SOURCE: Toporišič, Jože. “Kopitar's Grammar.” Papers in Slavic Philology 2 (1982): 76-97.

[In the following essay, Toporišič examines the morphological features of Kopitar's Grammar.]

In the scholarly literature Jernej Kopitar's Grammar¹ has been characterized as scientific, and that by right. Published in 1809 (with the year 1808, when the printing began, imprinted in it), it had two types of forerunners: that of Adam Bohorič (Arcticae horulae succisivae, de Latinocarniolana literatura, ad Latinae linguae analogiam accomodata … [Wittemberg, 1584, reprinted 1715, 1758]), and that of Marko Pohlin (Kraynska grammatika, das ist: Die kraynerische Grammatik oder Kunst die kraynerische Sprache regelrichtig zu reden, und zu schreiben … Zweyte verbesserte Auflage [Ljubljana, 1783; first edition, 1768]), to which Jurij Zelenko's (= Mihael Zagajšek's) from 1791 can be added, and even Ožbalt Gutsman's from 1777 (reprinted five times). Kopitar must have also known the manuscript grammars from Baron Zois' circle (by Kumerdej, Debevec, Japelj). Moreover, he himself mentioned dictionaries which had originated within this circle (Kumerdej, Debevec, Vodnik), from which he took lists of examples for morphological categories for his own Grammar; in it he also mentioned the 1791 grammar. About Gutsman's grammar he commented that Gutsman says little, yet nothing wrong,² while he treated Pohlin's very critically, especially as regards its graphic and orthographic systems. In many respects, however, he took it into account, as well as Gutsman's (the forms meno, tebo, sebo are probably taken from Gutsman). He knew Bohorič's grammar as well, and took certain ideas from it for his own morphology (e.g., supine). It is understandable that he considered it very interesting with regard to different writing systems.

The scholarly character of Kopitar's Grammar was achieved to a large extent by its limiting itself to the expressive and grammatical levels of language without going much into lexicology and syntax (on these levels he treated Slovene material as differing from German, but only occasionally). As far as formation is concerned he left out word formation, thus dealing only with form-formation and form-changing.³ But even in these thematically chosen fields he again overlooked general theoretical, e.g., non-Slovenistic questions,⁴ and treated only those which were of special interest to a Slovenist (only in this sense, namely, by differentiation or confirming, he took into account other Slavic languages, which he called dialects, and the non-Slavic ones—Greek, Latin, German, Italian, even Hebrew). In his Grammar Kopitar is scholarly also as a historian of the graphic and orthographic systems of the Slovene language,⁵ as a historian of Slovene literature (better to say the language of this literature) from the beginning to his time.⁶ Besides he appears also as a historian of that part of linguistics⁷ which discussed the expressive side of the Slovene language (especially graphic and morphophonemic).

The historic part of Kopitar's Grammar is beyond the frame of this paper. Those parts of the Grammar which treat the expressive form of the language⁸ will likewise not be dealt with, so that the paper will be centered on Kopitar as a scholar researching the morphology of the Slovene language. This was treated in the second part of his Grammar (213-366) entitled “Etymology” (entitled in the same manner in Bohorič's grammar); it concentrates on inflected word classes: the article (214-15), the noun (215-54), the adjective (254-71), the numeral (271-80), the pronoun (280-300), and the verb (300-66). The uninflected word classes were prepared by V. Vodnik,⁹ as a “register of uninflected word classes” (e.g., the adverb [366-74], the preposition [374-80], the conjunction [380-83], and the interjection [383-84]).¹⁰

Let us go ad rem.

In the Introduction, Kopitar wrote that “etymology … explores the laws of lexicology and inflection of words as such with the aim of determining on the basis of these laws what in the mouth of a speaker is not clear enough.” Which means that from the structural point of view it is possible to bring to consciousness certain facts, rearrange them, and prescribe norms.¹¹

In connection with the article¹² Kopitar says, agreeing with grammatical tradition, that it is not known in the Slovene language in the form which is found in the Germanizing texts of the Slovene literary tradition. It may appear however with adjectival noun phrases involving determination; cf. Katero kravo si drajši prodal, to pisano ali to černo? (“Which cow have you sold more dearly, the spotted or the black one?”). Here the declinable form of the article appears as opposed to the indeclinable one used today (ta pisano, ta črno).

What nouns¹³ are is stated indirectly by treating grammatical categories of gender, number, case, and declension separately, at the beginning of the corresponding chapter.¹⁴ Three genders are indicated by the nominative endings (Endigung): -a as a rule f., -o/-e, n., and C, m., which is grasped much better than in Gutsman. If the rule is to be efficient, all digressions have to be described relatively exactly. These digressions are as follows: for m. nouns endings -a, -o, and -e,¹⁵ for f. nouns ending -ø (here Kopitar enumerated examples only in part, but he could also have composed the list of corresponding suffix formants¹⁶).

Treating numbers, he was already aware of the usage of pl. instead of dual: “If arms, legs, ears, and other paired parts of the body are spoken about [he did not have in evidence yet other paired things, such as shoes, horses, etc.], perhaps because the number is understood by itself”; however, if necessary the dual may be used (e.g., z obema nogama ‘with both feet’).¹⁷ Kopitar took over Gutsman's arrangement for cases only in naming them (N G D A L I); later on he used N and V to denote the first case, which opens the question whether V in the Slovene language exists if words with their own tonemicity for this case are found.¹⁸ Declensions according to the G cases (very modern! and at the same time just like the Latins—and Gutsman and Pohlin¹⁹) are divided in three groups (-a, -e, -i); for greater clarity there is a further division of the first one in two groups on the basis of gender. This is pure grammatical morphemics. In the table of the endings of these four declensions²⁰ only the basic variants of end morphemes are taken into account; others such as -éma alongside -ima are not given. Such variants are registered only if they are very obvious. It is interesting that the zero ending, which means the ending without phonic representation, is graphically presented by a dash and not, as it might have been expected, by an empty space. In the m. declension Kopitar already indicated the category of animateness and even called it thus after Gutsman. When setting up declension patterns within this frame Kopitar was somehow lost, as he listed the following 16, all of them declined in all the 18 cases: (1) ràk, kràj, (2) posel, boben, petik (rabelna),²¹ (3) gospodar, pastir, mehur, (4) bratje,²²mož, bog, tat, (5) starešina, Noe, (6) očetom/očetam. For type (2) he himself doubted whether it represented a digression from the general pattern at all. It is true that posel and boben²³ are predictable morphonological variants of the main type, while petik could be explained by a morphonemic analysis: -k- r -ek-o. Kopitar did not realize that there was a difference between posel and boben on the one hand (the consonantal cluster NS which truly facilitates the “pronunciation”) and petik on the other (the consonantal cluster NN). When writing the unstable shwa he hesitated between i and e, although the following system could be discovered: as a rule it should be written by e, by i in the suffix -k- (but even here once with e as a variant: vosek/vosik), by i also in some cases with the suffix -c-, e.g., kosic, pevic, as well as in the cluster -nj- (oginj). He also already noted the permanent shwa (jazbic), “being less hard,” even in the type dedic (nowadays we would say “with one C preceding -c-). He would have been more successful in this respect had he all along considered information concerning morphology carried by the G case as a starting point.

The problems concerning the suprasegmental features were solved via facti, when alongside the less appropriate paradigmatic words (rak ráka)²⁴ he then listed also examples of the stem movable accent (mêdved -éda) or the end accent (bezeg -zgà). From his lists of examples the following morphonemic alternations of stems can be seen: berlòg -óga (hlèb hléba), bòj bôja, (dìm díma, gàd gáda, krùh krúha, and even skòk skôka), kôžuh -úha (mêdved -éda). The situation is, excepting a few examples, quite modern, and so to say without exception correct.²⁵

When dealing with endings Kopitar was partly less clear, to a large extent because of the nature of the things itself: in D sg. he quoted -u only, in L sg. -u and -i²⁶ (although inconsistently). In agreement with the Carniolan tradition he stated the endings -am and -ama for all case and number values.²⁷ He was critical about I pl. variants -ami/-mi/-i. He first did not set up a list for the subtype možjé, but he noted the pl. G ending -Q (for this reason he may have left out ljudje with the G ending -í (ljudi).

The accent is considered somehow implicitly.²⁸

Taken as a whole this is an extremely neat arrangement and a very exhaustive description of the first declension of the written Slovene of his time. Imperfections concerning nouns with G endings -á, -ú, or -ova are understandable, and will never be completely solved by Slovene linguistics unless the language itself settles them.

For the neuter gender declension Kopitar set the example with the -e ending in N sg. (as Pohlin had done before, but neither Bohorič nor Gutsman); the stem ends in j, nj, lj, and š, c, but Kopitar did not give an example with -e following č (there are probably no cases for ž in the Slovene language, if ože is declined ožeta, etc.). But Kopitar did not state the conditions which necessitate -e; the reason for placing this ending first was most probably the masculinization of neuters.²⁹ He registered examples for (as named in present terminology) fixed accent type (líce -a), mixed (srcé -á), and end accent type (kamnjè -à), given even for the -o ending: délo, blagó, dnò.³⁰ Kopitar formalized lengthening of stems: what is alive has -et-interpolated before the ending (Gutsman: “which refer to a young animal”), while those ending in -me get -en-.³¹ Here he registered also the stem movable accent (têle -éta), and the fixed one either on the first or on the last syllable of the stem (séme -na and imé -na). But he did not register the stem moving type among them (plême -éna). He recognized the -es- lengthening and listed examples for it. The endings are the same as for masc. excepting the N., while the L. pl. is -ih vs. L. dual -ah/-ih, the I. pl. -mi/-i for delo and -i only for srce. The N./A. sg. ending -o is supported by Spangenberg's Postilla (1578) and by Vodnik, and also by “consensus populorum”; the unstable vowel in the G. pl. is marked by e or i (i in naselij);³² implicitly the alternation ô—ó is given (ôkno—G. pl. óken).³³

In the 1st fem. declension Kopitar sets up two patterns: ríba -e and vôda -é. Different patterns according to the accent are Kopitar's contribution to Slovene morphophonology, for Pohlin stated them only if they differed phonemically as well (e.g., ljudém). The type vôda -é offers double non-sg. G., cf. vód and vodá.³⁴ The mixed accent type is not obligatory; there is also -eh in L. non-sg. (otherwise confirmed only in gospéh). N. dual vodé/vôdi.³⁵ The end accent type is indicated only by iglà (but is not developed all through its paradigm). Kopitar started the writing of unstable shwa as i instead of e when preceding j in G. pl. (ladij, zarij, Marí), but he wrongly believed that it was so in the N. sg. as well (ladija as Pohlin was supposed to have written by right).³⁶ The unstable vowel in the G. pl. (non-sg.) had up to this time not yet been recognized by the grammarians. The whole pattern of gospa is given correctly, the word beseda has the non-sg. G. -ø/-í.³⁷

In the 2nd fem. declension Kopitar explicitly stressed the influence of the accent type on endings, while examples for the fixed and mixed accent type had already been noted in Pohlin (with the example skrb), probably also Kopitar's subtype vas, which could well be illustrated by the D. against the G.: živáli živáli—vasí vasì—klopí klópi (but Kopitar did not come across pečí pêči). In G. pl. živál has the ending -ø, probably because of Kopitar's “love for systematics” (almost all fem. nouns have -ø in the G. pl.); both are found in Pohlin as well;³⁸ in D. pl. he mentioned the hesitation between -im and -am “because we write little and have not decided on one or the other yet.”³⁹ D./I. dual has -ima: -éma. It was Kopitar who first introduced the subtype with the unstable vowel (misel);⁴⁰ the second subtype, cf. cerkev, has a variant cérkva; I. sg. miseljo/mislijo and cerkovjo. Is his statement that the multisyllablic words prefer to combine with the endings -ama -am -ami rather than with -ima -im -imi acceptable? The third subtype, mati was introduced first by Gutsman,⁴¹ while Kopitar added hči. Had Kopitar been consistent in the classification according to the G. sg. ending, the subtypes bukov⁴² and mati should have been listed in the first feminine declension. Kopitar did not register (otherwise very rare) stem movable (senôžet -éti) and end accent types (vasì belongs to the mixed one). It is Kopitar's merit that compared with Pohlin correct lists of examples were added to particular types.

Like Pohlin⁴³ Kopitar lists pluralia tantum belonging to each declension as a rule.

Kopitar (as well as Pohlin)⁴⁴ emphasized the category of three forms concerning the gender of the Slovene adjective. For N./A. sg. n. he registered endings -o/-e together with the criteria for their distribution (-e follows j š č; earlier in Pohlin: adjectives ending in -še, -če, -je). The declension endings are of type -iga -imu⁴⁵ (-imu also in the L.), which remained into the 1850's. Generally speaking it was due to Kopitar that the indefinite form was taken as a primary form (in some instances he himself hesitated though).

Treating adjectives in -n- Kopitar was of the opinion that the pronunciation was the same regardless of the spelling: pravičen or pravični, but he gave priority to the indefinite form (Pohlin [1768 and 1783] had hesitated here).⁴⁶ It is Kopitar's merit also in setting the complete paradigm for the dual (for N. fem. he gives endings -i/-e, for the neut. gender -e/-a). The N./A. neut. pl. -e/-a.⁴⁷

Among accent types he mentions besides the ordinary fixed one the alternating one: bogàt -átega, the stem movable one (dêbel -éla -o) and mixed (lépo delo/lepó delo, delo je lepó—this is also in the predicative [vedró je]), and at least in the N. also the end one (gorìk, lepà—only for poetry). Kopitar stated three types for adjectives with -k- (gôrek gorìk gorák), two for -n- (môčen močán). The paradigms are only indicated and are not given in full. It seems to be particularly important that Kopitar discovered the so-called “connecting A” (i.e., vidiš lep kraj, jaz pa še lepšiga), which he named referential.⁴⁸ In determining adjective types which have only the definite form (the definitive adj.—božji, človeški, nebeški, sedanji),⁴⁹ there appears an idea about relational (vrstni) adjectives. The comparative form for fem. sg. ends in -i and is not declined; Kopitar thought that Dalmatin declined comparative forms “probably out of a grammarian's love for systems.” Contrary to Pohlin,⁵⁰ Kopitar did not explicitly mention the so-called zero adjective declension, but he included such examples in his list (peš indecl). He registered the masculinization of neuters “around Ljubljana”: grôzn' blat', dobr' jutr'.⁵¹

Dealing with the comparison of adjectives,⁵² Kopitar decided upon the free distribution between -ši and -ji (Pohlin had been uncertain about this, but he gave priority to -ši), and he also tried to determine those instances when -ejši/-iši was used. About alternations preceding -ši he took over from Pohlin d s j, but he noticed more complicated instances (grd/grši//grji) and added drag drajši (which is dialectal). Kopitar's merit is nevertheless that Pohlin's list of exceptions was made shorter;⁵³ he also mentioned the periphrastic comparison but not the elative which is found in Pohlin.⁵⁴ When dealing with comparatives Kopitar is somehow uncertain, which is rather understandable because of the complicated nature of things. Yet he was not aware of the system of jodization; he primarily enumerated examples instead. When he encountered the same problem in front of the suffix -en- (e.g., nositi nošen, roditi rojen), he solved it very elegantly with three groups: t—č, n—nj, and p—plj.⁵⁵

Kopitar grouped the numerals as follows: cardinal, ordinal, disjunctive,⁵⁶ distributive (po trije, “in threes”) and adverbial ones (ending in -ič). He practically dealt with them correctly; for type 5 he listed besides the variants petih/petéh, as well as a zero declension (s pet konji).⁵⁷ The difference between Pohlin/Gutsman's ena/eden, dva, tri … and Kopitar's eden/en ena eno, dva dvé dva, trije tri tri and štirje štiri speaks for the dualistic nature of cardinal numerals: Pohlin and Gutsman put down the forms for counting while Kopitar noted the adjectival ones. This is still not quite clear in the Slovar slovenskega knjižnega jezika.⁵⁸ Following Pohlin, Kopitar connected the declension of oba with dva (without including obadva), but he himself separated the declension of tri from štiri. Finally, Kopitar suggested the phonemic ending declension of the numerals of type pet; he was the first to have mentioned ones for tens from 20 onwards (enaindvajset …), again very exactly and reliably. The idea that in pet golobov the numeral pet is a noun was taken over from Gutsman.⁵⁹ The pl. of eden⁶⁰ should also be mentioned, which indicates the indefinite pronominal adjective is made out of a numeral. A pure noun out of the numeral is given in the example Ta mož ima tavžente.⁶¹

The classification of pronouns was taken over from tradition: (1) personal, (2) possessive, (3) demonstrative, (4) relative (kateri), (5) interrogative (kdo, kaj) with the relative ones (kdor, kar), adverbial (kadaj, kadar—tedaj) and the motivated ones: mnogiteri, vsakteri, (vsaki), nekteri, maloktiri, marsiktiri, nekdo, nekakšin, nihče, nič, kdor koli (kdor si bodi …), kèj.⁶² It seems that Kopitar did not set up a more uniform principle: he might have treated the nominal pronouns together followed by the adjectival ones, or classed the interrogative and de-interrogative ones for all orders, for example: kdo—nekdo, kakšen—nekakšen, kateri—neki, čigav—nekoga, etc. Yet he had listed examples for nearly all the eleven classes⁶³ which I have registered (he did not mention the pronouns of otherness of the type drug and those denoting sameness, while for the adjectival order he did not state the quantitative ones which are already found in Bohorič (inter. koliko, etc.).

Following Gutsman, Kopitar ranked among the personal pronouns also those for the 3rd person of all genders and numbers,⁶⁴ and he added tisti⁶⁵ to the demonstratives and leuni to Pohlin's le-ta, while he himself introduced the type tale, unile.⁶⁶ Kopitar listed only kateri and ki among the relative pronouns; he made the relative kdor and kar partly independent⁶⁷ and registered the relation among the adverbial interrogative-relative-demonstrative pronouns. He also listed personal possessive pronouns and at the end the derivatives from kdo and kateri (partly from kdor and kar). But he overlooked the relative optional pronoun kateri koli (yet he had kdor koli), and he half overlooked derivations with kakšen, which means he was not conscious of this pronominal order, just as, insofar as classes are concerned, he was not aware of the negative one or of those for totality. He simply tried to improve the Pohlin system.⁶⁸ Apparently this was a very difficult task, which has only recently been solved by exact differentiating of pronouns into nominal, adjectival, and adverbial ones having within them ten classes of motivated pronouns for all orders that have corresponding question words, e.g., kdo, kaj—kakšen, kateri, čigav, koliko—kam, kje, etc.⁶⁹

The declensions of pronouns are very thorough. Personal pronouns⁷⁰ should be mentioned separately, as in the G D A he strictly differentiated accented from unaccented ones: alongside the common accented forms in the A he also listed those connected to prepositions (mêne mé/-me me), yet not for 1st and 2nd pers. non-sing.⁷¹ He also gave the criteria for the usage of these forms. In this connection he disagreed with the Germanizing written tradition. The variants for the type menó in I. sing. were probably taken over from Gutsman (as already stated), while N dual forms were based on Pohlin. For G dual he looked back through history.⁷² He established parallels mene: me = jaz: φ.⁷³ He also dealt with the “metaphoric” usage of numbers and persons (addressing with “vi,” “oni,” partial “vi”).⁷⁴

There are other examples of how Kopitar contributed to a clearer approach to the rest of the pronouns, e.g., treating the relative variants kateri and ki (the former for solemn occasions, the latter in everyday life).⁷⁵ He had difficulties with kdo and kaj⁷⁶ which have not yet been overcome.

The last class Kopitar dealt with was the verb,⁷⁷ the word class which according to him (a) inflects in the noncomposed indicative, imperative, infinitive, active and passive participle, (b) forms an adverb, adjective, gerund, (c)—(g) is otherwise characterized by aspect, and (h) has the following forms: indicative—present, past, future tenses; conjunctive—past and past perfect tenses, imperative, infinitive, supine, participle (ending) in -oč, -l, and -n/-t.⁷⁸ Finally, besides the conjugation of the auxiliary sem he distinguished seven conjugations: I -am, II -ám, III -em, IV -èm, V -ém, VI -im, VII -ím.

Kopitar's conjugation endings are the same as today, except that in type -ím he thought it a special case when non-sing. endings were accented (thus delimò, etc.), while his gerund/participle endings in -č seem somehow not organic; cf. kličejoč, kažejoč, mečejoč⁷⁹ (besides the regular ones: dvigajoč, loveč, li(j)oč). It is Kopitar's merit that he finally discovered verbal aspect and consequently treated aspectual pairs apart from the inflected verb forms.⁸⁰ Verbs with perfective aspect were semantically determined by “do once and bring to an end” n “be engaged in an action without the connotation of finishing it.” He realized that aspectual pairs were made by the changing of suffixes (nesa, jesa, esi, VesVva) with an eventual alternation in the root (êrsír, ésá)—this is for the formation of imperfectives, while perfectives were formed by prefixes. There exist also cases with suppletion. He also introduced the diagnostic surroundings for aspectual pairs: Kaj delaš? ‘What do you do?’ for imperfectives, Kaj staviš, da? ‘What do you bet?’ for perfectives. For the latter he defended the present temporal state of the forms in -m as primary and rejected the future one which was found in North- and East-Slavic grammars (and languages).

For particular conjugations⁸¹ the following is to be mentioned. The auxiliary is treated just as it is today, except that he spelled it sim instead of sem; in the future tense he put bódem in the first place, bóm in the second, unlike Pohlin.⁸² From Pohlin he took over the imperative for 3rd pers. sing., but he himself added forms for 1st pers. sing. (Bodi len ali priden, nič mi ne pomaga).⁸³ He also accepted the imperative for 3rd pers. dual, but not Gutsman's for 3rd pers. pl.; he considered the forms with bi to be definite⁸⁴ and regretted that the gerunds were in decline (those ending in -v already dead, those in -č being rejected). It is characteristic that Kopitar did not quote the passive forms, tenses, moods, etc., except the participle ending in -n/-t.⁸⁵ He did not express his opinion on the passive at all, although its usage had been under a rather strong Germanizing influence.⁸⁶ However, he noticed the difference between the participles in -l and -n/-t: the former can only be used predicatively⁸⁷ and not attributively. He determined supine vs. infinitive. After Bohorič he was the first to register it.

The main part of Kopitar's treatment of verbs (besides the aspect) is the extensive morphophonemic treatment according to the present tense orders. The present tense form of the verb is, to a large extent, really indicative for the formation of other verbal forms. Nevertheless, the infinitive base often has to be referred to (just as in cases when the starting point is the infinitive stem).

The verbs in -am do not present a problem as their infinitive also ends in -ati. But complications appear when dealing with verbs in -ém, because on the one hand the athematic verbs had to be separated from the rest (Kopitar grasped them with the ending -em, whereas the athematic verbs belonging to the verbs in -am are characterized by tonemicity [dám—konč[UNK]m], but Kopitar did not know the tonemicity), while they also differed among themselves on the other hand: some had the theme φ in the infinitive, the others had a vowel theme, e.g., -a- as a rule (jém/jésti, grizem/gristi—berem/brati). Kopitar was partly inclined to think that in the infinitive all the verbs should be thematic (hence his griziti, although he knew this was not in use).⁸⁸ The verbs in -em preceded by j presented even more difficulties (Kopitar took into account the sound preceding jem, namely whether it was a V or a C).⁸⁹ The boundary between thematic and athematic infinitives was not always clearly defined (nor is it today): klati, orati—koljem, orjem.

For the infinitives of verbs in -im Kopitar based his classification mainly on the semantic criterion: those in -iti should denote action, while those in -eti/-ati, non-action (he knew of the exceptions as well).⁹⁰ I think that when classifying verbs in orders according to the infinitive there are not fewer but more difficulties regarding the predictability of the present stem, and that the classification according to the present stem is in a way a better one. The Slovenes should have retained it all along and not more or less discarded it from Metelko onwards.⁹¹

The awkwardness of Kopitar's classification—he himself admitted it in certain places—lies in the fact that it took into account too many of the supresegmental features with the result that the same things were dealt with several times (cf. verbs in -im and -ím). Yet certain morphophonemic subtypes were not predictable even from the above-mentioned seven orders (especially, for example, those with the so-called stem movable accent; cf. jókal -ála). Therefore, he could have set up fewer verb orders and dealt with classes in one place only; for example, (1) -am -al -ala, (2) -ám -àl -ála, (3) -am -al -ála, etc.

Where necessary he attributed classes to particular orders (e.g., verbs in -im had classes in -iti, -eti, -ati; verbs in -em were arranged according to the thematic or athematic character of the infinitive; and further: either on the basis of the consonant preceding -φ- or the thematic vowels themselves, mainly a and e). He correctly started the classification of verb orders of the verbs in -em on the basis of the characteristic consonants with the examples in z/s, but “for clearness' sake” he introduced the alphabetical order afterwards, from which neither clearness nor theory benefited. But it has to be admitted that within the given frames he discovered and registered practically everything, and he did it in an extremely reliable way. Let us mention the most important things.

For verbs in -am he registered the non-paradigmatic glej and imej, for dam the endings with s preceding t, etc.,⁹² while for verbs in -im he did not manage to grasp the exceptional boj se and stoj. Morever, he marked an i following the j; for verbs in -em practically everything is mentioned;⁹³ among the rest there also is the variant najdel/našel.⁹⁴ He also quoted the variants with thematic vowels, such as svetujem/svetvam.⁹⁵ At the same time he pursued also the variants based on prosodic features (as they might called nowadays). This can best be illustrated by the fact that for verbs in -im he differentiated three types: (1) mérim -iti -il, (2) hválim -iti/-íti -il, (3) gónim gôniti/goníti gônil, although he did not present all belonging to the same order in the same place but in different ones. There are, however, several mistakes, partly owing to printing errors (see séliti instead of sêliti/selíti).⁹⁶ There is a general deficiency in the fact that Kopitar did not quote the f. form of the participle in -l as well. Thus, he should have registered the non-sing. imperative as well, avoiding the wrong generalizations. It is interesting that for the verbs in -ím he quoted also the shift of the accent characteristic of the participle in -l, which is caused by the prefix (kosíl—pokósil);⁹⁷ and in connection with this he raised the question about certain analogous forms. For some variant endings he did not see the principle for distribution (e.g., when -eti and when -áti)⁹⁸ because he listed several examples with a following non-hush sybilant.⁹⁹

Considering that Kopitar was the first to treat Slovene morphophonemics, it makes us wonder how few mistakes he made and how complete his description is; he made mistakes only on occasion, when the shortness of vowels or the place of the accent is involved. To the contrary, he is almost always exact and correct.¹⁰⁰

The importance of Kopitar's Grammar lies not only in its scholarly character then, but is due to its practical applicability as well, as was confirmed by Prešeren (who otherwise rejected Kopitar's [i.e., Metelko's] attempted reform of the graphic and orthographic system). Kopitar's Grammar had only to be completed in those fields which had remained undescribed. This was done by Franc Metelko (Lehrgebäude der slowenischen Sprache im Königreiche Illyrien und in den benachbarten Provinzen. Nach dem Lehrgebäude der böhm. Sprache des Hrn Abbé Dobrowsky [Ljubljana, 1825]) in less than twenty years. Metelko presented word formation precisely, and to a certain extent he outlined even the syntax of the Slovene language. But he was unlucky in trying to introduce the reductional shwa, which caused the so-called alphabet war. With the new orthography being rejected, most of the advantages of Kopitar's Grammar were also rejected for almost a hundred years. It is true that as early as the 1870's, S. Škrabec had already begun reaching back to Kopitar and Metelko, but he had to do it mostly in his scholarly writings (the same applies to Pleteršnik's dictionary, where Kopitar's [and Metelko's] findings about the expressive side of the Slovene language were to a large extent incorporated). Present-day Slovene grammar likes to trace its origins to what Kopitar constituted in such an exemplary manner, and what Metelko and Škrabec continued. It is not based on its immediate forerunner, which was torn between the antagonisms of diachrony and synchrony and the wrong interpretation of the true natural basis of the literary language (the so-called all-Slovene scope); it also did not care much for pure scholarly methodology either (which, in my opinion, is above all of a formal nature where language is concerned). Kopitar himself had already anticipated this and discovered it to a large extent. It is high time for the Slovenes to revise certain negative attitudes toward Kopitar. When this has been done, we shall all be proud of our genius Dioskuri, Kopitar and Prešeren, the two representatives of the Slovene spiritual achievement, poetic and scholarly, in the early nineteenth century.

Notes

1. Grammatik der Slavischen Sprache in Krain, Kärnten und Steyermark, (Laibach: Wilhelm Korn, 1808), xlviii ＋ 462 pp. ＋ 19 (and 18) columns of texts (and noun irregularities) in the appendix to p. 157 (and p. 233); on p. 460: Wien, den 1. Hornung 1809, Kopitar.

2. Ibid., p. xliv: “der zwar wenig, aber doch nichts falsches sagt.”

3. Ibid., 213-366.

4. Ibid., xlvii-xlviii: “Der Verfasser denkt sich vorerst nur die Volkslehrer und irgend einen Slavischen Sprachforscher als Leser einer Krainischen Grammatik, folglich konnte er die grammatischen Vorbegriffe in Definitionen Kürze halber übergehen.”

5. Ibid., 1-160.

6. It is mostly a bibliographical registration of Slovene (printed) works and possible critical remarks in connection with Germanization (e.g., his quick statement of the “etymologische Grübeleien” in the revised new translation of the Bible, p. 156: “Dergleichen Flecken in dem revidierten Neuen Testamente, worin übrigens plurima nitent, sind z. B. kareg statt kręg (Zank), weil es von pokóra, von karam (ich strafe) herkomme.”)

7. Ibid., xxix-xlviii.

8. Ibid., 161-212.

9. Ibid., 388-89: “Prof Vodnik hatte die Güte, bey meiner Abreise von Laybach, die Besorgung der weitern Correctur (von Verbis angefangen), die Ausarbeitung der Verzeichnisse der indeclinablen Redetheile, und die Anzeige der Druck—(allenfals auch meiner eigenen Schreibe—) Fehler zu übernehmen.”

10. With small changes (e.g., that both kinds of nomina, i.e., nouns and adjectives, are treated together) such a treatment is found as early as 1584, except that Bohorič treated the participle as a separate class of words. Nouns and adjectives are treated as separate word classes (and not in the frame of the so-called nomina) for the first time by Gutsman. It must be added that Gutsman treats adverb, conjunction, and interjection in the syntax only, and not in the morphology, as Kopitar does.

11. In this there is an echo of rectifying the grammarians of the eighteenth century; Kopitar did it the same way, regardless of his severe remarks on behalf of such endeavors; cf. Grammar, p. xlvii: “selbst Kumerdey lässt sich manchmal von Systemliebe irre leiten: Grammatik aber ist analogischer historischer Bericht über eine Sprache; Facta entscheiden hier, nicht Räsonnements.”

12. Ibid., 214-15; about the articles and definiteness/indefiniteness in Slovene, see J. Toporišič, “Imenska določnost v slovenskem knjižnem jeziku,” Slavistična revija, 26 (1978), 287-304.

13. Ibid., 216-54.

14. Gutsman (1777) writes about gender after the verb only, pp. 70-71: “Die Wörter, welche eigne Namen, Würden, Aemter und Verrichtungen der Männer, oder sonst was nur den Männern zuständiges bedeuten … die Namen der monate, Winde, wie auch der Tiere und Vögel … sind des mannlichen Geschlects.” Following is a longer list of endings. “Des weiblichen Geschlechts sind die Wörter, welche eigne Namen der Weiber, der weiblichen Würden, Aemter und Verrichtungen, oder sonst etwas nur dem weiblichen Geschlechte zuständiges bedeuten.” Feminine is also what has the ending -a, and is not masculine, what ends in -ast, -est, -ust. Of the neuter gender is what ends in -e, -u. Exceptions are cited.

15. Here he based his rule on semantics: masculine names. Cf. Gutsman (1777, p. 70): “Von den Geschlechtern, die aus der Bedeutung des Wörter erkennet werden.”

16. He has a formal aspect when saying that to this group belong all polysyllabic words in -ast, -est, -ust (this had already been described by Gutsman), -azen, -ezen; among the enumerated examples are the following suffixes: -el, -tev, -t, -sen, -al, -l: kopel, obutuv, past, pesen, pišal, ral.

17. This cannot be found in Gutsman.

18. On the problem of the vocative as a special case in Slovene, indicated by the tonemicity, cf. J. Toporišič, “Pojmovanje tonemičnosti slovenskega jezika,” Slavistična revija, 15 (1967), p. 68 (now also available in my book Glasovna in naglasna podoba slovenskega jezika [Maribor: Založba Obzorja, 1978]). Pohlin (1783, pp. 31-32) thinks that Slovene has only five cases, but he preserved all six known from Latin, so he added to them the “associative ablative so as not to confuse people.” Pohlin knows that V and L are not necessary in Slovene, although there can be a difference in endings, e.g., tega sèrcâ against iz sèrca (we know that this is due to the preceding preposition). Pohlin even has as an exception when declining the word ljudji, the present-day six cases: ludji ludý ludéh ludém ludí ludmí; that means the whole and clean paradigm as it was then generalized by Gutsman in 1777.

19. Pohlin (1783), p. 32; Gutsman (1777), p. 16.

20. P. 220.

21. The lengthening with -n- had already appeared in Pohlin (1783, p. 32): Mihel Mihelna.

22. In the ending -je there might be a trace of the masculine personality (m[UNK]skoosobowość).

23. Pohlin (1783, p. 33), writes hlapc, kozl, pajk.

24. Škrabec later proposed korak because of its being without alternations when declined. It would be better to place the inanimate as first, but already in 1584 there is oče,(1768), p. 23, krayl, (1777), p. 8, gospud.

25. For example, the open o before v: pôuž; cf. also skòk skôka against potòp-ópa; while potrêsa, podmêta is odd.

26. Whether this reflects the true state of affairs remains unclear.

27. This fact is very inappropriately characterized by H. Orzechowska in her paper “Typologiczny aspekt postulatów slawizacyjnych B. Kopitara,” in Obdobja razsvetljenstva v slovenskem jeziku, književnosti in kulturi (Ljubljana, 1980).

28. In the lists of examples; as a criterion the accent in the G ending in -á, -ú (our mixed type) is taken. Generally the following (but unnamed) accent types are discerned: fixed on different syllables (Nóe Nóeta, Anžè Anžéta), those on the ending (bezeg bezgà—in the N possible bèzeg or bezèg), and stem movable ones (pôrok-óka).

29. Or it may be because of the fact that Pohlin had, as the second example, a noun in -e (znamìne); or because of the fact that Kopitar had still to choose between the ending forms -o and -u.

30. Pohlin (1783, p. 41), listed fixed and mixed types; with the example pleče maybe also the stem movable one.

31. Kopitar, p. 240; Gutsman, p. 18. Independently from Kopitar this formalization was done by the author of this paper; see Slovenski knjižni jezik, I, p. 180, with the remark that it includes the type Semè -éta. Pohlin (1783, p. 42) has persea and teleta but no example for the lengthening by n.

32. They are unknown to Pohlin and Gutsman.

33. P. 238.

34. Pohlin (1783) quotes, e.g., bervá, gospá, tetá, but not the whole type.

35. Pohlin (1783, p. 37) was of the opinion that the plural and the dual are “totally equal”; in this he was followed by Gutsman (pp. 12-13); it is Kopitar's merit for having posited the dual with the doublet ending -i in NA and with the ending -ama in the D and the I pl.

36. Pohlin (1783, p. 37), zarija, p. 38, ladija, hudobya hudobij. Kopitar does not have Pohlin's forms pešák, tresák (1783, p. 38), except ovác.

37. Pohlin (1783, p. 38) quotes only the G forms gospá, besedi.

38. Of course, if the fixed accent is not concerned where there is the ending -i (1783, p. 40).

39. Pohlin (1783, pp. 38-39) has only (perpovist) -am, -ah, -ami; Gutsman has the type skrb with the endings -im, -ih/-eh (-mi).

40. P. 251 with the remark that misl is not possible (as it was in Pohlin [1783], p. 39).

41. (1777, pp. 15-16); Pohlin (1783, p. 39); mate, misl—materjo, mislejo—matere misle, A sing. mater.

42. The type bukov is already in Pohlin (1783, p. 39), and Gutsman, p. 15: (1) cirkou, (2) cirkve, (4) cirkou, (6) cirkoujo/cirkvio; pl. (1) cirkve, (2) cirkvi.

43. Pohlin (1783): p. 36—otrobi …, p. 40—bukuvce …, p. 43—jetra. … Pohlin's remark that some nouns of the neuter gender are ordinarily used mostly in the sing. (sadje … “and those designating materials”) is interesting.

44. (1783), p. 44.

45. Pohlin and Gutsman had, of course, -ega, which for Pohlin meant [sga] and for Gutsman [ega].

46. Pohlin (1768, p. 36)—brumne, (1783, p. 45)—brumn; Kopitar: krogli, bridek/bridki against lev, praženj, pasij/pasji.

47. Pohlin (1783, p. 46): pl. -e, du. -a; Gutsman (1777, p. 22): -e.

48. Pp. 259-60: “so dass sich dieses darauf nur bezieht.”

49. Božji is also found in Gutsman.

50. P. 38—popolnema. Kopitar mentions as well res which is really a predicative; so is rad -a -o which as well can be found among Kopitar's examples for adjectives.

51. P. 262.

52. Pp. 267-71.

53. (1783, pp. 49-50): bulšè, dalšè, goršè, krajšè, lozejšè, manjšè, etc.

54. (1783, p. 49): formations with the prefix pre- or with drugače, zlo, močnu, samo, usega, cel, čez use: premodr, usegamogoč, drugači lep.

55. Pp. 359-60.

56. P. 278: eni -a -o, dvoji -a -e, peteri-a -o. Pohlin (1783, pp. 50-51) mixes them with the multiplicative ones, while Gutsman (1777) knows the cardinal and ordinal numerals only.

57. Pohlin (1768, p. 44): all cardinal numerals, except those for values from 1 to 4, are not declined (the same in Gutsman, p. 27). Pohlin and Gutsman have the plural form dvem for D/I du.

58. Cf. J. Toporišič, “Besednovrstna vprašanja slovenskega knjižnega jezika,” Jezik in slovstvo, 20 (1974/75), 33-39; “Esej o slovenskih besednih vrstah,” loc. cit., 295-305.

59. (1777), p. 88.

60. Ibid., p. 273.

61. P. 277.

62. Kopitar's originality can be seen in the fact that he, to a greater extent than Pohlin, discerned the basic forms from the motivated ones, which can be shown with kateri (variants ktéri/ktíri/tkíri) with the following derivatives: mnogiteri, vsaktiri (vsaki), nekteri, maloktiri, marsiktiri.

63. Cf. Jože Toporišič, “Slovenske zaimenske besede,” Jezik in slovstvo, 20 (1974/75), 117-19; Slovenska slovnica (1976). Kopitar has the following classes: (1) kdo, (2) kdor, (3) nekdo, (4) kàj, (5) mnogiteri, (6) kdorkoli, (7) vsaktir, (8) nihče, (11) ta. Pohlin (1783, p. 62): enkatir, tak, takršen, us, sledn, slehern, nikogra (G sing.).

64. Pohlin (and Bohorič, p. 81) mentions them among the demonstrative ones.

65. Pohlin mentions taisti only.

66. Pohlin (1783, p. 57): leta is known in Lower Carniola, tale in Upper Carniola.

67. Pohlin (1783, p. 61): “Instead of kaj sometimes kar is used.”

68. 1783, pp. 60, 62-63.

69. See Note 63.

70. He made, following Gutsman, a correction of Pohlin's impossible genitives, like: jest moj -a -e meni … Some disparateness appears in the D where nji (p. 281) and njé (p. 285) are found.

71. (1783) has only the following clitic forms: D mi, ti, A? sê; Gutsman, p. 29: D mi, ti, A me te jo (sing.) je (pl.), se, and for the n. jo.

72. Ibid., p. 281: “Bohorič has njiu … tiu … obeju.”

73. Ibid., p. 286; but Kopitar did not discover the parallel of me/te/ga with nas/vas/jih which can now be found in Toporišič, and earlier, at least partly, in Škrabec.

74. Ibid., 288-89; half vi-forms can be found in Pohlin (1768, p. 138) already; oni-forms instead of vi-forms are found in Gutsman (p. 92).

75. Ibid., p. 294.

76. Ibid., p. 295; Pohlin (1783, p. 60).

77. Ibid., 300-66.

78. In tabular and symbolized forms:

    Kopitar	A –m, –j, –ti, –t, –l, –n
    	B –joč, –joč –a –e, –nje
    Pohlin (1768)	A –m, –j, –ti/–t, –l, –n
    	B –joč, –joč –a –e, bi–l, je–l
    Gutsman (1777)	A –m, –m–l, –m–l–l, bom–l,–j
    	bi–l, bi–l–l, bom –l–l
    	B –joči–a –e, –l

79. Ibid., p. 309.

80. For example, on page 312:

    –am n	–nem	bi –al n bi –nil –ati n –niti
    		–ajoč n – – –
    sem –al	sem –nil
    bom –al	bom –nil

81. Kopitar thought that in Slovene we could have only one conjugation in -m, but in accordance with the tradition (from Bohorič on), relating to the foregoing vowel, he discerns three classes (-am, -em, -im) and divides them further on the basis of suprasegmental characteristics (-em, -èm, -ém, etc.). Pohlin had two orders in -em, one of them being -sm.

82. Ibid., p. 317; but the order in Pohlin and Bohorič (bom, bodem) was better. The forms sem n bóm are still found in the SSKJ but not correctly: it should be sèm, bóm n sem, bom.

83. Ibid., p. 319, especially the note on page 322. On page 321 Kopitar expresses his regrets because of the loss of the gerund forms in -ši.

84. Pp. 320, 323.

85. (1768) does not mention the participle in -n until treating the passive voice (p. 66): “The participle of the past tense of passive meaning” (1783, p. 76).

86. Bohorič (1584, p. 116): bom sekan secor; (p. 221): se sekam; (1768, p. 66): bom baran, (p. 48): (letu se bara).

87. Ibid., 331-32, which indicates its predicative nature. That the participle in -l has preterital meaning is mentioned in Pohlin (1768, p. 60) and Gutsman (1777, p. 40): “Das Mittelwort vergangener Zeit kann als ein Beywort nicht gebraucht werden.” None of them noticed the participium status in -l (usahel).

88. Page 329 griziti/grizti, p. 336 leziti/lezti, and the same way molziti, nesiti, pasiti, tresiti; however, p. 338, dolbsti; page 337: “Der contrahirte Infinitiv ist gebräuchlicher, als der uncontrahirte, so dass man sagen kann, die Verba in zem und sem contrahiren allzeit den Infinitiv.” Colloquially najdst—najdit, skubst can in fact be found. But this is younger.

89. Here too (p. 341), he wrongly supposed the thematic vowel, e.g., greti/grejti from grejiti.

90. Ibid., page 361.

91. Metelko (1825, p. 102): I A nesem nests, B. pijem pits; C. znam znats, II mahnem -nits, III živim -ets, IV živim -its, V maham -ats, VI omahujem -evats. In the same way, on page 102 for the I order, and on page 108 for the II, while for the III-VI orders the starting point is now the infinitive form.

92. Pp. 334, 335-36 with accents dáva/davà, dájo/dajò/dadó; in Lower Carniola dadé; about grem see pages 338-39.

93. For example, page 337, 3rd pers. pl.: nesó.

94. Ibid., p. 339.

95. Ibid., pp. 341-42, with all the problems of the quantity of the stressed vowels, for example igràl, kupovàl—kovál from igrám—kupújem, kújem.

96. Such a mistake is his remark to page 365 that imperative i when nonsingular is always accented, which would mean that it should either be replaced by s or be left out altogether. In hvalíti (besides hvaliti) the stress of the short infinitive (cf. gôniti from gônit) is noted.

97. Ibid., p. 362, zgúbil.

98. Ibid., p. 361, spati, p. 362, bati se, stati, scati, staying not after č ž š.

99. Kopitar has no examples with e after č ž š, for example, vršeti.

100. For example, with the verbs -ím -éti (p. 363): gôri -íte gôrel; sedì -íte hítel. But he does not quote forms which were not in usage, as imp. of bobnim. Cf. also the verbs in -im -ati, pp. 363-64: spà/mižì, tísi/lêži.