss. chattique (Glotta 21, 94 s.). — C'est aussd vers le Caucase que converge l'emploi du genre grammatical selon \V. Schmidt, Rassen und Volker in Vorgeschichte und Geschiclite des Abendlandes II (Luzern 1946), 259 s. On est tente d'expliquer d'une maniere analogue l'emploi de l'adj. poss. en slave et en tokh.
W. Brandenstein dit en effet (PWRE Supplbd.V 1 col. 170; WZKM 56, 265; Caucasica X 82 et 93: voy. aussi Kretschmer, Glotta 32, 174) qu'on substituait l'adj. poss. au gen. poss. en chattique et en lydien. L'etrusque aussi abonde en adj. poss. (Kretschmer, Glotta 28, 264), et les fitrusques sont venus de Lydiie. Nous avons done une preuve de l'existence de l'adj. poss. dans une couche de langues de l'Asie Mineure. Est-il trop hardi de ramener a cette meme couche aussi la frequence de l'adj. poss. en slave et en tokli.? Les suffixes sont i.-e., mais la frequence est due a une influence etrangere.
L'emploi de cet adj. en eolien est interessant. Si en effet les Grecs avaient apporte son emploi de leur habitat d'origine, on s'attendrait a trouver cet usage repandu dans tous les diulectes grecs, bien que leur immigration se soit produite en vagues suocessives. Comme, au contruire, l'emploi est limite a l'eolien, c. a d. au Nord du territoire grec, a la proximite de la Pontide, on a quolque droit de eonjecturer qu'il s'agit la d'un fait de substrat.
On ajoutera que O. Menghin a conclu a une influence protoehattique a l'Ouest de la Pontide (selon Kretschmer, Glotta 28, 259) et a la presence des Tokhariens dans le Kouban dans la premiere modtie du 2m<= millenaire (Weli-geschichte der Steinzeit p. 557).
Anton Grad
A CONTRIBUTION TO THE PROBLEM OF WORD-ORDER IN OLD AND MIDDLE ENGLISH
An interesting, though not very common phenomenon in Current English word-order is afforded by the occurrence of the inverted subject (inverted word-order, VS) in declarative sentences1 which is, on the one hand, far less frequent than in German and Scandinavian languages2 — in this respect English goes parallel with French —, and, on the other hand, also far less frequent than in Old and Middle English, both resorting to the inverted word-order in many cases in which it is 110 longer admitted in Current English.3 Thus we meet — among other cases of its use in older English — the inversion of the subject especially in the so-called introduced declarative sentences, e.g.:
a) after the stressed (emphatic) object in the front position of the sentence: pas lareoivas asende se eadiga papa Gregorius, Aelfric; mael-raeste geceas Byrhtnodes maeg, The Battle of Maldon; me sendon to pe saemen snelle, ib.; Hem of tok a menestral, Iacob and Iosep; Hym louede yung, him louede holde, Erl and barun, The Lay of Havelok the Dane 30: Vtlaroes and theues made he bynde, ib.; A fat sivan looed he best of any roost, Cliauc., Cant. Tales, Prologue 206: but hood, for jolitee, mered he noon, ib. 680, etc.
In Current English the frontal object does not occasion the inversion of the subject (expressed either by a noun or a personal pronoun), comp.: Numbers of the men she knew already, Thack.. Van. Fair, ch. 29; The great sea-chest none of us had eoer seen open, Stevenson, Treas. Island,
1	Comp. Jespersen, A Modern English Grammar on Historical Principles (MEG) VII 2.22 ff: Poutsma. A Grammar of Late Modern English, Part I, Ch. VIII 5 ff.
2	Jespersen, Growth and Structure of the English Language, 3rd ed., § 14. Id., MEG VII 2.18; Aronstein, Englische Stilistik, § 24; Poutsma, o. c., P. I, VIII 6; for German oomp. Behaghel, Deutsche Syntax IV, § 1466, Heyse, Deutsche Grammatik, 29. Aufl., pp. 546 f.
9 Jespersen. MEG VII 2.19, 2.49; Einenkel, Geschichte der englischen Spra-che II, Historische Syntax, § 59»?; Aronstein, I.e.; the same development is also stated for Old and Middle French, cp. Foulet, Petite syntaxe de I'ancien franfais, 3" ed.. § 450: Lerch, Historische franzdsische Syntax III, § 395 ff; Dauzat. Histoire de la langue francaise. § 514, ascribes the Old French inversion to the Genmajiic influence.
ch. 1; The second winter at Malta Joan enjoyed less, Baring, Darby and Joan, ch. 13; A man so overflowing with health and life he had seldom seen, Galsw., Maid in Waiting, ch. 4, etc.
The inversion, however, is the rule also in Current English after a negative frontal object: Not another word had he to say to me, Stev., Kidnapped, ch. 4; But no knob or spring could we discover, Haggard, King Solomon's Mines, ch. 17; They got into their cab, a?id not one word did they say the rvhole way home, Galsw., Maid in Wait., ch. 35; We abate nothing of our just demands; not one jot or title do we recede, W. Churchill, The Second World War II, ch. 10, etc.1
b) after the frontal adverbial adjunct: pa a ras sum aegleaw man, Gospel, Luke X 25; Ond paes on Eastron worhte Aelfred cyning lytle werede geweorc, Winchester Chr.; On pam dagum rixode Aepelbyrht cyning on Cantwarebyrig riclice, Aelfric, Lives of the Saints; On al pis yuele time heold Martin abbot his abbotrice XX wintre and half gaere, Peterborough Chr.; pis gaere for pe king Stephne ofer sae to Normandi, ib.; This thretty wynter... hath he gone and preched, Piers Plowman, etc.
In Current English the inversion of the subject is regularly brought about by negative frontal adverbs and adverbial adjuncts, as well as after some adverbs which are not negative in form, but are felt as negative, thrown by emphasis to the front position; inversion is, however, very frequent with intransitive verbs and in passive constructions when the subject, being a more weighty word than the predicate, is placed towards the end of the sentence; thus in: Never would she show him or that girl that they could hurt her, Galsw., Beyond II 3; in none of his plays had he ever come so near the truth, Deeping, The Bridge of Desire, ch. 36; No sooner does one expect you to go straight on... than you double round the corner, Hardy, Jude the Obscure VI 4; Not for a hundred years have the relations between Moslems and Hindus been so poisoned, Churchill, o. c. 1 5; for hardly had Ahab reached his i>erch; hardly was the rope belayed to its pin ... when ..., Melville, Moby Dick, oh. 134; Only in Richmond Park did she remember thai..Galsw., Indian Summer of a Forsyte, ch. 2; Only when they left the house... did his normal eloquence return, Cronin, Citadel I 4; Rarely has so generous a proposal encountered such a hostile reception, Churchill, o. c. 11 10; Finally came the parting with Miss Amelia..., Thack., Van. Fair, ch. 6;
* Poutsma, o. c., P. I, VIII 11; Grad, Affectivity and Inversion in Modern English, ill Zbornik filozofske fakultete II, Ljubljana 1955.
With Hubert's fate mas wrapped up the fate of her beloved home, Galsw., Maid in Wait., ch. 1; In Dinny flamed up compassion for young Croom, Id., Over the River, ch. 31, etc.5
c) after a pi-edicative thrown by emphasis to the front position: Ful worthy was he in his lordes werre, Chauc., Cant. Tales, Prol. 47; Whyt was his berd, ib. 331; Fair was this yonge wyf, ib., The Milleres Tale 3233, etc.; coinp. still Sh., R. Jul. Ill 1: Villain am I none.
In Current English the inversion is still observed with a substantival subject: Great was Mr. Tulliver's wonder, Eliot, The Mill on the Floss 111; An amazing fellow was Henry, A. Bennet, Riceyman Steps II 4, etc. But: Beautiful white it was, Galsw., Over the River, ch. 2; A poor lot they are, Id., The Island Pharisees, ch. 7, etc.6
As may be seen from the above instances of the old period, the second place in the sentence was occupied by the verb, attracted there by its modifier, i. e. the stressed frontal object, and adverbial adjunct respectively; in this way the subject, whose place in non-introduced declarative sentences was normally in front of the verb as early as that period, was thrown after it.
It must be admitted, however, that in the examples of the above kind the direct (non-inverted, regular) word-order, too, is often to be found in older Engbsh, a fact explicitly pointed out also by Mosse, Manuel de Tangleis du moyen age II, § 173, 3, who in a note states the non-compulsory use of the inversion in such cases, giving at the same time the following two instances: sume hi diden in crucethus, Peterborough Chr. 29; a I pou most sugge, The Fox and the Wolf 207.
Mosse puts too little stress on the fact that the direct arrangement is met with especially in cases in which the subject is expressed by a personal pronoun. Comp. also: nu ic neom wyrSe paet ic beo pin sun a nenuied, Gospel, Luke XV 19; Nu pu me forlaered haefst on mines herran hete, Genesis; Rihtlice hi sind Angle gehatene, Aelfric, Lives of the Saints; Aefter pysum wordum, he gewende to pam aerend-racan, ib.; sume treowu he watrode to pam paet hie py swift or sceoldon weaxan, Cura Past.; paet wif he onfeng fram hiere ieldrum, Bede's Eccl. Ilist.; A pousent shep ich habbe abiten, The Fox and the Wolf 203; Mani pusen hi drapen mid liungaer, Peterb. Chr.; Calibeorne his sweord he sweinde bi his side, Lajamon's Brut 21133; His spere he nom an honde,
5 Jespersen, MEG VII 2.41; Poutsma, o. c., P. I, VIII 7 a; Grad, o. c.
0 Jespersen, MEG VII 2.31; Poutsma, o. c., P. I, VIII 18; Grad, o. c.
ib. 21144; Ricth-wise men he louede a lie, The Lay of Havelok the Dane 37; Ful ofte lyme he hadde the bord bigonne, Chauc., Cant. Tal., Prol. 52 (frequent construction in Chaucer!); To herie knyf hue (= she) sette, King Horn 101; Nu se habbep iherd pane ende of Floriz, Floriz and Blancheflur 181, etc.
As early as that period, and probably under the influence of the numerous cases with the non-inverted personal pronoun, the direct word-order is also to be found in far less numerous examples with a substantival subject, as in: In pat time a I Hengelond perl Godrich hauede in his hond, Havelok the Dane 999; po pe bataile mas ido, The Chronicle of Rob. of Gloucester 108; The reule of seint Maure... This ilke mo7ik leet olde thinges pace, Chauc., C. T., Prol. 173; Therfore, in stede of rvep-ing and preyeres, Men moot ye De silver to the poore freres, ib. 231, etc.
As appears from the statistics given by Huchon, Eistoire de la langue anglaise I 257, the inverted word-order in such cases was still predominant in Old English (Chronicle 3 : 2, Cura Past. 3 : 2, Aelfric's Homilies 3:1), but the percentage of instances of the direct arrangement were already considerable in that period and were constantly increasing in the course of the following centuries, especially sharply when, with the collapse of declension, the direct arrangement was given a syntactic function, becoming, as a sort of "morpheme",7 vital to the understanding of sentences of the subject-verb-object pattern;8 the numerous examples of non-inversion, especially resorted to in cases with a pronominal subject, show that older English, as it were, overtook contemporary French where in similar cases even the pronominal subject had strictly to obey the Old French rule of inversion, comp.: Karlun aorum nus ja, Rol.; Vers Engletere passat il la mer salse, ib. 372; or sai jo oeirement que..., ib.; ja n'en descendrat il, Voyage de Charlemagne; Aultre chose ne sfay-je, Cent nouvelles nouvelles, No. 36; comp. as late as the 17th century: Une chose ai-je a dire, La Fontaine, Contos IIT 6, etc.9
7 J. Vendryes, Le Lan gage, p. 93.
* Comp. Jespersen's statistics for this pattern in older English: 40 % in Alfred's prose, only 16% in Beowulf (poetry!), hut 66% in Ancren Riwle (13"' cent.), 84 % in Chaucer's prose (Jespersen, MEG VII 2.18).
H. Reis states that one quarter of all principal clauses in Beowulf exhibits the object (adverbial adjunct) — subject — verb pattern, and sees in this word-order a survival of one of the three possible Old Germanic arrangements (II. Wunderlich-H. Reis, Der deutsche Satzbau, 3. Aufl, p. 94 ff).
" The pronoun-subject could be still omitted in the old period of French, cp. Foulet, o. c., § 457, Lerch, o. c., § 395, Note.
The strong tendency to use the pronominal subject in front of the verb in the English of that period is clearly proved by examples in translations and adaptations from French in which we already meet the direct word-order in the English construction in spite of the French inversion, comp.: Nou pou hest y-hyerd pe zennes, The Ayenbite of Inwyt (French original: Ore as-tu oi les pecchiez); And pervore pet zuyche zennes arizep commiuiliche ine taverne, pet is welle of zenne, peroore ich mylle a lite take of pe zennes, ib. (French: ... por ce voel je un poi toucher des pecchiez); pa$ pou ssoldest to cleve, ich nelle na^t lete asskapie pis mes, ib. (French: Se tu devoies crever, lie lairai-je pas ce mes eschaper). etc. Owing to such examples one cannot quite agree with the view of J. Delcourt10 that "1'anglais a du suivre en cela (sc. ordre des mots) l'exemple que lui avait montre le frangais qui de bonne heure avait substitue lord-re direct a l'inversion germanique", but, on the contrary, it may be assumed that already at this early date the tendency in English to conform to the direct arrangement in the so-called introduced sentences began to take root more intensively and probably prior to the similar development in French. It should also be observed that this tendency — probably first realised in the cases with a pronominal subject (euphonic and rhythmic considerations also playing an important part in this development11) — originated at a time when phonetic development had not yet reduced or eliminated the old declensional endings, a fact which argues in favour of the hypothesis expressed by Jespersen12 who supposes that the direct (logical) arrangement began to predominate already before the decay of the old declension, or, in other words, the origin of the new fixed direct word-order is not due, as is usually assumed, to the decay of the inflectional system in Old and Middle English.13
10	J. Delcourt, Initiation a Vetude historique de I'anglais, 1944, § 272.
11	Hirt, Handbuch des Urgermanisclien III 217, giving the example: niper he ahreas, even thinks this arrangement a survival of the older word-order when the unstressed ("tiefbetoot") personal pronoun used to precede the (also weakly stressed) verb and occupy the second place in the sentence, i. e. immediately after the stressed front-word ("Wackernagels Gesetz").
15 Jesipersen, Language, its Nature, Development and Origin XVIII, § 13; oomp. also Id., Progress in Language, § 75.
13 For French as well, Lerch, o. c. Ill 269 ff, denies any connection between the decay of the two-oase Old French declens-ion and the establishment of a fixed, uniform arrangement.
In connection with the problem of inversion, we shall deal in these pages with the question whether in the old stage of the language "the inversion of the subject also took place in the principal clause when this was preceded by a subordinate clause (usually an adverbial clause); inversion is, as known, rigorously observed in Modern German14 and Scandinavian languages, but it does not take place in Modern English — at least not with transitive verbs, nor with intransitive verbs with a pronominal subject" —, which in this respect again agrees with Modern French. Comp.:
Mod. E.: When I entered his room, he (my father) was reading a letter. Fr.: Quand j'entrai (je suis entre) dans sa chambre, il (mon pere) lisait une lettre. Ger.: Als ich in sein Zimmer trat, las er (mein Vater) einen Brief. Dan.: Da jeg tradte ind i h an s oaerelse, laeste han (min far) et brev. Sw.: Nar (Da) jag kom in i hans rum, Vaste han (min fader) et breo.
For older English the question is answered in the affirmative by Einenkel, o. c., § 59 r\\ »Wird irgend ein Satztei'l aus dem Gefiige des
14	Behaghel. o. c. IV. § 1472; Heyse, o. c., p. 547; for other languages (Gothic, Norse, Old Greek, Russian, etc.) oomp. also Hirt, IGr. 5. 346. and o. c. Ill 220 f; Behaghel, Btr. 53, 405; Delbriick, Synt. Forsch. 4, 14. Both Behaghel and Hirt suppose an Indo-European origin of this word-order, but they also admit other arrangements in our problem.
15	Jespersen, MEG VII 2.51. The inversion, however, is possible with an intransitive verb (be, come etc.) and a substantival subject, comp.: for, ere I Could dram to part them, mas stout Tybalt slain, Sli., R. Jul. Ill 1; Even mhile they mere talking came a male for adoice, Reade, The Cloister and the Hearth, ch. 52; As they debated the story, came a loud knock at the door, Thack., Van. Fair, ch. 23; While the government of the Tudors mas in its highest vigour, took place (= occurred) an event which..., Macauley, Hist. I. ch. 1 (quoted from Poutsma), etc. But usually the weak there ("grammatical subject") is used in the front position of the principal clause, e. g.: yls I mas sitting at breakfast this morning, there comes a knock at the door, Dick., Chuz., ch. 39, etc. (see below!).
Poutsma, o. c., P. I, VIII 8a, also only gives examples with intransitive verbs bearing on our problem; here are, however, two examples of inversion with transitive verbs: But just, mhen all mas fair, and I mas to see him safe aboard ship for Rome, if not to Rome itself, met us that son of a — the Lord Anthony of Burgundy, and his men, making for Flanders. Reade, The Cloister and the Hearth, ch. 48 (here, the inversion is perhaps due to the long apposition, following the subject, "Gesetz der wachsenden Glieder"!); "Hoo." Ringo said. "Me ask her? I bet if she stayed here mouldn't no Yankee nor nothing else bother that trunk, nor Marse John neither, if he knomed hit." W. Faulkner. The Unvanquished, ch. 2.
In Modern French the inversion with intransitive verbs in our construction is possible us well.
Satzes herausgehoben und emphatisch an die Spitze desselben gestellt, so ruft dies meist Inversion bervor. Das letztere ist im AE. hiiufig, gegen das Mod. seltener. Dieser Fall, gewohnlich wie er ist, braucht nicbt be-legt zu werden, nur der Sonderfall, in welchem, dem AE. unbekannt (?) ein Priip.-Adv. an die Spitze tritt... me. Out goon the smerdes, Ch., ne. Off goes his bonnet, Sh., Rich. II, mod. On they came, ua;10 sowie ein anderer, in welchem der an die Spitze gestellte Satzteil in g a n z e n (Adverbial-) Siitzen besteht, verdient Erwahnung: ae. SySSan he com ofer Waetlinga-straete, roorhton hi paet maeste yfel, Sax. Chr.; me. pa mile pe heo tmeoneden pus, clepede Membricius, La3.; mhanne Cristis chirclie prof, meren no sich pope and cardenals, and sipen pes prelatis meren conuin yn, regnide anticrist mip synne, Wycliif; ne. not as the ivorid gives, give I unto you, Sweet, Syntax; mod. (?) If this is not poetry, may the name perish, Acad. 17. Febr. 1900, p. 149.«
Similarly also Mosse, o. c. II, § 173: «L'ordre inverse verbe-sujet est des plus frequents en moyen-anglais. Yoici les principaux cas ou il se presente: ... 7. Apres les subordonnees conjonctives placees en tete: and so hi mere in po ssipe, so a ros a grat tempest e of minde, Sermon of Kent; Annd tatt te Laferrd Jesu Crist Wass borenn her to manne..., patt dide he forr to shaemenn srna Unnsezzenndliz mecnesse, Ormulum 5608.»
This is not the view of R. Huchon, o. c. I 257: «... et il est a remar-quer que, contrairement a Tusage de rallemand moderne, une proposition subordonnee precedent la principale n'exerce sur le verbe de cette derniere aucune influence decisive. Aussi ecrira-t-on sans inversion: gif pu ponne pis lytle bebod tobrecst, pu scealt deape smeltan, Horn. Th. I 14, bien que, <?a et la dans l'Orose, on ait pu noter quelques constructions inverses telles que: Aefter paem pe Romeburg getimbred maes //// hunde mintrum & XXVI, feng Alexander to Macedonia rice, Or., p. 122.»
It thuS seems that this question, an interesting one and still at issue for Old French as well,17 deserves more attention and research if we are to obtain more exact if not definitive judgments. These we shall try to present in these pages, in so far as our rather limited choice of older texts has allowed us to form them.
Of the examples quoted by Einenkel, his statement is unquestionably confirmed only by the first, Old English instance which in fact
16	For this construction see Jespersen, MEG VII 2.44; Grad, o. c.
17	Lerch, o. c. Ill, § 395 e; Grad, L'inversion du sujet dans la principale pre-cedee d'une subordonnee en ancien franfais, in Razprave SAZU, Ljubljana 1955.
2 Linguistic«	17

exhibits the inverted subject in the principal clause, with a transitive verb aind its object expressed; but unfortunately it is an isolated instance of inversion, other examples — which we came upon in our texts — with the preceding subordinate opening with sifiSan exhibit the direct word-order in the postponed principal clause (see below).18 In Einenkel's Middle English examples, however, we find the inverted arrangement with intransitive verbs, and as inversion with such verbs is still possible in Modern English as well, they cannot be considered good evidence for our problem. Nor can the two Modern English examples quoted by Ein-enkel satisfy us: in the first the inverted arrangement is undoubtedly occasioned by the frontal negative,19 and in the second the head-clause represents an optative sentence, expressed by means of the verb may, and, as is known, in such cases the inversion is commonly observed, comp.: May you be happy, etc.20
As for the instances quoted by Mosse we think too that they are not well chosen for the purpose of demonstrating the prevalence of the inverted word-order in our field: in his second example the inversion in the postponed principal clause must be ascribed to the frontal demonstrative object that (coanp. also: pat may 1 ful roel swere, bi God!, Havelok the Dane 252; this thoughte he roel ynoughe That..., Chauc.; still in Mod. English: On Lammas-eve at night shall she be fourteen; That shall she, marry, Sh., KJul. I 3; That shall 1 do, my liege, Id., Henry the Sixth II, 111 2; That mill I learn to-morrow, Scott, The Abbot, ch. 12,21 while in the first instance it is due to the frontal demonstrative (correlative) adverb so; in older English this, as well as some other
18 In: SySdan furpum roeox Of cildhade, symle cirde to him Aehte mine, Cynewulf's Elene 914, the pronoun-subject he is omitted both in the subordinate and the principal clause; in Ilolthauscn's view its place would be after sySdun, and symle respectively.
An example of the preceding object-clause with inversion in the following head-clause is furnished by the Lay of Ilavelok the Dane 29992: And hroou he meren mreken roel, Haue ich sey(d) you euerildel. But comp. also: IVher that he be. I can nat sothly seyn, Chaucer, Cant. Tal., The Milleres Tale.
" See also Grad, Affectioity and Inversion in Modern English.
20	Jespersen, MEG VII 2.22; Poutsma, o. c., P. I, VIII 21a.
21	Comp. Poutsina, o. c., P. I, VIII 8 k: "The word-order in the following examples, presumably common enough in older English, is now oidy occasionally met with, except, perhaps, in the language of the illiterate. "He speaks the truth", said a second voice firmly. "Ay, that doth he", said a third, Lytton. Rienzi I, ch. 3."
demonstratives, such as: pa(po), ponne, pan(ne), per22 are very often met with in this position, their function being to link the subordinate and the principal clauses; they appear almost regularly in Old English, less frequently in Middle English when the language began to get rid of them for stylistic reasons; as will be seen, they were far less in use after the subordinates opening with the conjunctions rvhen and if than after those introduced by an adverb-conjunction or a compound conjunction. Comp.:
pa he pa se cyning pas mord gehierde, pa het he hie bidan on paent ealande, Bede's Eccl. Hist.; Mid-py he pa Paulinus biscop Godes word bodade ond larde, ond se cyning elde pa-git to gelyfanne..pa waes sume daege se Godes wer ingongende to him, ib.; Betwux-pam-pe he clypode to Criste pa-git, pa tugon pa haepenan pone halgan to slaege, Aelfric's Lives of the Saints; pa hie fulle waeron, pa cwaep he to his leorning-cnihtum, Gospel, John VI 12; Ac sona-swa hie to Beam-fleote comon, ond paet geweorc geworct waes, swa hergode he on his rice, Winchester Chr., pa se cyng paet hierde, pa wende he hine west, ib.; and ponne se cyng waes on Normandige, ponne waes he maegester on pisum lande, Peterborough Chr.; ponne py ylcan daege pe hi hine to paem a de beran willafi, ponne todaelaS hi his feoh, Alfred's Geogr. Treatise; Ase hi ferden here wai in pe morewentide, penne seyen hi twolf 3ungemen after hem ride, Jacob and Iosep 406; Bot sone when he herd ascry pat king Edward was nere parby, pan durst he noght cum liere, Laurence Minot; And huanne pe mes byep y-come on efter pe oper, panne byep pe burdes and pe trufles vor entremes, Aycnbite of Inwyt (French original: ... lors sont les bordes e les truffes por entremes); And whan he drowgli to pe (lore, panne dymmed his eighen, Piers Plowman 356; Whaenne swa aei ferde funde to pan paerde, peonne fleo pa fu^eles feor i pan lufte, La3ainon's Brut 21748; Whan Zepliirus eek with his
swete breeth Inspired hath----The tendre croppes... Than longen folk
to goon on pilgrimages, Chauc., Cant. Tal., Prol. 5; And shortly, whan the sonne was to reste, So hadde I spoken with hem..., ib. 30; Whanne
22 See Mosse, o. c. I, § 191, II, § 162. — Quite similarly the adverbs so, da, and si (lors), opening the postponed principal clause in older German and older French (and other Romance languages), occasion the inversion of the subject, see Behaghel, o. c. IV, § 1472, H.Paul, Mittelhochdeutsche Gr., § 355, Lerch, o. c. Ill, § 395 e, Meycr-Liibke, Gram, der rom. Spraclien III, § 651.
Mosse's instance is a translation from French: Et si cum furent en la mier, si leva un grant torment.
2*
19
dame Prudence... liadde herd al that hir housonde lyked for to seye, thanne axed she of him..., ib., The Tale of Melibeus 15, etc., etc.23
With intransitive verbs, we often find the demonstrative local adverb there in the front position of the principal clause, sometimes still having its full local value, but usually already representing the weak there, "a mere anticipative element occupying the place of the subject which conies only later" (OED) (a sort of "grammatical subject"). Thus in: And ealie pa hrvile pe paet lie bid inne, paer sceal beon gedrync and plega..., Alfred's Geogr. Treatise; For if we luf God in al oure hert, par es na thyng in us thurgli pe tvhilk we serve to syn, Rolle of Hampole; And as I wente, tlier cam by mee A whelp..., Chauc., The Book of Ducliesse 388; And after pat he liadde endid alle his goodis, per fell a gret hungre in lond, Wyclif, etc. But the weak there may also be omitted, as seen from Einenkel's and other modern examples (see above!), and perhaps this is also the case of Huchon's example with inverted subject (see above!) aud intransitive use of the principal verb (fon to rice).24
In the same position and after a preceding concessive subordinate, we frequently meet in Middle English the adversative adverb yet which again occasions the inversion, e.g.: TI10U3 we culled pe catte, 3111 sholde per come another To cracchy us, Piers Plowman 185 (with inverted weak there); tliogli Argus... sete to rekene in hys countour, And rekened with his figures ten ... Yet shoulde he fayle to rekene even The wondres ..., Chauc., The Book of Duchess 435; For though a widwe liadde noght a sho..., Yet wolde he liaoe a ferthing, Id., Cant. Tales, Prol. 253; But
The introductory then with the following inversion is still occasionally met with in modern texts: When I have chased all thy foes from hend, Then mill I think upon a recompense, Sliak., Henry the Sixth I, I 2; When they are gone, then must I count my gains, Id., King Richard the Third 11;// the king offer not a sacrifice of a fair girl.... then shall he fall and his house, Haggard, King Sol. Mines, ch. 11; When the dance is over, then mill I speak to some of the great chiefs, ib., ch. 10.
2* Examples in Old and Early Middle French, similar to Huchon's inverted instance, owe the inverted word-order in the postponed head-clause to the fact that the first part of the compound conjunctions (for inst. apres ce in apres ce que, ainz in ainz que, pour ce in pour ce que, see Grad, o. c.) was in that period still felt as an adverbial adjunct belonging to the head-clause in which it occasioned — owing to its front-position — the inversion of the subject. Similarly, and with the same result, aefter paet (paern) in aefter paem pe could have been felt as part of the postponed head-clause, but other similar (unfortunately extremely rare) examples do not support such a supposition, e.g.: Aefter pan pe ic arise of deaSe gesunf, ic eoiv eft gemete, Aelfric (direct arrangement in the head-clause! see below!).
al be that he mas a philosophre, Yet hadde he but litel gold in cofre, ib. 297; And al-be-it so that it seme that thou art in siker place, yet shaltom almey do thy diligence, ib., The Tale of Melibeus 29; ib. 51, etc. Still in Mod. English: But, though thou art adjudged to the death, Yet mill I favour thee, Shak., The Comedy of Errors I 1; Although me fancy not the cardinal, Yet must me join mith him., Id., Henry the Sixth II, I 5.
The inversion of the subject in the principal clause of the above examples is due to the introductory (correlative) adverb, serving as a connecting link between the subordinate and the postponed principal clause and almost regularly occasioning the inversion (as pointed out by Huchon, 1. c., the inversion is especially the rule after short, monosyllabic adverbs; for some exceptions see below!). But as early as that period the language could do without such correlative adverbs in the front position of the principal clause, and — a most important fact concerning our problem! — in this case the principal regularly exhibits the direct, non-inverted arrangement, the whole hypotactical group thus already assuming its modern form. Examples with a pronominal subject are especially numerous, comp.:
OE: And peah man asette tmegen faetels full ealaS off fie maeteres, hy (= hie) gedod paet aegper bid oferfroren, sam hit sy sumor, sam minter, Alfred's Geogr. Treat.; Siddan ic liie d a geliornod haefde, swae-smae ic hie forstod, ond smae ic hie andgit fullicost areccean meahte, ic hie on Englisc amende, Alfred, Letter to Werferth; Aefter pan pe ic arise of deaSe gesunf, ic eom eft gemete, Aelfric (quoted from Einenkel, o. c., p. 27); Ac si ft ft an des Sin sunu com, de his spede micl miUestrum amierde, du of sloge him faett cealf, Gospel, Luke XV 30; And da hie ne meahton hine inn bringan for daere menige, hie openedon done hrof daer se Ilaelend maes, ib., Mark II 5; Sodlice da se Haelend geseah liiera geleafen, he cmaed to daem laman, ib.; da se Haelend daet on his gaste oncneom, daet hie sma betmeoh him dohton, he cmaed to him, ib.; And da he daem liuse genealaehte, he gehierde done smeg, ib.; da se Haelend Ids eagan upp ahof and geseah daet micel folc com to him, he cmaed to Philippe, ib., John VI 4; and pa he paet geseah, he hine forbeah, ib., Luke X 31; ib., Luke X 34; Matthew XII 2; paenne se yrdling unscend pa oxan, ic laede hi to laese, Dialogues of Callings, No. 4; ib., No. 15; Ond yrnb XXXI mintra paes-pe he rice haefde, he molde adraefan anne aepeling, Winchester Chr.
ME: And po he herde speken of mete, He molde blepeliche ben pare, Hie Vox and the Wolf; Als she shulde hise elopes hanclel On for to don,
and blame pe fir, She sam perinne a liht ful shir, Havelok the Dane 586; And as 1 lay and leonede and lokede on pe matres, I slumberde in a slepyng, Piers Plowman 9; ib. 14; As John pe apostel hit sys myth syst I syse pat cyty of gret renoun, The Pearl 985; As (John) pise stonez in mrit con nemme, I knem pe name after his tale, ib. 997; And as they sat, they herde a belle clynke, Chauc., Cant. Tales, The Pardoner's Tale 664; Axe not mhy, for though thou aske me, I mol not tellen goddes privetee, ib., The Milleres Tale; But er that he had maad al this array, He sente his knave ..., ib.; but ther — as thou hast told me heer — beforn, that..., I am al redy to chaunge... ib., The Tale of Melibeus; and mhyl I live a mayde, I mol thee serve, ib., The Knightes Tale; For tliogli the feend to hir y-coupled mere, She molde him overmacche, ib., The Marcliantes Tale; For that he scliold almay upon hir thinke, Sche yaf him such a maner love-drinke, Id. (quoted from Einenkel, o. c., p. 110); And sypen I have in pis hous hym pat al lykes, I schal mare my mliyle mel, Sir Gawain and the Green Knight 1234; Bot po sche ran so up and doun, Sche made many a monder soun, Gower, Conf. Am. 4097, pas pou ssold-est to cleve, ich nelle nas t lete asskapie pis mes (despite the French original with inverted subject: Se tu devoies crever, ne lairai-je pas ce mes eschaper), The Ayenbite of Inwyt.25
The direct word-order is especially very frequent after the temporal subordinate opening with the conjunction mlien whose use was spreading more and more in Middle English, presenting also numerous examples of the subordinate preceding the principal clause, as in: Ilman he felede hise foos, He made hem lurken, Havelok the Dane 67; ib. 1848; And quhen the kyngis houn.de has seyn Thai men assale his master sma, He lap till a ne..., The Bruce 451; huan he comp ayen he hep al pis oorlore, Ayenbite of Inwyt 69; And lohun pat ony gode knyglit pat mas hardy and noble cam to see this rialtee, he molde lede him into his parndys, Manderville's Travels; Hmen pu bilialdest te mon pu art in Eve point, Ancrene Riwle; Vor limanne hin longeth, ich him singe, The Owl and the Nightingale 890; ib. 894; Whan he ros up of his orysoun, He 3ede yn hys celle lip and domn, Handlyng Synne 4033; Whan sche pis vertu liadde sein, Sche let pe leste drope ... doun falle, Gower, Conf. Am. 4146;
2S Cases (with a personul pronoun us subject) of non-inverted word-order in our problem are also met with in Middle High German, see H. Paul, 1. c., und V. Michels, Mittelhochdeutsches Elementurbuch, 3. Aufl., § 313, Anin. 2. The lion-inverted arrangement disappeared towards the end of the Middle High Germun period.
Whan Melibens retourned was in-to his hous, and saugh al this meschief, he, lyk a mad man... gan to wepe, Chauc., The Tale of Melibeus 2163, etc., etc.
The non-inverted arrangement is also the rule after a conditional subordinate introduced by gif (if); numerous examples are afforded by Old and Middle English, comp.: Gif ge me ut adrifaS fram eowrum geferscipe, ge etaS wyrta eowre grene, Aelfric's Colloquy, Se Coc; gif ic wiste hwaet he waere, ic wolde licgan aet his forum, Id. (quoted from Einenkel, o. c., § 13 A); Gif ge so lice wiston hwaet is, Ic miile mildheort-nesee, Gen.; and gyf par man an ban finde unforbaerned, hi hit sceolan miclum gebetan, Alfred's Geogr. Treat.; gif he moste pa-gyt twa gear libban, lie haefde Yrlande gewunnon, Peterborough Chr.; sef hit nere for mi luue, He nere no^t fram his londe icome, Floriz and Blauncheflur 737; ib. 677; if ich pe liolde on mine uote... And pu were vt of pine rise, pu sholdest singe an oper wise, The Owl and the Nightingale 51; ib. 209; Aud if we grucche of his gamen he will greve us alle, Piers Plow. 155; Yif y late him lioes go, He milite me wirchen michel wo, Havelok the • Dane 509; If ihc come to live Ihc schal pe take to wyve, King Horn 559; if he wol nat tarie, ... we wol this tresor carie ..., Chauc., Cant. Tal., The Pardoner's Tale 799, etc. etc.26
[Inversion, however, seems to have been observed after the negative, thrown by emphasis to the front position of the principal clause, e.g.: Gif pu waere her, naere min broSor dead, Gospel, John (quoted from Einenkel, p. 27); Gif he nere yfel wyrcende, ne sealdon we hyne naefre pe, Ev. Nic. (Id., p. 28); if ich me holde in mine liegge, Ne recche icli neuer what pu segge. The Owl and the Nightingale 59; ib. 56; Though men me wolde al in-to peces tere, Ne slial I never... Biwrenge a word. C'h., C. T., The Sliipmannes Tale; the two Old English examples, quoted by Behaghel, o. c. IV, § 1472: gif hit Sanne cucu feoli waere, ne pearf he paet geldan, Lieberm. 34, 28; se Se slea his agene peowne esne, ne biS he ealies swa scyldig, ib. 32, 17, also owe the inversion in the head-clause to the stressed frontal negative (Behaghel ascribes it to the preceding subordinate); comp. also: peali sumum men gesceote laesse dael, ne biS swa-Seali na mare miht on Sam mar an daele ponne on Sam laessan, Aelfric, Eucharist.
28 In Old and Middle French as well, preceding subordinates opening with the conjunctions quand and si do not, as a rule, occasion inversion in the postponed head-clause, see Grad, o. c.
For the front-position of the negative, followed by the verb, see also Behaghel, o. c. IV 12, and Hirt, o. c. Ill 223.
Of course, examples — though not very numerous — with an introductory demonstrative and the inverted arrangement are to be found as well; thu6 in: Gif ge for-Sy me fram adrifaS, 8aet ge Sus don, Sonne beo ge ealle Sraelas, Aelfric's Colloquy, Se Coc (Latin text: ..., tunc eritis omnes coci); gif ure godo aenige mihte haefdon, ponne rooldan hie me ma fultumian, Bede's Eccl. Hist., and gyf me aenige bote gebidan sculan, ponne mote me paes to Gode earnian, Wulfstan, Homily; Gif me Sa gastlican mihte Saer-on iocnamaS, Sonne undergyte me ..., Aelfric, Eucharist; with the direct word-order in spite of the frontal adverb: Gif ic aenegum pegne peoden-madmas geara forgeafe, ... ponne he me na on leofran tid leanum ne me ah t e minre gife gyldan, Genesis.]
The direct word-order after the conditional subordinate might be brought about by analogy with examples in which the conditional subordinate (protasis) appears without the conjunction if and with the subject inverted (originally this construction represented an interrogative sentence), and with the non-inverted arrangement in the principal clause (apodosis, representing the answer to the preceding question), as in: A lite ic minra honda gemeorc ... ponne ic mid pys merode..., Caed., Gen. (from Einenkel, p. 43); ablinde pe lieorte, lieo is eS ouercumen, Ane. R. (ib.); Ich mihte liabbe bet i-don, hefde ich pen i-selSe, Poema Morale; for nadde they but a sliete... And a bak ..., They molde hem selle..., Chauc., C. T., The Chanouns Yemannes Tale, etc.27
We think, however, that such examples could not but strengthen the tendency to produce the direct word-order in our field — the same construction is to be found in German as well, and nevertheless the direct arrangement has not been introduced in the postponed principal clause ("Nachsatz", cp. Behaghel, o. c. IV, § 1472) — but not cause it directly, because it also appears very early after subordinates of other kinds, as seen from our examples above. A marked tendency to form the direct word-order is also evidenced by instances in which we meet the non-inverted subject even after a frontal object, and adverbial adjunct respectively, comp.: pa hi forS eodon, feores hi ne roliton, The Battle of Maldon; Hman he liaoede't greyped so, llaoelok pe yunge he dede perinne, Havelok the Dane 714; Whan Melibeus liadde herd that..., anoon he consented to liir conseiling, Chauc., C. T., The Tale of Melibeus
" This is the view of Behaghel, o. c. IV, § 1472.
2240; Whan th'orisoun was doon of Palamon, His sacrifice he dide, ib., The Knightes Tale, etc.
Parallel (and perhaps by analogy) with the numerous cases of the direct arrangement in the principal clause with a pronominal subject, we also see, as early as the old period and especially in Middle English, • the non-inverted substantival subject appear; thus in:
pa hit (sc. saed) up eode, seo sunne hit forswaelde, Gospel, Mark IV 5; Til pe list of day sprang, Ailniar him puste lang, King Horn 493; Annd sone anan se piss mass sessd purrh an off Godess enngless, A mi-kell here off enngleped Was cumenn ut off heffne, Ormulum 3368; Gif tma men oper III coman ridend to an tun, al pe tunscipe flugaen for heom, Peterborough Chr.; panne sho havede brouht pe mete, Haveloc anon bigan to ete, Havelok the Dane 649; ib. 474, 657, 1824; pa hit mes daei a marsen ..., ArSur pa up aras and strelite his aermes, Lag. Brut 14003; Whanne ich song was... meny ser hennes, My fader and my frendes founden me to scole, Piers Plow. VI 35; 31If se pis londe mid pefpe bep ifonge, 301/re dom is idemed, Iacob and Iosep 413; When pe nyhtegale singes, pe wodes waxen grene, Suete Lemmon; Whan they han goon nat fully half a mile,... An oold man and a poore with hem mette, Chauc., C. T., The Pardoner's Tale 711; And also sone as that he was gon, That oon (of hem) spak thus unto that oother, ib. 806; And whan this abbot had this wonder seyn, His salte teres trikled doun as reyn, ib., The Prioresses Tale; Whan that they comen... out of toune, This Somnour... gan to roune, ib., The Freres Tale; Whan that our pot is broke..., Every man chit, and halt him yvel apayd, ib., The Chanouns Yemannes Tale; Whan we came to the forest syde, Every man dide ryght anoon As to huntynge fil to doon, Id., The Book of Ducliesse 372; pas ho (= she) were burde brystest, pe burne in mynde hade pe lasse luf in his lode, Sir Gawain and the Green Knight 1283; And 3// whanne he was fer, his fadir sawe him, Wyclif, etc. (most examples are again furnished by the subordinates opening with the conjunctions when and if).
Thus the preceding pages may have shown. the following results concerning the problem of word-order in the principal clause preceded by an adverbial subordinate in Old and Middle English:28
28 It should be pointed out that such hypotactical constructions are not very frequent, especially in the oldest stage of the languuge, its characteristic feature being the loose association of clauses (parataxis; for this problem s. also Klaeber, Anglia Beibl. 52, 216 ff, and Anglia 25, 275 ff; 27, 246); only in Middle English
a)	The inverted word-order (VS) is the rule after an introductory demonstrative (correlative) adverb almost regularly used in Old, less frequently in Middle English in the front-position of the postponed (declarative) head-clause;
b)	whenever the demonstrative (correlative) adverb does not appear in this position — and such examples are far from being rare, they are almost the rule after subordinates opening with the conjunction when and especially if — it is the direct arrangement that had already asserted itself not only with a pronominal but also a substantival subject; some instances (especially with a pronominal subject) are to be found as early as Old English, and they become quite frequent in Middle English, proving that in this case, too, the tendency in English to resort to the direct (logical) word-order began to take root at this early date without having undergone any foreign influence, thus presenting an important divergence from other Germanic languages in this respect;29
c)	examples of the inverted arrangement in the non-introduced postponed head-clause can be found with intransitive verbs, which, however, already in that period, are frequently preceded by the weak "there"; in some cases the inverted word-order is occasioned by a negative, thrown by emphasis to the front-position of the postponed principal clause.
Povzetek
Prispevek k problemu besednega reda v stari in srednji angleščini
V začetku članka omenja avtor, da je raba invertiranega besednega reda (VS) v asortivnih stavkih v moderni angleščini na eni strani mnogo redkejša kot v drugih germanskih jezikih — v tem pogledu gre angleščina vzporedno s francoščino —, na drugi strani pa tudi redkejša 'kot v starem jeziku, v katerem je splošno nastopila inverzija subjekta v tako imenovanih uvedenih stavkih, t. j. takih, ki so se začenjali s kakim drugim sta včnim členom kot osebkom; tako n. pr. je lahko zaradi posebnega poudarka stopil na čelo stavka bodisi
when parataxis gives way to more precise indications of logical relationship and subordination (hypotaxis) do cases of our construction appear in a greater number.
29 For similar hypotaotical constructions in Middle High German, functional differences (concerning the subordinate) for cases mentioned under a), and those under b) respectively, are supposed by Karg, 11 y pot axe bei Hartmann von Aue, Gennanica (Festschrift fiir Sievers, 1925), p. 455.
objekt bodisi p ris lovno določilo in pritegnil k setbi (na drugo mesto v stavku) glagol, medtem ko je osebek moral zavzeti mesto za glagolom. Vendar avtor poudari dejstvo, da je tudi v uvedenih stavkih raba direktnega (logičnega) besednega reda (SV) že v stari dobi tudi možna in razmeroma kar pogostna, zlasti v primerih, ki imamo v njih osebek izražen z osebnim zaimkom. Taki primeri po avtorjevem mnenju potrjujejo Jespersenovo domnevo, da se je direktni red začel v angleščini uveljavljati že pred razpadom stare fleksije, oziroma da ni šele le-ta povzročil njegovega nastanka, kot se to splošno domneva. Zanimivo je, da so zadevni primeri v starejši angleščini pogostnejši kot v istodobni francoščini (ki je tudi poznala in zelo dosledno spoštovala pravilo o inverziji v takih primerih), kar — hkrati z dejstvom, da v prevodih tiste dobe iz francoščine, kjer v originalu srečamo še invertirani subjekt, že najdemo direktni besedni red v angleškem tekstu — nikakor ne govori za to, da bi bila angleščina prešla k logičnemu redu pod francoskim vplivom, kakor nekateri mislijo.
Predmet sprednje razpravice je še sporno vprašanje, ali je angleščina v stari dobi tudi poznala inverzijo subjekta v (asertivnem in neuvedenem) glavnem stavku, pred katerim je stal adverbialni podredili stavek, kar je strogo pravilo v moderni nemščini in v skandinavskih jezikih, ne pa zopet v francoščini. Na podlagi primerov, ki jih je našel v dosegljivih mu starih tekstih, je avtor glede tega problema prišel do naslednjih zaključkov:
a)	invertirani besedni red v zapostavljenem glavnem stavku je pravilo za uvodnim demonstrativnim (korelativnim) adverbom, ki se je v stari angleščini skoro dosledno, manj pa v srednji angleščini rabil na čelu glavnega stavka ter povzročil inverzijo. Taki demonstrativni adverbi so bili: pa, so, panne, pere (prim. nem. so, dann ter starofranc. si, lors v isti poziciji ter s prav tako sledečo inverzijo);
b)	kadar koli pa takega korelativnega adverba ni bilo — in taki primeri niso izjemni niti v najstarejši dobi in postajajo vse pogostnejši v srednji angleščini ter zlasti skoro ne poznajo izjem za odvisniki, uvedenimi z vezni kom mhen ter zlasti if — pa že srečamo direktni besedni red, in sicer zopet predvsem v primerili s pronominalnim subjektom ter, verjetno pod vplivom le-teli, tudi že v primerih s substantivniim subjektom; tudi tu je torej logični red brez tujih vplivov pognal močne korenine že v tej dobi ter ustvaril pomembno razliko k drugim germanskim jezikom;
c)	primeri obratnega besednega reda v našem problemu se najdejo pri intranzitivnih glagolih (kar je možno še danes), a so že v oni dobi često uvedeni s šibkim prislovom there (»gramatikalni« subjekt); v nekaterih primerih pa moremo pripisovati inverzijo niikalnici, ki jo je emfaza postavila na začetek glavnega stavka in ji je (kot še danes v nekaterih primerih) neposredno sledil glagol, a le-temu šele subjekt.
Podrobnosti ter navedbo strokovne literature glej v angleškem besedilu razprave oziroma v opombah.
Bojan Čop
ETYM A
Arm. azbn,
gen. azbin, instr. azbamb signifie selon Ciakciak «le fila dell'ordito*, (plus tard) «liccio». Le mot est inexplique.
Sa partie finale rappelle un autre mot armenien, a savoir skizbn, gen. skzban, instr. skzbamb cprincipio, origine». Celui-ci est derive du verbe sksanim «(in)cominciare», ce qui veut dire que la partie radicale skis- donne l'abstrait verbal au moyen d'un suffixe bizarre -bn; mais c'est Pedersen qui a eclaire d'une maniere suffisante cette anomalie (KZ. 38, 217): au lieu du suffixe -mn (i.-eur. *-men~), on a par dissimilation -on et le -v- a du donner une occlusive derriere une sifflante, qui elle-meme est devenue sonore. De la meme fagon, on a fait azbn. La ra-cine de ce mot est done a chercher dans az-. Nous posons une forme primitive *ant-s-mn et rap p roc h on s le groupe de mots que Pokorny, Idg. EW. 322 cite sous *ent- «anzetteln, weben>: skr. atka-s <vetement», av. adka-, a tka- «manteau», alb. ent, int «weben, das Gewebe anzetteln», gr. (att.) Snofiai «tisser». Pour les formes celtiques, voir Pokorny, 1. cit. — On posera plutot i.-eur. *ant-, a cause du grec Snap'... Eixpopitav 6t diaofia (Hesyche), dvtfjpiog' or.. (le meme) (voir Petersson, Heteroclita 262). Enfin, Hesyche cite un dir/vtao ' fiia&xo loidv.
Le mot armenien recouvre presque completenient le substantif grec Sofia que Suidas cite sous pafia c. ■ ■ orjftaivei 6i xal id iiaufia . . J. Mais il n'y a aucun rapport historique entre ces deux mot®, car le mot armenien repose sur un theme verbal *ant-s- tout coinme son consort skizbn sur un theme verbal skis-. D'autre part, il atteste de nouveau la racine jusqu'ici peu sure du groupe pre-cite.
ITitt. altanni-š
«Quelle, Brunnen* (Forrer, Glotta 26, 180 sq.; Friedricli, Heth. Wb. 20) rappelle par son suffixe un autre mot hittite presque synonyme: sakuni-s «Quelle, Sprudel, Quellteich*. Ce mot provient d'un theme verbal *šakn-, i.-eur. *snoq-u-, et presente le suffixe bien connu des abstraits indo-europeens *-ni-; pour les details, voir moil article dans Zbornik filoz. fak. (Ljubljana) II, a paraitre en 1955.
Si altanniš est forme de la meme facjon, on partira d'un theme verbal all- au sens de ccouler, jaillir». Ce theme repond a un des elargissements
dentaux de la racine *el~, *ol- «modrig sein, faulen» chez Pokorny, Idg. EJV. 305: *el-d- dans arm. alt «fange, ordure», aItimr, eltiror (selon Ciak-ciak) «luogo irriguo, prato irriguo; stagno, palude, valle» ou bien *el-dh-dans v. norr. uldna «schimmeln», v. h. ali. oltar «Sclimutzkrunie». Le groupe derive de cette racine est beaucoup plus vaste que ne le fait songer 1'exposition des exemples chez Pokorny, 1. cit.; voir Zioa antika III 183—186 et IV 146—148. Aussi le sens de ce groupe etait-il plus large: on constate les notions conime «humidite, eau», «etang, fleuve», «source, cavite pleine d'eau», etc., p. ex. lette aluots, aluogs «source», les noms de fleuve comme lit. Alanga (voir aussi Miihlenhach-Endzelin I 67), etc.
Pour le vocalisme de la racine du mot hittite, on peut se decider pour *old(h)- ou *ld(h)-. Le suffixe, a coup sur au degre plein, c.-a-d. i.-eur. *-oni-, permet tous les degres, cf. skr. kšipaniš «Peitschenschlag», oartaniš «Rollen, Lauf» et Brugmann, Grdr. II2 1, 289 sq.
Le mot montre la meme absence de h- que nous avons deja constatee dans son congenere aIpaš «nuage» (voir Živa antika III 183). Si le nom du fleuve Alta que cite Forrer, 1. cit., est reellement parent de altanniš, on a probablement un autre derive du meme theme verbal.
Hitt. hamešha-
et hamešhant- «Friihjahr» (voir Gotze, Lg. 27, 469 sq.; Friedrich, Fleth. Wb. 49 sq.) a ete objet de plusieurs essais d'explication, dont aucun ne peut passer pour 6uffisant:
a) L'explication de Sturtevant, Lg. 4, 163 sq., reprise plus tard par lui et les autres linguistes, fait sortir notre mot de la racine connue *(a)me- «moissonner» dans gr. dfiaco, v. h. all. maen. Dans ce cas, ha- du hittite et a- du verbe grec seraient provenus d'une initiale i.-eur. a la-ryngale *x-. Mais on objectera a cette hypothese comme a loute la theorie de6 laryngales i.-eur. que le a- du grec peut tout simplement representer la prothese vocalique; la racine i.-eur. ne devait etre aucuneinent disylla-bique. Puis, il seinble que hamešha- ait absorbe le sens du mot *uesr «printemps» en l'ajoutant a son sens primitif d"«ete>: car on sait que rindo-europeen commun n'avait pas de terme commun pour l'«ete»,1 tandis qu'il en comportait un pour le «printemps» et un pour l'«hiver». Le hittite conservait fidelement le dernier; s'il a perdu au contraire le nom i.-eur. du «printemps», c'est parce qu'il confondait les notions d'«ete»
1 Le mot *$3rw-, designant l'ete et l'an a la fois, passait d'usage des l'indo-europeen commun, comme le montrent gr. &l'qos, lat. aestas, etc.
et du «printemps», sans doute sous l'influence du different climat de 1 Asie Mineure; or, il serait peu probable que le mot hamešha-, s il signi-fiait a l'origine «saison de la moisson», ait elargi son sens assez specialise a la notion si large de toute la partie chaude de l'annee. Au contraire, on comprend facilement une pareille extension, si le mot hamešha- signi-fiait d'abord «la saison chaude, ete>: Les Hittites, venus dans les regions plus chaudes, ont pu se passer de la notion du «printemps», parce que cette saison ne se distinguait plus clairement de l'ete.
b) Nous jugeons impossible toute tentative d'expliquer hamešha- par le nom i.-eur. du «printemps», *uesr, etc. Pedersen, Hittitisch 197 (apres Kurylowicz, Symbolae Rozivadomski 1 101) tente d'expliquer le -m- hit-tite par le passage frequent (et peut-etre seulement graphique) du -u-a -m-; ce passage est au contraire restreint a quelques cas tout a fait differents de celui de hamešha-. La phonetique et la morphologie re-jettent l'hypothese de Gotze, Lg. 27, 471, qui derive le mot d'un *hant-uešha- = ital. prima-vera.
A toutes les exigences du sens, de la morphologie et de la phonetique repond au contraire notre hypothese que voici:
Le nom allemand du ble Emmer (masc.) «triticum dicoccum», v. h. all. amar(o), amari, designe une sorte de ble qu'on seme de preference pour la moisson dete; c'est ainsi qu'on a denomme cette sorte aussi Sommerdinkel, Sommerspelz. II est tres probable, que ce mot precede d'un nom tres ancien de l'ete, et que ce nom soit identique au hittite *hamešha-: on posera un theme i.-eur. *Haxm-es(H)- «saison chaude, ete»-1 Le mot allemand repose sur un adjectif derive de cette forme: *Haxm-es(H)o- «d'ete, estivab.
Un autre parent est a chercher, a notre avis, dans l'adjectif tokharien A omal, B emalle «chaud». Le mot est inexplique, car les etymologies enoncees jusqu'ici ne convainquent aucunement: Lane, Lg. 14, 32 a compare lat. amariw, skr. am/a- «anier>, ce qui heurte contre la phonetique, a- du latin ne correspondent jamais au tokharien B e-; Windekens, Lex. Tokh. 80 rapproche gr. ijftap et arm. amr «jour>, ce qui dit peu pour le sens et pour la phonetique. — Dans notre hypothese, nous ne nous trou-vons en face d'aucune difficulte, car le *-o- i.-eur. a donne -e- en tokharien B et -a- du dialecte A, repondant a cet e, passait plusieurs fois a o
1 De rail, arnar. on a derive le nom allemand d'oiseau Ammer, voir p. ex. Kluge, Deuisch. EW. (lie ed.) 15. Niedermann. Gl. 19, 10 note 1 a compare le nom grec 'AfiaovXXlg, mais celui-ci nous semble etre tire du theme dftapv-conserve dans le verbe d/iapvoaoj.
ou u sous l'influence labialisatoire des labiales subsequentes. Voir Peder-sen, Tocharisch 219 et 220 sq. On posera une forme primitive *Hom-xl-pour le tokharien et se decidera pour un -o- primitif aussi dans la forme germano-hittite.
Une racine inconnue jusqu'ici *Hem~, *Hom- «chaud» sera a 1'origine de toutes les formes. Dans un article separe, nous essaierons de montrer que le groupe germanique comporte encore un autre exemple de cette racine, a vocalisme e.
Ce qui est tres interessant est que le germanique et le hittite com-portent un nom commun de la saison chaude; on remarquera qu'il y a d'autres correspondances speciales ent.re ces deux dialectes; nous ne citons qu'une seule, mais assez remarquable: hittite ateš «Beil» = v. angl. aclo/esa cm/sens*, voir Zbornik Filoz. Fak. II, a paraitre en 1955.
Hitt. uarkant-
«gras» (en parlant des animaux, de la viande) — pour la forme colaterale roaggant-, nee par l'assimilation de -r- affaibli a la consonne subsequente, voir Friedrich, Heth. Wb. 241 — comporte le suffixe augmentatif/ collectif -ant- qu'on rencontre dans un n ombre assez grand d'adjectifs hittites: maklant- «maigre», humant- «tout», irmalant- «malade» en face de irmala- «m/sens», etc. Pour l'etymologie, on partira alors d'un *uarka-plus ancien. Si celui provient d'un * nor go- ou *urgo- i.-eur., on y cher-chera la racine i.-eur. *uerg- «strotzen, schwellen (vor Saft und Kraft oder Zorn)» chez Walde - Pokorny 1 289, cf. skr. urj-, urja «Saft und Kraft, Nahrung», urjayati «nahrt, kriiftigt», gr. ČQyr/, etc.
I.e sens du mot hittite resulte de la meme evolution que le sen6 du gr. ntaQ et skr. pioaras «gras» en face de skr. payate «etre gonfle», p ay as «suc, lait».
Arm. dur,
gen. droy, dri, instr. droD «scalpello, scarpello; pialla» (selon Ciakciak) est reste sans etymologic.
Nous le rapprochons du groupe de mots indo-europeens que Pokorny, Idg. EIV. 272 cite sous *dlio- «scharfen»: skr. dharu «Schneide, Schiirfe, Klinge>, grec 0o6g «aigu». Le mot armenien peut done provenir d'un nom indo-europeen *dho-ro-s et est par la presque identique au skr. dharu, si celui-ci repose sur un nom ti suffixe *-ra- et noil sur un theme en *-la-. Le sens du mot armenien est tout a fait clair: les outils qu'il designe («burin» et «rabot>) sont naturellement des objets aigus et leur fonction n'est pas assez differente de celle du couteau. Le mot skr. montre done
le sens plus ancien que comportait jadis le mot armenien lui-meme. — Nous remarquons en passant que la forme indo-europeenne sur Iaquelle repose le mot arm. peut etre definie autrement: *dhd-tro-, done un nom a suffixe connu des noms d'outils. Mais nous preferons le *dho-ro- deja propose.
Pour les parents ulterieurs de *dho-, voir Zbornik filozof, fakultete II (Ljubljana), qui va paraitre vers la fin de 1955. U11 autre parent sera publie dans Slavistična Revija VIII 2.
Arm. dar,
gen. dari, instr. dariv «luogo alto, eminente, scosceso, precipitoso, preci-pizio, colle, collina, argine, rocca» est reste obscur. Pour nous, le mot est parent de dur. II s'agit d'une forme a suffixe *-ri- que montre p. ex. skr. ašri-š «cote coupant, coin» et surtout lat. ocris «mons confragosus». Ce dernier explique immediatement le rapport semasiologique entre le groupe indo-europeen et le mot dar: ocris et dar (prim. *dho-ri-) ont du signifier tous les deux «objet aigu; eminence du sol pointue».
Le couple dur et dar repond lettre a lettre a un autre du sens presque identique: arm. sur «epee» et sair «le tranchant»; ceux-ci reposent sur *kd-ro- et *ka-ri- (p. ex. Pedersen, KZ. 39, 407). 11 est tout a fait clair que les deux groupes ont agit l'un sur l'autre; inais il est impossible de constater le point de depart de ces rapports. Nous remarquons que dar peut corroborer la parente de v. angl. darop m. «Spiess, Wurfspiess* (v. Pokorny, 1. cit.).
Povzetek
Arm. azbn »le fila dell'ordito« gre na ievr. *ant-s-mn in k ievr. *ent-, bolje ant- »tkati«; sem spada zlasti še glosa dirjvieio ' did^eto loidv (Plesycli.).
Hetit. altanniš »Quelle, Brunnen« gre na ievr. koren *eld- ali *eldh- »vlažen, moča: močvara, blato«, ef. Pokorny, Idg. EW 305 in dalje zL let. aluots, aluugs »izvir«.
Hetit. hamešha(nt)- »pomlad«, pravzaprav »celotna topla doba«, gre k nem. Emmer »triticum dicoccum«, ki je izveden od *IIomes(H)6- »poleten« (ta vrsta žita se seje zlasti za poletno žetev, zato ime Sommerdinkel, Sommerspelz). Osnova je gerinansko-hetit. * Homes- »topla letna doba«. Z drugim sufiksoan: toliar. A omal, B emalle »topel« (*Hom-xlo-).
Hetit. uarkant- »masten« gre na *nerg- »biti prepoln sile, soka« (gr. dpyry, sti. urj- itd. »Saft und Kraft« itd.).
Arm. dur »dleto, skobci j« je iz ievr. *dlio-ro- in ustreza sti. dliara »rezilo«.
Arin. dar »špik, skala« izvira iz istega korena *dh<>- »oster«, tokrut v reducirani obliki: *dha-ri-. Odnos med obema je isti kot pri arm. sur »meč« in sair »rezilo«.