Žodžių daryba rankraštinėje Kazimiero Kristupo Daukšos lietuvių kalbos gramatikoje

ACTA LINGUISTICA LITHUANICA XLIX (2003), 63 – 102 Word Formation in Kazimieras Kristupas Daukša’s Manuscript Lithuanian Grammar ANTANAS SMETONA Vilniaus universitetas Kazimieras Kristupas Daukša’s manuscript grammar, dating from the first half of the 19th century, is the earliest Lithuanian grammar written in Lithuanian. This makes it an important source for investigations into the history of grammar and grammatical terminology, and a testimony to the theoretical achievements of the science of language in Daukša’s times. Among other pre-Schleicherian grammars of Lithuanian (and other languages) it stands out by its exhaustive and theoretically sound treatment of word formation, which is given more prominence than in earlier grammars. Compounding and derivation are already distinguished and dealt with in separate sections. Daukša’s treatment of word formation can therefore be viewed as an autonomous part of the grammar, though its place in the grammatical system is not yet clearly defined. 1. HISTORICAL CONTEXT 1.1. Grammars of Lithuania Minor 1.1.1. Already in the earliest Lithuanian grammars, word formation is discussed alongside inflection. In Daniel Klein’s grammar Grammatica Litvanica, the noun and verb sections contain subsections on word formation: De nominis specie et figura (On the Form and Figure of the Noun) (Klein 1653: 64–69) and De verbi specie et figura (On the Form and Figure of the Verb) (Klein 1653: 129–131). There is no systematic treatment of word formation; only individual cases of suffixal formation and a few prefixed forms are mentioned. It is explained there that nouns are primary or derived, and that they are formed from nouns and from other parts of speech. Four types of derived nouns are formed from nouns: possessive adjectives – tėvains (: tėvas), namiškis (: namai), miesčionis (: miestas), laukininkas (: laukas), karaliaučionis (: Karaliaučius); patronymics – Adomaitis (: Adomas), klebonaitis (: klebonas); diminutives – ponelis (: ponas), žmогutis (: žmogus), silpnokas (: silpnas); and qualitative derivatives – geležinis (: geležis), teisybė (: teisus), grožybė, gražumas (: gražus). Nouns are also formed from other parts of speech: these include derivatives in the ending -a and the suffix -imas formed from verbs: valdžia (: valdau)1, pabaiga (: pabaigiu); išgulydymas, pamokinimas, sakymas. The variants of the suffix -imas are explained in somewhat greater detail. Verbs also form adjectives – šventas (: švenčiu), sveikas 1 Following the example of Latin grammars, older grammars take the first-person present tense as the base form. The presentation of examples varies somewhat in this study, because, to preserve authenticity, base words are not supplied where they were absent from the original. 64 | ANTANAS SMETONA (: sveikstu)2. Word formation from adverbs is also mentioned – tolus (: toli), artimas (: arti). The structure of the section on verb formation is similar: verbs are primary or derived – klausinėju (: klausau)3. There are three types: inceptive (pradedu mirti – merdami, pradedu pulti – puol inėju), frequentative (užpuolinėju, darbuliuoju, sakydavau, valgydavau)4, and diminutive (valginėju, vaikštinėju, klausinėju, mirinėju, girauju, girsnoju). It is worth noting that Klein sometimes classifies prefixed verbs as primary (pasenstu), and sometimes as compounds (suvadinu). The section on verb formation also mentions four types of compound verbs: noun + verb (karavedu, juokdarau), adverb + verb (gerdarau), preposition + verb (prieimi, sugrįžtu), and particle + verb (beesmi). There are no further sections on word formation. However, word-formation matters are also discussed in sections on inflection or the formation of grammatical forms. It is noteworthy that functional word formation was observed already in the first grammar. The section on participles (Klein 1653: 133) states that some participles become nouns under certain conditions: when a verb is combined with a negative particle (called an adverb by Klein) (nemokytas, negirdėtas), when they lose their tense meaning (my lėtinas, plaktinas, mokytas), and when they can be compared and graded (mokytas – mokyčiausias). Functional word formation also includes the adverbialization of the instrumental and locative cases (pulkais, vardu, atviromis, tylomis, viduj). Klein also includes the allative and illative under word formation; he does not consider them cases, but states that adverbs are formed with such endings (manimp, namuosna, jūsump5). In Klein’s Kompendium, diminutive formation is mentioned, with a note that more about it can be found in the grammar. The first Lithuanian grammar presents only the most general matters. Yet this grammar was the first, and was intended for practical use and to show that even the “language of the barbarians” was not as poor as many had thought (see Klein 1653: 14–15). 1.1.2. In Sapūnas and Schultz’s grammar Compendium grammaticae lithvanicae, word formation is likewise discussed in the noun and verb sections. There are sections on form and figure – Figura & Species (translator K. Eigminas renders species as “kind”) (Schultz 1997: 18). This grammar mentions the same types of word formation – possessive (tėvainis, mieščionis, laukininkas), patronymic (Adomaitis, Jokūbaitis), and diminutive (ponelis, vaik elis, žmogutis, lietutis, dukteraitė) forms (Schultz 1997: 18–20). It discusses the formation of nouns from nouns (geležinis, kūniškas, teisybė, viežlibumas, patogumas), separately mentioning derivatives with the suffix -omas and derivatives from other parts of speech. It also mentions participles that have become nouns (mokytas, pagirtas, priprastas, 2 This derivational association between disyllabic adjectives and verbs is important even for contemporary grammars (see Paulauskienė 1994: 181–182). 3 In fact, from klausiu. 4 As the examples show, Klein did not distinguish the grammatical iterative in -davo (the past frequentative tense) from lexical iteratives. 5 Kad senoviniai linksniai (kurie istoriškai ir nelabai seni – žr. Zinkevičius 1966: 200–202) Kleino laikais jau nebepriklausė linksnių paradigmai, rodo ir tai, kad jų postpozicinius elementus jis laikė prieveiksmiu, o ne linksniu darybos priemonėmis. Nors, kaip toliau matysime, Kazimieras Kristupas Daukša, būdamas rytų aukštaičių atstovas, iliatyvą dar įtraukė į linksnių paradigmą, o elementus -p, -pi jau laikė tam tikrais postpoziciniais priešdėlių variantais. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 65 išganytas, turįs, regis, tikįs) (Schultz 1997: 22). Apie sudėtinius žodžius (t. y. sudėtinius ir priešdėlinius) tik pasakoma, kad jų yra didelė įvairovė, kurioje lengva susigaudyti iš praktikos (Schultz 1997: 23). Veiksmažodžių daryba taip pat nedaug tesiskiria nuo Kleino gramatikos. Tik čia jau nustatomos keturios išvestinių veiksmažodžių rūšys (Schultz 1997: 41): dažniniai (puolinėju, siuntinėju), mažybiniai (pataikauju, vaikštinėju, girsnoju, garbinoju), leidžiamieji (juokinu, puldinu) ir intranzityviniai (myliu̇os, rašydinuos, be myliu, besi myliu̇os). Sudurtiniai veiksmažodžiai aptarti kiek išsamiau (Schultz 1997: 42): daromi veiksmažodžius jungiant su vardažodžiais (karavėdu, juokdarau, dalytėju)6, su prieveiksmiais (gierdarau, gandarau), su prielinksniais (prieimi, ateimi) ir su dalelytėmis (be eimi, be esmi). Dalyvio skyriuje minimas dalyvio virtimas vardažodžiu (Schultz 1997: 63–64) taip pat, kaip ir Kleino gramatikoje. In the grammar, Sapūnas distinguished the past frequentative tense from word formation and was the first in the history of Lithuanian grammars to include it in the tense system as a second perfect. However, he formed this tense not from the infinitive but from the first perfect, and therefore established two suffixes for the past frequentative tense: -davo (dirbdavo) and -ydavo (sakydavo). Like Klein, he considered the infinitive a derived form (a separate mood), so one suffix, -davo, was not enough for him to form a derived form from the first perfect. Moreover, his approach also lacked consistency: after all, the infinitive has three vowels before -ti: -ė-, -y-, -o-, so, to be consistent, Sapūnas should have distinguished not two but four past frequentative forms: -davo, -ydavo, -ėdavo and -odavo. Nevertheless, separating the past frequentative tense from word formation is an important theoretical observation, a discovery that took hold in many Lithuanian grammars and has persisted in contemporary practical as well as scholarly grammars. 1.1.3. Word formation is discussed in much greater detail in Povilas Friedrichas Ruigys’s (Paul Friedrich Ruhig) Anfangsgründe einer Littauischen Grammatick. Klein’s grammar influenced later authors, even up to August Schleicher (Schleicher), but German grammars may also have led to greater attention being paid to these works. At first, individual instances of word formation, especially compounding, are mentioned without any system in descriptions of certain parts of speech. However, an entire section is devoted to derivatives (three paragraphs). It mentions nomina derivata – denominativa: kantrumas (: kantrybė), puodžius (: puodas), girtuoklis (: girtas), geležinis (: geležis); gentilia: examples strikingly similar to Ostermejer’s (see the next paragraph); patronimica: Jokubaitis, Karalūnas; diminutiva: brolužis, broliukas, brolytį, and so on. Here, somewhat unclearly, the following derivatives are added: silpnokas, graudenimas, prašimas, atpirktotis, pridėtojis, daugumas, tarpininkas, priešininkas, pirmonis; in verba derivata – laupsinu (: laupsė), garbinu (: garbė), pasinaudoju (: nauda), atollinti (: toli), skirti (: skiru). There are, of course, many inaccuracies in all these examples, as well as a failure to understand the actual derivational relationships. 1.1.4. Word formation is described perhaps most comprehensively in Gotfridas Ostermejeris’s (Gottfried Ostermeyer) Neue littauische Grammatik. As many as 12 paragraphs are devoted to it. Clear- 6 Although both grammars distinguish compound verbs, there are no such verbs today. LKŽIII 96 includes gandaryti, but the examples are taken from old church books and the writings of Antanas Baranauskas, while karą vedu was probably already two words even then. 66 | ANTANAS SMETONA tly distinguish two methods of word formation—derivation and compounding. Derivatives are already described fairly systematically. First, noun suffixes are listed (-astis, -estis, -ybė, -yčia, -ystė, -immas, -ummas, -inkas, -innė, -innis, -ynė, -yklė, -ikas, -ėjas, -ėjis, -ėję, -ėna, -ojis, -ojė, -onas, -onis, -onė, -aukštis, -iškis, -iškė, -alas, -alus, -unas, -uwas, -uwe, -ėna, -ojis, -oje, -onas, -onis, -one, -alkštis, -iškis, -iškė, -alas, -alus, -unas, -uwas, -uwe, -gys, -lys, -wys, -orus, -žys, -ba, -džia, -kla, -we), followed by examples illustrating the parts of speech from which nouns are formed: from nouns—knygininkas (: knyga); from adjectives—plotis (: platus); from pronouns—saviškė (: savęs)7; from verbs—valdžia (: valdau), vagis (: vagiu), šėriks (: šėriau)—derivatives from the present, past, and future tenses and the infinitive are given; from adverbs—daugumas (: daug); from prepositions—tarpas (: tarp). The adjective is described in much the same way: its suffixes are listed, and examples are given of adjective formation from nouns, adjectives, pronouns, verbs, participles, adverbs, and prepositions. Ostermejeris does not list verb suffixes, but again provides examples of formations from nouns, adjectives, numerals (suvienyt), pronouns, verbs, adverbs, prepositions, and interjections. The discussion then strays somewhat from the logic of the classification and distinguishes several more individual cases of word formation: diminutives of nouns and adjectives, patronymics (Adomaitis – Adomo sūnus), followed by ethnonyms (lietuvininks, vokietis, tilžėnas, karaliaučionis, isrutiškis, klaipėdiškis, darkiemujis). Terms for people based on their place of residence (or origin) are repeated exactly as in Ruigys’s grammar; some also reappear from Klein’s grammar. Another, shorter section concerns composition. Here, compounds are classified according to their second component, as in contemporary grammars, with the first component understood to define or describe the second: noun + noun, adjective + noun, pronoun + noun, adverb + noun (pussiaužiemis, pussiauramžis, neganda, negalė, nežinė), preposition + noun (posunis, isunis, patevis, bedugnis, patvoris, pavasaris, priangis), adverb + adjective (negvys, nelab’s)8. 1.2. Grammars of Lithuania Major 1.2.1. Grammars written and published in Lithuania Major devote much less attention to word formation. One of the more important grammars, Universitas Lingvarum Litvaniae (1737), says nothing at all about derivation (mentioning only the formation of forms—degrees, adjective genders, tenses, and so on). Apparently, this grammar was not influenced in any way by works from Lithuania Minor; see Sabaliauskas (1979: 21). Their authors, one must assume, were influenced more by trends in German grammars. The influence of German grammars on authors from Lithuania Minor can likewise explain this divergent approach to matters of word formation. 1.2.2. No further grammars were published in Lithuania Major in the eighteenth century, so the grammars discussed below are by Daukša’s contemporaries. Unfortunately, they devote little more space to word formation than their predecessors. In Juozas Čiulda’s grammar, written in Polish, Krót-kie pomysły o prawidłach grammatycznych języka żmudzkiego, word formation is discussed in para- 7 It is in fact derived from savas, while savas alongside savo is derived from savęs. 8 Apparently, the particle ne is called an adverb. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 67 …written very little. In explaining the semantics of nouns, only sets of diminutive nouns are presented (Čiulda 1993: 133). That derivation is not understood here is shown by the juxtaposition of diminutive nouns with nouns that have a negative stylistic connotation and may also be morphologically unanalyzable (Pol. O imionach zgrubiałych i zdrobniałych), that is, they are classified according to semantic nuances. Derivation is mentioned again only in the section on syntax (Pol. O Składni): compounding, the addition of prefixes (also understood as compounding), and the addition of suffixes are compared with the syntactic joining and agreement of words. The fact that the basis of a compound noun is a syntactic construction may perhaps justify this approach, but here word formation is not understood as a relatively independent grammatical phenomenon. Rather, the beginnings of modern functional grammar can be discerned in this approach, since it explains such matters as these: time is expressed by the accusative and locative cases; an object made of a particular material is described by derivatives with the suffix -inis; and the suffix -ystė is used to express a collective. Indeed, even the title of the section on prefixed verbs emphasizes not derivation but function—o słowach dokonanych ‘on perfective verbs’ (Čiulda 1993: 191–197)—and explains the formation of prefixed verbs in terms of their function. Although placed in the section on syntax, word formation is perhaps discussed most effectively in two subsections: “On the formation of some simple nouns from others” and “On compound words” (Pol. O formowaniu się rzeczowników prostych jednych od drugich; O wyrazach złożonych) (Čiulda 1993: 191–193). Examples of the following types are presented, briefly described, and classified: 1) baltumas, aukštybė; 2) karalystė; 3) eglynas (collective nouns banda, būrys, krūva are also included with it, although the first two are not derivatives); 4) audimas; the second subsection contains: 1) pusgalvis, juodberis, balt-nugaris; 2) užtrobis, pasaulis, priekletis; 3) juokdarys, kaminšluostys, sandara, uždarbis; 4) girpel­nys, niekdžiugys; 5) šunbajoris, puspiemenis, pusšunis. The examples show that there is a considerable lack of a single basis for classification and, more generally, of a systematic presentation of word formation. Nevertheless, these subsections can already be regarded as the outlines of a science of word formation. 1.2.3. Another grammar from that period, also written in Polish, is Kalikstas Kasakauskis’s (Kalixt Kossakowski) Grammatyka języka żmudzkiego... This grammar was published in 1832, and so was probably available for Čiulda and Daukša to read. There is essentially no treatment of word formation in this grammar; only a few separate matters of derivation are mentioned in a handful of places, and these are presented haphazardly and without significance. For instance, derivation is mentioned in the subsection on adverbs, where it is explained that adverbs are formed from adjectives—giarai (: giras), didiey (: didis), gražiey (: gražus),—and that degrees of comparison of adverbs are formed from them—giariaus (: giarai), gražiaus (: gražiey), gražiausiey (: gražiaus). A little further on, another small group of adverb formations is distinguished; the claim is that these adverbs are formed from nouns (although in fact they are formed from denominal adjectives)—kalbiškai (: kalba), bitiškai (: bitė), poniškai (: ponas); kalbingai, poningai, akmeniškai, akmeningai (Kasakauskis 1832: 13–14). The section on the degrees of comparison of adjectives briefly explains how pronominal forms are made (Kasakauskis 1832: 57–58), while the sections classifying nouns and verbs identify the following types: žodey nesudiety – dedu, raszau (: Kasakauskis 1832: 27), sudiety – sudedu, įraszau (Kasakauskis 1832: 28), sudietiny – draugmokitinis, medutekans; pirmutiny – tiews, dirwa; iszeyny – tiewiszkas, dirwinis (Kasakauskis 1832: 15). 68 | ANTANAS SMETONA 1.2.4. Another manuscript grammar by an unknown author from the first half of the nineteenth century is X. D. K. P. S. Prawida języka Litewskiego Czyli Grammatyką Litewską. It mentions only the prefixal derivation of verbs, together with their inflectional paradigms. A detailed description of derivation (and other aspects of the language) was deliberately omitted. The author of the grammar states: “...I see no reason to expand this little book with those less necessary rules; for the student of the literature of this dead (or neglected) language is richly presented with the boundless treasures and beauty of the language, while the effects and interplay of most words, especially those that affect the mind, unfold; let us take, for example, the word bijau, bijausi: gądaw, gądaws, Jšsygądaw, pargądaw, nusygądaw, parsegądaw, parsegądynaw...” (X. D. K. P. S. 2002: 314). 2. GENERAL OVERVIEW OF THE PRESENTATION OF WORD FORMATION IN DAUKŠA’S GRAMMAR 2.1. Daukša’s grammar (hereafter DG) contains four sections specifically devoted to word formation: 1) On the formation of one part of speech from others (Apie tajsimų vienos dal’os kalbos iš kitų209/1)—on suffixal derivation (DG 209–211); 2) On compound words (Apie išreikš’kus sudetin’us212/12)—on compounds and prefixed words (DG 212–214); 3) On prepositions and prefixes (Apie prielinksnį131/1)—on prefixal derivation (DG 131–135); 4) On the formation of adverbs from adjectives and participles and their comparison (Apie tajsimų priežod’c’u iš pripultiniku ir vardžod’c’u ir apie ju žięngs’n’awimu174/14)—on one adverb-formation pattern (with the formant -ai) (DG 174). 2.2. SUFFIXAL DERIVATION has four subsections: 1) on the formation of diminutives (Apie tajsimų dajktiniku lapningu iš dajktininku abędnu (paprastin’u209/3); 2) on the formation of nouns (Apie tajsimų dajktiniku iš kitu dal’u kalbos210/1); 3) on the formation of adjectives (Apie tajsimu pripultiniku iš kitu dal’u kalbos210/19); 4) on the formation of verbs (Apie tajsimu žod’c’u iš kitu dal’u kalbos211/1). The section on the formation of diminutives is probably the most detailed: they are classified and their formation is briefly discussed. However, among the diminutive nouns there is also an interspersed group of adjectives with diminutive suffixes. This is entirely understandable, since in older grammars nouns and adjectives were often subsumed under a single category of “name.” In fact, the formation of diminutive nouns and adjectives can be equated only formally. Diminutive adjectives denote only a lesser degree of a quality than their simplex counterparts (cf. gražus ir gražčiukas, gražutėlis), and acquire an affectionate sense only when used alongside a diminutive noun. An adjective always agrees with a noun, so diminutive adjectives in fact express not the smallness of a quality but the smallness of the object possessing that quality, or an affectionate attitude toward it. The other subsections on suffixal derivation simply provide examples of derivational sets for the relevant parts of speech, classified according to their base parts of speech. There is no account of meaning; only formal examples of derivation are presented. One inconsistency is apparent in the section on suffixal noun formation: the formation of diminutives is repeated, even though it is entirely out of place in a section on formation from other parts of speech, since it is modification. This methodological inconsistency prevented the presentation of a comprehensive system of word formation. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 69 and the semantic relationship between base words and derivatives. However, the idea of presenting derivatives according to their derivational base is also quite valuable. That the discussion concerns precisely the prefixal formation of nouns from other parts of speech can be understood only from the examples. 2.3. COMPOUND WORDS In the grammar, they are presented not in a separate chapter but under the parts of speech, and are called not compounds but composite words (following Polish złożone). The grammar has three subsections: 1) on compound nouns (Apie daiktininkus sudetin’us212/2); 2) on compound adjectives (Apie pripultinikus sudetin’us214/1); 3) on compound verbs (Apie žod’čus sudetin’us214/29). These subsections not only provide examples of word formation, but also explain certain aspects of the use of compound words and indicate which inflectional type a derivative belongs to. The descriptions of the derivatives vary in detail: compound nouns are described most fully, while only examples are given for verbs. Like other authors of early grammars, Daukša also classifies prefixed words as derivatives, that is, he adopts the view that a prefixed word is compounded from a preposition and an independent word. This is neither an unexpected nor an unusual view, and it has a rational basis, since prepositions and prefixes are undoubtedly related9. Therefore, identifying a certain portion of prefixation with compounding is essentially acceptable for synchronic linguistics, since it is quite clear that both prefixed nouns and compounds have the same derivational formant—the ending (cf. prieš aušrą = priešaušris and grėblio kotas = grėbliakotis). 2.4. Despite the fact that prefixed words are regarded as compounds, it can nevertheless be argued that PREFIXAL FORMATION is also discussed separately in the grammar. In the chapter devoted to prepositions, Apie priev ardi131/1, the author treats prepositions and the prefixes derived from them as one part of speech, and therefore mentions prefixal formation here as well. This was a consequence of the grammar author's theoretical views on parts of speech: he follows the ancient tradition, in which preposition and prefix are called by the same name because they are related and because a preposition (especially one accompanying a noun) often becomes part of a word, e.g., po stalu – pastalė. The author sensed that these are one and the same thing. Finally, the equivalents of verbal prefixes are also prepositions, cf. išeiti iš..., įeiti į.... It was naturally understood that a prefix and a preposition are the same element, one free-standing and the other having become a constituent part of a word. However, this view of the nature of the prefix and preposition leaves many contradictions in the grammar: the prefix is written together with the word, which means that not only a word but also a part of a word is treated as a part of speech; prefixed words are considered compounds, yet in ordinary compounds neither element should be treated as a separate word, a separate part of speech; if prefixes derived from prepositions are equated with prepositions, there should be some explanation for those prefixes that have no corresponding prepositions. It is especially difficult to understand why the postpositional elements -p, -pi are also considered prepositions, although historical grammar can unite preposed and postposed elements in a single class of non-independent elements inclined to fuse with an independent wor- 9 Urbutis (1978: 284) teigia, kad priešėlinė daryba istorinėje kalbotyroje kartais laikoma kompozicijos rūšimi. Būtent taip yra ir Daukšos, ir daugelyje iki jo rašytų gramatikų, kur tik minimi sudurtiniai žodžiai. Netgi šiais laikais veiksmažodžių priešėliai klasifikuojami į prielinksnių ir dalelyčių kilmės priešėlius (be-, te-, nebe-, tebe-), o kad daiktavardžių priešėliai yra aiškūs prielinksniai, rašo daugelis tyrinėtojų ne tik Lietuvoje, bet ir užsienyje. Priešėlinius daiktavardžius sudurtiniais žodžiais siūlo laikyti ir Paulauskienė (1994: 64). 70 | ANTANAS SMETONA ds. This section of the DG lists prepositions, explains how they are classified according to the case they govern (an incomplete leaf out of the five devoted to this part of speech has been added: DG 131–135), while everything else is devoted to prefixes. This section can therefore be regarded as a chapter on prefixal formation, with an introduction on the origin of prepositions. Thus, ideas about prefixal formation can be found in the chapter on prepositions, where prepositions, prefixes, and locative enclitics (all of them, in the grammar author's terminology, prievardžiai) are divided into true and non-true forms, those written separately and clitics. These clitics are considered to be the endings of postpositional locatives -p, -pi (true prievardžiai) and prefixes (non-true prievardžiai). Little attention is given to true prepositions—they are simply classified according to the different cases they govern. Although the grammar author groups prepositions and prefixes together in one chapter, he emphasizes their essential difference: prepositions that have become part of a word lose their ability to govern a case (prievardž’ aji prilipėd’ ai ne tikrieji neva valdo prietik’ ais¹³¹/²¹). He then turns to explaining the use and formation of prefixes (non-true prievardžiai, clitics). First, they are classified according to origin: the first group comprises those like the true forms, since they have exact prepositional counterparts: až, ažu, in, i, iš, nu, nu, po, per, prie, pro, su¹³¹/²⁷; the second group consists of forms derived from actual prepositions, in other words, variants of prefixes: ap yr api – derived from apie; nu derives from nu; pa derives from po, par from per, pra from pro, pri from prie, sun from su¹³¹/²⁹; the third group comprises forms not derived from anything and meaningless on their own—that is, what we would now call prefixes of particle origin and those without prepositional counterparts: at, ata, ato; už, užu; be, tebe, nebe <…>from the Polish preposition do, which a Lithuanian pronounces as da¹³¹/³². Then comes a substantial, though unnamed, discussion of prefixal formation and prefix spelling (and usage)—almost four leaves are devoted to it (DG 132–135). The chapter Apie prievardį discusses all parts of speech that can take a prefix: nouns, verbs, and adjectives. In explaining formation with various prefixes, the grammar often specifies which part of speech serves as the base. This way of presenting word formation makes it very difficult to distinguish which material belongs to which part of speech. This confusion calls for the following plan for the subsequent description of prefixal formation: the main account of prefixal formation will be given in the chapter on prefixed verbs (since Daukša speaks mostly about verbs, and this method of word formation is generally most characteristic of verbs) (see 5.2.2), while the chapters on the formation of prefixed nouns (see 3.3.2) and adjectives (see 4.3.2) will mention only matters characteristic of those parts of speech, with references to the main chapter on prefixal formation (especially since, as recently noted, this grammar does not distinguish compounding from prefixal formation). 2.5. ADVERB FORMATION The chapter title Apie tajsimų priezod’cų iš pripulti- niku ir vardžod’cų ir apie ju žięngsn’avimu¹⁷⁴/¹⁴ shows that degree formation has not yet been distinguished from word formation here. Daukša regards the adverb as an invariable part of speech, and comparison as a derivational, not an inflectional, category. In general, all parts of speech are divided into inflecting and invariable ones, and the grammar is organized accordingly into two large chapters: On the Invariable Parts of Speech and On the Inflecting Parts of Speech; the adverb is described first among the invariable parts of speech, while the account of adverb formation appears in the chapter on inflecting parts of speech—perhaps because the title also mentions ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 71 gradation. In the section on adverb formation, an attempt is made to present as many as four cases of word formation; however, upon closer examination, it becomes clear that only one concerns the actual formation of adverbs (with the formant -ai, see 6.1.), while the other three groups of examples concern the formation of adjectives from adverbs (see 6.2), the formation of comparative and superlative adverbs, and the formation of the non-gendered stative form of an adjective. 2.6. The DG gives only a sparse description of SUFFIXAL WORD FORMATION. There are only one or two examples of non-prefixal suffixal derivatives, while prefixal suffixal derivatives are fairly consistently classified, according to their structure, in the sections on compound words. The grammar section describing the formation of feminine personal and ethnic names from masculine forms, e.g., anglė (: anglas). may be considered genuine suffixal formation. The section is entitled On forming the feminine from the masculine (Apie tajsimų gimties moteriškos iš vyriškos171/28). It in fact focuses more on inflection for gender, while one or two matters of word formation are discussed in the section on noun formation (see 3.4.1). 2.7. FRAGMENTS ON WORD FORMATION can also be found in other sections of the grammar, that is, wherever the logic of presentation calls for them. Among these fragments on word formation, mention should first be made of the fairly detailed account of verb-form formation in the introductory section on conjugation paradigms (DG 176–181). Three subsections on the formation of verb forms are presented there, organized according to the derivational base—the three principal verb forms. Some derivatives formed with verbal and nominal suffixes are also included among the verb forms. This is because in old grammars the formation of verb forms and of certain suffixal derivatives had not yet been fully distinguished. This can be understood if one considers how regularly, for example, -imas (-ymas) suffix derivatives are formed from verbs. Such regularity in word formation apparently did not allow the grammar’s author to omit precisely this case of word formation from the subsection of the phonetics section that explains in detail the spelling of the vowel i (DG 118–119). All suffixal derivatives in these sections are discussed in this study in the relevant parts on the suffixal formation of nouns and verbs. Thus, the DG devotes more attention to word formation overall than do other grammars written in Greater Lithuania. On the other hand, the greater attention to word formation was nevertheless not enough to break away from the predecessors’ understanding of matters of word formation. 3. NOUN FORMATION 3.1. Suffixal nouns 3.1.1. DERIVATIVES FROM NOUNS receive the most attention in the DG. The largest and most thoroughly described category of such noun formation is DIMINUTIVES (suffixal diminutive nouns). This opening to the description of suffixal noun formation has survived to the present day, although diminutives are stylistically marked and are not used in all styles. Two factors account for this: a) their frequent use in spoken language and b) their potential regularity of formation: a diminutive can be formed from every noun, even with several suffixes (Ulvy das, ed., 1965: 253–288). In the DG, diminutives are divided into three types of formation (the fourth type distinguished in the grammar is the formation of adjectives with diminutive suffixes, discussed in the corresponding section on adjective formation; see 4.1.2): 72 | ANTANAS SMETONA 3.1.2. FIRST DIMINUTIVE TYPE the following examples are given: darželis (: daržas), oželis (: ožis), arklelis (: arklis), sunelis (: sunus), žmogelis (: žmogus), šunelis (: šuva), dun- telis (: duntis), galvelė (: galva), dedelė (: dede), rikštelė (: rikštė), akelė (: akis), avelė (: avis), avinelis (: avinas), žilvitelis (: žilvitis), karalelis (: karal'us), akmenėnelis (: akmuo), menešelis (: menù), motinėla (: motina), ge.guže.la (: ge.guže), žibinkštelė (: žibinkštis)²⁰⁹/⁸. This group includes derivatives with two variants of the suffix: -elis, -ėlis. The choice of suffix is associated with the different number of syllables in the base word. It is stated that diminutives with the suffix -ėlis are formed from words that have more than three syllables (subalses²⁰⁹/⁸). This word-formation pattern is described in detail: the endings and forms of the base words from which derivation takes place are specified: from the nominative of nouns, with the endings -as, -is, -us, -a, -ea, -is removed, and from base words ending in -uo(-ū), -va, from the genitive, with the endings -ies, -o removed. In this way, old consonant-stem nouns are distinguished, that is, it is stated that suffixal derivatives of the nouns mėnuo, akmuo, šuva are formed from the genitive, not the nominative. The first two nouns are classified as trisyllabic¹⁰. This word-formation type also included the example davganora – davganorėlė²⁰⁹/¹², but it was deleted on the grounds that it is a compound word. Daukša had not yet distinguished direct from indirect motivation, or morphemic analysis from word-formation analysis, and it was still thought that the base of derivation could be either a simplex or a derivative (whether a suffixal or prefixal derivative, or a compound). At first, Daukša clearly saw that daugnorėlė is a diminutive noun and, regularly, should be derived from daugnora; however, because its stem structure contained two roots, he deleted daugnorėlę from the list of diminutives and assigned it to compound words. The form of the suffix itself is not discernible because of orthographic peculiarities: galvelė – ge.guže.la. The grapheme ea is polyphonic, that is, it is written where, in the author’s view, the sound varies across dialects. The grammar’s author probably pronounces the suffixal consonant l as hard, but the spelling does not reveal this, since hard l is marked inconsistently: ge.guže.la, žibinkštėlea. Or perhaps no single notation was found that would reflect the situation in different dialects. At the end of a word, where present-day standard language has ė, the letter ea is systematically written. Written this way, the e variants after l (both hard and soft) can no longer be distinguished. This merely confirms Z. Zinkevičius’s claim that after hard l the vowel ė varies (Zinkevičius 1966: 162–163), but it is impossible to ascertain the quality of the variants. Nor is it possible to verify the quality of both vowels in the suffix -elė (original -edlea). Although the grammar’s author did not equate the suffixes -elis and -ėlis, the grammar’s orthographic conventions in some cases prevent the reader from distinguishing them; at times these suffixes may have been entirely identical. 3.1.3. SECOND AND THIRD DIMINUTIVE TYPES: avin’ukas, brol’ukas, šun’ukas, sun’u-kas, dunt’ukas, gajd’c’ukas²⁰⁹/¹⁷, runkitėa (: runka), deditea (: dedea), karvitea (: karvea), avitea (: avis), dukitea, dukteritea, duk’ritea (: duktea), sešitea, sešearitea (: sesu, sesuo)²⁰⁹/¹⁹ The author’s explanation—that from masculine base words ending in -as, -is, -us²⁰⁹/¹⁴, derivation is made with the suffix -iukas, while from those ending in—made it possible to combine these types. 10 The consonant-stem paradigm is not named separately: on the basis of the genitive it is grouped with i-stem nouns—as in the present-day Dabartinės lietuvių kalbos gramatikoje (1997). ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 73 endings -a, -eₐ (-ea), -is, -ù (-uo)209/19—with the suffix -ytė. Both the examples and the endings show that these types are clearly distinguished by gender, and are used to name living beings whose gender is motivated. Base words with the same root are given—avinas and avis—aviniukas and avytė; or, in another subsection on noun formation, brol’ukas, brolitea210/4. This shows that the grammar’s author assigns words to word-formation types not only according to word-formation formatives, but also according to semantic relations between gender forms. Although it is doubtful whether such an association of suffixes with the distinctive feature of gender reflects actual usage, since diminutives of both genders can in fact be formed with the same suffix from base words of different genders; cf. brolyt is, brolyte. There is insufficient evidence to claim that such a strict differentiation of suffixes by gender is a dialect feature. Although the author does not state this, it may also be one of the grammar’s prescriptive facts. When discussing the forms of base words (nominative or genitive), in the masculine group the grammar’s author explains the word-formation peculiarities of the consonant stem and derives šuniuka from the genitive. In the feminine group, however, such detailed explanations of word formation are absent, although more examples of noun formation from the consonant stem are provided: from duktė and sesuo – duktytė, dukterytė, dukrytė, seserytė, sesytė. No explanation is provided, but this probably gives a comprehensive account of the usage both in the author’s own dialect and in dialects familiar to the author. The suffix -ukas is systematically written with an apostrophe, -’ukas, and it is emphatically stressed that the vowel -u- is supposedly always fronted (in Daukša’s term, soft), while the end of the base word’s stem is always soft. As the author notes, for this reason the consonants d and t, when occurring before the fronted u, inevitably become the soft affricates dž and č: dančiukas, gaidžiukas. No mention is made of a consonant that does not undergo softening before the suffix, although pronunciation varies considerably in dialects and in the standard language (Ambrazas, ed., 1997: 91). 3.1.4. OTHER WORD-FORMATION MODELS OF NOUN DERIVATIVES To illustrate these, the grammar’s author draws on the tėvas ‘father’ and brolis ‘brother’ derivational families: from tėvas – tėvainis, tėvainystė, tėvikšt’a, tėvinė, tėviškė, tėvelis, tėvutis, tėvuta or tavta; from brol’is – brolėnas, bralajt’a, brol’ukas, brolel’is, brolitea210/2. Thus, diminutives are also assigned to other word-formation types. These two derivational families raise no major difficulties. In the brolis family, brolėnas, broliukas, brolelis and brolystė are regular derivatives; the two feminine examples, brolaičia and brolytė, are somewhat less usual. These are names for kinship relations or participants in weddings, although they are not included in the first volume of LKŽ. Without sentences, one can only conjecture that they are equivalent to brolaitė (LKŽ, 1068). Brolaičia is the result of the mixing of ē and iā stems in East Aukštaitian dialects, as with Bronia, Teklia, eil ia, špūlia, kandžia (Zinkevičius 1966: 226, 247): the suffix -aitė changes into -aičia. The tėvas family is somewhat more complex. Here the examples are divided into three groups (separated by semicolons): the first—tėvainis and its abstract derivative tėvainystė; the second—tėvynė, its derivative tėvykščia, and tėviškė; the third—diminutives tėvelis, tėvutis, tėvuta and its synonym tavta. The latter two synonyms are the most difficult to understand. In the grammar, the diphthong au is written av—as in tavit’āj137/22 Vitavtas137/32. However, three considerations rule out identifying tavta with present-day tauta: first, its classification as a diminutive; second, its synonymy with tėvuta; and third, tauta is in fact not derived from tėvas (Fraenkel 1965: 1069). Finally, this may simply be a spelling error: not tavta, but *tavuta – a variant of tevuta. The form tevuta itself can probably be identified with tėvutis (LKŽXVI 157). Incidentally, some of the derivatives of tėvas- 74 | ANTANAS SMETONA The meanings of the nouns are explained in the section on the classification of nominal stems: Tevikšt’a, Tevinė, Teviškė – (Oczyzna), Tevaj^niste (dziedzictwo), Tevajnis (dziedzic)138/12. In the Polish–Lithuanian dictionary (Vaitkevičiūtė 1994) oczyžyna is translated as homeland, birthplace and arch. tėviškė; dziedzictwo – inheritance, inherited estate; dziedzic – heir, landowner. Thus, some derivatives can be semantically linked to formations already recorded in dictionaries: tėvaynistė – tėvonystė ‘inheritance, patrimony’; tėvajnis – tėvonis ‘2. inheritor of his father’s property, heir’ (Keinys 2000: 842). 3.1.5. DAIKTAVARDŽIŲ DAR ribai IŠ VEIKSMAŽODŽIŲ parodyti pasitelkta veiksmažodis rašyti ir trys šaknies mok- veiksmažodžiai – mokyti, mokti, mokėti: rašyti – rašalas, rašiklas, rašiklis, rašilas, rašimas, rašytojas, rašituvas, rašituvėa, raštas, raštinikas; mokiti11, mokti, moketi – mokitojas, mokitinis, mokituvėa, mokstas, mokimas, mokėjimas210/10 (Vedinių iš daiktavardžių ir veiksmažodžių pateikimas rodo, kad Daukša suvokė, jog darant naujus žodžius susiformuoja bendrasaknių žodžių lizdas, kurio žodžiai skiriasi savo struktūra, bet yra jungiami tos pačios šaknies formos ir semantikos.) Rašyti lizdas išsamus ir aiškus. Visi vediniai padaryti iš veiksmažodžio, bet nenurodyta iš kurių jo formų. Truputį neįprastai pateikiamas mokti lizdas, nes pateikiami net trys pamatiniai veiksmažodžiai, kurie irgi tarpusavyje susiję darybiniais ryšiais: mokyti ir mokėti yra vediniai iš mokti. Autorius darybos lizde nenurodo, kokie vediniai iš kokio veiksmažodžio padaryti. Veiksmažodis mokti nepriešdėlio dabartinėje kalboje nevartojamas, bet Lietuvių kalbos žodyne pateikiamas iš Aušros, Antano Juškos ir Barono žodynų: Kūdikis moksta mylėti artimą. Temoksti jis visa, ja nor pramokti išmokti. Baigti atmintinai mokti (LKŽVIII, 333). Dabar tesakome išmokti, pramokti, Daukšos laikais jo tarmėje mokti galėjo būti vartojamas ir be priešdėlio. Todėl racionaliausia būtų buvę mokti laikyti šaknies mok- lizdo pagrindiniu žodžiu ir nebūtų reikėję šiam lizdui trijų pamatinių žodžių: mokti, mokyti ir mokėti. Pastarąsias dvi formas dabar priskirtume mišriosios darybos tipui, kur ne visos pagrindinės formos išlaiko priesagą (mokyti, moko, mokė; mokėti, moka, mokėjo). Gali būti, kad esamasis laikas be priesagos lėmė tai, kad Daukša visas formas darybiniu atžvilgiu laikė lygiavertėmis (plg.: moksta, moko, moka; nė viena esamojo laiko forma neturi priesagos). 3.1.6. As already mentioned in 2.7, suffixal noun derivation is also discussed in the sections on verb inflection. The introduction to the section on verb conjugation first establishes four inflectional verb classes and explains that, once these are known, all the moods, tenses, numbers, persons and participles, as well as nouns derived from verbs, can supposedly be identified from the principal forms of a verb175/33. This is why the section also includes some more regular cases of suffixal verb derivation. The author then presents three subsections which, alongside the formation of verb forms, also describe the derivation of certain suffixed verbs and nouns from the first-person present, first-person simple past, and infinitive. 3.1.7. DG states that nouns with the suffixes -imas and -ėjas/-ėjųs are formed from the first-person simple past. Derivatives with the suffix -imas are described most fully (and most confusingly). The text makes clear that the grammar’s author did not distinguish between the suffix variants -imas and -ymas. This is characteristic of eastern dialects: unstressed long vowels ū and y in the middle of a word are shortened by the Panevėžys speakers and pronounced like the short vowels u and i. 11 The manuscript is underlined with a dashed line. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 75 in the middle of a word, and pronounced like the short vowels u and i; moreover, the lengthening of short stressed u and i must also be borne in mind (Zinkevičius 1994: 62). Since the text is unaccented, it is now difficult to determine where the accents were, but the numerous corrections and convoluted explanation show that the grammar’s author himself was also uncertain about the pronunciation and spelling of the suffix -imas. Here are several examples from the grammar as a whole: darimų181/21 dari-mo142/6 grasimų137/1 išgul(d)imo193/21 išskaitimų129/7 paliginimo140/17 parašimas126/29 pataj-simo166/21 perskírimas128/1 rašimas117/19 skajtimo112/29 sugrįžimu134/21 šavkymu136/30 iškal-bejymas117/17 išmeatymas125/10 perskiryymas129/19 suminkštinymas116/29 išradimas133/8 išradymas133/17. These examples yield a very clear conclusion: the letters y and i are not intended to indicate sound length. Daukša himself explains the difference between these letters in the section on the pronunciation and spelling of letters, using the same examples with the suffix -imas. Yet the explanation there, too, is very convoluted (DG 117–121). It is natural that derivatives with the suffix -imas appear in morphology and other sections of the grammar, since these derivatives are regular, like inflectional forms. In dictionaries they are generally not entered as headwords, but are presented within the entries for the corresponding verbs. All the grammar author’s explanations appear to address the problem of palatalizing the stem-final consonants of the base word when they occur before -imas. Daukša states that in this position two variants of the vowel [i] may be written: a soft i (softness is marked either with a dot above i, a breve beside the preceding palatalized consonant, or both together) and a hard i (hardness is marked either by the absence of a dot above i or by writing the letter y): šal’imas – šatys ‘having become cold’178/8 vadin’imas, sodin’ymas, pažin’i-mas – pažin’is, va/so-din’is ‘having recognized, called, planted’178/9. As exceptions in which the stem-final consonants are not palatalized, infinitives of verbs with the suffix i (y) and consonants preceding it are listed: -biti, -piti, -giti, -kiti, -riti, -siti, -siti178/25. This may also be evidence that Daukša did in fact distinguish the suffix variants -imas and -ymas, since it is precisely the variant with the long vowel that occurs most often among the exceptions: mokimas, plikimas, kinkimas, minkimas, mirkimas, n’urkimas, nukimas, dulkinimas, spirgimas, skubimas, vilbimas, kusimas178/2712 3.1.8. The grammar states that nouns with the suffixes -ėjas and -ėjus are supposedly formed from the first-person simple-past form; more precisely, they are formed with a single suffix, -ėj-, but the derivatives belong to different inflectional paradigms: pirkejas, pirkejus – pirkeja, pirkejį (: pirkav); kasejas, -jus, kaseja, -ji (: kas’av)179/5. Nouns with the suffixes -tojas, -tojus, and -smas are formed from the infinitive. The suffixes -tojas and -tojus resemble -ėjas and -ėjus in their mixing of inflectional paradigms. The grammar gives the following examples of derivatives with these suffixes: garbintojas, garbintojus, garbintoja, garbintoji (: garbinti); artojas, artojus, artoja, artoji (: arti)181/8. It is important that Daukša already recognized that nouns are not formed only from infinitives. The examples suggest that in the mid-nineteenth century, derivatives with ia- and iu-stems bearing these suffixes were used interchangeably among the northern Panevėžys speakers and in the dialects known to the grammar’s author. Zigmas Zinkevičius (1966: 255) likewise observes that stem-paradigm mixing is particularly characteristic of derivatives with the suffixes -ėjas and -tojas. Incidentally, the same source states that in the Panevėžys dialect the ia- and iu-stems “practically merged into a single paradigm” (1966: 244). The manuscript 12 In general, all this has more to do with spelling and phonetics than with derivation, and is therefore described in greater detail in the sections on spelling and phonetics. 76 | ANTANAS SMETONA The grammatical data show that by the mid-nineteenth century the stems had not yet merged into one, although they were already being confused to a considerable extent. The most unusual feminine forms are pirkėjai, kasei, garbintoji and artoji. Alongside marti and pati, Zinkevičius also mentions the Samogitian forms poni and viešni, see Zinkevičius (1966: 223–224); however, the grammar gives personal nouns that can be of either gender, more precisely, forms in the feminine are derived from those in the masculine. Different gender forms of the same noun are aligned with the adjective stems in -u and -io, cf. gražus, graži; smalsus, smalsi – pirkėjus, pirkėji. There are inconsistencies as well: alongside gražios, graži, the grammar also uses giedrus, giedra (rather than giedri!). Both types of derivatives (-tojas, -tojus; -ėjas, -ėjus) are treated by the grammar’s author as variants, and both are acceptable. More derivatives with these suffixes are given in the section on the formation of compound nouns (see 3.2.3), where they are presented as the basis for forming compounds, e.g. mėsėdis (: mėsos ėdėjas, ėstojas). The regular presentation of derivatives with both suffixes suggests that the grammar’s author probably regards them as equivalent, but the forms with an -iu stem are not mentioned there. These are names for people based on the action expressed by the base word. The regularity of the formation indicates that suffixed personal nouns of this type were formed in the same way with both suffixes, from the infinitive or the past simple tense: kirtėjas – kirstojas, ėdėjas – ėstojas, davėjas – duotojas, pjovėjas – pjautojas. Only žuvėtojas lacks a derivational pair; this should probably be taken to reflect actual usage among the Northern Panevėžys dialect speakers. The section on derivatives from the infinitive in DG also mentions the formation of nouns with the suffix -smas. It states that nouns are formed from some verbs by replacing -ti with -smas: šauksmas (: šaukti), reksmas (: rekti), d’c’avgsmas (: d’c’avgtiesi), gr’avsmas (: gr’avsti). Daukša observed that not many nouns with the suffix -smas are formed, and only from certain verbs (and the Dictionary of Contemporary Lithuanian Language lists only about 50 of them). In the section on derivation from the present-tense stem, the grammar discusses only the formation of verb forms and provides no examples of new verb formations. 3.1.9. The following section on suffixed nouns presents NOUN FORMATION FROM ADJECTIVES with the suffixes -ybė, -umas, -uomenė, -ulis: didybė, didumas, didumė, didumėna, did’c’unas, did’c’ulis (: from did’c’o)₂₁₀/₅. Two derivatives are somewhat unusual: didumenė and didžiulis. The word didumenė is written with u, not the diphthong uo. Daukša consistently used the letter ū to represent the diphthong uo, and the grammar contains many proofreading errors, especially omissions of diacritics; on the other hand, this could also be a variant form. For example, the Dictionary of the Lithuanian Language lists didumenė and didūmenė (LKŽ₍ₙ₎ 486). It is unclear whether didumenė could be regarded as a Samogitian form; cf. Zinkevičius (1966: 85). The other derivative is didžiulis. Today we understand it only as an adjective, but Daukša entered it among nouns. If this is not an error, he may have had examples of this word used as a noun, but he does not elaborate on this. 3.1.10. The grammar’s author claims that there is a separate group of SUFFIXAL DERIVATIVES FROM PRONOUNS: iš manies, tavies, savies, musu, jūsu – maniškis, taviškis, saviškis, musiškis, jušiškis; savibė, savitnumas, nūsavistėₐ₂₁₀/₇. Today, derivatives with the suffix -iškis, maniškis, taviškis, saviškis, musiškis, jušiškis, are interpreted in three ways: a) as derived possessive pronouns (e.g. maniškis berniukas – this is how most grammars classify it); b) as pronominal adjectives used substantivally (e.g. Maniškis jau parvažiavo) (Keinys 2000: 381): ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 77 c) kaip iš įvardžių padaryti būdvardžiai (Paulauskienė 1994: 187). Iš tiesų tai yra įvardiniai būdvardžiai, bet dėl įvardžio kamieno specifikos jie, ir gavę būdvardžio priesagą, pasilieka įvardžio klasėje. Labai dažnai jie vartojami kaip daiktavardžiai, pvz.: Mūsiškiai šįkart pralaimėjo. O ar taviškis dar nesugrįžo? (galvoje turima žmonos vyras). Taigi Daukšai daiktavardinė vartosena pirmiausia krito į akis ir todėl šiuos įvardinės kilmės būdvardžius jis tiesiog vadino daiktavardžiais. Daukša himself used only one derivative with this suffix in the grammar’s text: Vardaj savišk’aj arba tikri Lietuviški (imiona swojskie albo prawdziwe litewskie)¹³⁸/²⁷—and in the position of an adjective, not a noun. Thus, although the author classified words of this type (maniškis, taviškis, etc.) as nouns, he provided no convincing examples. Daukša was right to distinguish maniškįfrom maniško, which clearly expresses a quality (on the suffix -iškas, see 4.1.3). Savybė, savitumas, nuosavybė are abstract nouns. Savybė, with the suffix -ybė, is derived from savas, which can also be interpreted grammatically as an adjective. Therefore, savybė belongs to the same derivational type as aukštybė, didybė, galybė, tolybė (the Dictionary of Contemporary Lithuanian Language lists about 200 such derivatives). Daukša’s neologism savitnumas is not a direct derivative of savęs; it is derived from the adjective savitnas. The group of derivatives from savęsalso includes the only prefixed noun in this section, nuosavystė. It belongs in this derivational nest, but the full derivational chain is not revealed. Moreover, it contradicts some of the manuscript grammar’s theoretical principles: first, although prefixed derivatives are considered compounds, this prefixed word is included in the list of derivatives; second, the adjectival origin of this suffixed derivative is not explained; third, the actual formation of the word is nevertheless not established, and it is entered a second time in the section on compound nouns (see 3.3.1). 3.1.11. DG states that NOUNS WITH SUFFIXES ARE ALSO FORMED FROM ADVERBS, INTERJECTIONS, AND CONJUNCTIONS. According to the grammar’s author, abstract nouns are formed from adverbs with the suffix -umas: ūkstumas (: ūksti), tolumac (: toli), daugumas, daugybė (: daug)²¹⁰/¹⁵. The formation of daugumasand daugybėfrom an adverb is generally not disputed today—at least, grammars do not question it. Tolumasand ankstumasmay also be derived from the corresponding adverbs, but the possibility that they derive from the corresponding adjectives tolus, ankstusmust not be overlooked (Ambrazas, ed., 1997: 412). The formation of abstract nouns with the suffix -umasfrom gradable adjectives is just as regular as that of nouns with the suffix -imasfrom verbs; therefore, ankstumas and tolumaccould not be exceptions. Furthermore, the grammar’s author himself uses the adjectives ankstus and tolus, though he too derives them from adverbs—see 6.2. The Grammar of Contemporary Lithuaniansays nothing about the possibility of also linking daugumąand daugybęto the adjective or noun daugus, daugis(LKŽ 326, 322). In addition to anksti, toli, the adverb form daugiis also used in dialects (Zinkevičius 1966: 392). 3.1.12. Daukša claims that nouns can also be formed from conjunctions. The meaning of a conjunction is wholly incompatible with that of an independent word, so a conjunction cannot serve as the base for noun formation. This is an error in DG’s theory, but this monograph analyzes what is written. The author gives the following examples: kadaj, kad’ a (: kad)²¹⁰/¹⁷ It is impossible to regard the word kadajas a noun. It is an adverb of time— 78 ANTANAS SMETONA mis, whose derivational base is probably kad: “perhaps an attached particle -ai is present here” (Zinkevičius 1966: 408). An even less clear example is kadžia. Nevertheless, it may be understood as a noun. The following example is given in the manuscript of Daukša’s contemporary Ambraziejus Kašarauskas: Kuri tat kadė! (LKŽ, 44) – kadė ‘time, hour’. The transformation of kadės into kadžia in Eastern Aukštaitian dialects through the mixing of ė- and ia stems is entirely regular (Zinkevičius 1966: 226, 247) (see also 3.1.4). However, the derivational association of the noun kadžia ‘time, hour’ with the conjunction kad is in fact erroneous. The semantics of a conjunction do not allow it to take nominal derivational devices (suffixes), and therefore nouns cannot be derived from conjunctions: a suffix must change lexical meaning, whereas a conjunction has no lexical meaning. Its function is grammatical. It is a sentence morpheme. 3.2. Compound nouns 3.2.1. The examples presented in the section on the formation of compound nouns in the grammar are perhaps commented on by the author more than any others. First, compounds formed from two nouns are discussed. The grammar’s author states that the basis for this formation is the genitive plural of one noun and another freely inflected noun. It is unclear why the genitive plural is considered the base form of the first component of the compound noun; it is claimed that the genitive plural ending -u must be changed to -a or omitted altogether. One thing is clear: this avoids having to explain changes to the consonant stem, e.g., šun-pėdis. Changes to the second component of the compound noun and the choice of the compound’s ending are not explained as smoothly: masculine endings are changed to -is, while the endings of consonant-stem nouns are changed to -enis. From the perspective of morphemic segmentation, this is not entirely precise, but it is quite convenient for practical teaching purposes: ryt-met-is, mišk-a-žmog-is, galvij-piem-enis. Connecting vowels are not discussed. The author states that the ending of a feminine noun is changed to -e: bal-puš-ė. The author observes this kind of change in endings only in compounds formed from two nouns, and even then not always, since, as will be seen below, the choice of ending does not correspond to the gender of the base word, e.g., ritmėtis²¹⁷ and darbmėtė²¹²/³¹—the second component of both compounds is metas. The claim that if a base word of the corresponding gender has the ending -is or -ė, it remains unchanged in the compound is also rather questionable: vilka-žolė. This is true only formally, since what is actually involved here is the morpheme homonymy characteristic of Lithuanian¹³. For nearly all compounds, the declension class to which they belong is indicated. After describing the word-formation pattern and the inflectional paradigm of the compound, the meanings of some compounds and their underlying word combinations are also explained. The following examples are used: Ritmėtis (: Rito metas), Mišk ažmogis (: Miško žmogus), Galvijpiemėnis (: Gal’ 13 This theoretical inconsistency is not always fully understood even in contemporary linguistics, so it is worth explaining separately. It would clearly be wrong to claim that the words ož-k-a, ei-k-i and Graž-k-a contain the same suffix: these are three different suffixes, whose forms merely coincide. Likewise, the endings al-us (nom. sg., u-stem) and vaik-us (acc. pl., a-stem) are entirely different. It follows that the endings of žol-ė and vilkžol-ė differ in exactly the same way as those of puš-is and balpuš-ė. Therefore, one cannot claim that the ending in the word vilkžolė is unchanged. The ending of the word žolė denotes nom. sg.; that of the word vilkžolė denotes nom. sg. + derivational formant. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE 79 cattle herder), Vilkažolė, vilkžolė (: Wolf’s grass), Šilažemė, šilžemė (: Pine-forest soil), Šunpedis, šuntakis (: Dog’s foot, dog’s path), Balpušė (: Bog pine) and so on212/17. It is clear that the author selects examples so as to illustrate the relationship between the compound noun and its underlying syntactic construction; that is, the author observes that the derivational basis is not just any arbitrary nouns, but nouns that form a meaningful phrase. (Of course, it is not always possible to split up a compound and obtain a phrase with the same meaning: dviratis is not two wheels; kupranugaris is not a hump and a back; a compound often becomes semantically detached from its base, functions in the language as an independent word, and acquires new and figurative meanings.) Daukša’s examples are not given arbitrarily; rather, he explains how a compound should be divided or which syntactic construction corresponds to it. It is worth quoting the grammar’s author here: Compound nouns are often expressed by two nouns: the first is used in the singular genitive case, while the second is inflected for case and number according to its declension212/13. This interpretation suggests that Daukša may sometimes have failed to distinguish the direction of derivation. Moreover, the underlying syntactic construction and the compound are not regarded as complete synonyms; a semantic difference is perceived: the compound word has a more negative connotation, while the syntactic construction has a more positive one: all nouns expressed by two nouns denote a degree of reverence (cf. Polish poszanować ‘to honor’), but when expressed by a single compound noun, they denote greater reproach212/24. 3.2.2. Daukša goes on to identify several more compound models and comments on their formation and inflection: šienapjūtis, rug’apjūtis, vasarojpjūtis, mežlavežtis212/28 are inflected like uncompounded nouns according to the seventh paradigm established in the grammar, which is feminine. It is said of the compounds naktigonėa, darbimėtea, naktidavža212/31 that the bases of their first components are shortened singular instrumental forms, although it was just stated that the first component of a compound is based on the genitive plural (see 3.2.1). This solution is apparently the only one possible when connecting vowels are not identified. The gender of some compounds in these groups is unclear. The gender of nouns denoting inanimate objects is unmotivated; consequently, the gender of some of them has not become fixed even today and varies across dialects. The grammar states that the gender of a compound is determined by the gender of the second base word (see 3.2.1). The examples are, of course, few, but if the feminine compounds šienapjūtis, rug’apjūtis, vasarojpjūtis, mežlavežtis are formed from two nouns, it follows that the stated rule for selecting the ending -ė for feminine compounds is not observed. The base nouns were probably pjūtis, vežtis or važtis14—their gender is unclear. If we assume that these compounds are formed from a noun and a verb, the choice of gender is unmotivated; however, the emergence of the suffix -t- then becomes unclear. In general, such variation in the choice of gender (more precisely, stem class) is entirely regular in present-day dialects as well (Zinkevičius 1966: 219); this means that the phenomenon was also current in the Eastern Aukštaitian dialects of the mid-nineteenth century. 3.2.3. Compounds denoting persons are described separately: šienapjuvis, šienpjūvis (: hay cutter, reaper); žmog ažudis, žmogžudis (: žmogavs or žmon’u 14 Incidentally, the nouns vasarojpjūtis and vežtis / važtis are not recorded in the Dictionary of the Lithuanian Language. The absence of a meaning referring to a period of time (as with pjūtės) rules out associating it with važta. Therefore, mežlavežtis / mešlavaitės may also be a case of analogical formation. 80 | ANTANAS SMETONA žuditojas); dünadavis, dündavis (: giver of dünos, dūtojas); malkakirtis (: cutter of malkų, kirstojas); mesedis (: eater of mėsos, estojas)213/1. Only masculine forms are given. The grammar states that compounds of this type are formed in the same way as two simple nouns, except that more components are supposedly removed from the deverbal noun: which ones and how is not explained. In fact, these may be compounds with a deverbal second component: šenapjūvis ‘one who cuts hay’, žmogžudīs ‘one who kills a person’, duonadavis ‘one who gives bread’, and so on. Šienapjūvis, žmogžudīs, duonadavis may have their base components in reverse order: žmogžudys (: to kill a person), but it may be assumed that when compounds of this type were formed, word order in the sentence was different: duondavis (: to give bread), or perhaps it was already formed from a different construction: duonadavis (: giver of bread; descriptive genitive + deverbal noun) ‘one who gives bread’. The latter explanation seems more convincing—the second step from the verb, that is, an agent noun derived from a verb, davėjas (: to give), is further narrowed in meaning by a descriptive genitive, duonos davėjas; in compounding, the suffix and ending are removed, yielding duondavis. There are quite a few such words, and Daukša sees precisely this as their formation. Variants of some compounds are given both with and without a linking vowel—this presentation most likely reflects actual usage tendencies. Among the specific examples worth mentioning is šienpjūvis, with the verb-root vowel u (probably long or half-long). LKŽxiv 750 gives examples only from written sources and the Western and Southern Aukštaitian dialects; there are no examples from the Eastern Aukštaitian dialects. Another example is mėsėdis ‘one who eats meat’. Its structure is usual, but on the same page another compound formed from the same base words is also mentioned—mesiedas213/1. It should be assumed that the period from Christmas to Shrove Tuesday is meant. In the Northern Panevėžys dialect (around Biržai), the change of ē to ie is regular (LKŽviii 44). Its different ending distinguishes it from the agent noun mėsėdis. The base words in this group of compounds are derivatives with the suffixes -tojas and -ėjas. In DG they are presented in the section on derivation from the principal forms of the verb, whereas in this study they are discussed in the section on suffixed nouns (see 3.1.8.). 3.2.4. The grammar describes separately compounds containing what are now called prefixoids and suffixoids—elements of compound words that, because of their regular derivational behavior, increasingly resemble prefixes or suffixes, e.g., pusrytīs arba laibgalis. The recognition and grouping of these phenomena demonstrate the grammar author’s considerable linguistic insight. Compounds with pus-, vid- and -gal- are discussed separately: pusrytis, pusbernis, pusbrolis, pusdienis, pusžuvis, pus’vanktais, pus’avžiemis, pus’avgavenis, pus’avdienis, vidudienis, vidunaktis, viduvasaris, vidužiemis, vidupievis (vidulavkis arba vidulavke15), drutgalis, laibgalis, tvoragalis, tvorugalis, sienogalis213/27, 213/29, 213/30. Examples with the elements pus- and pusiau- show (and the grammar’s author notes) that the derivational base may be not only the noun pusė, but also the adverb pusiau. The grammar also states the derivational meaning of this type: it denotes half of the thing named by the second component—pusęą rito, pusęą berno213/20. Of course, this is not a very precise generalization; the actual derivational meaning of pusbernio is ‘not yet- 15 The words vidulavkis and vidulavke are struck through in the manuscript. Žodžių daryba K. K. Daukšos gramatikoje | 81 grown to a certain age, pusšašio—‘approximately half of the thing named by the second component’, pusbrolio—‘a child of one’s father’s or mother’s sister or brother’, and so on. Two examples first entered in the section on compound adjectives and later crossed out may also be mentioned: pus’žmogis, pus’ponis214/17 (compound adjectives with the element pus- are discussed in 4.2.3). Their derivational meaning is much more generalized and varies less; perhaps for that reason, the corresponding variety of meanings among compound nouns was not noticed. It is also worth noting adjectives formed from pus- and a noun, since today they have become strongly nominalized—for more on this, see 4.2.3. DG already recognizes a difference in the derivational meaning of compounds with vid-: vidudienis, vidunak-tis, viduvasaris, vidužiemis mean ‘the middle of the day, night, summer, winter’, that is, ‘the middle of the period of time named by the second component of the compound’. vidupievės does not belong to this semantic group—it means “situated among meadows between fields”“213/25. The meaning is similarly described in the dictionary: ‘a meadow situated among fields or arable land’ (LKŽXIX 96). It is worth noting that the grammar’s author crossed out vidulaukis and vidulaukė from this group of compounds. And indeed, vidulaukis does not have the meaning ‘being situated between something’. Compounds with the root -gal- form two groups, distinguished by the first component of the compound—an adjective or a noun. The following examples are given of compounds formed from adjectives: drūtgailis ir laibgalis. The grammar states that such compounds cannot be formed from every adjective, but only from an adjective atžvalg(i)ngo? (względnego)213/26. With only two examples, it is difficult to say now what the author had in mind, since it is clear that he himself was unsure about the term atžvalg(i)ngas, and therefore even supplied a Polish equivalent. Today we understand this Polish term as santykinis (Vaitkevičiūtė 1994: 924). Thus, Daukša argued that compounds with galas are formed only from relational adjectives. Of course, not all relational adjectives form them (for example, gražus + galas, greitas + galas do not combine), but the examples given do not contradict this. In addition, the base phrases drūtasaj galas, lajbasaj galas213/27 are also provided, showing that the base adjectives are definite. Another group consists of compounds whose first component is a noun: tvoragalis, tvorugalis, sienojgalis. The derivational meanings of the examples given differ considerably: tvoragalis means ‘the end or a piece of a fence’213/29, whereas tvorugalis means ‘the end of a fence, where the fence ends’213/30, and sienojgalis means ‘the end of a fence or a piece, and a small fence’213/31. The author’s statement also indicates a generalization of meanings: all “equivalent forms” have the same meaning. Thus, Daukša considers that in the case of tvoragalio / tvorugalio it is specifically the linking vowel that causes the difference in meaning. It is not known whether this is merely a dialect feature, since no further material on such cases could be found. 3.3. Prefixed nouns 3.3.1. Prefixed deverbal nouns are mentioned in the section on compound nouns. In the author's view, they are compounds consisting of a prefix and a verb: priedas, apirašas, atarašas, ažurašas, įrašas, išrašas, parašas, perrašas, prirašas, surašas, užrašas213/3 As discussed in 2.3, here derivatives of prefixed verbs formed with endings are conflated with prefixed nouns, which can in fact be considered compounds. This classification was also prompted by the fact that derivatives formed with endings are not mentioned at all anywhere in the grammar. More precisely, the author does not discuss word formation here. He simply states the fact, 82 | ANTANAS SMETONA that “derivatives from prefixed verbs” inflect like non-prefixed nouns of the corresponding declensions (išėjungtinė nu žodž’ų sudetin’u [...] atsimaino kaip n’ėa sudeti₂₁₃/₂). Therefore, the appearance of these examples in the section on compounds would suggest that the grammar's author did not distinguish between the formation types apirašas and apyrankė. The same section dispels these doubts: prefixed nouns formed from a preposition and a noun are also mentioned separately: pamiškė, palavkė, paupė; apinak­tis, apižiemis; priešienis₂₁₃/₃₄. The derivational meaning of these formations is also given: žink­lina artumų sudetin’o prie n’ėasudetin’uj egsunt’0₂₁₃/₃₂ – the formation indicates proximity to the entity denoted by the base word. The section on compound nouns also mentions the prefixed noun nūsavistė₂₁₃/₃₆. It is stated to be formed from a preposition and a pronoun. This formation certainly belongs in the derivational family of possessive pronouns, but because of the complex derivational chain, the grammar's author was unable to determine it: this formation is also mentioned in the section on suffixed nouns (see 3.1.10). 3.3.2. More about the formation of prefixed nouns is discussed in the preposition section of the manuscript grammar (see 2.4). Although most of it is devoted to prefixed verbs, it contains many examples from other parts of speech, including prefixed nouns. NOUNS WITH THE PREFIX AT- and all its variants are few in number. Verbs are most often formed with this prefix (see 5.2.7), and the use of the prefix variants is explained. The author states that the prefix ato- is used with nouns derived from verbs, within the noun class itself, and sometimes with adjectives: atož’ura, atotvanas₁₃₃/₄. No further examples are provided, but the author himself uses atožvalga₁₅₉/₃₄ several more times and atomajna₁₅₀/₁₉ once (this may also be a proofreading error, since atamaina is used everywhere else). In the grammar author's language, only one noun with the prefix at-: atmaina ‘paradigm, declension, conjugation, ending’ is used – atamajnos, atmajnas, atomajna. THE PREFIXES PA- AND PO- are also mentioned in the preposition section. It is explained there that these prefixes (prepositions in the grammar's terminology) are most often added to verbs and much less often to nouns and adjectives. No verb examples are given at all; only the less common usage is illustrated. The following nouns are given: poil’sis, posunis, povajkis, podukra₁₃₃/₄. The use of these prefixes is closely associated with nouns and adjectives, as well as with verbs formed systematically, although examples of the latter are least numerous. In the nineteenth-century grammar, prefixed derivatives are not yet systematized by parts of speech; moreover, perhaps this was impossible to do, since, for example, adjectives are also presented with the same prefix variants: padidis, pabaltis and so on. Therefore, for a fuller picture, it is more appropriate to discuss all derivatives with the prefix pa- in a single section. This is done in the section on verbal prefixation, 5.2.3. THERE ARE ONLY A COUPLE OF NOUNS WITH THE PREFIXES UŽ-, AŽ- (ažurašas, užrašas₂₁₃/₃), and these are formed from prefixed verbs; therefore, their characteristics and differences are discussed in detail in section 5.2.8 on verbal prefixation. There are also only a couple of nouns with the prefix į-: islenis₁₃₂/₁₁ (: valley), irašas₂₁₃/₃ (: to write in). The grammar's author rarely uses derivatives with this prefix and discusses them rather briefly- ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 83 …ria. The issues surrounding derivatives with this prefix are revealed more clearly through prefixed verb derivatives, so it is discussed in the section on verb formation, 5.2.4. At first glance, the PREPOSITION AND PREFIX PAIR WITH SU AND SAN- (SĄ-) formed in the prepositions section seems entirely unjustified. This was apparently prompted both by the altered vocalism of the prefix są-, san- (pronounced su-, sun- in the northern Panevėžys dialect), and by the variants of the preposition su known to the grammar’s author (sa in some Western Aukštaitian dialects and in Latvian)—that is, because of the purely phonetic similarity, he may have equated this preposition with the prefix. The derivatives with the prefix są- in the grammar are: suntarmeₐ, suņdara, suņžineₐ₁₃₃/₅, sužineₐ₁₂₄/₁₆, sunkalbinis₁₄₃/₃₁, surašas₂₁₃/₄; derivatives with the prefix su- are: sustatimas₁₁₂/₃₆, subalʼseₐ₁₂₁/₁₂, as well as two verbs—sudėtas₁₃₅/₁₆, sudūti₂₁₄/₃₅ (for the formation of verbs with są- and su-, see 5.2.5). Two derivatives are open to question. Surašas may be a suffixal derivative of the prefixed verb surašyti (cf. aprašas from aprašyti), although the grammar’s spelling does not preclude interpreting surašas as sąrašas. Subalsę can only be understood as a derivative with są-—sąbalsę—because there is no corresponding form of the same verb with the prefix su-. All these examples nevertheless suggest that, because the prefixes są- and su- had the same vocalism in the northern Panevėžys dialect, speakers may have equated them. 3.4. Nouns—suffixal derivatives 3.4.1. The DG does not identify suffixal derivation, but Daukša nevertheless inevitably has to discuss suffixal derivatives when explaining The formation of the feminine from the masculine (Apie tajsimų gimties moteriškos iš vyriškos₁₇₁/₂₈). Two word groups are presented here. The first group is described as follows: feminine forms are made from masculine nouns derived from other parts of speech by changing the ending -us to -ė: undęₐneₐ (: undęₐnʼus), tiknėₐ (: tiknʼus), kitrėₐ (: kitrus)₁₇₂/₁₆. This is already a presentation of inflectional word formation. Although the words tiknius and kitrus can also be used as adjectives, we must rely on the grammar’s author’s statement that they are nouns, since without examples there is little else one can infer. The second group is characterized even more briefly: a feminine ethnonym is formed from a masculine ethnonym by changing the ending -as to -ė: Lietuvʼnikeₐ (: Lietuvʼnikas), Rusėₐ (: Rusas), Lienkėₐ (: Lienkas), tʼcʼekėₐ (: tʼcʼekas), Latvēₐ (: Latvīs), Ingleₐ (: Inglis)₁₇₂/₁₈. It is hard to imagine that this could have been understood differently in the mid-nineteenth century. After all, only the ending is actually changed here. Even today, it is unclear whether this constitutes a clear derivational opposition or whether these are equivalent substantiva mobilia forms, not linked to one another by derivational relations. Judita Džežulskienė’s dissertation states that these pairs should be regarded as “a separate derivational category of persons (animals), encompassing names of agents and bearers of qualities” (Džežulskienė 2001: 5). Thus, it is hardly possible to speak of derivational relations between the members of the pair. This derivational pattern does not seem to have been as usual or as developed in the dialect as it is today. It applied only to personal names. Having presented seven paradigms for the declension of personal names, the author also lists fourteen groups of words that in one way or another do not fit these paradigms: some lack plural forms, others singular forms; some have singular forms declined according to one paradigm and plural forms according to another, and so on. The list of such irregularities includes animal and bird names that belong to one grammatical gender but can designate either males or females (e.g., meška ‘bear’, kiškis ‘hare’). The author 84 | ANTANAS SMETONA does not suggest extending the derivational pattern under discussion or using any suffixes to form both gender forms (e.g., meškinas, kiškė), but instead uses the auxiliary words patinas and patytė, and does not even make them agree with the nouns: bajkštus kiškis patinas, bajkštus kiškis patytė; kėbli meška patinas, kėbli meška patytė; pilkas žvirblis patinas, pilkas žvirblis patytė; marga krėgždė patinas, marga krėgždė patytė169/16. 3.4.2. The section on noun formation also gives examples of words derived from interjections: aimana, ajman’a210/18. The Dictionary of the Lithuanian Language states that the interjection ai-man arose from the merging of ai and man (LKŽ, 38), so it is entirely possible that such words are classified as compounds in the DG. However, aimana is listed in the section on suffixal nouns, even though this word has no suffix. Nouns of this type are in fact inflectional derivatives from verbs; cf. rauda (: raudoti), malda (: maldauti), etc. In the sections on verb and adjective formation, the interjection aiman again serves as a derivational base (see 5.1.4)—aimanuoti and aimaningas. Indeed, the interjection is the central word in this word-formation nest, but first a verb is formed from the interjection, and then all other possible derivatives are formed from that verb. For example, derivatives with the suffix -ingas are formed from a noun: aimaningas (: aimana). The grammar gives two variants of the noun aimana, with stems of the ā and iā types: aimana and aimania. This probably reflects actual usage among Eastern Aukštaitians of the period. 4. ADJECTIVE FORMATION In his grammar, Daukša devoted two sections to adjective formation, just as he did to noun formation: one on suffixal adjectives and one on compound adjectives. Prefixed adjectives appear in two sections—compound adjectives and prepositions. The presentation of adjective formation differs neither from that of verb formation nor from that of noun formation. 4.1. Suffixal adjectives 4.1.1. The formation of suffixal adjectives from other parts of speech is described separately: Apie tajsimu pripultininku iš kitu dal’u kalbos210/19. It is stated that adjectives are formed from nouns, adjectives, pronouns, participles, verbs, adverbs, prepositions, and interjections, and examples of this formation are provided without comment. Some suffixal adjectives are also mentioned in the section on the suffixal formation of nouns (see 3.1.1) and constitute the fourth type of diminutives. This is entirely understandable, since even in the list of parts of speech, nouns are not distinguished from adjectives, and both are called names—a feature of the Latin grammatical tradition. 4.1.2. As in the section on the formation of nouns, the discussion of suffixal adjective formation begins with adjectives bearing diminutive suffixes: baltutis, baltutė, baltutelis, baltutelėa (: baltas, balta), nėdidutis, nėdidutėa, nėdidutelis, nėdidutelėa (: nėdidis, nėdidė), gražutis, gražutėa, gražutelis, gražutelėa (: gražus, graži)209/23. This type of formation is presented very briefly, without comment. It is stated that the adjective endings -as, -is, -us, -a, -i, -ė209/22 are changed to the suffixes -utis and -utelis (-utis, -utėa, -utelis, -utelėa209/22). The derivatives are entirely systematic and clear. The suffix -utelė is considered a single independent suffix, rather than a combination of two suffixes. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 85 4.1.3. The parts of speech from which adjectives are formed are then listed. From nouns: iš brol’o – broliškas, brolinis; from adjectives: iš balto – balsganas, balsvas ir visi susmulkinti – baltutis, baltutėlis210/20; from pronouns: iš manies, tavies, savies, musu, jusu eina: maniškas, taviškas, saviškas, musiškas, jusiškas210/21; from participles: iš pirkto – pirktinis, rašito – rašitinis16; from verbs: iš duoti – davingas, davus, dūšlingas, dūšlus, dūsningas, dūs(n)us210/25,26; from adverbs: iš toli – totus, tol’imas; iš unks(t)i – unks(t)us, unks(t)ivas; iš dabar – dabarikšt’čas, dabarikštis210/28; from prepositions: iš prieš – priešingas; iš lig – ligus210/30; from interjections: iš ajman – ajmaningas, ajmanus210/31. The examples provided also give an idea of the author's theoretical thinking. Of course, derivatives presented without sentences or word combinations do not allow their meaning to be understood precisely. The formation of adjectives from nouns, adjectives, and pronouns raises no doubts or problems. The examples of formation from adjectives are divided into two groups: ordinary derivatives (balsganas, balsvas) and diminutives (baltutis, baltutėlis). Daukša considers derivatives maniškas, taviškas, saviškas, broliškas, and others, to be adjectives. This is entirely regular, since only adjectives have the suffix -iškas (cf. žemiškas, žmogiškas, pieniškas, and others). Very similar derivatives maniškis, taviškis, saviškis, and others, are classified as nouns in the grammar (see 3.1.10). Although in fact derivatives with the suffix -iškis could also be adjectives (cf. vilniškis laikraštis), maniškis, taviškis are most often used as nouns, or may even have already become fully nominalized. The examples chosen for the formation of adjectives from interjections are unsuitable—it is clear that aimanus and aimaningas are genuine members of the derivational family based on the interjection aiman, but they are not formed directly from the interjection; rather, they are secondary derivatives formed from the corresponding noun—see 3.4.2 and 5.1.4. 4.1.4. Daukša described the formation of adjectives from participles—the suffixal derivatives pirktinis, rašytinis. Lietuvių kalbos gramatikoje (Ulvdas, ed., 1965: 571) proposes that such derivatives can be linked to the infinitive and analyzed as having the suffix -tinis. Nevertheless, this reveals Daukša's insight, which enabled him to discern in the derivative both a relative adjective meaning and the meaning of a passive participle: even today, there are scholars who believe that pirktinis, austinis, dirbtinis and similar adjectives are formed from the corresponding passive past participles (pirtkas, austas, dirbtas) with the suffix -inis (Paulauskienė 1994: 183). The grammar's author had also written in a pair here, rastas – rastinis, but for some unknown reason crossed it out. 4.1.5. The derivational family based on duoti is quite complex. In general, the formation of adjectives from verbs is common, but the examples in the derivational family are presented as equivalent, without regard to the hierarchy among the derivatives: davingas, davus, dūslingas, dūslus, dūsningas, dūsnus (: dūti)210/26. Only two derivatives here can be formed from duoti: dūsnus and dūslus. Only davus can be formed from the past tense, and it is an ending derivative, not a suffixal one. The base forms of the verb are not established in this family, although elsewhere Daukša clearly distinguishes derivation from different principal verb forms—see 2.7. Derivatives with the suffix -ingas are formed from the corresponding adjectives. The meanings of some derivatives are very unclear, so there is even reason to doubt whether they are attested. Davus, dūsnus and dūslus are synonyms (LKŽ₁₁ 343). There is no doubt about the synonymy of dūsnus and dūslus, their suffixal variants, or their meanings, since in the dictionary pa- 16 Here the following was crossed out: iš rasto – rastinis. 86 | ANTANAS SMETONA sufficient examples are provided (LKŽII 614). However, beside davus there is no example, only sources are indicated—Jurgis Šlapelis and Ambraziejus Kašarauskas. One fact, however, casts doubt on the reliability of these sources: both had the DG and the dictionary in their hands and copied examples from them (see Būga 1923: 344–345; Balčikonis 1955); consequently, these are not two different sources, but essentially one unchecked source is being relied upon, and it may contain forms that are not real. For the same reason, doubts also arise about the noun kadžia, kadė ‘time, hour’ (see 3.1.12). Another rather complicated issue is the spelling of the root vowel (diphthong?): dūti, dūšlingas, dūslus, dūšningas, dūs-nus. This polyphonetic grapheme could be understood in as many as three ways. First, it is indeed a polyphonetic grapheme, meaning that the grammar’s author heard various sounds in different dialects, e.g.: duoti, duti, dūti (cf. Zinkevičius 1966: 84–88); this is also supported by the grammar’s statement: in some places, Aukštaitians pronounce the diphthong intact, whereas Samogitians pronounce only the first vowel or reverse the order of both sounds and pronounce ou (ov) instead of uo114/35. Incidentally, we see that Daukša explains the alternation of the diphthong uo as a difference between Aukštaitian and Samogitian, although this diphthong is also monophthongized in the eastern Aukštaitian dialects themselves. Second, this may be a matter not of polyphonetic but of morphological spelling: the grammar’s author heard no pronunciation variants, but wrote according to the base word; one is required to write avti, gavti, mavti, because ave, gavo, move116/37. Third, the author heard Samogitians diphthongizing the long root vowel in the words dosnus, doslus, dosningas, doslingas and therefore wrote the polyphonetic grapheme ū; however, this is the least likely explanation, since when discussing the pronunciation of the vowel o, the author says nothing about its diphthongization. 4.1.6. Adjective formation from adverbs: tolus, tol’imas (: toli), ūnkstus, ūnkstivas (: ūnksti), dabarikšt’as, dabarikštis (: dabar)210/28. Adjective formation from adverbs is again illustrated with adverbs toli, anksti. This is discussed further in 3.1.11 and 6.2. If the direction of derivation toli – tolus and anksti – ankstus is not considered, everything here is understandable. The suffixal derivatives tolimas and ankstyvas are, of course, formed not from adverbs but from the corresponding adjectives. The formation of the derivatives dabarykščias and dabarykštis from an adverb is not in doubt. Derivatives with different nominative singular endings are presented in alternating order. It seems that the derivative with a nominal ending was used more often in Eastern Aukštaitian, since it is listed first. 4.1.7. Adjective formation from prepositions: priešingas (: prieš), ligus (: lig)210/30. A nominal word cannot be formed from prepositions, just as it cannot be formed from conjunctions (see 3.1.12). Prepositions have no root to which derivational formants could attach. Consequently, examples of this type of formation are incorrect. Priešingas could be formed from the adverb priešais (there was also the adjective priešus) or the noun priešas, as are many derivatives with the suffix -ingas (Ambrazas, ed., 1997: 198), whereas lygus should be considered a simple word altogether. Thus, the pattern for adjective formation from prepositions remains without examples and should be regarded as an error on the part of the grammar’s author. 4.2. Compound adjectives 4.2.1. The section on compound adjectives is considerably shorter than the section on compound nouns. Together with the section on compound verbs, it fits on a single page and ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE 87 titled On Compound Adjectives (Apie pripultinikus sudėtin’ūs214/1). Compound adjectives are classified according to their derivational base; brief explanations are given for some of them: adjectives formed from two nouns – linaziedis, šunžardis214/2; from an adjective and a noun – žmogus baltaplaukis or baltaplaukis214/6; from a pronoun and a noun – kasdienis, kasdieninis, kasmetinis214/9; from a noun and a verb – šalnakunis214/11; from two adjectives – judmargis, judberis, žal’margis214/18; from an adjective and a participle – Visogalis214/21; from an adverb and a participle – davgalis214/22. The noun pusė is mentioned as a separate compounding component, with which nouns and adjectives can be combined, e.g.: pus’ūnžis, pus’bal̃tis214/15, 214/16. 4.2.2. Many derivatives raise no doubts or questions. An attempt is made to explain the derivational meaning of some derivatives, as with compound nouns: judmargis, judberis, žal’margis denote two things mixed together in one place, as these are expressed in the examples – two colors214/18; baltaplaukis or baltaplaukis denotes this: a person having white hair or a person with white hair or a person of white hair214/6. However, what is explained here is not the derivational meaning but the individual lexical meaning; moreover, synonymous syntactic equivalents are provided – a case form, a prepositional construction, and a phrase with a detached participial modifier; linaziedis, šunžardis denote adjectives [būdvardį] having the properties of those nouns214/2. The meaning of the other groups of compound adjectives is not explained; it is merely stated from which parts of speech they are formed. The grammar states that the derivatives judmargis, judberis, žal’margis are formed from two adjectives. Indeed, the roots of both are adjectival. If Daukša had tried to find a syntactic equivalent (as in the case of compound nouns), it might have become clear that this is a derivative of an adjectival adverb and an adjective – juodai margas, juodai bėras, žalial! margas (not žalai – this is shown by the softness mark next to l). The Academic Grammar of the Lithuanian Language does not recognize an adverb in the first constituent, but instead claims that there are two adjective stems (Ulvy das, ed., 1965: 600). Some other sources (Paulauskienė 1994: 197; Ambrazas, ed., 1997: 234) take the same view as Daukša, although in fact the structure can vary: juodmargis, juodberis, žal’margis are indeed formed from an adverb and an adjective, but tamsžalis may also derive from tamsus žalias, since such a combination is entirely possible in the language: I bought a dark green suit. 4.2.3. Another type of compound adjective formation involves a component of nominal origin, pus-. Compound nouns are also formed with this prefixoid (see 3.2.4.). The grammar clearly distinguishes two derivational types according to the second base constituent (an adjective or a noun). The first group of derivatives strongly resembles compound nouns: pus’ūnžis, pus’gal’vis, pus’bernis, pus’kvortis214/15. Especially pusbernis and puskvortis. Pusbernis was even included in the section on compound nouns. Daukša explains the meaning of these derivatives and convincingly shows that they can be understood as adjectives: having half of all that: the age, head, and qualities of a young man214/15, i.e. having half the age, half the head, and half the qualities of a young man. In certain circumstances all these derivatives can be used substantivally, and may even become fully lexicalized as nouns. Unfortunately, old grammars contain very few examples, especially sentences or phrases. Yet the classes of these derivatives can only be discussed in the clear context of phrases (e.g.: the past half-century was peaceful – noun; the middle-aged man was peaceful – adjective; only a middle-aged person can understand and do this – an adjective used substantivally; or an example with the word, 88 ANTANAS SMETONA taken from DG – kopūsto pusgalvis; pusgalvis žmogus; ko tas pusgalvis nori). Thus, theoretically, it is entirely possible that the compounds mentioned here may be nouns, adjectives, or adjectives used substantivally; however, the reader can only rely on the grammar’s author and believe that he knows the adjectives mentioned here—the grammar’s material does not allow this to be verified. Incidentally, pušžmogis and pusponis, which were included in this group but later deleted, show that the grammar’s author was also uncertain and may have found it difficult to determine the part of speech, since in actual usage many adjectives formed in this way are used as nouns more often than as adjectives. All the claims made in this paragraph also apply to another very similar group of words: the prefixed adjectives called compounds in the grammar, formed from the function word be and a noun: bęprotis, bęgalvis, bępirštis214/12 (they are discussed in the section on prefixed adjectives). Another group of compounds with pus- consists of compounds with an adjective as the second base constituent: pus’bal­tis, pus’judis, pus’grąžtis, pus’girtis, pus’žabalis, pus’kurtis – having half the quality expressed by the base: whiteness, agility, toughness214/16. The grammar’s author describes the meaning succinctly and precisely: having half of the quality expressed by the second base word. Here one can only note that the softness of the final root consonant in the first constituent is marked very systematically—the quality of the vowel following it is disregarded: pus’greitis, pus’girtis – in fact, s may be soft17, but pus’baltis, pus’žabalis, pus’kurtis should not have a soft s before back vowels. Given Daukša’s tendency toward morphological rather than phonetic spelling, it is entirely possible that here he simply marks softness according to the softness of the base word: pusė, pus’av (or perhaps the softness here actually derives from the base word). It is also possible that the apostrophe indicates that the compounded components have not fused completely. This section does not mention the base adverb pusiau, although compounding this adverb with another word is entirely possible in an adjectival compound. 4.3. Adjectives with prefixes 4.3.1. In DG, adjectives with prefixes are presented in the section on compound adjectives, since the theoretical assumption is adopted (as noted in 2.3) that different parts of speech can be compounded with a function word to form a compound. An attempt is made to characterize the derivational meaning: bęprotis, bęgalvis, bępirštis, i.e. having no mind, head, or finger, or ‘not having the thing denoted by the base word’ (see also 5.2.10); apigrąžtis, pagrąžtis, apidis, apisal’dis, padidis, pasal’dis, etc., indicate a little of that property214/23—that is, they indicate part or a small degree of the property expressed by the base word; pailgas, pergaras214/26 and nūsavis214/28—their meaning is not explained. In fact, the adjectives beprotis, begalvis can be regarded as compounds, since they have a clear derivational base, a prepositional construction—be proto, be galvos. However, apygeris remains questionable even today: the ending has changed, there is a prefix, but the derivational base is difficult to establish. Perhaps it is only understood, not expressed (apygeris—more or less good, about good), but the grammar consistently maintains that adjectives of this type are regularly formed from all qualitative adjectives: apygeris (: geras). 17Incidentally, here we see a disregard for the phonetic patterns of the language: the consonants k, g in a consonant cluster softened by a front vowel block the palatalization of s. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE 89 However, Daukša does not lump them all together in one place. Although the section is entitled On compound adjectives (Apie pripultinikus sudėtin’us214/1), its examples are grouped according to their different derivational bases: adjective + adjective (žalmarģis), noun + noun (linažiedis), preposition + noun (begalvis), preposition + adjective (apygeris), and so on. Some prefixed adjectives merit separate mention. First, pagrájtis, padidis and pasal’dis, with the prefix pa-. The adjective pailgas is presented separately because its derivation and meaning are clearly distinguished as somewhat different. The Grammar of Contemporary Lithuanian does not generally consider the adjective pailgas a derived word from the perspective of the contemporary language (Ambrazas, ed., 1997: 226), but when compared with padidis, it could also be a prefixal derivative—at least, Daukša may have thought so. THE PREFIXES PA- AND PO- are also mentioned in the section on prepositions. It is explained there that they are most often attached to verbs and much less frequently to nouns and adjectives. This section gives no examples of verbs at all, and only one adjective: pabalítis132/15. The use and relationships of these prefixes are discussed in greater detail in 5.2.3. At first glance, pergeras seems unjustifiably classified as a derived form. Today, per geras is treated as two words. But Daukša’s reasoning for treating it as one word can also be understood. If we have geras, gerutis, apygeris (various degrees of goodness), then an excess of goodness can also be marked by a prefix, especially since in DG the prefix is considered the same linguistic element, written separately in some cases and together in others. Here pergeras is understood like pogeris, that is, as an element of the same derivational and semantic class. There are other examples of this kind: potrętʿcʿaj133/21—po trętʿcʿaj134/8. In the section on prepositions, the author recognizes that elements of the same class are, in some cases, true prepositions and, in others, clitics. From the perspective of modern linguistics, this is not far from the truth—we must remember that the endings -p and -pi of postpositional locatives are also included alongside them, so the concept of clitics may encompass all of this. Nor can we rule out the possibility that this spelling reflects not a failure to distinguish prefixes from prepositions, but simply unsettled conventions of writing words together or separately. 4.3.2. More on prefixed adjectives are discussed in the prepositions section of the manuscript grammar (addressed in 2.4). Once again, the meanings of words with the prefix be- are explained. The first meaning is clearly associated with prefixed verbs (besidžiaugiantis), while the second again repeats and discusses adjectives or nouns of the bekojis132/21 type. All this is also linked to the prefixes nebe- and tebe-. Daukša does not divide the issue of words with the prefix be- by parts of speech, but links individual derivational cases so closely that separating them into different parts of speech and discussing them in separate sections would now be inappropriate and perhaps impossible. Therefore, a more detailed analysis of words with the prefix be- is presented in 5.2.10. For the same reasons, the variants of the PREFIXES AP- AND API- (APY-) are discussed in 5.2.6. Daukša used neither the long y nor, in essence, any orthographic marks to distinguish vowel length. Therefore, only two variants are discussed: ap- and api-. The grammar author's claim that only api- is used with adjectives probably refers to the prefix apy-: apibaltis, apibaltė, apididis, apididė, apig’aris, apig’area, apijūdis, apijūdė133/12. Daukša states that verbs are most often formed with the PREFIX AT-, but that nouns and adjectives can also be formed (see 5.2.7). However, examples of adjectives 90 | ANTANAS SMETONA he does not provide, although he himself uses once the adverb atžagar’ą114/37, regularly formed from an adjective (: atžagarus). 5. VERB FORMATION The sections on verb formation are not as clear as those on nouns and adjectives; moreover, they are much shorter. Modern grammars identify two ways of forming verbs: by prefixation and suffixation. Prefixation is far more flexible than suffixation. Each verb has many meanings, and prefixes are attached to the verb as used in each meaning (and not one at a time, but several; often all the prefixes can be attached to the same root). Daukša regarded a prefix as a preposition in every case and its addition as compounding; consequently, compound verbs appear in his grammar. Suffixal verb formation is somewhat clearer. The formation of verb forms and the formation of new words are often confused. For all these reasons, the account of verb formation is rather disorganized and contradictory. Yet, if we disregard the confusion of terminology and word-formation phenomena, as well as the distinctive interpretation of compounding and prefixation, and isolate the strictly word-formation-related material, we can arrive at a fairly clear picture of verb formation. 5.1. Suffixal verbs 5.1.1. Suffixal verb formation is described using the same method as for nominals: attention is paid to the categorical meaning of the suffix, and the parts of speech from which the verb is formed are identified. Verb formation by suffixation could also be described according to suffixes, since the same suffix, e.g. -auti, can form verbs from adjectives—gudrauti (: gudrus)—and from verbs—šūkauti (: šaukti). However, Daukša can be understood as a grammar author striving for consistency: 1) the suffix indicates the part of speech to which a word belongs; 2) the suffix is the principal means by which words belonging to one part of speech are formed from words belonging to other parts of speech. Accordingly, the grammar states that verbs are formed from various parts of speech: nouns, adjectives, pronouns, verbs, adverbs, prepositions, and interjections. 5.1.2. In the section on suffixal derivatives, DG gives considerably more examples of FORMATION FROM NOUNS than from other parts of speech, with most attention devoted to the suffix -auti. The examples provided in the grammar illustrate three word-formation patterns (the author himself does not classify them): 1) ‘to gather the things denoted by the base word’: gribavti (: gribę, gribę̇s), ūgauti (: ūga -os), kupolavti (: kupolai)₂₁₁/₂; 2) ‘to perform an action involving the state denoted by the base word’: keḷ’avti (: keḷ’as)₂₁₁/₄; 3) ‘to provide oneself with the thing denoted by the base word, or to prepare that thing’: šienavti (: šienas -no)₂₁₁/₃—the latter may be regarded as a variant of the grybauti type grybauti. Far fewer examples are given for other suffixes that form verbs from nouns. There is only one example with the suffix -oti: med’c’oti (: med’c’as)₂₁₁/₄ ‘to do something in the place named by the base word’. A pair of examples with the suffix -inti: sidabrinti (: sidabras)₂₁₁/₄ ‘to cover something with the thing denoted by the base word’; garbinti (: garbę, garba)₂₁₁/₅ ‘to perform actions intended to bring about the state denoted by the base word’. Some base forms of suffixal verbs are presented somewhat differently ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE 91 — not only in the nominative but also in the genitive. The plural nominative kūpoliai appearing alongside such forms suggests that the plural form was more usual among Eastern Aukštaitians, or perhaps that the noun itself was formerly plural (LKŽ does not give a plural form). Two variants of the noun garbė are also given: garbė and garba. This may reflect the mixing of different stems. Daukša’s examples suggest that Northern Panevėžys dialect speakers used both forms concurrently in the mid-nineteenth century. The section on the formation of compound verbs gives a pair of examples with the fairly productive suffix -(i)uoti. Of course, verbs are not formed by compounding. Both derivatives with the suffix -iuoti that Daukša provides are formed from compound and prefixed nouns: piktažodc'uoti214/30 and pravardc'uoti214/31. By assigning them to the category of compound verbs, the grammar’s author classified them not by their formation but by the morphological structure of the derivatives. 5.1.3. The section on suffixal verbs contains only one word-formation set of verbs derived from verbs: rekavti, rekdti, rekdinti, rekinti, rektereti (: rekti)211/9. There are no comments, but the examples show that there are three meanings here: iterative (rėkauti), causative (rėkdyti, rėkdinti, rėkinti), and momentary (rėkterėti). Moreover, the examples show that -yti, -dyti, -inti have the same meaning and may be considered variants of a single suffix. Of course, judgments about the meanings of words presented without context can be mistaken. Here, however, they are treated as variants of one suffix on the basis of their current meaning and dictionary evidence. Another group of suffixal verbs is discussed in the introduction to the section on verb conjugation, which explains the entire verbal system; however, it also includes several suffixal nouns (see 3.1.7 and 3.1.8) and verbs with a verbal stem. Examples of suffixal verb formation are provided in the subsections on infinitive and simple past stems. Derivatives with the suffix -terėti: šavkterėti ir šukterėti (: šavkti), skustėrėti (: skusti), begterėti (: begti), lavkterėti ir luktėrėti (: lavkti), travkterėti ir truktėrėti (: travktie), spavstėrėti ir spustėrėti (: spavsti), stavgtėrėti ir stugtėrėti (: stavgti)181/11. The grammar’s author draws attention to two groups of verbs: those with the diphthong au in the root (šaukti, laukti) and others (skusti, bėgti). Two word-formation patterns are distinguished accordingly. Only one example each is given for skusti and bėgti, whereas two are given for verbs with au in the root (with and without a change in the root vowel): šaukt erėti ir šukt erėti. Of the nine variants of this suffix, the text itself uses only -terėti. Necessitative and causative verbs with the suffix -dinti (a variant of the suffix -inti): dirbdinti ir dirbditi (: dirbti), s’u dinti ir s’u diti (: s’uti), valgidinti ir valgiditi (: valgyti), migdinti ir migditi (: migti), rikdinti ir rikditi (: rikti)181/23. The examples show that both suffixal forms are used systematically: -dinti and -dyti. This group also repeats the forms of the verb rikdyti from the section on suffixal verb formation. Necessitative participles: vardžodc'aj jiemingi n’apžinklint o lajko, adding -nas, -na, e.g.: dirbtinas, dirbtina181/3. In general, there is still debate as to whether the necessitative participle belongs to the grammatical category of verbs or adjectives (Paulauskienė 1994: 184); Daukša’s position on this issue should therefore also be significant. As can be seen from the sentence just quoted, the necessitative participle is assigned to the class of passive participles as the unmarked member with respect to the category of tense. This theoretical position 92 | ANTANAS SMETONA This also makes it possible to classify the half-participle among the active participles, likewise as a member lacking a tense category. The grammar’s author sees the suffix -nas, -na, although in fact it is the suffix -tinas, -tina. It seemed to him that -nas was added directly to the infinitive (without removing -ti). The author conflates the infinitive formant with the first syllable of the suffix. From this one may infer that, from a theoretical point of view, it had not yet been decided that derivatives formed from the infinitive first lose the infinitive formant, and only then are suffixes and endings added. On the other hand, this authorial approach can also be justified: after all, the infinitive formant is not an ending; in function and form it is closer to a suffix, and the combination of two or even three suffixes is a common phenomenon. This formal approach is followed quite systematically. For instance, derivatives with the suffix -inėti: first it is shown that the past-tense ending together with the stem-final element (-au) is replaced by the suffix -inėti – arinėjau (: ariau)18. When -in- appears before the stem-final element, the rule is immediately corrected: -au is changed not to -inėti but to -ėti: in verbs ending in -inav, -av is changed to -ėjav, etc. Vadinav – vadinėja v179/11. Other examples of derivatives with this suffix (without any suffix variants) are: dirbinėjav (: dirbav), davinėjav (: dav’av), arinėja v (: ar’av), valginejav (: valg’av), kalbinejav (: kalbėjav), vaikštinejav (: vaikšt’ojav), ieškinėjav (: ieškojav)179/8. No comments are provided, so the only point that raises doubts is the choice of derivational base: it is clear that kalb-inėjau (: kalb-a, kalbėj-o), vaikšt-inėjau (: vaikšt-o, vaikščio-j-o), iešk-inėjau (: iešk-o, ieškoj-o) are not formed from the past frequentative form. On the other hand, kalbinėti ‘to talk at work’ can also be understood as derived from the past frequentative, since the grammar includes the infinitive kalbti141/28. 5.1.4. The chapter on suffixal verbs also presents DERIVATIVES FROM ADJECTIVES, PRONOUNS, ADVERBS, PREPOSITIONS, AND INTERJECTIONS: bal’ti, bal’kti, bal’diti, baltinti, balti-ti (: from balto)211/6; savinti, prisavinti (: from savies)211/8; davginti (: davg), tolinti (: toli)211/10; priešintiės‘ (: prieš), ligi, liginti, ligintis‘ (: lig)211/11; ajmanuti, ajman’oti (: ajman’)211/12. The examples given allow conclusions only about the derivational patterns being proposed. The same base words are used as examples across all derivational categories, although this is often incorrect; for example, a noun is derived from an interjection—aimana (actually from the verb ajmanuoti); an adjective from an interjection—ajmanus (actually from the noun aimana, see also 3.4.2); and a verb from an interjection—ajmanuoti (this one is correct). Derivation from adverbs is explained in exactly the same way (verbs—dauginti, tolinti; adjective—tolus; nouns—tolumas, daugumas, daugybė), or from prepositions (priešingas, priešintis), even when this is not always correct from the standpoint of derivation. Presenting derivation across different parts of speech in this way is probably more convenient and clearer from a pedagogical-methodological perspective—we should not forget that Daukša was a parish-school teacher. However, this does not reflect the actual hierarchical relationships between derivatives. Neither the adjective nor the verb derivation chapters mention derivation from conjunctions (it was mentioned only in the chapter on noun derivation; see 3.1.12). Many suffixal derivatives from adjectives are conventional: bal’ti, bal’kti, bal’diti, baltinti, baltiti211/6; savinti211/8. Daukša derives the two-syllable verb balti from the adjective baltas. Both these words have a simple structure. Therefore, the grammar author’s recognition of a derivational relationship between these words should be highly valued from a theoretical point of view. After all, when comparing baltas 18The first person is taken as the principal form, following the example of Latin grammars. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 93 and balti no suffix is discernible: the two words have the same structure. But Daukša perceived an internal derivational relationship—a motivation: balti ‘to become white’. Even today, the relationships between such words are not entirely clear (Paulauskienė 1994: 190), since bidirectional motivation is also possible here: baltas ‘having turned white’ and balti ‘to become white’. Thus baltas can be understood as the result of becoming white. It is important that this relationship between words of simple structure was already noted in a grammar from the mid-nineteenth century, and it remains unexplained in full even today. A different case is baltinti, baldinti, savinti. Here the suffix -inti is clear. Doubts arise about the verb prisavinti, which is assigned to this group, since its final derivational step is the addition of a prefix—after all, a prefix cannot first be added to the pronoun savęs, because the categorical meaning of the pronoun does not permit this. It seems that to Daukša prisavinti was clearer and more familiar than savinti. Even today, the reflexive form savintis is used more frequently. 5.2. Prefixed verbs 5.2.1. As in the chapters on compound nouns and adjectives, the chapter on compound verbs also discusses prefixed verbs. Yet it gives not a single example of compound-verb formation—the entire section consists specifically of prefixed verbs. The following examples are presented: 1) prefixed reflexive verbs (in Daukša’s words, formed from a prefix, a reflexive marker, and a verb) apsėjti, apsiz’ureti, pasikalbeti214/32; 2) prefixed verbs apidūti, apdūti, atadūti, ažūdūti, aždūti, idūti, išdūti, nudūti, padūti, pardūti, perdūti, pradūti, pridūti, sudūti, užūdūti, uždūti214/33. One verb, duoti, is given with almost all the prefixes (api-, ap-, ata-, ažu-, až-, į-, iš-, nu-, pa-, par-, per-, pra-, pri-, su-, užu-, už-). Modern Lithuanian has twelve verbal prefixes, while DG has sixteen, since it also presents variants of the same prefix as separate prefixes. The prefixes api- and ap- are regular variants arising for phonetic reasons. The prefix at- should also be included, but is probably omitted here because the chosen example begins with the consonant d. In general, both variants are used elsewhere in the grammar: at- and ata-; no example with ati- was found. The examples show that Daukša pays attention only to word structure. A visible prefix and a nominal or verbal root lead him to conclude that the word is a compound. Reflexive verbs are mentioned separately. For the most part, the examples are formed in entirely regular ways and require no comment. The form pradūti is probably not a real form. Forms with the prefixes až- and už- differ semantically (aždengti – Polish zakryć, uždengti – Polish nakryć), and the grammar’s author discusses them in the chapter on prepositions (see 6.1). 5.2.2. In discussing the derivation of prefixed verbs, the grammar chapter devoted to prepositions merits separate attention (see 2.4). The grammar’s author systematically equates this part of speech with prefixes, so most of this chapter is devoted to derivational matters. As already noted, the chapter attempts to discuss all parts of speech that can take prefixes; consequently, the derivation of prefixed nouns and adjectives is transferred to the relevant sections (nouns—3.3.2, adjectives—4.3.2). In the manuscript grammar’s chapter on prepositions, prefixes are discussed in no particular order. Although the chapter brings together all parts of speech that can take prefixes, the gram- 94 | ANTANAS SMETONA The author of the grammar understands the differences between the prefixal derivation of nouns and verbs. Here is how derivation with certain prefixes is described: prepositions ato, nuo, po, prie, pro, san are most often written with nouns derived from verbs and with compound nouns, but are sometimes also written with adjectives, e.g.: atož’ura, atovanas, poilsis, priešpauda <...> posunis, povaikis, podukra, priekletis, priešienis, produris, prokritis133/1. This statement shows that deverbal prefixal nouns, such as priespauda (: prispausti), are clearly distinguished from what the author of the grammar calls compound nouns, such as priekletis (: prie klėties). Elsewhere, the syntactic equivalents of true prefixal (compound) nouns formed in this way are even explained: bebarzdžis – means the same as neturintis barzdos or be barzdos132/22. The author of the grammar could not yet be entirely consistent. In the section on compound-word formation, he nevertheless calls prefixed verbs compound words, although this does not invalidate the fact that he distinguished the derivation of nouns from prefixed verbs from the derivation of prefixed nouns within the noun class. Treating prefixed words as compounds in the later sections of the grammar (especially those devoted to compound words) also helps avoid a theoretical error that is still often made today: nowhere is it claimed that a derived word can be formed from another part of speech by adding a prefix—the prefix does not change the word’s categorical meaning, after all. For example, the section on compound adjectives states that an adjective can be formed by combining a noun and the preposition be, which is quite different from claiming that an adjective is formed from a noun by adding a prefix. Paulauskienė (1994: 192–193) explains this phenomenon in greater detail. It is mentioned several times that in reflexive derivatives formed with prefixes, the reflexive particle occurs between the prefix and the root. The author of the manuscript grammar understands that individual parts of speech have their own prefixes and that prefixal derivation does not change a word’s categorical meaning. He lists the prefixes used to form verbs only: ap, at, ata, až, da, in, iš, nu, par, per, pro, pri, su, už, užu – dedasi su žod’c’ajs yr visajs išreikš’c’ajs nū žod’c’u išėjunt’c’ajs132/6. This is, of course, a peculiar error in the conception of derivation, since, for example, atvaž̌iavimas is formed not from važiavimas and a prefix, but from atvažiuoti and a suffix. However, Daukša consistently focuses on the structure of words everywhere, rather than on the relationship between a base word and its derivative. Grammars of the period in question did not distinguish between morphemic and derivational analysis. Daukša therefore could not think otherwise. He saw only what is present in the word, and on this basis the same derivative can readily be assigned to words formed by either prefixal or suffixal derivation. The prefix and the preposition are related, so Daukša can claim that verbs, nouns, and even adjectives are formed with the same prefixes: až̌uupis arba až̌uupė, až̌ukalnis arba až̌ukalnė, išlenis, nėbilis, nėgalea, n’apikunta, nėkrikštas, netikėlis,... apiname, paname, apibaltis, pabalts, apibalt’c’aj, pabalt’c’aj132/8;132/14. Such statements and examples raise the question of whether the use of certain prefix variants differs. This question arises not only for readers of the grammar but for the author himself. He then attempts to distinguish the use of some prefix variants in practice: not only on formal orthographic grounds, but also by discerning differences in meaning. The grammar discusses the following prefix variants separately: ap- api-, at- ata-, až- ažu-, už- užu-, par- per-. Par- and per- are, of course, different prefixes, not variants of the same prefix, but they have been placed together with pairs of variants for phonetic reasons— ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE 95 because of phonetic considerations. On the other hand, až- and už- are not regarded as variants and are strictly distinguished. The pairs of prefixes i- and in-, są- and su-, po- and pa- are not discussed at length, but the examples, notes, and mentions elsewhere in the sections on word formation make it clear that they, too, are in some way or other being linked semantically. For this reason, all of them are discussed below. 5.2.3. PO- AND PA- are described relatively thoroughly, although there is no separate subsection devoted to these prefixes. Their usage is defined in two places. Daukša states that po- is most often used with nouns derived from verbs (such as poilsis from pailsėti). Since the author of the grammar considers pa- and po- to be the same prefix, it follows that he is already able to unite allomorphs into a single morpheme (morpheme identification). The phonetic change in the prefix is regular. Such a change is usually an accompanying feature of word-final derivation. The author of the grammar also recognizes a separate derivational pattern based on a syntactic construction (preposition + noun). The nouns posūnis, povaikis are considered compounds. Adjectives are sometimes formed as well, but no examples are given, although later in this subsection we see that padidis, pasaldis are indeed used. The prefix pa- is used with verbs, words and forms derived from them, and sometimes with nouns, adjectives, and adverbs: pa- dedasi su veiksmažodžiais ir visais iš jų padarytais žodžiais, taip pat dedasi ir su daiktavardžiais, būdvardžiais ir prieveiksmiais: panamė, pabal tis pabal tʿaj132/12. No verb examples are given; only rarer forms—a noun and an adjective—are presented. More words with these prefixes appear in the manuscript grammar’s sections on compound nouns (see 3.3.1) and adjectives (see 4.3.1), and they occur throughout the grammar text as well. To clarify what the author of the grammar thought about the prefixes pa- and po-, we can review their usage throughout the grammar: the prefix po- occurs 9 times: povaikis, posūnis, podukra133/6, povaik iš207/1 ‘example’, porit130/16, poligintinʿaj136/16 ‘comparative forms’, poilʿsis133/4, žimės po išrėškines arba poštarinės129/1 ‘sentence punctuation marks’; the prefix pa- is used much more frequently (in several hundred examples, though these are mostly verbs and their forms). The grammar’s section on compound adjectives first gives pagraj tis, padidis and pasalʿdis214/23. Other examples include: pabajgos114/16, paavgo128/18, pabajgtinʿaj142/9, pabal tis132/15, pažisti125/22, pauntrins126/20, pašalijie130/30, pa-pėjktinʿaj138/4, paprastinʿis205/23, pamiškė, palavkė, paupė213/34, panamė132/14, and so on.19 The examples given show that the use of the prefixes pa- and po- does not differ from present-day usage, except for adjectival prefixal adjectives pagraj tis, padidis, pasalʿdis, pabal tis and some prefixal adjectives with the suffix -inis pabajgtinʿaj, papejktinʿaj, paprastinʿis – poligintinʿaj, po išrėškines, poštarinės. In general, Eastern Aukštaitians may render stressed o as a (Zinkevičius 1966: 69), so several inconsistencies can be discerned in DG. First, it is unclear why the polyphonetic form is not used— 19 In the Grammar of Contemporary Lithuanian, in the noun and adjective derivational systems, the prefix pa- is not considered a variant of po-—the two are treated as distinct prefixes (Ambrazas, ed., 1997: 145, 224). This leaves unexplained the change, albeit not entirely regular, of the prefix pa- into the prefix po- in the formation of nouns from verbs: potraukis (: patraukti), postūmis (: pastumti), poilsis (: pailsėti), pokalbis (: pakalbėti), and so on. There are very many such derivatives, perhaps as many as pagalba (: pagelbėti), paroda (: parodyti), pažyma (: pažymėti). The manuscript grammar therefore also adopts the view that the prefixes pa- and po- are phonetic variants; Aldona Paulauskienė likewise does not object to this, regarding these prefixes as phonetic variants because of their identical relationship to the preposition po (Paulauskienė 1994: 96). 96 | ANTANAS SMETONA letter, because the examples show variation in pronunciation. Second, there is no system in the variants of derivatives with the suffix -inis. From this it can be concluded that the author of the grammar either failed to notice or did not understand certain dialectal features. 5.2.4. Į- AND IN-. The author of the grammar established that these prefixes are related, but did not emphasize their semantic identity. The grammar contains the following derivatives with the prefix į-: in the examples – įsikalbėti, įsirišti, įkalbėti, įidūti (there are also a couple of nouns – islėnis, įrašas); in the author's language – įtikti, įvežtu, intiktu, įneštus, imėstos. The following tendency is apparent: the preposition į and its variants in, ing, ieng are mentioned repeatedly in descriptions of prepositions, but are not used at all by the author himself, while the prefix in- is used only once (indeed, both forms occur: įtikti and intikti), and the prefix į- is used 10 more times. The reason is clear: northern Panevėžys speakers systematically replace the preposition į with the preposition ant (Zinkevičius 1966: 416). This is confirmed by the author of the grammar himself: sometimes Aukštaitians too follow the Samogitians, but often use the preposition ant instead of į, e.g., ‘unt miestų ejnu’; however, this expression has a slightly different meaning, since it means ‘at the outskirts, near, close to the town, at the town, by the town’, but not ‘into the town’, because ‘ejnu miestan or ing miestų’ means ‘inside, into the middle of the town’. The prefix ant- is not used even once in the grammar and, as the examples just given show, is not confused with the prefix į-. 5.2.5. SĄ- AND SU-. Most examples with these prefixes are nouns, so they are described in this study in the section on prefixed nouns. The verbs and their forms given in the grammar are: sudetas, sudūti. These examples raise no doubts, but comparison with the noun prefixes shows that, because the vowels of the prefixes są- and su- coincide, northern Panevėžys speakers may also have confused them (see 3.3.2 for more detail). 5.2.6. AP- AND API-. These prefix variants are described very briefly; no semantic difference between them is apparent. First, recommendations are given as to when one form is preferred (ap- is used with reflexive verbs) and when the other is preferred (api- is used primarily with verbs beginning with b, p); it is then explained that in all other cases they are used synonymously (apdėti or apidėti; apdūti or apidūti; apgauti or apigauti; apjükti or apijükti). Exactly the same is said about the prefixes až-, ažu- and už-, užu-. Only ažu- and užu- are recommended with verbs beginning with the consonants š, ž. According to the author, only api- should be used with adjectives. This most likely means that, because of orthographic features, the grammar conflates the different parts-of-speech prefixes api- and apy- (see 4.3.2). 5.2.7. AT- AND ATA-. It appears that no semantic distinction is made between these prefixes. Thus, in the grammar the difference between them is purely an orthographic problem: as many as five cases are specified in which at- should be written, and three in which ata- should be written. At- is written: 1) with reflexive verbs – atsikalbėti, atsidūti; 2) with verbs beginning with vowels – atarti, atarditi, atesti, atęti, atieškoti, atirti; 3) with infinitives of verbs beginning with the consonants b, g, j, k, l, m, n, p, r, s, š, t, v, z, ž, and forms derived from them, except passive past-tense participles – ataneštas, atnešti, atnaujinti, atpešti, atpilti, atrasti, atrašyti, atsakyti, atšaukti, atvežti, atžingti, atžibti, atrasti, atsakiti, atvariti, atžlobinti; atnešti, atpešti, atšaukti, atžėlti; 4) with active present and simple past participles – atnešis, atpe- ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 97 šis₁₃₃,₃; 5) and, in general, in the infinitive more often, at- rather than ata- comes out, especially before n, p, š, s, e.g., atnešti, atpėšti, atpilti, atšavkti₁₃₃/₃₃. Ata- is written: 1) with verbs beginning with the consonants d and t, and their derivatives – atadirbti, atadirbimas, atadėti, atadejimas, atadūti, atadavimas, atatravkti, atatravkimas, atatvėrti, atatvėrimas, though atdirbti -bimas, atdėti -dejimas, atduoti -davimas₁₃₄/₃ are also found (we see that the rule is not strict, apparently based on facts of spoken language; that is, the dialect does not very strictly avoid an awkward consonant cluster); 2) more often with first-conjugation verbs in the present and simple past before the consonants n, p, š, s – ataneša -eₐ, atapeša²⁰ -eₐ, atašavk’a -kea, but one also says atšavk’a, atšavkėa – or, speaking briefly, before š both at- and ata- come out equally often₁₃₄/₆ (essentially, synonymous use of the prefix variants is being asserted); 3) with passive past-tense participles and some other verb forms (there is no need to give examples, since one form is asserted above (with ata-), but immediately afterwards it is acknowledged that another form (with at-) is also used). Thus, we see that the attempt to distinguish and strictly define the spelling of these prefix variants is very unclear. In fact, the presence of a prefix before the root has no connection whatsoever with the grammatical form of the verb, which is indicated by suffixes or endings, while the author of the grammar conflates phonetic matters with grammatical ones, which have no bearing at all on the use and spelling of the prefix variants. From this complicated argumentation, one can say only that Daukša’s dialect used the longer prefix form as well – atanešas, atanešiau, but that the shorter forms were used more often, as is asserted in the fifth case for at- and in all cases for ata-. These rules for the use of at- and ata- do not include the prefix variant ato-. The author himself uses it rarely and mentions only once, separately, that it is written with nouns (see 3.3.2). Another variant of this prefix, ati-, is neither used nor commented on at all. In the language of the author of the grammar himself, there are very few instances of derivatives with the prefix at-, but even these show that the prefix variants are not distinguished systematically in any way. Only a few derivatives with the prefix at- and its variants are used, so they can all be listed: atmėsti – atameta, atametam’o s, atametamunt, atmėsdami, atmėasti, atmėtimas, atmėtus; atimti – atjiemis; atmainyti ‘to change’ – atmainamas, atmainimu, atmaino; atrasti – atrastie, atradimų; atsakyti – atsakiti, atsakunt’c’u. 5.2.8. The relationship between DG prefixes až- AND už- is much more complicated. The author of the grammar strictly warns that confusing these prefixes is a very serious mistake. The semantic motivation for this requirement is as follows: one means ‘to cover’ (zakryć) the eyes of someone standing, while something quite different means ‘to cover’ (nakryć) the eyes of someone lying down₁₃₅/₅; or one means ‘to put something on top of someone/something located above (on top, upward)’ (not from, not off the top), while something quite different means ‘to put something onto someone/something from the side, from the flank (from the beginning, end, middle, from, from below, from above, from the edge or from the edge)’₁₃₅/₂₈. Another motivation is the different government of derivatives with these prefixes. Daukša’s explanations are complicated and confusing because scholarly language and terminology had not yet taken shape, so the distinction is best understood from the examples given. In his view, derivatives with the prefix už- govern the construction with the preposition ant, the accusative, and the illative: užjoti ant kalno, užeiti ant bažnit’os, uždėti peilį ant stalo, užstoti ant kel’ō, uždėti akmenį ant kel’ō, užžengą dangun, uždava sieną viršun₁₃₅/₁₀. The prefix až- de- 20 In the words ataneša and atapeša, the prefix at- was corrected to ata-. 98 | ANTANAS SMETONA … compounds govern the preposition až construction, the accusative and instrumental: ażoti až kalno/-u, ažėti až bažnit’os/-u, aždeti peili až stalo/u, ažstoti až keł’ol-u, ažvariti galvijus až kal-no/-u, až keł’ol-u; <...> ažstoti kedų; ażmegsti akmenimis keł’ų, ažtajisiti žabais pili-mų135/19. Here it is very useful to examine not the author’s examples presented in the grammar, but how he himself uses these prefixes and prepositions: in all, they occur 32 times; the preposition už, užu is not used at all; the preposition až, ažu occurs 4 times (gal’ame skaityti ažu klajdy124/7, ažu toki skaityti131/33 prijemtu ažu ligu140/21 prijemtas ažu pirmų182/3); derivatives with the prefixes až-, ažu- occur 25 times (ażginti149/30, ażgre-simo130/23, ažlajkidami149/17, ažrakintin’ąj136/13, ażsimajno125/10, ażsitil’a114/26, ażtil’a113/26, ažutil’a166/16, ažsitūr’a187/3, ažutur’a174/24, ažudravaimu136/29, ažujiema139/35, ažulajko158/14, ažustoja167/5, the noun ažuvardis129/21); derivatives with the prefixes už-, užu- occur 3 times (užžymėtos128/24 and two adjectives, užgamin’aji137/22, užgamin’ūsie172/17). Thus, Zigmas Zinkevičius’s research also accords with the description of the preposition až, ažu and prefixes až-, ažu- and už-, užu- as used by Eastern Aukštaitians (Zinkevičius 1966: 428). Of course, Daukša’s description reflects the dialect situation of the mid-nineteenth century: besides the meaning ‘on’, už- also covers the illative meaning – užjoti ant kalno, užžengti dangun, užduoti viršun. The two derivatives with už- used by the author of the grammar are systematic as well: užžymėtos refers to accent marks written ant the tops of letters; vardas užgaminis – from užgamas ‘people, kin’, is meant as a people’s name, an ethnic name and is probably associated with užgimęs ant Lietuvos. Two verbal nouns that have both of these prefixes should also be mentioned: ažurašas, užrašas213/3. In accordance with Daukša’s rule, ažurašas should regularly mean ‘heading’, while užrašas probably corresponds to the modern meaning of užrašą. 5.2.9. It seems that PAR- and PER- were juxtaposed probably only because of their superficial resemblance, since in some dialects they coincide: parlaužti – perlaužti, parkirsti – perkirsti. Here, too, an attempt is made to explain the derivational meanings of derivatives with these prefixes, which indicates that the prefixes are not variants of a single prefix; therefore, one can only surmise that Daukša knew about the confusion of these prefixes but clearly distinguished them in his grammar for purposes of standardization. Indeed, we find no warning (and warnings are generally frequent here) that one pronunciation variant or another would be a serious error; from this it can be concluded that Daukša himself clearly distinguished derivatives with these prefixes and had not heard them confused. On the other hand, the spelling of the prefix and preposition per – pe,r – – indicates that it may be pronounced in different ways: both par and per. The grammar distinguishes two derivational meanings of derivatives with the prefix par-: movement toward one’s home and an action with an individual meaning: 1) par is added only to verbs denoting some kind of movement, and denotes the return or bringing back of a person or thing134/19 – parvažiuoti, pareiti, parjoti, parkakti – and it can be formed not with all verbs, but only with verbs of direction; 2) parduoti, išparduoti ką ant turgaus. Four derivational meanings are established for derivatives with the prefix per-: it denotes passage over something [1], through something [2], or doing something excessively [3]134/28, repetition, completion, or redoing [4]134/33 – pe,reiti, pe,rjoti, pe,rvaž’uoti, pe,rbegti, pe,rdauti, pe,rarti, pe,rskajiti, pe,rdirbti134/29, 134/31, 134/34. As already noted, the polyphonetic spelling pe,r does not allow the actual form of the prefix and preposition to be determined. The grammar manuscript also contains errors. The prefix and preposition per is written incorrectly as par only once134/32. From this error one might tentatively infer that the author of the grammar himself used the form per as par, although, as noted, he clearly distinguished the meanings. ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE 99 5.2.10. The meanings of the prefixes BE-, NEBE- AND TEBE- are also discussed separately in the grammar. First of all, the grammatical insight is striking—one that researchers sometimes still lack even today. This refers to the identification of two meanings of derivatives with the prefix be-: first, a purely formal, functional role of be- is recognized—not to change the meaning of the verb and its derivatives, but to move the reflexive marker away from the ending to make inflection easier: Bėa is added to verbs whose meaning it usually does not change, though sometimes lots means do – or – gi is most often added to reflexive verbs132/16. Although the section on prefixes (prepositions) provides no example of such use of be-, this statement is repeated with examples later, in the section on participle inflection: dᶜ’avg’untisis‘ or beᵃsidᶜ’avg’untis; dᶜ’avg’untisi or beᵃsidᶜ’avg’unti; dᶜ’avg’untisi or beᵃsidᶜ’avg’unti207/26. Or another example: dedamas (beᵃsideᵃduntᶜ’as)130/1. Juozas Balčikonis did not discern this function of the prefix be- until much later. Responding to a similar claim in a dissertation, he wrote: “The assertion that ‘be- is even necessary for the reflexive participle form’ is a living heresy. The particle be- gives a certain meaning to the verb.” (Balčikonis 1978: 333). Later, Paulauskienė and Tarvydaitė (1986: 154–156) convincingly demonstrated that the prefix be- nevertheless has the function of moving the reflexive marker away from the inflected ending. Another meaning of derivatives with the prefix be-: beᵃ with nouns denotes lack, or is expressed by the same preposition with the genitive, e.g.: beᵃkojis means the same as having no leg, legs, or without a leg, legs132/21; that is, a derivative with the prefix be- indicates the absence of the thing denoted by the base word and fully corresponds to a syntactic construction with the preposition be. In other words: besidžiaugianti is a prefixed verb, whereas begalvis is a compressed syntactic construction (a compound). The grammar distinguishes the independent prefixes nebe- and tebe-. Their difference in meaning is clearly visible: nebe- means almost the same as jau ne-, while tebe- means almost the same as dar – Nebedarašau, ką neᵃ tiek pat kaip jau n’ėa rašav, tebedarašav, tebebeg’a-ras, ką n’ėa tiek pat’ kaip – do rašav132/26, 132/30. The author of the grammar also recognizes the synonymy of constructions: alongside jau nerašau one can also say jau neberašau, and alongside dar rašau one can also say dar teberašau. In this way, Daukša seems to acknowledge that the meanings of the prefixes nebe- and tebe- are inadequately characterized. 6. ADVERB FORMATION 6.1. As noted in the general overview of word formation (see 2.5), the first group of adverb formation comprises regularly formed adverbs with the formant -ai. In fact, at first the author seems not to recognize this formant, claiming that adverbs are formed from the nominative of adjectives and participles ending in -as by changing -s to -j: g’araj (: g’aras), mil’a-maj (: mil’amas)174/20. This is a formal approach, since the ending is segmented, although in fact an ending cannot be segmented as a morpheme. Once a morpheme is segmented, one automatically moves to the level of non-meaning-bearing linguistic units (phonemes). This is probably done to present the material in the textbook as clearly and memorably as possible. The formation examples of this type given below show that the author did distinguish the adverb formant -aj. It is stated that adverbs are formed from adjectives and participles ending in -is and -us from the genitive, by replacing the endings -o or -aus with the formant -aj: didᶜ’aj (: didis, didᶜ’o), 100 | ANTANAS SMETONA gal’unt’ė aj (: gal’untis, gal’unt’o), nęatikus’aj (: nęatikis, nęatikus’o), plat’c’aj (: platus, plataus), gard’aj (: gardus, gardavs), giedr’aj (: giedrus, giedravs)174/24. It is further specified that the final consonant of the stem is always palatalized, and t, d become č, dž. Several things here are unusual. First, it is proposed that adverbs be formed from the genitive, although it has been explained that stem-final t, d become the affricates č, dž. The reasoning was apparently as follows: the genitive most often already has č, dž, so the rule becomes shorter, clearer, and easier to understand. However, the examples revealed that this change does not occur in the genitive of u-stems (platus, plataus); consequently, it was necessary to clarify the palatalization of the stem-final consonant transferred into the adverb category, and the change of t, d into č, dž. Another claim that seems somewhat unusual at first glance is that adverbs are formed from adjectives and participles ending in -is. In fact, there is only one such adjective—didis—and this can be considered an exception. However, in DG the masculine singular nominative ending of the active voice past simple participle is systematically written -is: netikis, atnešis, atpešis, matis and has become completely identical to the masculine singular nominative ending of the present tense, e.g.: galiuntis, dirbuntis. Adverbs can sometimes also be formed from these participles, if the meaning of the base word allows it, e.g., netikusiai (: netikęs, netikusio), although an adverb formed from these forms is often considered an error today. 6.2. Another group of adverbs mentioned in the section on adverb formation has the ending -i: arti, toli, anksti. It is stated that adjectives are formed from them: tolus, tolimas (: toli), artus, artimas (: arti), unkštus, unkštis (: unkšti)174/29. Considering the derivatives from arti, toli described in other sections of the grammar as well, one may assume that Daukša considers arti, toli, anksti to be base words. This is correct, but arti, toli, anksti are historically derivatives, most likely from adjectives. Moreover, the grammar’s author discusses adjective formation in the section on adverb formation. Although he himself uses the adjectives artus, ankstus, tolus, he does not notice the formation of -umas and -yvas derivatives from adjectives. There is enough material to draw conclusions and identify formation types, but no generalization is provided. In closing, the grammar’s author notes that these adverbs can be further shortened: tol’, art’, unkšt’174/30. In general, formation from the adverbs anksti, toli, arti is mentioned in several places: in the section on noun formation (see 3.1.11), adjective formation (see 4.1.6), and verb formation (see 5.1.4). 6.3. Adverb formation is also discussed, when the opportunity arises, in other parts of the grammar. For example, in discussing the synonymy of the adverb formants -aj and -u: grajtokaj aba grajtoku, tolokaj aba toloku129/29, 129/30. It is not entirely clear here whether these really are adverbs, since functionally they may also be indeclinable adjective forms; cf.: man gerai – man geru – man gera. The grammar’s author himself uses both adverbs and genderless stative forms with the formant -u, corresponding to the formant -a in the present-day standard language: kas ir reg’as buti labai doru157/15—an unquestionably indeclinable adjective form; kadu164/28, tadu158/14, slapt’ų168/22—unquestionable adverbs. Thus, there is no reason to doubt the grammar’s author’s claim that the forms grajtokaj aba grajtoku129/29, tolokaj aba toloku129/30 are synonymous, and that adverbs with the formant -u are a marked eastern Lithuanian feature. On the other hand, it should be remembered that the third group in the adverb formation section of DG includes precisely the genderless stative adjective form šaltu arba šalta175/5. Elsewhere, in a discussion of the inflectional features of nominal words, mention is made of the adverbial use of the instrumental: some of them, though they have all the cases, acquire a different meaning in the “broken” cases, as follows: slapt’ų, šalin (from slaptis, šalis), tolin, artin’, etc. All such nouns can be ŽODŽIŲ DARYBA K. K. DAUKŠOS GRAMATIKOJE | 101 called adverbs...168/23, that is, the meaning of the instrumental (and illative) can change so much that those nominal words even become adverbs. This is a very important observation, demonstrating the diversity of adverb formation. In the older paradigms, the singular instrumental of both genders in DG most often has the ending -u. It is even emphatically stressed that saying sukalba, su rikšte arba su rikšti, su širdc’a arba su širdi; su kumede aba su kumedži, su žole aba su žoly124/ is a serious error—the correct forms are su rikšt’ų, šird’ų, kumedų, žolų, kalbu. The section on adverb formation in DG (although in fact it discusses the formation of forms, not words) also mentions the adverbial formant -yn: šiltėlin (: šiltėlėsnis), šiltin (: šiltėsnis), while in the inflectional paradigms of nominal words this formant is called the illative ending.175/9 7. SUMMARY Kazimieras Kristupas Daukša’s grammar devotes considerably more space to word formation than the Lithuanian grammars written before it. Although word formation is not yet conceived there as a coherent section of grammar, a clear distinction between derivation and composition, and their treatment in separate sections, is already evident. Therefore, word formation in Daukša’s grammar can already be discussed as an independent section of grammar, and the following conclusions can be drawn from how it is understood and presented: • Compared with the grammars of his predecessors and contemporaries, Daukša’s manuscript grammar devotes incomparably more attention to word formation, and does so at a higher theoretical level. • The place of word formation in grammar has not yet been determined. It is conceived in three ways: derivation and composition are named and described in separate sections; some aspects of word formation are described as integral parts of the accounts of parts of speech (the description of verb inflection begins with notes on word formation, while most of the account of prepositions is devoted to word formation); and some aspects of word formation have no place at all and are presented in separate short sections in incidental locations (the formation of pronominal forms and the formation of gender forms). • Suffixal formation is abandoned (or not conceived at all): pure suffixal formation of nouns (stalius) is not mentioned at all, while prefixed derivatives from verbs are assigned to composition. • Composition and derivation are discussed not in general, but by parts of speech; that is, the entirely different word-formation patterns of individual parts of speech are already discerned. • Prefixal formation of nouns is conceived as a type of composition, and this is supported theoretically and with examples. • It is already understood theoretically that different parts of speech have different sets of word-formation formants. • As in all the grammars of his predecessors and contemporaries, research and the interpretation of claims are greatly hindered by the small number of examples and the lack of broader context, without which some phenomena cannot be evaluated at all. • Non-gender-marked adjective and participle forms are quite rightly classified as adverbs. • The sections on word formation also address matters of language standardization: various variants of the word-formation norm are sometimes assessed with justification, while at other times the assessment relies only on linguistic intuition. 102 | ANTANAS SMETONA ŠALTINIAI ČIULDA, J. Trumpi samprotavimai apie žemaičių kalbos gramatikos taisykles. Parengė Giedrius Subačius, Vilnius: Mokslo ir enciklopedijų leidykla, 1993. (Lietuvių atgimimo istorijos studijos, 6.) DAUKŠA, K. K. Trumpa kalbmoakslea li’žuv’o li’etuvįsko li’nkiškaj išguldīta pear Kaz’ų (Kazimier’ų) Kris’u (Krištap’ų) Davkši. Rankraštis, VU rankraštynas, F 119–516. KASAKAUSKIS, K. Grammatyką języka żmudzkiego ułożona przez X. Kalixte Kossakowskiego: Kalbrieda ležuwio žiamaytiszko, Wilno, 1832. KLEIN, D. Gramatica Litvanica, 1653. Cituojama pagal: Pirmoji lietuvių kalbos gramatika: 1653 metai, red. Jonas Kruopas, Vilnius: Valstybinė politinės ir mokslinės literatūros leidykla, 1957. OSTERMEYER, G. Neue littauische Grammatik, Königsberg, 1791. RUHIG, P. Anfangsgründe einer Littauischen Grammatic, Königsberg, 1747. SCHULTZ, T. Compendium Grammaticae Lithvanicae. Cituojama pagal: Sapūno ir Šulco gramatika, spec. red. Bonifacas Stundžia, Vilnius: Mokslo ir enciklopedijų leidybos institutas, 1997. Universitas Linguarum Litvaniae, red. Kazimieras Eigminas, Vilnius: Mokslas, 1981. X. D. K. P. S. Prawidła Języka Litewskiego Czyli Grammatyką Litewską. Cituojama pagal: X. D. K. P. S. lietuvių kalbos gramatika: XIX amžiaus pradžia. Parengė Giedrius Subačius, Vilnius: Mokslo ir enciklopedijų leidybos institutas, 2002. LITERATŪRA AMBRAZAS, V., red., 1997: Dabartinės lietuvių kalbos gramatika, Vilnius: Mokslo ir enciklopedijų leidybos institutas. BALČIKONIS, J. 1955: Surastas kalbos mokslo dokumentas. Tiesa 1955 04 16. BALČIKONIS, J. 1978: Rinktiniai raštai 1, Vilnius: Mokslas. BŪGA, K. 1923: Medžiaga lietuvių kalbos žodynui ir šnektoms tirti. Tauta ir žodis 1, 344–345. DŽEŽULSKIENĖ, J. 2001: Lietuvių kalbos substantiva mobilia (mobilieji daiktavardžiai). Daktaro disertacija, Vilnius. FRAENKEL, E. 1962, 1965: Litauisches etymologisches Wörterbuch 1–2, Heidelberg: Carl Winter, Goettingen: Vandenhoeck & Ruprecht. KEINYS, S. 2000: Dabartinės lietuvių kalbos žodynas, Vilnius: Mokslo ir enciklopedijų leidybos institutas. LKŽ₁–ₓₓ: Lietuvių kalbos žodynas, Vilnius. PAULAUSKIENĖ, A., TARYDAITĖ, D. 1986: Gramatikos normos ir dabartinė vartosena, Kaunas: Šviesa. PAULAUSKIENĖ, A. 1994: Lietuvių kalbos morfologija (Paskaitos lituanistams), Vilnius: Mokslo ir enciklopedijų leidykla. SABALIAUSKAS, A. 1979: Lietuvių kalbos tyrinėjimo istorija iki 1940 m., Vilnius: Mokslas. ULVYDAS, K., red., 1965: Lietuvių kalbos gramatika 1, Vilnius: Mintis. URBUTIS, V. 1978: Žodžių darybos teorija, Vilnius: Mokslas. VAITKEVIČIŪTĖ, V. 1994: Lenkų–lietuvių kalbų žodynas, Vilnius: Enciklopedija. ZINKEVIČIUS, Z. 1966: Lietuvių dialektologija, Vilnius: Mintis. ZINKEVIČIUS, Z. 1994: Lietuvių kalbos dialektologija, Vilnius: Mokslo ir enciklopedijų leidykla. Antanas Smetona Lietuvių kalbos katedra Vilniaus Universitetas Universiteto g. 5, 01122 Vilnius, Lietuva Gauta 2003 09 15