FUNDAMENTALS OF SLOVENIAN PAREMIOLOGY
PETER GRZYBEK
This article discusses some major theoretical issues concerning the documentation and study of proverbs in general, and of Slovenian proverbs in particular. The author argues in favor of a close interplay between paremiography and paremiology, and places particular emphasis on empirical methods in paremiology. Some initial results of an empirical study on familiarity with Slovenian proverbs are presented and discussed with regard to their relevance not only for paremiography, but also for linguistic, poetic, and other analyses.
Keywords: Proverb (theory, definition, classification), paremiography, paremiology, empirical approaches.
Razprava obravnava nekatera glavna vprašanja dokumentacije in raziskovanja pregovorov na splošno in še posebej slovenskih pregovorov. Avtor zagovarja tesno vzajemno povezavo med paremiografijo in paremiologijo. Pri tem prikaže in komentira prve izsledke empiričnega študija seznanjenosti s slovenskimi pregovori ne le z ozirom na njihov pomen za paremiografijo, temveč tudi za jezikoslovne, poetične in druge analize.
Ključne besede: pregovor (teorija, definicija, klasifikacija), paremiografija, paremiologija, empirični prijemi, metode.
INTRODUCTION
Slovenian proverbs have been the object of numerous scholarly studies. On closer examination, most of these studies have focused on the history of documenting and collecting Slovenian proverbs; that is, they are primarily oriented towards paremiography [cf., among others, Grafenauer 1952, Rode 1989, Stanonik 2000]. In comparison, paremiological studies (i.e., scholarly analyses of proverbs themselves) represent only a minority of approaches; in addition, they are positivistic and symptomatic, rather than systematic — for example, Forster's [1992] linguistic analyses. In this sense, Grafenauer's [1952: 17] early statement that the scholarly study of Slovenian proverbs has not yet been carried out satisfactorily remains true.
This state of the art is quite surprising, considering that ultimately the definition, classification, and selection of proverbs are three interrelated categories, making paremio-graphical and paremiological issues closely related, rather than separate faces of the same coin [cf. Grzybek 1992].
Given this perspective, this article presents some general theoretical problems related to the interplay between paremiography and paremiology, with specific reference to Slovenian proverbs. Particular emphasis is placed on empirical methods in paremiology and their relevance for paremiography. The initial results of an empirical study on familiarity with
traditiones, 37/1, 2008, 23-46
_razprave in razgledi / articles and essays_i
slovenian proverbs are presented and discussed with regard to their relevance not only for paremiography, but also for linguistic, poetic, and other analyses. Because little comparable empirical work is available, we begin with a short history of slovenian paremiography and place our approach in this framework.
the history of slovenian paremiography
Regarding the origin of slovenian paremiography, there is no definitive answer to the question of when the earliest documentation of slovenian proverbs took place. some consider the proverb collection by Janez Mihelič [1780] to be the beginning of slovenian paremiography; however, this collection has not been preserved.
Likewise, considering Hieronymus Megiser's [1592/1605] Pareomiologia the first (preserved) collection of slovenian proverbs is problematic because it has repeatedly been questioned whether these twenty proverbs are of slovenian origin or Croatian instead [cf. Radics 1882; Levstik 1895; Rode 1989]. More recent studies in this area have shown that the necessarily meticulous philological work has been deficient; in addition, the argument has been raised that, from a historical point of view, this material may be documented in both slovenian and čakavian Croatian, suggesting that the origin might be in the Istrian border region [cf. Eismann 1997]. Finally, the overall uncertainty as to the indigenous character of Megiser's proverbs has caused scholars to consider the approximately 150 proverbs published in oswald Gutsmann's Windische Sprachlehre [1777] and in his Deutsch-Windisches Wörterbuch [1789] to be the oldest preserved collection of slovenian proverbs [cf. Grafenauer 1935].
Regardless of possible additional insights into the origin of slovenian paremiography, aside from a number of smaller collections from the eighteenth and nineteenth centuries [cf. the bibliographical data in Klančar 1948], Kocbek's [1887] Pregovori, prilike in reki undoubtedly represents the first comprehensive, separate proverb collection. It is not only a milestone but, in fact, the foundation of slovenian paremiography. It goes without saying that this collection, too, refers to previous smaller editions of proverbs. However, it is of particular importance because it provided an important basis not only for its later expansion [kocbek & Sašelj 1934], but also for subsequent collections such those by Bojc [1974] and prek [1970].1
A different approach was pursued by ethnologist Marija Makarovič in her booklet Pregovori — Življenjske resnice [1975]. Although this work contains only 100 proverbs, and although no information is given regarding the origin and source of these proverbs, this book deserves particular mention here because, in contrast to the collections cited above, it
1 The various later editions of Bojc's and prek's collections speak strongly in favor of their popularity; however, it should be taken into account that the more recent editions contain not only popular proverbs but also, to a large degree, aphorisms and auctorial winged words, which are far from being of slovenian origin.
combines paremiography with empirical research on these proverbs' familiarity. Unfortunately, the two parts are rather disconnected, so that the results of the empirical study have no consequences for the paremiographic part. Instead, the paremiographic and paremiological aspects remain parallel to each other, which is characteristic not only for Slovenian. This is one of the reasons why Makarovic's approach deserves greater attention.
EMPIRICAL PAREMIOLOGY The Empirical Beginning of Slovenian Paremiology
One of the objectives of this article is to analyze Makarovic's approach in greater detail, trying to outline the relevance of empirical work in paremiology. This is to say that, without attempting to reconstruct the origin of the proverb sources included, the empirical approach as such (and its results) is discussed and placed into a contemporary framework of empirical methods in paremiology. Following a short presentation of Makarovic's approach, its results, and the author's interpretation, her material is re-analyzed. Her work is also methodologically evaluated by way of this re-analysis. After this presentation, some initial results of our own empirical study are presented in a second step based on Makarovic's material, but not using her method.
Makarovic's study was conducted in 1972 and 1973. A questionnaire with 100 proverbs was given to 64 persons from 57 different places in Slovenia; detailed information is given regarding the concrete places. Makarovic or one of her associates asked all of the subjects about three points: a) Familiarity with the 100 selected items; b) Ability to explain these proverbs' meanings; and c) An indication of proverbs known in addition to the 100 presented. The entire study is basically composed of three parts:
a)	The largest part [pp. 1—168] is represented by the 100 proverbs that served as the material basis of the study; the texts are accompanied by short explanations of the proverbs' meanings based on the relevant answers given by the informants.
b)	The second part [pp. 172—178] contains shorts comments on the proverbs, with some hints at the history of Slovenian paremiography and paremiology, and with some terminological comments. In addition, some results of the empirical study are reviewed.
c)	The third part [pp. 184-204] presents the results of the empirical study in greater detail, arranged by individual informant; in addition to personal data (age, sex, education, etc.), this includes each person's familiarity with the individual proverbs.
As already mentioned, and according to Makarovic, it was not her primary intention to provide an empirical study on the familiarity of Slovenian proverbs. Rather, she wanted to add some information about the proverbs' meanings based on the subjects' information; nonetheless, Makarovic's approach can serve as a good starting point for methodological discussion on studying proverb familiarity.
Empirical Approaches to Proverb Familiarity
A few words are necessary regarding the notion of "familiarity." First and foremost, it is important to distinguish familiarity from frequency. Although both categories are interrelated [cf. Grzybek 2008a, 2008b; Grzybek & Chlosta 2008], these concern basically different concepts. By way of a general characterization, one may say that frequency-oriented studies are text-based, whereas familiarity-oriented studies are knowledge-based and thus subject-dependent. In this context, an additional terminological differentiation seems necessary and appropriate: whereas (individual or collective) proverb knowledge refers to a person's (or a group's) subjective acquaintance with proverbs, proverb familiarity refers to either individual proverbs or to a group of proverbs — in any case, denoting average familiarity in a given population. Thus, regardless of the fact that, within a given group, the result of collective knowledge of particular proverb material coincides with collective familiarity, two essentially different perspectives are concerned. Frequency-oriented studies are based on text analysis (i.e., either written or spoken sources), whereas familiarity-oriented and knowledge-oriented methods are based on asking (or testing) persons, in one way or another. Consequently, the following discussion focuses solely on familiarity.
Regarding the concrete method by which proverb familiarity can be established, a variety of options are available [cf. Grzybek 2008a, 2008b]:
a)	In attempts to simply find some well-known proverbs, it has been regarded as sufficient to ask a particular number of subjects to write down the proverbs that spontaneously come to mind. In fact, this method may suffice for this specific interest. However, this approach usually yields only a limited number of prototypical proverbs (usually not more than 30 to 50 proverbs per person, differing on the time available and situati-onal context). This is because proverbs, by definition, tend to be used and therefore remembered only with regard to a specific situation.
b)	As an alternative, more or less comprehensive lists of selected proverbs have been presented to subjects and they are asked to introspectively state whether the proverbs are familiar to them or not. This method (called "full text presentation" or FTP in this article) thus demands that subjects make a clear binary yes/no decision; it has the advantage that people will also recognize and recall proverbs that do not spontaneously come to mind. On the other hand, this method has two obvious disadvantages: first, the results of introspection-based studies may be misleading because subjects may only think that they know a given proverb (or appreciate it as being "correct"), but in fact do not. Second, subjects may know a given proverb in a more or less divergent verbal form, and in this case may therefore be unsure what the correct answer is.
c)	In principle, the same objections hold true for scaling techniques, in which subjects are presented with a list of proverbs and a scale (e.g., from 1 to 7) on which familiarity with the individual proverb must be rated. This method is termed "full text rating" (FTR) in this article. In addition to the problems listed above (b), individual differences in rating may come into play and they must also be controlled very carefully.
d) A method that tries to avoid the problems outlined includes the presentation of only the beginning of a given proverb; the subjects' task is then to complete the text (e.g., Out of sight, ...). This method is called "partial text presentation" (PTP) here.
some Re-Analyses
The empirical results reported by Makarovic and the scant statistical analyses accompanying them do not go beyond an analysis of individual proverb knowledge: for each person, information is given on which proverbs they classified as "known" and on the overall number of proverbs known by each person. On the basis of the data given, it would have been quite easy to at least calculate some additional statistics on individual proverb knowledge. This minimal statistical re-analysis is presented here as a first step.
Because 64 informants each looked at 100 proverbs, there are 6,400 data points to analyze. The method used was full text presentation, and so two analytical categories seem to be sufficient at first sight: (0) unfamiliar and (1) familiar. However, given the fact that some persons wrote down variants, a third category (2) must be introduced for when subjects answered with variants only. As a result, of the total of 6,400 reactions, 3,263 (50.98%) units must be classified as "familiar." 3,102 units (48.47%) are unfamiliar, and 35 answers (0.55%), involved variants.
The relatively small number of variants implies that variants only play a minor role. Analyzing these variants in detail, it turns out that in most cases these are lexical variants, which do not significantly change the proverb's overall meaning. In addition to the fact that the percentage of variants is rather small, this is an additional argument in favor of combining categories (1) and (2). In doing so, it is important to note that there is no accumulation of variants, neither in the case of individual persons nor with regard to individual proverbs: 46 persons (71.88%) did not mention any variant, and a maximum of four variants was given by a single person (cf. Table 1).
Table 1: Number of variants given by subjects in the Makarovic study
Number of variants given	number of subjects	F( %)
0	46	71.88
1	9	14.06
2	2	3.13
3	6	9.38
4	1	1.56
Similarly, it can be shown that the variants given do not cluster on specific proverbs: for 75 of the 100 proverbs, no variant is given; for 18 proverbs, there is one variant, and for four proverbs there are two variants; the maximum of three variants involves one proverb only. As a consequence, all answers including only variants are counted as familiar here;
under these circumstances, 3,298 of the 6,400 possible answers need to be interpreted as an indication of familiarity with a given proverb.
It would go beyond the limits of this article to discuss the relevance of factors such as age, sex, place of residence, and so on; therefore, only some very general observations are reported here. Generally speaking, individual proverb knowledge ranges from 28 to 98 proverbs2 of the material presented, with an average proverb knowledge of % = 51.53 (s = 13.02). Only 35 of the 64 informants knew more than 50% of the proverbs presented; nine persons knew more than two-thirds of the material, and only one person more than 90%.
Figure 1 shows the degree of individual proverb knowledge, ranked in decreasing order:
With respect to proverb familiarity, the results are analyzed from a different perspective, focusing on the familiarity of individual proverbs rather than on individual persons' proverb knowledge. From this perspective, 52 of the 100 proverbs are familiar to at least 50%, 40 proverbs are known by at least two-thirds of the informants, 17 proverbs are known by 90% or more, and one proverb (Kdor visoko leta, nizko pade 'He that flies high falls low') is known by all 64 persons. Figure 2 shows the degree of proverb familiarity, ranked in decreasing order.
As can be seen, Makarovic's data thus provide interesting information, not only for paremiography but also for statistical analysis in paremiology.
It has already been noted that both sample size and the sociological structure of Makarovic's study provide only a poor basis for further analyses. Yet, another point shall be emphasized here: calling into mind the different options for studying proverb familiarity
2 Because exactly 100 proverbs were presented, the absolute number of proverbs corresponds to the percentage in this case.
Figure 2
(see above), Makarovic's approach would have to be classified as full-text presentation (FTP). This explains why there are relatively few variants in her study, suggesting a homogeneity of the individual proverbs' texts. Therefore, an attempt is made here to compare Makarovic's results with those achieved by another method: partial text presentation (PTP). Only a few selected examples can be presented here because a complete analysis would exceed the scope of this article.
The relevance of this comparison may be highlighted with two examples. Proverb no. 53 from Makarovic's collection was 100% known: Kdor visoko leta, nizko pade. One of her informants gave an alternative: Kdor visoko leta, nizko sede 'He that flies high sits low'. Interestingly enough, two very similar variants are given (as the only variants) in Kocbek's collection (Kdor visoko leta, nizko se usede 'He that flies high sinks low'), as well as in the Kocbek & Saselj collection (Kdor visoko leta, nizko seda 'He that flies high sits low') — the ...nizkopada 'falls low' version, however, is not contained in these two collections. How can one know which of these versions is most likely to be known? One may object that this concerns linguistic or philological details, but the problem behind this observation is quite relevant and important not only for linguistics, but for paremiology as well. After all, two crucial questions are at stake here. The first question is which variant(s) should be included in a future proverb collection. The second question is which of these variants is familiar to what degree. This question, too, is relevant for paremiography, if the inclusion of obsolete or individual variants is to be avoided, and only current variants are to be included.
In this context, then, a theoretical definition of the proverb and of proverb variation comes into play, which must be discussed here. After all, using the PTP method, it has to be decided whether each individual completion concerns a variation of a given proverb or
one of its possible variants, or perhaps even a different proverb, or no proverb at all, in the extreme case. For example, item no. 42 from Makarovičs data reads Kdor ne dela, naj tudi ne je 'He that doesn't work also doesn't eat'. The Kocbek & Sašelj variant is Kdor ne dela, je brez jela 'He that doesn't work is without food', which may definitely be considered a variant of the proverb, with an additional rhyming ending, as well as a different syntactic and lexical structure. However, what about the two variants given in the Kocbek edition Kdor ne dela, mora krasti ali krave pasti 'He that doesn't work must steal or herd cows' (no. 893), and Kdor ne dela, pa platno prodaje 'He that doesn't work sells linen' (no. 894)? What kind of criteria should be applied here to decide to what degree a modification of an expected version should be regarded as a variant? After all, this question is related to the basic question of paremiology; namely, what is a proverb?
WHAT IS A PROVERB? PAREMIOLOGICAL SEMIOTICS IN A NUTSHELL
A definition of the proverb as a folkloristic genre is therefore a necessary precondition for any proverb classification. One might expect paremiology to easily answer the seemingly elementary question of what a proverb is. However, paremiology is no discipline in its own right, or with its own method; instead, proverbs are studied by various disciplines such as linguistics, anthropology, education, sociology, cultural history, folklore studies, literary scholarship, ethnography, etc. This is why one obtains at least as many answers to the question of what a proverb is as there are approaches to the proverb.
Furthermore, this is why there are so many concurring statements such as "Proverbs are the wisdom of the street," "Proverbs are condensed experiences," "Proverbs contain (at least a kernel of) truth," "Proverbs are true," and so forth. From a semiotic perspective, relevant questions such as "Are proverbs true?" or "Which proverbs are true?" are incorrectly asked; rather they should ask "When are proverbs true?" or, even better, "When do they fit?" The slight difference between these wordings may be seen, among others, if one takes into account the existence of antonymous proverbs, such as Out of sight, out of mind, and Absence makes the heart grow fonder. Both proverbs may turn out to be true, or contain a kernel of truth, when they refer to a corresponding situation. In this respect, as has been repeatedly shown, different notions of "situation" must clearly be kept apart, as was first noted by Seitel [1969]:
a)	The interaction situation in which a proverb is uttered, with a communicative act between some speaker(s) and hearer(s);
b)	The reference situation that the proverb refers to;
c)	The proverb situation, verbally expressed in and by the proverb itself.
The "correct" usage, or rather application, of a proverb thus implies, as Seitel has shown, two aspects (cf. Figure 3):
a)	The use of a proverb in a given interaction situation must satisfy all pragmatic needs and restrictions;
b)	The proverb situation and reference situation must be adequately related to each other.
Ultimately, the distinction of proverbs from proverbial sayings and phrasemes is inherent in the notion of "situation" if one follows Permjakov's ideas on understanding a proverbial situation as the denotation and/or modeling of two paremic concepts. In this sense, then, items such as igrati z ognjem 'to play with fire' or odkriti/razkriti svoje karte 'to reveal one's hand' would not be understood as "situation," denoting one concept only and therefore belonging to the category of phrasemes. Compared to this, expressions such as izganjati hudica z belcebubom 'to exorcise a devil with Beelzebub' or priti z dezja pod kap 'to come out of the rain under the eaves (= out of the frying pan into the fire)' would be classified as proverbial expressions — although they contain some kind of situation, they are neither linguistically nor logically closed, but need an explicit referential completion, logically denoted by existential quantification. Proverbs such as Ni dela, ni jela 'No work, no food' or Ni pravila brez izjeme 'There is no rule without an exception', as opposed to this, are logically self-contained statements that also tend to be linguistically closed statements involving universal quantification. Arvo Krikmann [1984] has shown the relevant differences in logical notation, focusing on the two different kinds of quantification, existential (3x) and universal (Vx), on the most relevant concept of implication (c), and on the phraseological and/or paremiological information along with it (Px, Qx): Proverbs:	Vx(Px c Qx)
Proverbial Sayings: 3x1(Px c Qx) Phrasemes:	3 x1(Px)
The distinctions mentioned are highly relevant in this context, if one takes into consideration the fact that there are quite a number of non-proverbs among Makarovic's 100 items; for example, Wellerisms (85), phraseological comparisons (36, 49, 51), and proverbial expressions (40).
Be this as it may — following Seitel's schema, which pays attention to the literal meaning of the proverb text only, it has been suggested that ultimately proverb semantics should be explained by taking into account two levels of signification [cf. Grzybek 1984, 1987/94]. on the first level, a literal meaning is conveyed, whereas the second level aims at the deeper
proverbial meaning of what is actually "meant," not what is "said" explicitly on the verbal surface.3
The elaboration of this concept has developed and significantly changed over the years. Because it has been thoroughly described previously [Grzybek 2000], only the essentials need be mentioned here.
First, elaborating on Seitel's schema, the concept of double analogy has been suggested [cf. Grzybek 1984/87], involving (a) the analogical integration of the first and second level of signification into some abstract meaning, and (b) the integration of this abstract meaning with the reference situation on the other (cf. Figure 4).
Later, it was argued [cf. Grzybek 1998a, 1998b, 1998c] that this scheme of double analogy needs to be more differentiated and, in fact, extended. To this end, two additional concepts of "situation" have been introduced; namely, the situation model on the one hand, and the model situation on the other. The reason behind this concept is the assumption that two processes of abstraction are involved:
a) Some kind of general meaning must be deduced from the literal meaning; it seems reasonable to use the term "situation model" for the result of this abstraction; The individual and unique reference situation must be attributed to some general type, or class, of situation, which may be termed the "model situation." As a result, the schema depicted in Figure 5 has been suggested.
b)
3 It should be mentioned that this concept neither implies any need for static concepts of denotation and connotation, nor any assumptions regarding possible stepwise (successive) literal-first models of understanding. Rather, it concerns dynamic semiotic processes that, theoretically speaking, are infinite in principle but tentatively interrupted in order to guarantee understanding and communication.
However, even this schema is still problematic because it gives rise to the impression that some kind of symmetry characterizes the entire modeling process. In fact, Figure 5 characterizes some synchronous final state, which might suggest that a proverb's meaning can be derived from the verbal surface structure without knowledge of the reference situation or model situation. This may be possible, but only a posteriori; that is, only if one knows the referential conditions and restrictions — or, in other words, when one has constructed some model situation from a number of applications (i.e., referentializations). Thus, in order not to reflect the final state, but the modeling process in its genesis, a schema like the one in Figure 6 is necessary.
■4 A-B(i^)"|-
i m i
	--i c		D' <		
					
					
10,-0,1 1 C3~D, 1 1 C3 ~ D; | 1 C4~D, C„-D„ 1					
Proverb situation Situation m«del
Model situation
Keferenee situation^)
This schema illustrates the essential interrelation between semantics and pragmatics by emphasizing that it is not possible to generate a model situation without repeated re-ferentialization (or its semiotic mediation), and by making clear that the entire modeling process cannot be reduced to the second level of signification. This framework thus yields the following definition:
A proverb may be considered a model of a situation denoted by it if— possibly in a given interaction situation (I) — such a situation model (lib) may be derived from the proverb situation (Ila), which stands in an isological relation to a model situation (Illb) abstracted from a concrete reference situation (Ilia).
With this definition in mind, it is much easier understand what is variable and what is invariable with regard to proverbs. The proverb situation and reference situation are variable, whereas the situation model and model situation are invariable — if not, then either one is dealing with a different proverb or the proverb is inappropriately applied (according to the given culture's standards). In other words: the concrete verbal surface of a given proverb (i.e., the proverb situation) may be modified as long the situation model attributed to it remains unchanged; likewise, the usage of a proverb may be regarded as "correct" as long as the proverb is applied to reference situations that a given culture understands as variants of one and the same model situation.
This concept thus only includes a working definition of the proverb; it allows for a better distinction of proverbs from proverbial sayings, idioms, phrasemes, and so on, and it
also opens the door for a semiotically based proverb classification. Both questions are very relevant for both paremiologists and paremiographers.
FROM DEFINITION TO CLASSIFICATION
Knowing (at least approximately) what a proverb is makes it possible to tackle the question of classification. Here we do not mean alphabetical arrangement (e.g., in Kocbek or Kocbek & Saselj) or "thematic" classification (e.g., in Bojc and Prek, among others), which puts all proverbs about "eating," for example, into one category. Such classifications have their advantages: alphabetical arrangement, for example, can facilitate retrieval of a given proverb in a proverb collection, but only if one knows its exact wording (which may vary). The same holds true for thematic classifications based on surface imagery, where one can find all proverbs about "eating" (for example) in one category, regardless of whether eating has something to do with a deeper, proverbial meaning or only serves as a vehicle on the proverb's verbal surface. Unfortunately, many editors of proverb collections do not sufficiently distinguish between these two kinds of thematic classifications, disregarding the theoretical distinctions outlined above.
This is not the case with the M6 system, originally based on Kuusi's [1972] International Type System of Proverbs, and perfected by Outi Lauhakangas [2001]. This system also strives for thematic classification, but it clearly aims at deeper, abstract meaning.
To give but two examples, again by way of illustration, with the notion of "eating" involved: The proverb If the fish are not biting, the worms are spared is classified in category (E) "Sense of proportion." This may seem odd at first sight, but there is a subgroup here termed "Relativity of ranking / the essential unity of differing things" (E1), and finally, within this subgroup, another one termed "Things look different when seen from different perspectives, paradoxes concerning luck in misfortune" (Elk). This classification therefore makes some sense, although the question remains whether the classification under "sense of proportion" still meets the intuitions of users of proverb dictionaries. A second example, Cleanliness is half the food, shows how difficult it is to keep apart the two levels of signification when dealing with thematic classification. In the M6 system, this is categorized among D4d "Eating habits and norms" — eating, however, is not at all what is at stake here; instead, it deals with cleanliness, which is rewarded.
Given these disadvantages, it seems reasonable to pursue Permjakov's original idea to strictly separate thematic and logical classification, and to classify proverbs not only by theme [cf. Grzybek 2000]. This approach recalls the above discussion of the notion of situation(s) because, ultimately, Permjakov's classification of proverb meanings coincides with the differentiation of specific types of situations modeled in proverbs. In this respect, it is important to keep in mind that Permjakov did not distinguish different types of situations as strictly as has been outlined above; therefore, some of his classifi-
cations are rather deep-structure oriented, whereas others are clearly surface-based. In any case, a classification strictly oriented towards deeper meaning is quite possible [cf. Chlosta & Grzybek 2000].
According to Permjakov, there are four basic types of situations (IA, IB, IIA, IIB), depending on the complexity of the relationships between the entities modeled in a given proverb: IA and IB model relationships either between two entities (E) or one entity and (one of) its qualities (Q); in comparison, IIA and IIB model the dependency between the relationships of entities and the relationship of their qualities. Disregarding the subtle subdivision of further logical classes and subclasses yields the following four basic groups:
IA E ^ Ee Q
Every entity has a particular quality (or trait). IB Ej ^ E2
If one entity (Ej) exists, another (E2) exists as well.
IIA (E1 - E2) ^ (E1 e Q ^ E2 e Q2)
If two entities are in a particular interrelation, the qualities of these two entities correspond to each other (or not).
IIB [(E1 e Q1 a E2 e Q2) a (Q > Q2)] ^ (E1 > E2)
If an entity has a particular quality, and another one has another quality or the same to a different degree, then one of the two entities is preferable as compared to the other.
This logical classification is then completed, by way of a complementary addition, by a thematic classification, and the proverb's meaning ultimately derives from a combination of both. Binary oppositions such as male—female, friend—non-friend, good—bad, and so on serve as an instrument to fill the thematic slots. These oppositions can also be subcategorized into various classes (which cannot be discussed in detail here).4
This system can be illustrated using some Slovenian examples. The proverb Roka roko umije 'One hand washes the other' is a typical representative of type IB, if one assumes that it is about the reciprocity of "actions" as the entities dealt with ("The existence/ presence of one entity depends on the existence/presence of another entity"). There are two options for the semantic filling:
a) The relevant actions are regarded as rather neutral — this is thematic opposition III1 ("Action — Reaction").
Quite obviously, this approach owes much to structuralist ideas that, in recent years, have been submitted to legitimate theoretical critique. It goes without saying that, from a philosophical point of view, much of this general critique is justified if claims are made about the innate binarism of thought, authoritative determination of meaning, closeness of the descriptive system, and so on. However, such criticism can relate to any other kind of classification, and Permjakov's system need not necessarily imply the claim of these criteria. As long as one does not generally deny the existence of cultural agreements on proverb meanings, one may well understand Permjakov's approach as an attempt to describe proverb meaning from a dynamic perspective, as a tentative description of its semantic potential.
4
b) If one restricts this proverb to negative actions only, as is done in some cultures, an additional pair would have to be introduced, namely IIa10 (Good—Bad). The resulting meaning is therefore "If there is a particular (good) action, there will also be a corresponding (good) reaction."
Such differences are highly unconscious in the simple use of proverbs, and even pare-miologists probably do not always pay attention to such fine differences.
A second proverb may serve to illustrate the relevance of cultural knowledge for the semantic classification of proverbs, and thus the need to take into consideration culturally accepted and unacceptable reference situations. From an external point of view lacking sufficient cultural knowledge, a proverb such as Voda tece, mlini meljejo 'Water flows, mills grind' may be interpreted in two different ways:
a)	One may interpret it in terms of a statement about particular qualities of things; in this case, the proverb belongs to category IA (E ^ Ee Q): Every entity has a particular quality (or trait). The statement would sound like "water" flows, "mills" grind — everything has its own characteristic, and it behaves correspondingly.
b)	By way of an alternative, the proverb might be interpreted in terms of a logical implication between one entity and another, depending on the first: If "water" flows, then (water) "mills" can grind; the proverb would then be classified in category IB (Ej ^ E2): If the one entity (El) exists, the other one (E2) exists as well — if the necessary conditions are fulfilled, things go as they should.
Of course, any native speaker of Slovenian, or any person enculturated in a Slovenian-language environment, will reject one of these two options as incorrect, if not impossible; this kind of reaction is quite typical from an internal point of view, but from an external perspective both options are available, with equal right.
Given what a proverb is and a way to classify them, two further questions remain to be answered:
a)	How does one know which proverbs should be included in a proverb collection?
b)	How does one know which linguistic variant of these proverbs should be included?
It is obvious that only empirical studies can answer these questions.5 This comes back to the empirical study by Makarovic discussed above. Based on the above discussion, the importance of empirical approaches for paremiography and paremiology should be obvious. With this in mind, we now turn to some preliminary results of our own study on proverb familiarity using Makarovic's data, but not her method.
For a general presentation and discussion of empirical approaches in paremiology, including methodological options and adequate statistical procedures, see Grzybek et al. [2009].
FROM THEORY TO PRACTICE (RETURN TICKET INCLUDED) The Experiment
In order to see how the PTP method works for Slovenian and to test Makarovic's material, we conducted an empirical study in Slovenia. 103 informants took part in this study, all of them native Slovenian speakers. They were presented with only the beginning of the 100 items, and their task was to complete the remaining text based on their knowledge.
Here we can offer neither a complete analysis of the overall performance, nor all of the individual proverbs systematically in detail. Nonetheless, some general tendencies and especially a number of illuminating examples may show the relevance of the suggested method.
Sample Structure
The subjects' average age 39.67 years (s = 14.62), with no significant difference between the 39 male and the 64 females (cf. Table 1).
Table 1: Age Structure of the Sample
		A	ge
Sex	n	x	s
Male	39	39.18	14.61
Female	64	39.97	14.77
Total	103	39.67	14.62
The informants came from various Slovenian regions. The majority (58.25%) came from Styria, 18.45% from the Littoral, and 13.59% from Upper Carniola. For further details, see Table 2.
Table 2: Regional Structure of the Sample
Region of origin	n	%
Styria	60	58.25
Littoral	19	18.45
Upper Carniola	14	13.59
Other	6	5.83
Unknown	4	3.88
Total	103	
Regarding the educational level of the participants, almost half of the informants had a medium education, and about a third a high education; for details, see Table 3.
Table 3: Educational Structure of the Sample
Education level	n	%
Low	14	13.59
Medium	50	48.54
High	36	34.95
Unknown	3	2.91
Total	103	
All of the participants in this study were presented with the 100 proverbs from the Makarovic study. In contrast to her study, only the beginning of each proverb text was given in written form to the participants, who were asked to supply the missing ending.
Taming the Modifications
Because verbal modifications and variations are to be expected in the completions, it would not be appropriate to accept only completions without any variation as "familiar" and to ignore any kind of linguistic variation. However, there are different kinds of variations, ranging from minor orthographic errors, to lexical and syntactic alternatives, to the "invention" of "new" proverbs, which have nothing to do with the items presented. Within empirical paremiology, the first aim is to document all of the completions and their occurrence frequencies, which requires a detailed classification key. In addition to such meticulous documentation — which may seem hypertrophic but is a good way to gain insight into the variation spectrum of a proverb and to determine the most frequently used forms — this classification schema should also allow some kind of placement in superordinate categories. To this end, variations may be classified along a number of major modification categories:
a)	zero modifications are cases in which a proverb is completed in the "expected" form. With regard to this "zero variant" (V0), it is important to note that in empirical pare-miology this form does not imply any kind of a priori prejudice regarding frequency, as opposed to traditional paremiography and paremiology, in which terms such as "standard form," "basic variant," and so on imply that this form is the most common form in comparison to all kinds of variations. As a matter of fact, such a V0 may well turn out to be the most common or most frequent one, but only a posteriori; that is, as a result of empirical study.
b)	First-degree modifications of the V0 (also) indicate familiarity with the given zero variant; such modifications are termed "variations" hereinafter; these variations include, for example, the omission of single words that are not crucial to the proverb's semantics, orthographic or morphological variations, and lexical or syntactic variations, as long as the modification does not result in a second-degree variation (see below).
c)	second-degree modifications of the V0 result in a proverb form that no longer implies knowledge of the zero variant, but of the proverb type (in the paremiological meaning
of this word) underlying it. Such modifications are termed "variants." These variants include shortenings or prolongations, lexical variations (other than lexical synonymy), and so on.
d) Third-degree modifications imply neither knowledge of the V0 nor of the proverb type underlying it. In some cases, these may involve a different or "new" proverb, and in other cases nonsense answers or other modifications. Some other kinds of third-degree modifications have been attributed to categories in their own right; for example, completions representing explicit negations of the zero variant, unreadable answers, and so on.
Each of these basic categories is further subdivided so that each individual completion may be adequately classified. The entire classification schema has been presented in detail elsewhere and need not be presented here [cf. Grzybek, Chlosta, & Roos 1994; Chlosta & Grzybek 1995, 2005]. With regard to the question of proverb familiarity, it is thus important to note here that, in addition to all completions in the form of the v0, all first- and second-degree modifications may be recoded as "familiar." The results of applying this classification schema to the completions of all 100 proverbs will published in detail elsewhere; therefore, we focus here on some illustrative examples as well as on some general findings and conclusions.
Some General Results
On the whole, as the possible (theoretical) maximum, there were 10,300 (100 x 103) completions. 2,170 proverbs (21.07%) were completed in the form of the V0 and can therefore safely be regarded as familiar. 1,099 completions (i.e., 10.67% of all completions) were first-degree variations, and 451 (4.38% of all completions) were second-degree modifications. Accepting all these first- and second-degree modifications as "known," a total of 3,720 completions (i.e., 36.12%) are calculated as "familiar" under the PTP condition, as compared to 50.98% familiar items in the Makarovic study.
Among all the 3,720 completions are counted as familiar, there were 1,550 variants and variations, summing up to 41.66%. In addition to the fact that the PTP method yields reliable results, as a general result variants and variations play a much greater role in proverb use and proverb knowledge than has hitherto been assumed. Recall that there were only 35 variants (0.55%) in Makarovic's study.
5,554 (53.92%) items were classified as "unknown" because they remained uncompleted by the informants. Comparing these results to those of Makarovic's FTP study, the percentage of definitely unknown proverbs is quite comparable, at 48.47% in the Makarovic study as compared to 53.92% uncompleted items in our PTP study.
As mentioned, the overall result for familiarity was x = 51.53 in the Makarovic FTP study, as compared to 36.12% in our PTP study. This result is not really surprising, showing that overall performance is lower under the PTP condition, and implying that the PTP task has higher distinction. Nonetheless, it would be interesting to see exactly how
the performances under both conditions relate to each other. A first approach would be to test for correlation; because the results of both studies are not normally distributed, the calculation of the Pearson correlation coefficient must be replaced by the non-parametric Spearman rank correlation coefficient, which in our case, at p = 0.892, is highly significant (p < 0.001). This is to say that the higher a proverb's score on the FTP condition, the higher its score on the PTP, but at a lower level. In other words: both methods do indeed measure the same phenomenon, but obviously on different levels: the PTP method is sharper and more demanding.
Interestingly, for 78 of the 100 items, the FTP method yields higher results, whereas in 22 cases this is the reverse. Nonetheless, the extent of the differences is not the same in both cases: if FTP > PTP, then the differences amount to about 22% on average, but only to about 8% for PTP > FTP. One would have to check these differences for significance, of course, and test them for each individual item — which clearly goes beyond the scope of this article.
Individual Proverbs — Some Examples
Regarding the individual proverbs, 9 of the 100 proverbs were familiar to more than 95% of the informants, 19 of the 100 had degree a familiarity > 90%, 30 proverbs were familiar to more than two-thirds of the informants, and 51 proverbs were familiar to more than half.
We now turn to the analysis of some examples, first concentrating on the most familiar items. Only one item attained a familiarity of 100%; interestingly enough, it was completed by all 103 informants in one and the same form and, in addition, in form of the V0: Iz te moke | ne bo kruha 'This flour will make no bread'. On closer inspection, however, it turns out that this is not a proverb.
A preliminary analysis of the most common proverbs yields some interesting results. The proverbs with a familiarity of more than 95% among the informants numbered nine items; not surprisingly, they had a high degree of familiarity in the Makarovic study as well. For eight of these, it was clearly the V0 that was the favorite completion.
Table 4: Highly Familiar Proverbs
No.	Proverb beginning	Proverb ending (according to Makarovič)	Familiar ratings (abs.) with Vn in parentheses	Familiar ratings
20	Laž ima	kratke noge.	102 (100)	62
53	Kdor visoko leta	nizko pade.	102 (94)	63
87	Kjer se prepirata dva,	tretji dobiček ima.	102 (101)	54
99	Jabolko ne pade	daleč od drevesa.	102 (101)	61
46	Kadar mačke ni doma,	š. -S' šš	101 (93)	60
57	Ni vse zlato	kar se sveti.	101 (100)	59
70	Po toči	zvoniti je prepozno.	99 (64)	63
72	Tudi slepa kura	zrno najde.	99 (84)	59
A first interesting observation among these highly familiar items is proverb no. 70 After the hail it's too late to ring the bell (= It's no use crying over spilt milk)'; here in addition to the V0, a second variant seems to be common (Table 5 does not show all given variants, but only the two most frequent ones for this proverb). In fact, the V0 with its 64 completions is still the dominant one, at 62.10%, as compared to the syntactic variant with 27.2%:
Table 5: Proverb no. 70
No.	Proverb beginning	Proverb ending	Cat.	f	f(%)
70	Po toči	zvoniti je prepozno	1000	64	62.10
		je prepozno zvoniti	1701	28	27.10
This fact is the more interesting, considering that the Kocbek & Saselj edition (for example) has the variant Po toči je zastonj zvoniti 'After the hail there's no point in ringing the bell', which turns out to be out of use (regardless of whether it ever was in use).
One might object that this is only a minor deviation; however, such modifications are always systematic, at least as soon as they are accompanied by a particular frequency with which they occur. This distinguishes them from guessing, instantaneous associations, ad-hoc solutions, word plays, and so on (which, taken individually and not as a group, tend to occur rather rarely).
One might also object that such orthographic, phonological, morphological, syntactic, and other differences may seem to be the hypertrophic result of linguistic or philological interests. However, such differences can have a significant impact on the word or sentence length of a given proverb (i.e., on its overall linguistic, metric, or rhythmic structure), shaping not only its external form, but consequently also its memorability, usage, frequency, and so on.
Among these highly familiar proverbs, there is even one for which another modification turns out to be more frequent than the V0: This is proverb no. 59 'Do not praise the day before it is over', for which the V0 (also found in the Kocbek & Sašelj collection) is given in 32 cases only, as compared to 66 completions with a lexical variation:
Table 6: Proverb no. 59
No.	Proverb beginning	Proverb ending	Cat.	f	f(%)
59	Ne hvali dneva	pred večerom.	1000	32	31.10
		pred nočjo.	2601	66	64.10
This is one case in which a second-degree modification turns out to be the most frequent, and this is a good example for showing the relevance of distinguishing both levels: Of course, in both cases the proverb meaning does not change, it is one and the same proverb type in the sense defined above. The reason for this is that večer 'evening' and noč 'night' here, in opposition to dan 'day', both serve as metonymies to refer to the abstract concepts of 'beginning' and 'ending'; thus, within this proverb, večer and noč represent a synonymous
pair; however, without this concrete context, they are definitely not. This is why this is a second-degree modification.
With this in mind, let us take a look at some more examples. The first proverb in Makarovičs sample is Podarjenemu konju se ne gleda na zobe 'Don't look a gift horse in the mouth'. In the original Kocbek collection, it has a different preposition (Podarjenemu konju se ne gleda na zobe), and in the Kocbek & Sašelj edition an imperative (Podarjenemu konju ne glej na zobe). Interestingly enough, our results show that none of these three versions turns out to be the most frequent one — which, in fact, is a combination of all three (see Table 7).
Table 7: Proverb no. 1
No.	Proverb beginning	Proverb ending	Cat.	f	f(%)
1	Podarjenemu konju	se ne gleda na zobe	1000	22	21.36
		se ne gleda pod zobe	1501	4	3.88
		se ne gleda v zobe	1502	43	41.75
		ne glej na zobe	1702	12	11.65
		ne gledam v zobe	1802	1	0.97
		ne gledamo v zobe	1804	2	1.94
		ne gledamo zobe	1805	1	0.97
		ne gledamo zobov	1806	1	0.97
		ne glej pod zobe	1807	1	0.97
		ne glej v zobe	1808	4	3.88
		ne smeš gledati na zobe	1810	1	0.97
		se ne gleda zob	1811	1	0.97
		se ne gleda v usta	2701	1	0.97
		ne smeš gledat na zobe	2711	1	0.97
		—	0	8	7.77
				103	100
Interestingly enough, the obvious (and of course "unconscious," in the Jakobsonian understanding of this term) shaping may include even "extensions" (a term that, strictly speaking, is incorrect in this context because it would still refer to some codified basic form whose relevance was called into question above).
For proverb no. 93, for example, Makarovičs version Kar se Janezek nauči, | Janez zna 'Johnny learns what John knows', was completed by only seven persons (6.80%) in this form. In contrast, 33 informants (32.04%) completed it with the first-degree variant ...to Janez zna; that is, with an additional to 'that'; even more surprisingly, the majority of 44 subjects (42.72%) filled in ...to Janezek zna 'Johnny knows that'; that is, neutralizing the crucial diminutive, resulting both in an extension and a significant change of meaning.
Let us conclude with one final example that has already been discussed above: What
would the informants do with the beginning Kdor ne dela____'He that does not work'? We
already mentioned that, as compared to Makarovic's suggestion ...naj tudi ne je 'also should not eat', there are different versions in the Kocbek and Kocbek & Saselj editions. In fact, none of these versions was offered by any particular informant. However, more strikingly, Makarovic's version was also completed by only one person; as opposed to this, no fewer than 86 persons (83.80%) filled in ...naj ne je 'should not eat', thus clearly paralleling the rhythmic structure of the proverb's beginning.
CONCLUSIONS
The foregoing discussion has shown how closely interrelated paremiography and paremiol-ogy are, and how important the interplay between theoretical and empirical approaches is. The definition of the proverb is not only a necessary precondition for separating related forms, such as proverbial expressions, idioms, phraseological fusions, and the like; it is also necessary for classification. Empirical work comes into play both in the decision about what to include in a proverb collection, and which variant(s) to include: more often than not, compilers of proverb collections simply copied from previous collections; in other more rare cases, when proverbs have been collected anew, the collected items have been recorded and documented with painstaking effort, but hardly ever have studies been undertaken to determine the overall familiarity of the collected items in a given culture. Therefore, hardly any reliable data are available about common linguistic forms, which is crucial for any further analysis in this direction. Of course, much more empirical work and systematic analysis is necessary than could be presented in this article. As a consequence, even the first systematic analyses of Slovenian proverbs, concentrating on linguistic form with regard to Slovenian proverbs, initial analyses of linguistic form, such as word length [Grzybek 2004] and sentence length Grzybek [1999], may consequently be partly misleading because these analyses were based on traditional dictionary proverb material. In order to obtain reliable data, future linguistic and paremiographic researchers would therefore be well advised to base their analyses not only on a well established (empirically validated) selection of proverbs, but also on well established (empirically tested) variants and variations of these proverbs.
REFERENCES
Bojc, Etbin
1974 Pregovori in reki na Slovenskem. Ljubljana [21980, 31987]. Chlosta, Christoph & Peter Grzybek
1995 Empirical and Folkloristic Paremiology: Two to Quarrel of to Tango? Proverbium. International Yearbook of Proverb Scholarship, 12: 67-85.
2000 Versuch macht klug? Logisch-semiotische Klassifikation bekannter deutscher Sprichwörter. In: Peter Grzybek (ed.), Semiotische Studien zum Sprichwort - Simple Forms Reconsidered I, 169-199. Tübingen: Narr.
2005 Varianten und Variationen anglo-amerikanischer Sprichwörter - Dokumentation einer empirischen Untersuchung. ELiSe [Essener Linguistische Skripte elektronisch], 5/2: 63-145.
Eismann, Wolfgang
1997 3aMeTKH k cöopHHKy ^0CA0BH^ Meracepa "Paroemiologia polyglottos" 1592 h 1605. In: Peter Durčo (ed.), Frazeologicke študieII, 63-77. Bratislava: Esprima.
Grafenauer, Ivan
1935 Gutsmannov besednjak in njegova zbirka pregovorov, rekov in prilik. Časopis za zgodovino in narodopisje 30: 1-29.
Grzybek, Peter
1984 Überlegungen zur semiotischen Sprichwortforschung. In: Peter Grzybek (ed.), Proverbium. Yearbook of International Proverb Scholarship 4 (1987), 39-85. Repr. in Wolfgang Mieder (ed.), Wise Words. Essays on the Proverb, 31-71. New York: Garland.
1999	Wie lang sind slowenische Sprichwörter? Zur Häufigkeitsverteilung von (in Worten berechneten) Satzlängen slowenischer Sprichwörter. Anzeiger für Slavische Philologie 27: 87-108.
2004 Zur Wortlänge und ihrer Häufigkeitsverteilung in Sprichwörtern (Am Beispiel slowenischer Sprichwörter, mit einer Re-Analyse estnischer Sprichwörter). In: Christine Palm-Meister (ed.), Europhras 2000. Internationale Tagung zur Phraseologie vom 15.-18. Juni 2000 in Aske / Schweden, 161-171. Tübingen: Stauffenburg.
2008a Proverb Familiarity and Proverb Frequency: Two Faces of One Coin or False Friends? Presentation at the International Phraseological Conference "Europhras." Helsinki, Finland.
2008b Proverb Familiarity in Theory and Practice. Presentation at the 2nd International Symposium on Proverbs. Tavira, Portugal.
Grzybek, Peter (ed.)
1984 Semiotische Studien zum Sprichwort - Simple Forms Reconsidered I, 215-250. Tübingen: Narr.
2000	Die Grammatik der sprichwörtlichen Weisheit von G. L. Permjakov. Mit einer Analyse bekannter deutscher Sprichwörter. Hohengehren: Schneider.
Grzybek, Peter, & Christoph Chlosta
2008	Some Essentials on the Popularity of (American) Proverbs. In: Kevin McKenna (ed.), Festschrift on the Occasion of WolfgangMieder's 65th Birthday. Burlington, VT: University of Vermont.
Grzybek, Peter, Christoph Chlosta & Rüdiger Grotjahn
2009	Empirische Sprichwortforschung: Methoden - Probleme - Status - Perspektiven. (In print).
Klančar, Anthony J.
1948 A Tentative Bibliography on the Slovene Proverb. Journal of American Folklore 61: 194-200.
Kocbek, Fran
1887 Pregovori, prilike in reki. Ljubljana: A. Trstenjak.
Kocbek, Fran, & Ivan Šašelj
1934 Slovenski pregovori, reki in prilike. Celje: Družba Sv. Mohorja.
Krikmann, Arvo
1984 1001 Frage zur logischen Struktur der Sprichwörter. In: Peter Grzybek (ed.), Semiotische Studien zum Sprichwort - Simple Forms Reconsidered I, 387-408. Tübingen: Narr.
Kuusi, Matti
1972 Towards an International Type-System of Proverbs. Helsinki: Suomalaien Tiedeakatemia. Lauhakangas, Outi
2001 The Matti Kuusi International Type System of Proverbs. Helsinki: Suomalainen Tiedeakatemia. Levstik, Fran
1895 Zbirka slovenskih pregovorov iz leta 1592. In: Levstikovi zbrani spisi III., 254-257. Ljubljana: I. Kleinmayr & F. Bamberg.
Mieder, Wolfgang
1988	Einleitung. In: Die deutschen Sprichwörter. Gesammelt von Karl Simrock, 7-15. Stuttgart: Reclam.
Mieder, Wolfgang
1989	Moderne Sprichwörterforschung zwischen Mündlichkeit und Schriftlichkeit. In: Lutz Röhrich & Erika
Lindig (eds.), Volksdichtung zwischen Mündlichkeit und Schriftlichkeit, 187-208. Tübingen: Narr.
Permjakov, Grigorij L.
1979 Die Grammatik der sprichwörtlichen Weisheit. In: Peter Grzybek (ed.) Semiotische Studien zum Sprichwort - Simple Forms Reconsidered I, 295-344. Tübingen: Narr.
Prek, Stanko
1972 Ljudska modrost. Pregovori, domislice in reki. Maribor: Obzorja. [Ljubljana, 21974, 31982, 41986, 51996].
Radics, Peter
1882 Zbirka slovenskih pregovorov iz leta 1592. Kres 2: 332-334. Rode, Matej
1989 Slovenski pregovori in srednji vek. In: Jože Toporišič (ed.), Obdobje srednjega veka v slovenskem jeziku, književnosti in kulturi, 155-158. Ljubljana: Filozofska fakulteta.
Seitel, Peter
1969 Proverbs: A Social Use of Metaphor. Genre 2: 143-161. Stanonik, Marija
2000 Slovene Proverbs through the Time [sic]. Acta ethnographica Hungarica, 45(3/4): 379-391. Toporišič, Jože
1984 Inhaltliche Aspekte der komparativen Phraseologeme in slowenischen Sprichwortsammlungen. In: Harald Burger & Robert Zett (eds.), Aktuelle Probleme der Phraseologie, 291-321. Bern: Peter Lang.
OSNOVE SLOVENSKE PAREMIOLOGIJE
Prispevek obravnava osnovne probleme dokumentacije in raziskovanja pregovorov na splošno in še posebej slovenskih. Posebna pozornost je namenjena empiričnim metodam v paremiologiji. Pri tem velja najprej razlikovati med obravnavami, ki se nanašajo na besedilo in preiskavo frekvenc na eni, ter raziskovanjem informatorjevegapoznavanja in znanja o pregovorih na drugi strani.
Pri raziskovanju podatkov, ki se nanašajo na informatorje, je treba ločiti med (a1) posameznikovim ali (a2) kolektivnim poznavanjem pregovora in (b1) poznanostjo posamičnega pregovora oz. (b2) množice pregovorov.
Avtor predstavlja prve vzorčne in ilustrativne rezultate empirične raziskave znanja in poznavanja pregovorov na Slovenskem: 103 informatorjem s slovenščino kot materinščino je bilo v dopolnitev predloženih 100 pregovorov iz zbirke Marije Makarovič. Naloga spraševanca je bila začetek pregovora, npr. Kdor visoko leta., dopolniti z ustrezno izjavo.
Razprava pokaže, kako tesno sta povezani paremiografija in paremiologija in kako pomembna je medigra med teoretičnim in empiričnim prijemom. Definicija pregovora ni le nujni predpogoj za razločevanje sorodnih oblik, npr. pregovornih izrazov, idiomov, frazeoloških fuzij ipd., je neogibna tudi za klasifikacijo. Empirično delo je pomembno pri odločitvah o tem, kaj in katere variante vključiti v zbirko pregovorov; pogosteje kakor ne so sestavljalci zbirk pregovorov preprosto kopirali iz prejšnjih objavljenih zbirk, le redki so pregovore tudi dokumentirali, še manj pa so preučili njihovo poznanost v danem okolju. Tako je malo zanesljivih podatkov o splošnih jezikovnih oblikah, ki so pomembne za katero koli nadaljnjo analizo. Nekatere študije jezikovnih oblik, npr. dolžine besed [Grzybek 2004] ali stavkov Grzybek [1999], so lahko zato tudi zavajajoče, saj slonijo na tradicionalnih zbirkah pregovorov. Raziskovalci jezika in pregovorov bodo zato morali svoje analize opreti ne le na dobro empirično zbran izbor pregovorov, temveč tudi na empirično veljavne variante in variacije teh pregovorov.
Prof. dr. Peter Grzybek, Universität Graz, Institut für Slawistik, Merangasse 70, A-8010 Graz, Austria