Search

This study analyzes periphrastic permissive and factitive causative constructions in a corpus of 16 th and 17 th century Old Lithuanian texts. In contrast to modern use, permissive constructions with duoti play a central role in many texts, while those with leisti are only more frequent in some sources and appear to have spread from the east to the west. Due to the influence of bifunctional German lassen constructions, duoti is used not only in permissive constructions, but also in factitive constructions, especially in Prussian Lithuanian. The permittees in duoti and leisti constructions are usually marked as dative, although the accusative is also attested due to the influence of German lassen + acc constructions; however, in the case of leisti , the accusative may sometimes be interpreted as archaic, marking the direct object of the source construction leisti ‘release’ + acc . In addition to permissive constructions with duoti and leisti , this paper also discusses rare cases of the archaic verb (pa-)velti and the borrowed Slavic permissives pa-velyti and pa-/pri-zvalyti . The majority of reflexive permissive constructions are based on duoti and contain a reflexive affixal marker on the matrix predicate, but constructions with an additional marker on the subordinate infinitive are also well-attested. In general, factitive constructions are less frequent than permissive ones and, just as in Modern Lithuanian, the most common factitive is (pri-)versti , but borrowed Slavic (pri-)sylyti is also attested in some sources.

Lietuvių kalbos būtojo dažninio (habitualinio) laiko afiksas -dav- kilmės požiūriu gali būti laikomas iteratyvine priesaga -dau-, kuri ilgainiui buvo imta vartoti tik būtajame laike (Fraenkel 1936). Tarmėse ir senuosiuose tekstuose šalia -dav- vartojamos ir priesagos -lav- ir -dlav-, kurios taip pat galėjo išriedėti iš atitinkamų iteratyvinių afiksų (Fraenkel 1936) – taip habitualinių lyčių yra atsiradę ir įvairiose kitose kalbose (Bybee et al. 1994). Naudojantis elektroniniu Lietuvių kalbos žodynu pavyko rasti šiek tiek galimų priesagos -dau- ir kitų afiksų iteratyvinių vedinių, kurie leidžia daryti tokias pagrindines išvadas. (1) Dažninės priesagos imtos vartoti tik būtajame laike prieš pasirodant pirmiesiems lietuvių rašytiniams tekstams (iki XVI a. vid.), o atitinkami esamojo laiko ir bendraties kamienai jau buvo pasitraukę iš vartosenos; jei taip nebūtų buvę, senuosiuose tekstuose ir tarmėse turėjo išlikti daugiau kalbamųjų priesagų vedinių. (2) Iteratyvai su būsimosiomis dažninėmis priesagomis turėjo būti pasiekę tam tikrą produktyvumo laipsnį būtent tose tarmėse (ar jų dalyse), kurios juos sugramatino kaip būt. d. l. rodiklius. Jei tie iteratyvai būtų buvę produktyvūs visame lietuvių kalbos plote, jų turėjo išlikti daugiau ten, kur kalbamosios priesagos nevirto būt. d. l. afiksais. Straipsnyje taip pat keliama mintis, kad susidarant lietuvių kalbos būtajam dažniniam laikui turėjo veikti formos ir dažnio principas (Haspelmath 2008; 2014; 2017). Dėl jo ilgesni afiksai (su -d-, -l- ir -dl-) buvo pasirinkti žymėti dažninėms (įprastinėms) situacijoms, mat jos yra retesnis iteratyvinių situacijų potipis ir turėtų būti stipriau pažymimos, o pãčios dažninės formos ilgainiui imtos vartoti tik būtajame laike, nes habitualumas yra viena iš galimų (dažnesnių) nežymėtų esamojo laiko formų reikšmių. Lithuanian has regular past habitual forms with the suffix -dav-, which can be explained as an originally iterative suffix -dau- restricted to the past tense (Fraenkel 1936). Dialectal and Old Lithuanian, in addition to -dav-, also feature habituals with the suffixes -lav- and -dlav-, which could have followed the same path of development (Fraenkel 1936), as evidenced by a number of diverse languages (Bybee et al. 1994). Using an electronic edition of Lietuvių kalbos žodynas (The Dictionary of Lithuanian) as the data source, a limited number of possible iteratives with -dau- and other related suffixes were found, which has led to two main conclusions. (1) Habituals were restricted to the past tense before the appearance of the first written Lithuanian texts (mid-16th c.) and the present and the infinitive stems went out of use. If this had not been the case, more corresponding verbal formations should have remained. (2) Iteratives with the habitual-to-be suffixes had to be productive to some extent in the dialects, which grammaticalized them as past habituals. If these formations had been productive in all dialects of Lithuanian, more iteratives should have been found in the areas that did not grammaticalize them as past habituals. It is also suggested that the form-frequency correspondence principle (Haspelmath 2008; 2014; 2017) should have operated in the formation of the Lithuanian habitual. Longer suffixes were chosen to mark habitual situations as a less frequent subtype of iterative situations and habitual forms were restricted to the past tense because habituality is one of the default (more frequent) readings of the present and hence the habituals in the past tend to be marked explicitly (Bybee et al. 1994).

Once more on the origin of Lith. auka This article aims to show that the origin of the words aukauti ‘sacrificeʼ, auka ‘sacrifice, donationʼ and aukuras ‘sacrificial siteʼ goes back to the 19th century. These are the transformations and new formations by Simonas Daukantas. The word aukuoti ‘rock, heave’ was assigned the contextual meaning of ‘to make a heave offering’ by the editors of Bretkūnas’ Bible. Daukantas made use of the words aukuoti , aukavimas in this particular sense; he is likely to have taken them as substitutes for Slavisms apieravoti , apiera from Milkus’ (Mielcke) dictionary. Daukantas unconsciously altered the infinitive form of the verb ( aukauti ) and provided a peculiar interpretation of the word itself, he saw a different motivation for the act of aukauti , that of calming, pacifying the gods (the purpose of the offering rather than its manner). This is shown by the specific collocability of aukauti coinciding with the valency of maldyti , raminti ‘appease, soothe’ as well as some word collocations used by Daukantas. However, already in the Istorija žemaitiska (Samogitian History) where Daukantas’ remake aukauti was first used, a different valency (the same as in the modern language) of this verb was encountered; it was determined by the influence of synonymous Slavisms apiera , apieravoti . The article also discusses the formation of the neologisms auka , aukas and aukuras .

Straipsnis – atsakas į M. Villanuevos Svenssono pateiktą V. Jakulio monografijos ,,Lietuvių kalbos tekėti, teka tipo veiksmažodžiai“ recenziją. Straipsnio autorius atsako į monografijos recenzentui kilusius klausimus, pagrįstai ir išsamiai apgina savo monografijos teiginius: 1. Tekėti, teka tipo veiksmažodžiai yra tik rytų baltų inovacija. Prūsų kalboje tokio tipo nebuvo, o slavų kalbose tipų perdirbimo procesai greičiausiai vyko dėl kitų motyvų ir kitokiu keliu, todėl panašios struktūros veiksmažodžiai negali būti laikomi bendros baltų ir slavų inovacijos padariniu. 2. Seniausi šio struktūrinio tipo veiksmažodžiai susiformavo iš šakninių veiksmažodžių. Pokyčiai prasidėjo ne bendratyje, o preterite: prie preterito kamiengalio -ē pridėjus dar vieną kamiengalį -ā ir susiformavo baigmuo *-ē ā. Tas pats preterito kamiengalis -ē- apibendrintas ir bendračiai (*tekti → *tekē+ti* = *tekēti). The article is a response to the review of the monograph of V. Jakulis Lithuanian Verbs of “tekėti, teka” Type produced by M. Villanueva Svensson. The author of the article answers the questions that arose to the reviewer and defends the statements of his monograph reasonably and thoroughly: 1. The Verbs of Tekėti, teka type are exclusively the innovation of the Eastern Balts. The Prussian language did not contain this type and type-reconstruction processes in Slavonic languages were most likely caused by other motives and happened in another manner; therefore, verbs of similar structure cannot be considered the consequence of joint Baltic and Slavonic innovation. 2. The oldest verbs of this type were formed from stem verbs. Changes did not begin in the infinitive but in the preterite form: a thematic –ā was added to the thematic –ē and an ending *-ē ā was formed. The same thematic -ē- of the preterite was generalized in the infinitive as well (*tekti → *tekē+ti* = *tekēti).

THE SECOND AND THIRD CONJUGATIONS IN THE EARLIEST LATVIAN GRAMMARSSummaryThe earliest writers on Latvian inherited much of their framework from Latin, but many ques­tions required independent empirical analysis, especially in verb morphology. The present tense contrast between dzied-u and strād-āju proved particularly troublesome.Rehehusen (1644) seems to have been hardly aware of the problem. Büchner (c. 1670) did little more than note the two types. Langius (1685) did not refer to the problem at all. Dreszell (1685) recognized the distinction between monosyllabic and polysyllabic infinitives, but his treatment of polysyllables was highly confused. Adolphi (1685) provided the first extensive treatment, but yet left many details unresolved. The „Dispositio imperfecti...“ (1732) offered a new theoretical ap­proach in which the present (not the infinitive) serves as the starting-point, but this innovation, both in practice and in theory, proved to be unworkable. The next grammar, "Lotavica Gram-matica” (1737), reverted to a re-statement of Adolphi’s views.It is only with Stender (1761), who, ironically, expressed grave doubts about the organization of polysyllables, that further real progress is made. His description is inferior to the modern theory in minor respects only. His work stands as the culmination of the first phase of research on the Latvian verb.

PECULIARITIES OF VERB STRUCTURE IN OLD PRUSSIANSummaryTwo types of verbs with a suffix variation have been reconstructed on the basis of Old Prussian catechisms.Verbs of type 1 have the suffix -ī- in the Infinitive and its derivatives (the vowel ī in the dialects of the catechisms is slightly diphthongized, pronounced as eī and recorded as ī, ij, ei, i), and the suffixes -ē- and -ā- in the Present and Past tenses, e.g. billīt, billītwei ‘to speak’, billīton ‘is said’, bille ‘I speak, you speak, he speaks’, billē / billā, billa ‘speaks’, billēmai ‘we speak’, billai ‘I spoke’, billa, byla, bela, billāts, belats, bilats, bylaczt / billē ‘he spoke’.Verbs of this type can correspond to verb forms with different structures in East Baltic languages or have no correspondences at all. In the linguistic literature the suffix -ī- of the Infinitive is usually presented as being derived from *-ē- on the basis of the theoretically reconstructed thematic stems of the Present and Past tenses. Yet the existing materials do not support this kind of reconstruction,Verbs of type 2 have the suffix *-ē- in the Infinitive and its derivatives (the vowel ē in the dialects of the catechisms became narrower and was pronounced almost as ī), and the suffix -ī̆- at the end of the Present tense stem (the Past tense forms have not been recorded); e.g. turīt, turrit, turrītwei, turrettwey, turryetwei (*tur-ē-) ‘to have’, turri ‘I have’, tur, turri, turei ‘you have’, turri, turei, turret,ture ‘he has, they have’, turrimai ‘we have’, turriti ‘you have’. In East Baltic languages their equivalents are i-stem verbs with the suffix -ē- in the Infinitive.Type 1 is very copious and productive, while type 2 ought to be considered a relic. Yet the domi­nance of type 1 should be accounted for not by phonological, but rather by morphological reasons, as in live Old Prussian *ē > ī and *i > eī could not have merged.

DENOMINATIVE STATIVES IN -ėti IN MODERN LITHUANIAN Summary Denominative verbs in -ėti, -ėja are predominantly inchoative in Modern Lithuanian ( senėti, -ėja ‘to become old’ : senas ‘old’, akmenėti, -ėja ‘to harden, to turn into stone’ : akmuo, -ens ‘stone’). There is a relatively small group of synchronically or diachronically analyzable denominatives in -ėti which have stative meaning. I have tried to compile a preliminary list of these verbs from the electronic edition of Dabartinės lietuvių kalbos žodynas ( Dictionary of Modern Lithuanian, 2003, DŽ 5 ). The present stem of denominative statives in -ėti varies and can be one of the following four: (1) ja -stem in -ė-ja ( gars-ė-ti, gars-ė-ja ‘is known, famous’ : garsus ‘known, famous’), (2) i- stem ( sykst-ė-ti, sykst-i ‘is stingy’ : sykstus ‘stingy’), (3) i -stem plus reflexive marker ( bjaur-ė-ti-s, bjaur-i-si ‘be disgusted’ : bjaurus ‘disgusting’), (4) a -stem ( vert-ė-ti, vert-a ‘to be worth’ : verta ‘it is worth’). I have used a frequency dictionary (DDRKŽ, 1998) to separate rare and dialectal cases from the verbs of the main lexicon of Modern Lithuanian (henceforth “rare list” and „main list“). The main list of type 1 ( -ėja ) includes garsėti : garsus ‘to be known, famous’ and malonėti ‘to be kind’ : malonus ‘kind’. Rare or dialectal are badėti ‘to starve’ : badas ‘starvation’ (= Modern Lithuanian badauti ), dagėti ‘(figurative) to be prickly, biting’ : dagus ‘prickly, biting’ (cf. also dagys ‘thistle’), ganėti ‘to be enough’ : gana ‘enough’ (cf. OCS goneti, gonejetъ ) , gardėti ‘to be delicious, tasty’ : gardus ‘delicious, tasty’, grožėti ‘to be nice, beautiful’ : gražus ‘nice, beautiful’, grožis ‘beauty’, padorėtis ‘to dear, to consider appropriate (of clothing, behaviour)’ : padorus ‘appropriate, decent’, skanėti ‘to be delicious, tasty’ : skanus ‘delicious, tasty’. Type 2 ( i -stem) posits some problems, because it is not always easy to prove the denominative origin of some verbs. The synchronic interpretation is usually obscured by the general tendency of Modern Lithuanian to derive deverbative u -stem adjectives from primary or secondary (the so called „mixed type“ which includes presents in -a, -ia, -i and -o beside suffixed infinitive stems) verbs. The main list of less problematic cases of type 2 includes gail ė ti, gail-i ‘to feel sorry, pity’ : gaila ‘(impers.) it is a pity’, myl ė ti, myl-i : dial., Old Lithuanian mylas (cf. Latvian milet : mīls, miļs, Old Prussian milijt : mijls, Bulgarian милея, Macedonian милее beside common Slavic *milъ ), sykst ė ti, syksti-i ‘to be stingy’ : sykstus ‘stingy’, vies ė ti, viės-i ‘to be on a visit’ : (rare, archaic) viesis, -ys ‘guest’ (cf. Latvian vieset(ies) 2 : viesis). Problematic (at least from the synchronic point of view) cases can be exemplified by tyl ė ti, tyl-i and ting ė ti, ting-i. The first one could be deverbative in origin if we accept the comparative evidence from Slavic (OCS tьleti ‘modern, vergehen’), Germanic (Gothic þulan, -aiþ ‘ertragen, dulden’) and Celtic (Old Irish -tuili ‘schlaft’). The second one could be denominative if Lithuanian tingus is considered archaic based on correspondences in Slavic (*tegъkъ, cf. OCS otegъciti, -ati; note also prefixed inchoative *-ē- verb otežati ‘to become burdened’) and Germanic (Old Icelandic þungr ) . The list of rare verbs includes non-problematic skon ė ti, skon-i, skom ė ti, skom-i ‘to be delicious, tasty’ (: skonis, skomas ‘taste’), a lengthened grade deverbative gor ė ti, gor-i ‘to experience a bitter taste’ (beside gar ė ti, gar-i ) , an etymologically unclear spul ė ti, spul-i (denomina­tive derivation could be implied by the existence of gorus, spulus, but note the aforementioned general tendency to form deverbative adjectives in Modern Lithuanian) and skard ė ti, skard-i ‘to echo, to sound’ (related to the family of kerd-/kird- verbs or simply denominative beside skardus ‘sono­rous’). Type 3 ( i -stem plus reflexive marker) covers a group of verbs expressing a certain attitude or evaluation and can be paraphrased as ‘to consider something / someone (or a situation in general) to have the quality of the base adjective or noun’, cf.: jis slykst-i-si (inf. slykstėtis ) tuo valgiu ‘he considers that dish disgusting’ (: slykstus ‘disgusting’). Type 3 is almost unproblematic, because the possibility of deverbative formation of adjectives is blocked semantically in most of the cases. There is no model of adjective formation in Lithuanian which could be paraphrased as ‘the one who / which is evaluated by the activity of the speaker described by the base verb’ (vs. productive model of ‘the one who / which verbs’, e. g. pavydus ‘the one who envies’ : pavydėti, pavydi ‘envy’, but slykstus ≠ ‘the one which is disgusted at’). Lithuanian drovėtis, drov-i-si ‘to be shy’ also belongs to type 3 because of its morphology, but is probably to be analyzed as a deverbative formation with root vowel lengthening (cf. Latvian druveties ‘erschrecken, sich furchten’, drudet ‘zittern’). Type 4 (a-stem) has only one verb in standard Modern Lithuanian: vertėti ‘(impers.) to be worth’ : verta ‘it is worth’. Some other verbs based on predicatively used neuter adjective forms or nouns could have probably belonged to this type (cf. ganėti, -ėja : gana, gailėti, gail-i : gaila ) , if their nominal ending was reinterpreted as a 3rd present indicative (cf. the deverbative type tekėti, teka ) , but was later substituted by stems in -i- and -ė-ja. My main conclusion is that denominative statives in -ėti are an unproductive derivational category in Modern Lithuanian. It is still possible to identify some isolated groups characterized by presents in -ė-ja, -i, -i-si or -a. Predominant inchoative semantics of Lithuanian denominatives in -ėti is probably to be considered an innovation. As inchoatives in -ėti, -ėja started to gain ground, at least some statives adopted i -stem presents (a typical feature of deverbative statives). If Indo-European formations in *-eh 1 - primarily had presents in *-eh 1 -ie/o- (as argued by Jasanoff 200212003, 147; cf. reconstructions of athematic inflection *-eh 1 -ti by Watkins 1971, 83; Jasanof f 1978, 124; Ruijgh 2004, 56 and zero-grade suffix in *-e-h 1 -ie/o- suggested by Rasmussen 1993, 481 and Hardarson 1998, 337), the more archaic morphology was taken over by the dominating inchoative type in Lithuanian, while the possibly archaic stative type lost the morphological uniformity of the present stem and drifted towards deverbative i -presents. Many of the latter ones became derivationally obscure as verbs in -ėti, -i can be considered as bases for the productive formation of deverbative u -stem adjectives in Modern Lithuanian. The space left by (formerly productive?) statives in -ėti was taken over by new denominative suffixes, cf.: gudr-auti ‘to act using tricks’ : gudrus ‘clever, cunning’, draug-auti ‘to be friends’ : draugas ‘friend’, balt-uoti ‘to be white’ : baltas ‘white’, snip-inėti ‘to spy’: snipas ‘spy’.

THE ALLATIVE CASE IN MIKALOJUS DAUKSA’S Postilla Summary A semantic analysis of the allative case in Dauksa’s Postilla (1599) is performed in order to discuss various types of its use. The following 12 functions of allative are distin­guished, which reflect well-established patterns of grammaticalization of goal-marking morphemes (Rice et al. 2007). I. Spatial functions. 1.1. Inanimate Goal of motion. The movement of the Figure (a verb of motion) is directed to the Goal expressed by the allative. This is the most basic meaning of the allative, e.g. Luke 24,28: du i ǯ g mokitiniu […] priârtinos mieſtelo ṕ kurio ṕ eio 18931, Dwa ʒ vcniow […] prʒybliyli ſido miaſtecka / do ktorego sli 19811, ‘And they drew nigh unto the village, whither they went’. 1.2. Goal of motion, when the Figure acquires a location in physical proximity to the Goal, e.g. Luke 24,24: Ir nueio nekurie i ġ muſ grâbo 18925, Y chod ʒ ili niektor ʒ y ʒ nasych do grobu 1985, ‘And certain of them which were with us went to the sepulchre’. This spatial meaning of the allative is frequent and is traditionally considered as the main one. Proximity, however, is determined by the context in the case of three dimen­sional Grounds, whereas with two-, one-dimensional or dimensionless Grounds it is deter-mined by the lexical meaning of the reference object. 1.3. Animate Goal of motion. The Figure acquires a location in the physical sphere of the person, e.g. John 1,11: Sawumpiump ateio / ir ſawieii io ne prieme 434, Do wlaſności pr ʒ ysedl / a ſwoi go nie pr ʒ yili 4219, ‘He came unto his own, and his own received him not’. 1.4. Goal of direction or orientation. The Figure is static. It is motivated by the trans­formations Figure’s motion towards Ground → Figure’s orientation towards Ground and Figure’s motion towards Ground → Figure’s extension towards Ground : itieſeu iſumṗ rkas ſawȧs 964, wycigalem ku wam rce ſwoie 9716, ‘I have spread my arms towards you’. II. Non-spatial functions: 2. Purpose (purposes are destinations). This use type is very common. Allatives of this type correlate with the infinitive and dative of purpose. [žmonės] nesa kriksto ṗ S. maus waikelus 34813, nosdo Chrʒtu świtego male ditki 35927, ‘People carry little children to baptize’. 3. Abstract Goal of motion (state or emotion). The Figure (a person) moves towards or into a psychological condition, e.g. John 13,20: Wiſſokes nſſkuris piktai daro nopkncʒe swieſos / ir noteiṫ swieſoſṕ 24036, ktory ʒle cyni / nienawidʒi światlości / y nie prʒychodʒi na światlośc 24915, ‘For every one that doeth evil hateth the light, neither cometh to the light’. 4. Addressee, e.g. Luke 1,34: Tare ta ḋ Maria Angelo ṕ 4389, R ʒ ekla tedy Marya do Anyola 4547f, ‘Then said Mary unto the angel’. 5. Emotional target. The Figure (emotion) is directed to a Ground (a person, God, or an inanimate object). 6. Additive allative: Neſstoṗ purwoṗ emes / silgiwâtos primâise 4064, do tego blota iemſkiego / śilywotnprʒymiesal 42011, ‘Power of life was added to the mud of the Earth’. 7. Standard of comparison. With prilyginti ‘equate’, pridėti ‘add’ the allative is used as the standard of comparative constructions. 8. Possessor. With užgulėti , prigulėti , prieitis ‘belong, fit’ (Polish nalezec ) the allative is almost exclusively used with inanimate possessors. 9. Discourse organizing type. With eiti ‘go’, prieiti(s) ‘approach’, sugrįžti ‘return’ the allative denotes the topic of the discourse: Bet’ ſugrskime Simeono ṗ 43013, Ale ſiwrocmy do Symeona 44528, ‘Let’s go back to Simon’. 10. Allative of time (time is space). Usually the allative denotes a border of time ap­proached by the Figure. 11. Recipient. With atsiųsti ‘send’, atnesti , nunesti ‘bring’, nesiojimas ‘bearing’ the allative is a person-recipient. 12. Benefactive. The allative is not a clearly spatial case. Though its local meaning constitutes almost a quarter of the instances, it is quite heterogeneous: four spatial functions are distin­guished. The schematic meaning of the allative can be defined as Goal. This motivates the vast majority of its abstract functions. The traditional meaning of proximity is not fundamental for the allative; it is determined by the context and the lexical meaning of the reference object.

WERE THERE IN THE OLD PRUSSIAN LANGUAGE ANY VERBS ANALOGOUS TO LITH. tekėti, teka VERBAL TYPE? S um mary The article seeks to determine whether the Old Prussian language could have had verbs analogous to the Lithuanian type ending in -ėti in the infinitive, -a in the present tense and -ėjo in the past tense. Vytautas Mažiulis’ Prussian Etymological Dictionary directly attributes only one reconstructed verb to this type, *skeletvei, *skela ‘to owe, to be at fault’, and indirectly connects several suffixal verb forms, *peldetvei ‘to earn; to obtain’, *kabetvei ‘to hang’, *bilet ‘to tell’ and *kalset ‘to ring’. The OPr. verbs under discussion, however, belong to another verbal type (with variable suffix) - inf. -i-, praes./praet. -a-/-e-. The tekėti, teka verbal type arose from the past tense stem, *-e , of root verbs with the addition of thematic personal endings, i.e., a secondary stem *-ia (*tekeia). The past tense stem *-eia suffix *-e was then introduced into the infinitive (*tekti > *tek-e-ti). The Prussian language added plural personal endings to vowel suffixes athematically (directly after the suffix vowel), and because the first, second and third person singular usually had a zero ending – the final vowel was that of the suffix (see Kaukienė 2000b, 58 f.). Therefore there was no basis for tekėti, teka verbs to arise. Hence we may conclude that tekėti, teka verbal type is an East Baltic innovation.

COMPLEMENTATION OF THE VERB matyti (‘SEE’): INTERACTION OF FORM AND MEANING Summary The paper sets out to disclose the workings of syntax-semantics interaction using as an example the analysis of complementation of the Lithuanian verb matyti (‘see’). The focus is on the structural differences of basic types of complements and their impact upon the meaning of the matrix verb. Following Lyons’ classification of entities (Lyons 1977,1991) and the typology of perception-verb complements based on the hierarchical structure of the clause developed by Dik and Hengeveld (1991), the present analysis suggests considering the feature of simultaneity to be crucial in distinguishing between MP and IPSoA constructions among non-finite (participial/gerundial) complements of this verb of seeing in Lithuanian. Types of complements containing the forms of the Lithuanian participle and gerund marked for the future or past tense found in matyti (‘see’) complementation denote propositions (3rd-order entities). The given MP constructions can be also said to express an evidential qualification. The meaning extension of matyti (‘see’) is the same as that of see in English, from direct to indirect or mental perception, and the ‘cognitive format’ of expression becomes fully acceptable to the verb. A few present tense personal forms like matot(e) (‘see’2ndPlPrs), matome (‘see’1 PlPrs), matau (‘see’lSgPrs) that function parenthetically preserve an evidential qualification and thus, can be regarded as undergoing pragmaticalization (Aijmer 1997), while the infinitive form matyti (‘see’ Inf) undergoes grammaticalization: its shortened form matyt (‘evidently’) develops into a parenthetical word with the modal meaning of epistemic possibility/uncertainty.

ETYMOLOGICAL REMARKS. PART V: Postverbal derivatives of the type indas, pergas, iseiga Summary Certain two-syllable Lithuanian nouns at first glance seem to be primary formations, but in fact they are not. Cf. padas like vadas, pradžia like vadžia,, parda like vada, apstas like mastas, proga like uoga, etc. What separates these quasiprimary words is the presence of the prefixes pa-, pra-, par-, ap-, pro-. Pursuant to this observation it has been proposed to etymologize the Lithuanian words with the help of a model suggested by Vedic composita with a final zero-grade root member, e.g. duḥ-sth-a-, su-ṣṭh-u- etc. Applied to the Lithuanian material, this hypothesis yields the following analyses: pa-d-as, pra-d-jā with -d- from */d h h 1 V-/ related to */d h eh 1 / „put“; par-d-a, with -d- from */dh 3 -V-/ to */deh 3 -/ „give“, ap-st-as, with -st- from */sth 2 -V-/ to */steh 2 -/ „stand“; pro-g-a, with -g- from */g w h 2 -V-/ to */g w eh 2 7 „go”. The problem, however, is: (1) the Vedic material offers no parallel for adding preverbs to the zero grade of a verbal root to form a nomen compositum, (2) the compounds quoted from Vedic with su- and dus- are rather rare in other ancient languages, and in Balto-Slavic are not attested at all (apart from the hypo­thetical instances mentioned here). Hence we should look at the quasiprimary Lithuanian words from another angle. The author’s suggestion is that they are postverbal derivatives formed by replacing the verbal suffixes -ė(ti), -uo(ti) or -o(ti) by the nominal suffixes -a-, -ā-, -jā-, -ė-. Cf. pad-as žad-as, nežad-as; byl-oti => byl-a. The hypothesis of postverbal derivation from preverb compounds also convincingly explains a number of other quasiprimary nouns, e.g. samd-as from samd-yti