Das unflektierte Partizip des präteritums auf „-us“ im Litauischen

LIETUVIŲ KALBOTYROS KLAUSIMAI, VI LIETUVOS TSR MOKSLŲ AKADEMIJA THE UNINFLECTED PRETERITE PARTICIPLE IN -US IN LITHUANIAN G. BENSE 1. Introduction 1.1 The inventory of forms of the uninflected participles (unfl. part.)1, also called gerunds or verbal adverbs (Lith. pãdalývis; Russ. деепричастие), in the modern Lithuanian literary language comprises a form of the active present participle in -nt, e.g. sùkant, šaũkiant, mỹlint; a form of the active future participle in -siant, e.g. sùksiant, šaũksiant, mylėśiant; a form of the active preterite participle in -us, e.g. sùkus, gỹdžius, mylė́jus; and a form of the active preterite participle II in -davus, e.g. sùkdavus, šaũkdavus, mylė́davus. Reflexive forms of verbs without prefixes are formed with the ending -is, e.g. juõkiantis, juõkusis. With prefixed verbs, the reflexive particle -si- occurs between the prefix and the verbal stem, e.g. nesijuõkiant. nusisùkus. 1.2 With regard to syntactic function, two application types of the uninflected participles can generally be distinguished, clearly differentiated from one another and strikingly opposed to their relatively uniform form. 1.21 The two application types can be briefly characterized as follows: I. Skylant atomams, išsiskiria energija (with the agent of the participial form expressed); Auštant tėvas išvažiavo (without the agent of the participial form expressed). These constructions with the uninflected participle serve to express a secondary action (“accompanying action”) alongside the main action, without grammatical attachment to any constituent of the sentence as a whole. 1 By uninflected participles are meant those nominal verbal forms that take no number or case markers, but are formally differentiated with respect to tense (in contrast to the infinitive) and, to a limited extent, with respect to voice—provided that one may regard the diathesis reflexive: non-reflexive as a manifestation of the opposition active: passive. 191 II. Visi mato (mašīna) atvažiuojant (the so-called accusative construction following the verbs “sentiendi et declarandi”). Such constructions serve to express one action as the object of another action. 1.22 It should be noted that the distribution of the individual forms of the uninflected participle across the two application types is not uniform: while forms of the present, preterite and future occur in the so-called accusative construction, within the first application type one encounters only forms of the present and the preterite2. 1.3 In modern Latvian literary language, constructions with the uninflected participle can be divided into two corresponding application types, e.g. Saulei lēcot, visi putni dzied; Metot harpūnu, dzīvnieks tika ievainots: first application type; Visi redz (mašīnu) atbraucam: second application type. In Latvian, the two different forms of the uninflected participle, those ending in -ot and those in -am/ām, are now, in their essential features, each assigned to one of the two application types (overlap being possible only in the second application type). Latvian has—apart from the innovations in -am/-ām developed from the accusative singular of the present passive participle—only one uninflected active present participle in -ot (=-uot from -ant) and one uninflected active future participle in -šot (=-šuot from -siant). The latter form occurs only in the so-called “modus relativus,” which is structurally close to the second application type3. 1.4 Thus, only the uninflected form of the active present participle (from the -nt stem) and the uninflected form of the active future participle (with the stem-forming element -siant-4) can be regarded as formations common to Lithuanian and Latvian (I shall not discuss here the few Old Prussian attestations). The development of an uninflected active preterite participle (from the -us stem) into a special grammatical form is therefore to be understood, in this sense, as a Lithuanian innovation. 2. On the history of constructions with the uninflected participle in Lithuanian (Baltic) 2.1 The mere fact that the uninflected preterite participle in -us is to be regarded as a Lithuanian innovation indicates that the problem of the origin of this formation cannot readily be identified with the problem of the origin of uninflected participles in Baltic as a whole. There are, however, further reasons that make a separate development of the uninflected -us participles probable. To present the attested— 2 The uninflected active preterite participle II in -davus is in fact almost purely a grammarian’s fiction, since it is practically never encountered. 3 Latvian “reported speech mood.” Ultimately, this construction arose from the so-called accusative construction, although it also shows connections with constructions using inflected participles to express an indefinite statement (e.g. Lithuanian Sako miškas didelis buvęs). 4 The attestations of possible future formations in -sint- can be disregarded here. 192 I may be permitted, in view of the problems that arise, to briefly discuss the relevant findings of my “Investigations into the origin of the indeclinable participle (gerund) in the Baltic languages”5. 2.2 First, it becomes clear that the emergence of indecl. participles, especially in Lithuanian but also in Latvian, cannot by any means be regarded as a single, closed process, but rather as the bringing together of several originally independent morphological and syntactic phenomena. This is demonstrated by the emergence of indecl. participles in -am/-ām in Latvian—a process that only became operative after the beginning of written records. This is probably the most recent phase in the emergence of indecl. participles in the Baltic area. On the basis of the evidence in the oldest Lithuanian texts—both formally and with regard to the distinction between the two types of usage mentioned above—it also follows for Lithuanian that the formation of accusative constructions with indecl. participles represents the most recent phase in the emergence of the indecl. participle as a specific grammatical form6. Thus, observations on constructions with indecl. participles, insofar as they occur as accusative constructions after verbs of “sensing and declaring,” may be disregarded when treating the formal and functional characteristics of the indecl. participle within the first type of usage. 2.3 Within the first type of usage, the use of indecl. participles can be traced back to older construction patterns. Here, corresponding to the distinction between constructions with an expressed agent and those in which the agent of the participial form is not expressed, comparisons may be drawn, on the one hand, with the so-called absolute participial constructions of other Indo-European languages7 and, on the other, with nominal verbal forms occurring independently8. 2.31 In their development within the individual languages, the Lithuanian (and Latvian) constructions are particularly notable in this respect: the verbal (predicative) component of the absolute participial constructions is formally identical to the nominal verbal form occurring independently9; in both positions, the indecl. participle occurs. 5 Inaugural dissertation, Halle, 1961 (typed manuscript). 6 This is essentially a differentiation between accusative constructions with a declined participle and those with an indecl. participle, based on the tendency to construe the action expressed by the participle more strongly as the direct object of another action in the case of the indecl. participle, in contrast to the strongly attributive relation between the declined participle and the nominal accusative object. 7 For example, the Old Indic “locativus absolutus,” the Greek “genitivus absolutus,” the Latin “ablativus absolutus,” and the “dativus absolutus” in Gothic and Slavic. 8 For example, the Old Indic absolutive in -tvā, -(t)ya, the gerund in Slavic, the gérondif in French, and similar formations in other Romance languages. 9 The formal identity corresponds to a specific functional load borne by the indecl. participle in both constructions, as is clear from the relationship between them: they differ not in their position in the sentence as a whole (as an “absolute case” or “subject-related secondary action”), but primarily in whether the agent of the indecl. participle is present or absent. 13. Kalbotyros klausimai VI. 193 2.32 The specific feature of an absolute participial construction is that it represents a fixed combination of a nominal case and an inflected (agreeing) participle, independent of the main verb and every other clause constituent10, which serves to express a primarily temporal determination in the sentence. If a participle functioning as a closer (attributive) modifier of a noun in the locative is added, then, owing to the verbal nature of the participle, the predicate is also further specified11. The relation of this now no longer accidental locative noun–participle combination to the predicate thus becomes closer than is the case with the simple locative (as the case of spatial and temporal location, the “grammatically unbound” adverbial case par excellence). Within the sentence as a whole, a syntactic division arises between a main and a secondary action, each with a different agent. And the temporal (not “local and/or temporal”!) character of this “absolute” participial construction does not result from the lexical and grammatical/categorial meaning of the individual elements, but from the syntactic structure of the sentence construction with two verbal actions, one expressed in the predicate and the other in the participle, which is to be understood not as punctual (simultaneity), but rather as linear (sequence, succession). The absolute participial construction thus receives its fundamentally temporal character not simply from the actualized expressive content of its two components, but from the purely syntactic qualification of the main and secondary actions. The specific determination of a temporal relation in Lithuanian through the obligatory distinction between the temporal values conveyed by the -nt- and -us participles is secondary from this point of view as well (all the more remarkable, then, that Latvian seems not to have this distinction in the absolute participial construction!). The use of active perfect participles in the absolute participial construction, for example in Greek, and of perfect passive participles in Latin likewise represents a later development, oriented toward the verbal system of the individual language in question, quite apart from the situation with the absolute participial constructions in Slavic and Germanic, whose age and origins are problematic12. 2.4 Since the beginning of the written record of Lithuanian and Latvian, an absolute dative (abs. Dat.) has been attested: a combination of a (nominal) dative and an uninflected participial- 10 Vgl. Hermann Hirt, Indogermanische Grammatik, Band VII, Heidelberg, 1937, S. 92. 11 Vgl. Kühner/Gerth, Ausführliche Grammatik der griechischen Sprache. Satzlehre, Teil II⁴, Hannover, 1955, S. 77. 12 Vgl. Rudolf Růžička, Struktur und Echtheit des altslavischen dativus absolutus (in „Zeitschrift für Slawistik“, VI, Berlin, 1961, S. 588—596). 194 form13 of the original consonantal stem of the -nt participles. However, this exclusively dative interpretation may well be secondary, especially since it is now based on the nominal component of the construction. From the case relations in Baltic14, as well as from the attested forms of the uninflected participle (-nt < -nti; alongside -ntie < *-ntei), an absolute participial construction based on a locative case can be inferred for (East) Baltic. This assumption is supported, among other things, by the fact that the dative can never be a temporal case in Baltic. Within this construction, the uninflected present active participle in -nt(i) arose when, because of its special syntactic position, it left the group of inflected participles, while the nominal case, in the course of the emergence of the new Baltic locative, retained its connection with the case system as a dative form, owing to its formal identity (or similarity) with the dative. 2.5 Constructions without an expressed agent (absolutive constructions15) in Lithuanian and Latvian are to be subsumed under the absolute participial construction, notwithstanding the different developments in these two languages16. 3. The position of the uninflected preterite participle in -us within constructions with uninflected participles in Lithuanian 3.1 In terms of the possibility of forming constructions, the uninflected preterite participle in -us is today almost completely equivalent to the uninflected present participle in -nt. The two uninflected participles merely indicate different relative time levels. The present form expresses simultaneity, the preterite form anteriority of the verbal predication contained in the participial form in relation to another utter- 13 The abs. Dat. constructions with an inflected participle of the participial jo-stem prove to be formations lying outside this development, probably hypercorrect. For a different view, see V. Ambrāzas, Дательный самостоятельный в памятниках литовского языка XVI—XVII вв. Автореферат диссертации..., Vilnius, 1958, p. 4. 14 The peculiarities of the Baltic case system that are of interest in our context include, above all, the formation of an East Baltic locative with a postposition, absent in Old Prussian; the presence of so-called residual locative cases such as the illative, allative, and adessive (with clear traces of older locative formations in these “residual cases”; see also A. Laigonaitė, Pašalio vietininkai dabartinėje lietuvių kalboje, „Gramatikos klausimai“ Vilnius, 1957, pp. 21—39); and the behavior of the dative with respect to prepositions. 15 In what follows, I use “absolutive construction” simply as a purely grammatical term for absolute participial constructions without an expressed agent (= constructions with an uninflected participle standing alone) in Lithuanian. 16 Whereas absolutive constructions in Lithuanian generally cannot be related to the subject case (“different subjects”), this is entirely possible with corresponding constructions in Latvian, for example Atgriežoties (or atgriezdamies!) no ekskursijas, jaunieši runāja par gūtajiem iespaidiem (quoted from Gailums/ Vilāns, Latviešu valodas gramatika VIII—XI klasei, Rīgā, 1958, p. 265). 195 Thus, the problem is why in Lithuanian (but not in Latvian) these two temporal stages occur both in the absolute dative and in the absolute construction, which are formally strictly separate means of expression (within this problem, the possible uses of the uninflected participle in the so-called accusative construction are irrelevant). The uninflected participle in Lithuanian denotes a specific, fixed temporal value in both constructions. 3.2 In their expression of temporal values, uninflected participles behave similarly to inflected (active) participles, which, regardless of their syntactic position, indicate solely through their form a particular relative temporal relation (“simultaneity,” “anteriority,” or “posteriority”) between the verbal predication they contain and another (verbally or nominally formulated) predication. 3.21 Given that the uninflected participle and the inflected participle share certain features in the syntactic function “expression of a subordinate action in relation to the main action”17, one might, of course, infer that the uninflected participle also agrees with the inflected participle in expressing a fixed temporal value; compare, for example: eidamas suklupo — nereikia einant suklupti; atėjęs rado visus namie — nelengva atėjus visus rasti namie. It is likely that this agreement results from a secondary convergence. In Lithuanian, the absolute construction, with regard to its relation to the bearer of the main action (“difference of subject”), is determined by the character of the absolute dative, i.e. the absolute participial construction, and is thus distinguished from constructions with inflected participles. Conversely, however, the uninflected participle, in its capacity as a predicative component of the absolute participial construction, also acquires the possibility of expressing a fixed temporal value, though only as an alternative between “simultaneity” and “anteriority.” 3.22 This process of secondary convergence, which leads to the present-day situation, can be illustrated by a simple statistical analysis that I can compile on the basis of the Old Lithuanian examples I have consulted18. Specifically, whereas 210 absolute datives with forms of the 17 Constructions with an uninflected participle and those with an inflected participle, which serve to express a subordinate action in relation to the main action, are syntactically related construction types whose distinguishing feature essentially lies in the position occupied in the sentence as a whole by the respective agent of the verbal predication expressed by the participial form. 18 In addition to the Old Lithuanian texts listed in the bibliography of texts consulted, only Daukša’s catechism (Литовский катихизис Н. Даукши перепечатанном Э. А. Вольтером, С-Петербург, 1886) was used for this statistical analysis. The examples of -norint- in the Old Lithuanian texts cited above are not included in this statistical analysis. 196 formed from the present participial stem, 92 absolute datives contain forms of the preterite participial stem19. By contrast, among the absolute constructions there are 132 combinations with present forms and 101 with preterite forms. This ratio indicates that the standalone uninflected participles, in their distribution of present and preterite forms, correspond to the “connected” inflected (active) participles, whose present and preterite forms each account for approximately the same percentage. 3.3 More specifically, the statistics also show the following: A large proportion of the absolute constructions are phrasal expressions already used with an uninflected participle, such as ißskirus, ißbėmus, pradeius, among others, and parenthetical expressions such as sakant, kalbant, netarent, etc. In the remaining, relatively few absolute constructions (64 in total), preterite forms greatly predominate, for example: ir numirus dubias mūsų Dievui afierawojɔ “after our death he offers our souls to God” (DP 516); Nėssa tarp ławōno žmogaus ir maitós ßunes/ne kokio numirus skiriaus ne rasi “For between the corpse of a human being and the carcass of a dog one finds no difference when (the human being and the dog) have died” (DP 583); Bet priminus Diewa mus per wargus wadinanti pakutospi/weisdekem... “Remembering that...” (BrP II 479); ...adunt žmones mokietu ne tiktay ne ápłavst dáŕbu ižgānimo/del pasipeneimo/bet ir pati walgimu áp taydus skubinties ßwentump dayktump: (Si II 177; the only example of this kind I have found at all in Širvyd!). The following sentences may serve as examples of absolute constructions formed with the uninflected present participle: ...kokis ten’ būwo gałfėjimas Wiespaties Jėsaus ręgint’ teip’ primistų motiną sawą... (DP 174); Musu budas irá/taymiey giwenant/ir sweykays esant/ape Pona Diewa maž ad-boti... (KNSu 238); Tay ptáčiey išguldìt galesis kas čionay trumpay ažu-sirakiną/mokštu duodunt kaliniams/ku turi weykti kaliney (Si I 42/3); ...at-tad ûti dani/tatai est/cziese alba kartais renkant mesława (BrP II 503). 3.4 The observation that preterite forms occur relatively frequently in the absolute constructions of Old Lithuanian texts (in relation to present forms in the absolute construction, or in absolute participial constructions generally) supports the conclusion that these absolute constructions with the uninflected preterite participle represent a construction type that must be particularly close to the original structural model for the use of unflected -us participles. This structural model therefore cannot have been an absolute (i.e. originally attributive) participial construction; consequently, it may have been a participial construction that would conventionally be described as predicative. Thus— 19 V. Ambrazas, loc. cit., p. 7, arrives at a similar assessment; idem, Absoliutinis naudininkas su dalyviu XVI—XVII aa. lietuvių kalbos paminkluose, “Lietuvos TSR Mokslų akademijos darbai,” Serija A 2(5), 1958, p. 152. 197 This provides a fact from which one can start in order to shed some light on the history of the development of the uninflected past participle in -us, found within Baltic only in Lithuanian, which can occur both in the absolute dative and in the absolutive construction. 4. Remarks on the formation of the active present participle and the active past participle in Baltic 4.1 First, some basic features of the formation of the active present participle and the active past participle in Baltic that are relevant to our discussion must be set out. The Baltic active present participle, with the formative element -nt-, is an old Indo-European inheritance; it is formed from the verb’s present stem and exhibits such archaic forms as sant- from the root es-, “be” (cf. Sanskrit sant-/sat-, Greek ōn, etc.), ent- from the root ei-/iā, “go” (with an unexplained ablaut relationship²⁰). In Baltic, the active present participle is clearly marked as a present formation in all its forms. The situation is different with the Baltic active past participle, which, in terms of its mode of formation, is a Baltic innovation. This is supported by the generalization of one ablaut grade of the suffix -ves-/-us-, which forms nominal verbal forms from perfect stems in Old Indian and Greek²¹, as well as by the attachment of this suffix to verbal stems of widely differing types. Both are also found in Slavic. However, Baltic and Slavic agree no further than in the general outlines of this formation. Just as the Baltic preterite is a specifically Baltic innovation, comparable with Slavic or other Indo-European languages only in individual forms of certain verbs, the -us- participle, too, follows its own course in its development and integration into the verbal system. 4.2 Judging from the Baltic evidence, which does not allow one to establish that Baltic had “lost” the Indo-European perfect and the Indo-European aorist²² (one must rather assume that Baltic never truly developed these formations), the suffix -us- was probably not originally a “perfect suffix.” At first, it should be understood in Baltic merely as a means of nominal verbal derivation with its own semantic character. The connection between these us- formations and the Baltic preterite is, of course, obvious. The ²⁰ Siehe J. O t r ę b s k i, Gramatyka języka litewskiego, III, Warszawa, 1956, S. 251; A. S a b a l i a u s k a s, Atematiniai lietuvių kalbos veiksmažodžiai, „Gramatikos klausimai“, Vilnius, 1957, S. 99 ff. ²¹ Cf. K. B r u g m a n n, Grundriss der vergleichenden Grammatik der indogermanischen Sprachen II, 1, Strassburg, 1889, S. 410 ff. ²² On a possible participial perfect form in Old Latvian buwis, būwis from *būwens (?), which can be compared with Slavic byvъ, Sanskrit babhūván, cf. J. E n d z e l i n, Lettische Grammatik, Heidelberg, 1923, S. 727; Chr. S. S t a n g, Das slavische und baltische Verbum, Oslo, 1942, S. 196/209. 198 This does not, however, yet mean that derivation with the suffix -us- could not also have taken place independently of preterite formation, especially since this principle applies within the corresponding framework precisely to the old Indo-European participial formation. In very general terms, it can be said that the Baltic participle in -us- is formed from a verbal stem that must not have a present-tense marker (present ablaut, nasal infixation, -st--infix, ie/o stem formation, and the like); compare, for example, krentant-: kritus-; renkant-: rinkus-; rūkstant-: rīkus-; āriant-: ārus-. 4.3 For Lithuanian (today), the rule is that the active past participle is formed from the preterite stem. The Baltic preterite stems in -ē and -ā, in turn, are based on a complex interweaving of different kinds of verbal stem formations, which originally did not possess an exclusively temporal character. The form of the verbal root in this stem formation can be compared with various old Indo-European verbal root ablaut types and, in some circumstances, with other stem formations as well²³, but not with those of the old present tense. In Baltic, the infinitive stem and the preterite stem are not identical. 4.31 Furthermore, certain differences in preterite stem formation are reflected in the Lithuanian participial forms in -us- found today, although they no longer appear in finite verb forms: compare sãkius- with sakýti (infinitive), sãko (present), sãkė (preterite), as against baigus- with baigti, baigia, baigė and dėgus- with dègti, dėga, dėgė. The 1st person singular is in each case sakiau, baigiau, degiau. Thus, alongside preterites in ē that are identical today, the participle has -ius- in one case and -us- in the other. In our context, this difference may reinforce the view that the Baltic -us- participle did not necessarily become part of a preterite formation within the Baltic tense system at the same time as the verbal ē stems²⁴. 4.4 In terms of root and stem structure, the active past participle is marked less strongly and by fewer distinctive features than the active present participle. Moreover, within the complex of -us- participles, Baltic also has forms of the -nt- participles, namely in the masculine nominative singular (nešęs) and in the masculine nominative plural and neuter nominative singular (nešę). For it is rightly maintained in more recent scholarship that these forms derive from the -nt- participle of the verbal ē stem: thus nešęs from *nešent-s, just as nešąs from *nešant-s, and so on²⁵. These forms should be regarded as remnants of the verb’s ē inflection from a time when the temporal (preterital) function of this formation had not yet developed. In the course of the development of Baltic ²³ Chr. Stang, l. c., S. 189/190. ²⁴ For the discussion of the various verbal stems underlying the Baltic preterite, see most recently Chr. Stang, l. c., S. 150 f., 209 f. ²⁵ J. Otrębski, l. c., S. 257 f.; Chr. Stang, l. c., S. 208. 199 Ultimately, these -nt-forms of the preterite have been incorporated into the complex of -us participles. 4.5 This “heteroclisis” of the active preterite participle can therefore readily be considered probable, since the nominative of the active present participles also involves a point at which, owing to syntactic-functional conditions, the inflectional paradigm of the participle is, as it were, disrupted. In the nominative of the present participle (which, needless to say, must refer to a noun in the subject case), the distinction between attributive and adverbial/circumstantial function corresponds to a formally differentiated mode of expression; the -nt participle and the -dama participle occur side by side (nešąs — nešdamas). In the nominative of the preterite participle, we have a “heteroclitic” “formation (nešęs — nešusi, nešę — nešusios), but no double mode of marking the various functions of the participle. It may be assumed that the “heteroclitic” forms of the active preterite participle reflect an attempt at differentiation similar to that found in the present. Compare the situation in Lithuanian: 1. e i d a m a s namo nuvargo — parėjęs namo pailsėjo (adverbial/circumstantial) 2. p r i d e r ą s (priderantis) elgesys — n ū v a r g ę s žmogus (attributive) 4.6 In the oblique cases, the participial inflection offers no possibility of formally distinguishing the attributive function from the adverbial/circumstantial function. It is entirely possible that the situation in the nominative, which can realize this distinction, may also influence the oblique cases. In the course of the development of the East Baltic languages, such differentiation was in fact made largely possible with the aid of the uninflected participle. Compare the situation in Lithuanian: 1. Pasakė pareinantiems vyrams — pasakė parėjusiems vyrams (attributiv) 2. Tai padaryta vyro, einančio namo — Tai padaryta vyro, parėjusio namo („appositiv“, nicht adverbial/zirkumstantial) 3. Tai padaryta vyro einant namo — Tai padaryta vyro, parėjus namo (adverbial/zirkumstantial)26 26 It does not conflict with the guiding principles of the argument presented that the agent of the -dama participle in the adverbial/circumstantial construction is invariably expressed in the subject case or its equivalent, whereas the agent of the adverbial/circumstantial absolutive construction may be expressed in any oblique case, but may also remain entirely unexpressed. The following interpretations are possible in the example above: 1. Tai padaryta vyro (jam pačiam) einant namo; 2. Tai padaryta vyro (kitiems) einant namo; 3. Tai padaryta vyro (jam pačiam ir kitiems) einant namo. 200 4.7 Alongside this tendency to make the distinction “attributive—not attributive (or adverbial/circumstantial)” differentiable also by means of form, there are, on the other hand, efforts to make the present and preterite participial paradigms as similar as possible with regard to the ways in which such distinctions can be expressed (quite apart from the absolute temporal value denoted by each form of one or the other participial paradigm). Lithuanian and Latvian have taken different paths in this respect: Latvian has, in a sense, extended the indeclinable present participle of the -nt stem to the nominative and thus aligned the paradigm of present participles with that of the preterite participle. In the present tense, it neutralized formal distinctions at the functional level (cf. the example sentence in Note 16) and, as in the preterite, subordinated the characterization of the various participial functions27 to the context, albeit with certain limitations28. By contrast, Lithuanian created additional formal means of distinction by developing the indeclinable preterite participle of the -us stem, so that functional conditions could also be taken into account in the preterite. In the context concerning the origin of the indeclinable -us participle, however, particular attention must be paid to the situation in the nominative of the preterite participle (with its “heteroclitic” forms). 5. The feminine singular nominative of the active preterite participle 5.1 A review of the situation just described within the preterite participle, and of the special status of the nominative in particular, strengthens the supposition that the point of departure for explaining the origin of the indeclinable preterite participle in -us in Lithuanian is to be sought here, also in light of the conclusions reached in Section 3.4 concerning the original structural model of constructions with the indeclinable preterite participle. The nominative of the active preterite participle can therefore in any case express a secondary action accompanying the main action when the actor is the same (whereas in the nominative of the active present participle, in this particular construction, the -dama participle is used, restricting the functional range of the -nt participle in the nominative). Indeed, examining any arbitrary longer Lithuanian text shows that the proportion of nominatives of the active preterite participle used adverbially/circumstantially is usually considerably greater than that of nominatives used attributively. 27 Latvian uses only inflected forms of the preterite participle even in the so-called “relative mood,” that is, in the same syntactic environment in which the indeclinable participle occurs in the present tense. This is, incidentally, probably further confirmation that Lithuanian indeclinable participles in -us did not originate in the so-called accusative construction. 28 Vgl. Mūsdienu latviešu literārās valodas gramatika, I, Rīgā, 1959, S. 662/3. 201 5.2 Of the nominative forms of the active preterite participle, the feminine singular nominative forms are of particular interest to us here. The feminine singular nominative of the -us participle normally ends in -usi and contains the old feminine formation of the participle in -ī. This form originally and properly belongs to the -us participles, whereas the masculine and neuter nominatives in -ęs and -ę derive from a different participial paradigm (cf. Section 4.4 above). The feminine singular nominative of the -us participles differs in turn from that of the -nt participles in that the ending -usi shows a considerably stronger tendency to lose its final -i (from -ī) than does the ending -nti (-i from -ī)²⁹. Such forms of the feminine singular nominative of -us participles without an ending are found in modern Lithuanian literary language just as in old Lithuanian texts, for example:...ir užžaugūs padārė wáišių Bimterioką (DP 97);...karalienė...ižgirdūs Ißmiti Salomonolui regėdama aną didį ápstią...apie jį nustėbus bėčio: (DP 323); Mergaitė sėdėjo prie lango, padėjus ant rankų galvą, ir verkė — (T); Ji nepaprastai išaugusi, nebe mergytė, bet subrendusi mergina, pasi-puošus ir sustambėjus (P). It is easy to establish that forms without -i occur predominantly when participles are used adverbially/circumstantially (and in the compound verbal form, which is not considered here); attributive forms without -i are much rarer. No rule can be given for the distribution of forms without -i and those with -i in non-attributive use³⁰. With regard to Old Lithuanian written monuments, it can generally be said that texts which tend to omit the final -i in the infinitive behave in the same way with forms in -usi of the feminine singular nominative of the active preterite participle (e.g. Daukša, Širvyd), whereas, conversely, texts that preserve the final -i in the infinitive also show more full forms in -usi (e.g. Bretkun). 5.3 In the present study, I aim to show that, by a quite different route from the uninflected present participle in -nt, a participial form that must be regarded as inflected acquires the function of “expressing an accompanying action alongside the main action”; from the beginning of written records to the present day, it has been formally indistinguishable from the uninflected past participle in -us. In such cas- ²⁹ Endingless forms of the feminine singular nominative of the active present participle in -nt are found occasionally (according to information from Professor J. Kruopas) in Old Lithuanian Calvinist texts, and more frequently in recent times in East Aukštaitian dialects, especially northeastern ones—that is, in dialects which, unlike, for example, those forming the basis of today’s Lithuanian standard literary language, otherwise also show more extensive loss of endings. By contrast, loss of the final -i in the ending -usi of the feminine singular nominative of the active past participle is possible probably throughout the entire Lithuanian-speaking area. ³⁰ Nor can a general rule be formulated for the occurrence of infinitives without i and infinitives ending in -i in arbitrary Lithuanian texts whose language permits both possibilities; their distribution appears often to be governed by euphonic considerations. 202 Only the determination of whether the main and accompanying actions have the same agent can decide whether a particular form in -us is to be construed as the feminine singular nominative of the active past participle or as an uninflected past participle; compare sentences in which the subject and the participial form are far apart: Kayp tay rayßku ugnij/kuri (= ugnis) ne pirmiaus azudega medžiu/net ižžio-wins iump ßłapiby/želiby/ir ßalti/ir kitas priepuolebes/kurios priešīnasi ios prigimimuy/tay padarius/tuoiaus iż medžio daro ugni (Si I 142); Apkurti-na (=Welinas) idant ne girdetu (=dūsia) ir ne klausitu Diewo bałso sąwo širdiy/ney Diewo žodžio/kurii žgirdus (to be read as -kuri iżgirdus-) lautus piktibiu (Si II 50/51). 5.4 All this suggests the conclusion that the feminine singular nominative of the active past participle was a significant factor in the development of the uninflected past participle, not only functionally but also purely in terms of form. If one assumes that the Baltic -us participle could not originally be used in the absolute participial construction, the general prerequisite is lacking for explaining the “absolute” use of uninflected forms of these participles on the basis of a locative or dative case.31 5.41 In addition, deriving the uninflected past participle from an oblique case of the -us participles also presents other difficulties. The -us participle is usually regarded as an old consonant-stem declension type. It should be noted in this connection that no securely attested consonant-stem declensional form of these participles can be found in the Baltic record32, since the nominatives in -ęs and -ę, which are certainly forms of the consonant-stem declension, must be interpreted as formations of an -nt participle. And one cannot rely on the sole remaining piece of evidence, namely the uninflected past participles in Lithuanian. These are generally regarded as the masculine and neuter singular dative—possibly locative—of the old consonant-stem participial stem; however, this view can be supported only by the particular conception of the nature of the uninflected -us form. 5.42 The dative interpretation of the uninflected past participle in -us is generally supported by its occurrence in the absolute dative. In the case of the -nt present participles, the grammatical isolation of the participle as a predicative component of the absolute participial construction seems either to have preceded or at least to have accompanied the dative interpretation. Even if, for this period of Baltic absolute participial constructions, one already assumes the inclusion of the -us participles in the- 31 A Baltic form in -usi could, for example, correspond to the locative forms attested in Old Indian; cf. OI viduṣi. 32 The Lithuanian and Latvian masculine singular accusative in -į (-i), the Old Prussian form in -in, and the Old Prussian plural accusative in -ins are indistinguishable from forms of i-stems. Nevertheless, the existence of consonant-stem inflectional endings for the Baltic -us participles cannot be categorically denied. 203 If one wished to assume the same, it would have to seem doubtful whether the dative-locative form of the -us participles was then still a consonantal declension form. 5.5 In terms of form, no more objections can surely be raised against explaining uninflected -us forms from the Nom. sg. fem. in -us(i) (from -usī) than there are reservations about the emergence of uninflected -us participles from an oblique case of the relevant paradigm. The reflexive form in -usis, too, largely fits this nominative explanation. In most Lithuanian dialects, including most Old Lithuanian dialects, and in the modern Lithuanian literary language, the reflexive form of the Nom. sg. fem. of the active past participle and the reflexive form of the uninflected past participle are identical: in both cases the form is nešusis (or nu-si-nešus(i) with a prefixed verb for the nominative form, and nu-si-nešus for the uninflected participle). Compare, for example, in Old Lithuanian: stóiusis išg ußpakalo kóiosiam[p] io/pradėio aßaromis wilgit... (DP 476), kuriol' ii ertės uždariusis medes... (DP 599). 5.51 Attention must also be paid to the vocalic reflex immediately before the reflexive particle -s(i) in the case of the uninflected active past participle, since Lithuanian (Baltic) often has the “unshortened” inflectional ending, generally considered original, in the position immediately before the reflexive particle, e.g. in the finite verb 1st sg. pres. nešu: reflexive nešuos(i). If the identity of the reflexive forms mentioned above is indeed intended to reinforce the derivation of the uninflected past participle from the Nom. sg. fem. of the active past participle, then no other vocalic reflex should be expected in the Lithuanian-speaking area. In the case of the uninflected present participle, the Lithuanian dialect ending -nties (e.g. juókianties³³) and the Latvian ending -ties³⁴ (e.g. mazgājoties) point to a diphthongal inflectional ending (-ie from *-ei) of the participial form, which is indeed not older or more original than the ending -nti (cf. Lithuanian reflexive -ntis, e.g. juókiantis)³⁵, but can nevertheless be regarded as an (old) variant form. However, reflexive forms of the uninflected past participle in -usies, e.g. mušusies, dirbusies, etc., occur only in Žemaitian dialects, 33 J. Otrębski, l. c., S. 277. 34 It may, however, be argued that the evidential force of the Latvian reflexive ending -ies in the uninflected present participle is weakened insofar as it is not otherwise based in every case on continuous development, but often results from generalization (in favor of uniform marking of grammatical categories). Compare, for example, just the Latvian reflexive form of the uninflected participle in -am/-ām, e.g. ceļamies, mazgājamies (-am/-ām developed from *-amā via -amu/-āmu; see J. Endzelin, l. c., S. 715). 35 In both Lithuanian and Latvian, the endingless uninflected form in -nt or -ot can be explained more readily from a participial form with the ending -i than from a form with a diphthongal ending. 204 with36; these forms seem to argue against the hypothesis that the uninflected past participle originated in the nominative. The significance of these Lithuanian dialectal attestations, however, can be properly assessed only once it is possible to establish (which I cannot yet do) the relationship in the dialects between the forms mentioned and the reflexive Nom. sg. fem. of the active past participle; that is, what status the reflexive ending -ies has in the respective linguistic systems (of the dialects) (cf. the generalization of the reflexive ending -ies in Latvian; see note 34). 5.52 I can attest a non-reflexive form of the uninflected past participle in -usi—that is, one with an inflectional ending that could shed light on the original form of the participial form in question—from the Old Lithuanian texts available to me only in the so-called accusative construction. Such forms in -usi can scarcely be regarded as a “fuller” form of the uninflected past participle, since, in keeping with the usage found in the respective Old Lithuanian texts, they must ultimately be interpreted as the Acc. sg. masc. of the -us participle (cf. section 2.2 above). Forms in -usi within the first application type of the uninflected participles in Lithuanian, if such forms can be attested37, would not constitute an irregularity in terms of explaining the uninflected past participle from the Nom. sg. fem. of the active past participle. 5.6 Initially, it is difficult to explain the transfer of the feminine form to the masculine and to the entire plural. In Old Lithuanian and in Neo-Lithuanian dialects, one does occasionally find masculine forms used in the Nom. plur. fem.38, but no genuine transfer in the singular, nor the use of feminine forms for the masculine. One might also object that it is not self-evident that one and the same grammatical form—in this case, the Nom. sg. fem. of the active past participle—should be used as a generalized (and no longer inflectable) form, that is, as an uninflected past participle, while at the same time persisting as the specific feminine form of the inflected participle, that is, remaining in the position from which generalization proceeds. 5.61 To support the assumption that the Nom. sg. fem. was generalized, one might point to a similar development in Slavic, for instance in Russian: the present gerund in -uči is formally derived from the Nom. sg. fem. It thus functionally represents a generalization of this case form for all genders and numbers 36 Such endings -ies (that is, their Žemaitian reflexes) are recorded in J. Aleksandavičius, Kretingos tarmės veiksmažodis, “Vilniaus Valstybinio Pedagoginio instituto Darbai,” VIII, Vilnius, 1959 (unfortunately not yet accessible to me), and in P. Jonikas, Pagramančio tarmė, Kaunas, 1939. The material presented on pp. 67–68 in the latter work shows, however, so far only that the reflexive form of the uninflected past participle -mūšūsi˙s- is identical in sound to the reflexive form of the inflected past participle -mūšūsi˙s-. 37 Vgl. J. Jablonskis, Rinktiniai raštai, I, Vilnius, 1957, S. 323. 38 Vgl. A. Senn, Kleine litauische Sprachlehre, Heidelberg, 1929, S. 209. 205 meri dar. This gerund in -uči still survives in Russian today as a residual form39 alongside the dominant form in -ja, which is derived from the Nom. sg. masc. of the so-called short form of the participle40. However, the agent of the Russian gerund must be the same as that of the main action (apart from some phrasal expressions). The degree of generalization is therefore lower than would have to be assumed for the Lithuanian uninflected participle in -us. Moreover, the Slavic forms in -uči are, of course, the old -nt participle. The development of the Slavic forms in -uči and the emergence of the uninflected past participle in Lithuanian from a Nom. sg. fem. cannot, therefore, be regarded as parallel developments or as syntactic borrowing. Nevertheless, a possibility of comparison remains for the use of certain syntactic constructions with active participles. 5.62 Finally, the fact that the masculine nominative forms in -ęs and -e are unsuitable for generalization because they do not display a characteristic Baltic preterite marker in the stem-forming participial suffix may also support the generalization of a feminine form of the active past participle into an uninflected participial form with a special syntactic specialization. Since, however, even after it has acquired its specification as an invariable nominal verb form in the grammatical system of Lithuanian, the uninflected participle in -us serves as a means of expressing the anteriority of one verbal utterance in relation to another statement, an unambiguous formal preterite marker is surely necessary for this formation. 6. On the question of an old endingless nominative of the active past participle 6.1 Despite everything, the development of the uninflected participle in -us from the Nom. sg. fem. of the active past participle remains questionable (the aforementioned derivation of the Slavic gerund in -uči is by no means certain either). The system of the Lithuanian participial paradigm itself seems to present a considerable contradiction to the hypothesis explained in the preceding sections, insofar as the feminine never functions in this system as an unmarked member and therefore cannot be used as a “neutral” form; at least, such use is contrary to its very nature. No less hypothetical, but more in keeping with the Lithuanian participial system, is the attempt to invoke masculine or neuter forms to explain the origin of the uninflected form in -us. 39E.g. будучи, читаючи, едучи, живучи, cited from R. Trautmann, Kurzgefasste russische Grammatik, Leipzig, 1948, p. 102. 40 P. J. Tschernych, Historische Grammatik der russischen Sprache, Halle, 1957, S. 239 ff.; В. В. Иванов, Развитие грамматического строя русского языка, Москва, 1960, S. 121 ff. 206 6.2 For one must ultimately ask whether a formation of the -us participle should originally also be posited for the masculine41. The observations concerning the “heteroclitic” formation of the nominative forms of the active past participle do not preclude this—in fact, they may strengthen the case for it—namely, the assumption of older, unattested formations from the participial -us stem for the masculine as well. One could imagine, for example, that these -us formations did not appear to be sufficiently characterized formally (and functionally) and were therefore “replaced” in all syntactic positions by the forms in -ęs and -ę, which may initially have been used only to mark either the attributive or the adverbial/circumstantial function of the active past participle. In this sense, formal characterization by gender, number, and case would have proved stronger than the unambiguous formal characterization of the absolute temporal value or the specific expressive content of the Baltic -us participle. It nevertheless seems difficult to find concrete evidence for a development of this kind or a similar one. 6.3 Slavic, in turn, offers a clue in support of a corresponding argument concerning the Lithuanian (Baltic) development: the past gerund in -vъ (e.g. znavъ) is derived from a Nom. sg. masc. in *-(u)us4243. A Slavic form *znavus would thus correspond to a Lithuanian žinojus. It cannot be proved with certainty that such a masculine form ever existed in Baltic (except on the basis of the functionally motivated formation possibilities I have presented here), but neither can it be disproved. If the conjecture nevertheless proved correct—that a form such as žinojus was a masculine nominative—we would have a striking Balto-Slavic (or, more precisely, Lithuanian-Slavic) correspondence, a kind of partial innovation within Balto-Slavic. 6.4 This formation of the type žinojus is to be understood as an endingless form of the -us participle, possibly originally a neuter form. It may be noted that the Nom. plur. masc. of the -nt participles of the nešą type is likewise traced back to a neuter form, that is, to an endingless form. The situation is also quite complicated with the -nt participle; for example, the (later) use of the endingless form in the plural is not easy to explain. And yet the explanation of the plural forms in -ą, -į, -ę as (identical) neuter forms seems to be the only plausible one. 41The attempt to explain the nominatives in -ęs and -ę as formations with the strong stem grade of the participial suffix -ves\-us is, after all, based above all on the assumption of a single paradigm for the -ves\-us participles, in support of which considerable irregularities in the sound development are accepted. 42The generalization of the stem in -uus- (= -vus-) is a Slavic innovation. 43Cf. V. Ivanov, l. c., p. 123; J. Otrębski, l. c., p. 257 regards the Slavic form in -us as an originally neuter form that penetrated the masculine. 207 Thus, it is surely not entirely far-fetched to assume the existence of a form of the -us participle without an ending. In Baltic, this formation in -us may perhaps have progressed no further than the stage of a “neutral” mode of expression. 6.5 It is therefore possible to assume, however bold this may seem, that the Lithuanian uninflected past participle, as shaped within an individual language, is an old form of the -us participle without an ending. This form must have been nominative in Lithuanian, and may have belonged to the masculine gender after the loss of the (Baltic) neuter. It is entirely conceivable that, as this form dropped out of the paradigm of inflected participles of the -us stem, the nominatives in -ęs and -e were adopted. On the other hand, the “heteroclitic” formation of the nominative of the active past participle might be seen as the reason for the grammatical isolation of a possible masculine nominative in -us, which in this case must have fulfilled the participle’s adverbial-circumstantial function. This also readily explains the survival of the nominative singular feminine in -us(i), which is very similar in form to the uninflected past participle. The reflexive formation of the uninflected past participle in -usis may be dismissed as an analogical formation, a formal leveling, perhaps modeled on the infinitive, the uninflected present participle, the masculine nominative singular of the active present and past participles (e.g., dirbąsis, dirbęs(is)), or the masculine nominative singular of the -dama-participle (e.g., nėsdamasis44). However, one would again not expect forms of the uninflected past participle with a diphthongal element before the reflexive particle. Yet, as long as the Žemaitian forms in -usies remain unexplained as to their actual character, from the perspective of an internal linguistic reconstruction of the respective dialects themselves (cf. section 5.51 above), they can neither undermine nor confirm the hypothesis just mentioned, or any other hypothesis concerning the origin of the uninflected past participle. 7. The further development of the endingless form of the past -ūs participle 7.1 It is therefore possible to propose the hypothesis that Baltic had, from the outset, an uninflected form of the -us participle, which, in the course of the (language-specific) development of the participial paradigm in Lithuanian, acquired exclusively the function “designation of an accompanying action preceding the main action.” A form of this kind, with such a function, could easily become an (uninflected) adverb-like form, incapable of expressing gender and number distinctions, and with no need to express them. This account naturally makes it unnecessary to explain the “adoption” of the uninflected -us form across all genders and numbers. 44 dirbąsis instead of *dirbąs-si, dirbęsis instead of *dirbes-si, nėsdamasis instead of *nėdamas-si, cf. J. Otrębski, l. c., pp. 252 f., 258 f., 271. 208 7.2 To illustrate the position of this uninflected -us form in relation to the uninflected present participles in -nt, as well as the subsequent “alignment” of these two uninflected participles into a grammatical formation characteristic of Lithuanian and possessing its own distinctive character, in such a way that the particular nature of this phenomenon becomes apparent, it must be stated here that a corresponding derivation from a nominative form is not possible for the uninflected present participle in -nt, not even as a parallel development to the origin of the uninflected present participle in -nt within an absolute participial construction (with these formations subsequently merging), as presented above in Chapter 2. Two reasons in particular can be given for this: 7.21 An endingless form of the -nt participle occurs in the masculine nominative plural or neuter nominative singular, e.g., neša, saka, myli. These forms are conventionally traced back to *nešant, *sakōnt, *mylint. According to the Baltic laws of word-final sounds, a Lithuanian form in -nt can only have developed from a longer form, whereas Lithuanian word-final -us can be traced back to Baltic -us; compare, for example, the masculine nominative singular of the u stems, sūnūs, saldus, etc. 7.22 The final -i of the feminine nominative singular of the -nt participles is not dropped, with only a few exceptions (cf. note 29), anywhere in the Lithuanian-speaking area, whether the (attributive) participle precedes or follows, nor normally in the compound verbal form. (Here we may disregard certain phonetic conditions arising in the course of speech production that could cause the final -i to be dropped, e.g., elision before an initial vowel, and the like.) Besides purely phonetic and historical sound-related factors, the fact that the feminine form in -nti is used predominantly attributively and in the compound verbal form (the -dama participle is available to express the adverbial/circumstantial function) must also play a role here; thus, like adjectival attributes, it must express case marking in its form. 7.3 In Lithuanian, the uninflected present participle in -us underwent further grammatical development on the model of and under the influence of the uninflected present participle in -nt, particularly with regard to its use in the absolute dative and under the conditions of so-called “subject dissimilarity.” However, it cannot be entirely ruled out that the tendency toward “(partial) subject identity” in constructions with uninflected participles in modern Lithuanian literary language (e.g., Išėjus iš girios, aš, tikra to žodžio prasme, parkritau ant griovio krašto (išėjau aš ir dar keli vaikai) I. Simonait ytė,...o buvo taip, Vilnius, 1960, p. 321; Einant palei dailiai pakirptus krūmus, ji įsipainiojo į voratinklį P 1960/12 p. 75) may be due to the fact that it is precisely the past uninflected form in -us that, by origin, is a “subject-identical” form— 14. Kalbotyros klausimai VI. 209 …expression has been, and under certain circumstances may never have entirely abandoned this possibility of construction45. 7.4 To summarize, let it be emphasized once again that the indeclinable past participle in -us, which occurs only in Lithuanian, did not arise at the same time or by the same route as the indeclinable present participle in -nt. While the latter, which may be assumed to be Baltic, owes its origin to the special conditions of the absolute participial construction in Baltic, the former consists of originally nominative forms that occurred in an adverbial/circumstantial function and either stood apart from the (inflected) participial paradigm from the outset or left it relatively early and assumed a kind of special status. 7.5 The circumstances described in this way apply only to the first application type of the indeclinable participle (see above, section 1.21). For the development of the indeclinable participle in -us within the second application type, as with the indeclinable participle in -nt, one must reckon with a morphological and syntactic intermingling of accusative constructions with an inflected participle and constructions with an indeclinable participle of the first application type. In both cases, the area of overlap probably comprised above all the masculine singular accusative of the inflected participles on the one hand and the form of the indeclinable participle on the other. Since in the so-called accusative construction Latvian uses only forms of the indeclinable participle in -nt (Latvian -ot), apart from the Latvian indeclinable participle in -am/-ām, whose use need not be discussed in this context, the inclusion of the indeclinable participle in -us in the so-called accusative construction must also be regarded as a phenomenon that developed specifically in Lithuanian. In Old Lithuanian texts, one can observe that endingless -us forms in the so-called accusative construction occur only very sporadically with the masculine singular accusative (that is, -us alongside the more frequent -usi, -usį as the inflected form), but hardly at all in other environments; in any case, my examination of the Old Lithuanian texts available to me yielded no different picture. By contrast, endingless -nt forms in the so-called accusative construction are attested in all environments—albeit with varying frequency. This, too, allows us to conclude that, at the time of the earliest written records, the indeclinable past participle in -us, even within the first 45 However, in these sentences with “partial” subject identity, one encounters both the indeclinable present participle and the indeclinable past participle, even in the absolute dative! Examples include: Kalbant apie tai, paklausių mokinių, kuo skiriasi gyvas žmogus ar gyvulys nuo mirusių TMa 1959/11 S. 20; Dabar anksti aušė, anksti kėlė, jautėsi sveikas, valgė daug, kaip studijų laikais gavus stipendiją M. Sluckis, Vėjų pagairėje, Vilnius, 1958, S. 6; Kolūkiečiai vieningai nutarė — Lietuvos TSR dvidešimtmetį sutikti įvykdžius pusantros metinės užduoties T 1960 VI 16;...paklausė Adomėnienė Barvainio, jiems išėjus į aikštę P 1960/12 S. 44. The tendency toward subject identity may therefore have other causes; in addition to the aforementioned position of the indeclinable participle in -us, it certainly has other causes as well, such as the spread of certain formulaic expressions, for example atidarant mitingą (kongresą, spaudos konferenciją u. a.), atsisveikinant, baigiant (with or without an object), and so on. 210 …application type had not yet been placed on an equal footing with the indeclinable present participle in -nt to the same extent as is the case in the modern Lithuanian literary language. Halle/Saale List of texts used, insofar as they were not quoted in full in the paper BrP I—II — Postilla tatai esti Trumpas ir Prastas Ischuldimas Euangeliju... Per Jana Bretkuna Lietuvos Plebona, Karalaucziuje... 1591, Teil I und II. DP — Daukšos Postilė, Fotografuotinis leidimas, Kaunas, 1926. Si I—II — Šyrwids Punktay Sakimu, Herausgegeben von F. Specht, Göttingen, 1929, Teil I und II. KNSn — Kniga Nobaznistos Kirkščioniškos... antrą kartą perdrukawota, Kiedaynise, 1653; Summa arba Trumpas iszguldimas Ewangeliju Szwentu... T — Tiesa (Lietuvos Komunistų partijos Centro Komiteto, Lietuvos TSR Aukščiausiosios Tarybos Prezidiumo, Lietuvos TSR Ministrų Tarybos organas), Vilnius. P — Pergalė (Literatūros, meno ir kritikos žurnalas; Lietuvos TSR Tarybinių rašytojų Sąjungos organas), Vilnius. TMa — Tarybinė mokykla (Lietuvos TSR Švietimo ministerijos organas), Vilnius. (All scholarly literature used for this article—insofar as it is not cited in the paper—is listed in the author’s dissertation bibliography. THE LITHUANIAN PAST PARTICIPLE IN -us Г. БЕНЗЕ Abstract On the basis of the formal and syntactic features of the past adverbial participle, the article proposes that this participle differs in origin from the present adverbial participle, which the author regards as a fossilized locative of a consonant stem. The past adverbial participle can be derived from the reconstructed nominative singular neuter form of the past participle in -us, i.e. without an ending (cf. Lith. žinojus and Slavic znawъ<*znawus). The ending -i characteristic of reflexive adverbial participles (e.g., praūšusi-s) may have arisen by analogy with the infinitive, the present adverbial participle, and the nominative singular masculine forms of the present and past iterative participles. 211 THE LITHUANIAN PAST ADVERBIAL PARTICIPLE IN -US G. BENZĖ Abstract On the basis of the form and use of the past adverbial participle, the article proposes that this participle differs in origin from the present adverbial participle, which the author regards as a fossilized locative of a consonant stem. The past iterative adverbial participle may be derived from the reconstructed nominative singular neuter form of the past iterative participle with *-us, i.e. without an ending (cf. Lithuanian žinojus and Slavic znawъ < *znawus). The ending -i characteristic of reflexive adverbial participles (e.g., praūsusi-s) may have arisen by analogy with the infinitive, the present adverbial participle, and the nominative singular masculine forms of the present and past iterative participles.