ŽODŽIŲ FORMOS IR JŲ VARTOSENA LIETUVOS TSR MOKSLŲ AKADEMIJA

V. AMBRAZAS

THE STRUCTURE AND USE OF THE ACCUSATIVE WITH PARTICIPLE (ACCUSATIVUS CUM PARTICIPIO) IN OLD LITHUANIAN

The accusative with participle (hereafter — GD) is one of the few participial constructions that changed markedly already during the period of the written tradition. The grammatical form of GD in Old Lithuanian differed from that of the present day; its sphere of use was also much broader in the past. Some features of GD found in sixteenth- and seventeenth-century texts help shed light on the ancient structure of the simple sentence and its development. Therefore, in studying this construction, attention must first be paid to its structure in the oldest monuments of the Lithuanian language.

GD in Lithuanian was first described in D. Klein’s grammars¹. Considerable data on this construction are found in later grammars², especially in the works of J. Jablonskis³. From the perspective of historical grammar, GD was discussed by A. Potebnia⁴ and E. Fraenkel⁵; separate observations and data on this subject appear in the works of A. Bezzenberger⁶, B. Delbrück⁷, and other researchers. E. Tangl examined GD in a special study, of which, unfortunately, only an abstract was published⁸. It contains perceptive observations and significant ideas about the origin and use of this construction, but cannot take the place of a comprehensive description of GD’s structure. In the past decade, the relationship between GD and other constructions has continued to be discussed in I. Marvan’s articles⁹, while J. Pali-

¹ Grammatica Litvanica… edita à M. Daniele Klein… Regiomonti… 1653, p. 163, 165; M. Danielis Kleinii Compendium Litvanico—Germanicum… Königsberg, 1654, p. 111.

² Plg. Universitas Lingvarum Litvaniae… 1737, Vilnae, p. 27, 52; W. Haack, Vocabularium Litthanico—Germanicum et Germanico—Litthavnicum… Nebst einem Anhang einer kurtzgefaßten Litthauschen Grammatic. Halle, 1730, p. 336,; P. F. Ruhig, Anfangsgründe einer Littauischen Grammatik, Königsberg, 1747, p. 150—151 (Ch. G. Mielcke savo gramatikoje p. 188—189 pakartoja tuos pačius duomenis); G. Ostermeyer, Neue Litauische Grammatik ans Licht gestellt, Königsberg, 1791, p. 177; F. Kurschat, Grammatik der littauischen Sprache, Halle a. S., 1876, p. 366, 376, 423, 424; A. Schleicher, Litauische Grammatik, Prag, 1856, p. 263, 316, 321—322, 332.

³ J. Jablonskis, Rinktiniai raštai, red. J. Palionis, I, Vilnius, 1959, p. 303, 321—322, 472 tt., 614—615.

⁴ А. Потебня, Из записок по русской грамматике, 1—11. М., 1958, 314 тт.

⁵ E. Fraenkel, Syntax der litauischen Kasus, Kaunas, 1928 (hereafter — Kas.), p. 154—155.

⁶ A. Bezzenberger, Beiträge zur Geschichte der litauischen Sprache, Göttingen, 1877, p. 223, 227—228, 262.

⁷ B. Delbrück, Vergleichende Syntax der indogermanischen Sprachen, (hereafter — Vergl. Synt.) II, Strassburg, 1897, p. 490—491.

⁸ E. Tangl, Der Accusativus und Nominativus cum Participio im Altltauischen, Weimar, 1928 (hereafter — Acc.).

⁹ И. Марван, О некоторых функциях литовского причастия „Kalbotyra (Языкознание)“ IV, Vilnius, 1962, p. 33—43; К вопросу о лингво—географическом положении балто—славянских языков, „Baltistica“ V (1), Vilnius, 1969, p. 17—19.

13*

195

...’s monograph presents new data on the use of GD in sixteenth- and seventeenth-century texts¹⁰.

The aim of this study is to describe the structure and use of GD in Old Lithuanian and to clarify the direction and regularities of the development of this construction. The study is based on data collected exhaustively from all sixteenth- and seventeenth-century texts currently available (excluding dictionaries); translated texts have been systematically compared with the original texts available¹¹.

1. Double accusative with nouns and participles

1.1. GD usually consists of three main constituents—a verbal (V), a nominal (N), and a participial (P) one. N and P depend on V (are subordinated to it) and take the accusative form; moreover, P agrees with N in gender and number. This construction differs from other combinations of an accusative noun with a participle in that neither of its dependent constituents, taken separately, can replace the entire NP combination while retaining its syntactic meaning. For example, in the sentence (welinas regeia ana išchalkusi (BrP I 271), the combinations regeia ana and regeia išchalkusi can be distinguished, but neither accounts for the meaning characteristic of the construction as a whole: “he saw that she was hungry.” Although N and P are in a dependent case form, in this construction they are linked by a semi-predicative relation, that is, a predicative relation that does not form the basis of sentence structure¹². Such a relation between the dependent constituents of the construction is possible only after verbs of a certain meaning (verba percipiendi, dicendi, etc.).

The same semi-predicative relations also arise in accusative combinations with adjectives and nouns, cf.: regeiot ghi vschžengenti Mž 299; regeiau tą Moterischkę girtą BrB Apr 17, 6 and regėjome jį … wirą sopulų DP 169⁵². Since the time of J. Jablonskis, these constructions have been called the double accusative. They are rare in the modern language¹³ and are formed only with a small number of verbs (see, find, etc.), but are used more widely in old texts and traditional folklore. Since the double accusative with nouns and adjectives helps us better understand how GD arose, it is useful to discuss it separately.

A. Potebnia’s ideas about the relationship between the double accusative and the broader use of the object accusative in ancient times are highly important for the history of GD in Lithuanian. Since direct and indirect objects were once not formally differentiated, the accusative with verbs of perception, speech, and the like was linked in the same way as

¹⁰ J. Palionis, Lietuvių literatūrinė kalba XVI–XVII a., Vilnius, 1967, p. 159–161.

¹¹ For data on the distribution of GD in the sixteenth and seventeenth centuries and its relationship to constructions in the originals, see the author, LKK XI, pp. 41–58. Excerpts from the writings of J. Bretkūnas and B. Vilentas were compared with the 1549 edition of Luther’s Bible: Biblia: Das ist: die gantze heilige Schrifft… D. Mart. Luth. Wittemberg. Gedruckt durch Hans Lufft. 1549. The following editions were used for comparison with the Vulgate: Biblia Sacra Vulgatae editionis Sixti Qvinti Pont. Max. ivssv recognita atqve edita Romae… MDXCIII; Novum testamentum graece et latine… curavit Eberhard Nestle, Stuttgart, 6¹⁹²¹. Abbreviations are not indicated in quotations from these texts in this study.

¹² The term semi-predicative relations corresponds approximately to the term sociative relations used by A. Muchin; see A. M. Muchin, Структура предложений и их модели, Л., 1968, p. 92 ff.

¹³ Because this construction is rarely used and is not described entirely accurately in A. Schleicher’s grammar (he thought that in the utterance sakė tave šiokią, sakė tave tokią the linking verb esant or esančią had been omitted; see A. Schleicher, Litauische Grammatik, p. 263), B. Delbrück considered the double accusative to have almost completely disappeared from Lithuanian; cf. B. Delbrück, Vergl. Synt., I, p. 378.

196

with factitive constructions. On this basis, A. Potebnia pointed out that, for example, in the Old Church Slavonic construction mbnjasėtb otbca svoiego živa “I consider my father alive,” the verb’s relation to the first accusative is the same as in the sentence postavi Mefodbja episkopa “he appointed Methodius bishop”¹⁴.

Applying A. Potebnia’s approach to Lithuanian, E. Tangl pointed to instances in which the verbs regėti, sakyti in old texts are used without a preposition with an accusative denoting a more remote (looser) object, e.g.: pakáiu iėmus sáko DP 203 (= pokoy im opowiada W 211)¹⁵. The Lithuanian double-accusative construction developed from such accusative objects with an accompanying predicative¹⁶.

1.2. A broader survey of data from sixteenth- and seventeenth-century texts fully confirms the genetic connection, observed by A. Potebnia and E. Tangl, between the double accusative and the verb-governed accusative denoting a more remote object.

The accusative of a more remote object occurs with verbs of speaking in many old texts, especially with the verb sakyti.

I speak all miracles Mž 519¹⁷. I shall speak his law Mž 506. She gave birth… to the one whom Gabriel had spoken of before Mž 179¹⁸. I spoke to you of earthly things SE 136⁸. The interpreter… commanded them to speak repentance and forgiveness of sins MT LŽ 14, 17—22. Ans… told us of your longing ChB II Kor 7, 7. The least… have those who speak him ChB Apd 15, 21. He spoke to everyone… the baptism of repentance ChB Apd 13, 24. He states the reason MP 205⁵. …His own wife / whom he speaks of through the prophet MP II 354^a25.

J. Bretkūnas uses the verb sakyti with an accusative when translating constructions in the German original with the verb predigen, meaning “to prophesy, proclaim, preach about something,” cf.:

I would preach deliverance to the captives Iz 61, 1¹⁹ — Zu predigen (praedicarem) den Gefangenen eine Erledigung. Ir wardą Wieschpaties sakisiu Ps 116, 17 — Ich wil… des HERRN namen predigen (invocabo). …Kursai pakaiu saka Nah 1, 15 — der da Frieden predigt (annuntiantis pacem).

This meaning of the verb sakyti is also shown by its juxtaposition with skelbti, cf.: God… was spoken of and proclaimed among the pagans WP 137; the latter verb, in turn, may also take an unprepositioned accusative: they proclaimed God from it DP 562⁶ (stawili stad Bogą W 956).

In M. Daukša’s Postil, phrases with sakyti translate expressions with various verbs, namely opowiadać (sako praszokusius dabar sanczius ir atęisęczius daiktus 474²⁴ — opowiada przeszłe |ninieysze| y przyszłe rzeczy W 794; ii mumus tikrai sako 409⁴⁴ — ie nam… opowiedą W 671²⁰), powiedać (daiktus žęmiškus sakiau 452⁵ — rzeczy ziemskie powiedat W 755), zwiastować (Angelas… dziauksmą passake 440²⁴ — wesele zwiastowal W 718) and even dawać: To kauliimo norint’ tuli tulas priežastis sako 496¹⁰ — rožne przyczyny dawaiją W 836. In the Polish text of Sirvydas’s “Punkty,” such phrases with sakyti correspond to constructions with spowiadać (Sako sawo piktibes — Spowiad aia się swych złości SP I 160²⁶) and powiadać (Iszminti tieğ

¹⁴ A. A. Потебня, op. cit., pp. 295—301. For examples from Lithuanian and Latvian, see pp. 304—305, 306—307, 315 ff.

¹⁵ E. Tangl, op. cit., p. 42.

¹⁶ Ibid., p. 45.

¹⁷ Similarly, Mž 333, 397, BrGD LXXVI (132).

¹⁸ Ibid. BrGD II, RG 13.

¹⁹ Ibid. Iz 61, 2.

²⁰ Ibid. DP 203₇ — W 343; DP 452 — W 755.

197

we say — We proclaim wisdom SP II 215₁₀); similarly in Petkevičius’s catechism, the 1600 Postil, and the 1674 Gospels, cf.: I shall proclaim your goodness — I shall tell of your holy goodness PK 89; proclaiming the forgiveness of sins — declaring the forgiveness of sins PK 171; I spoke of earthly things — he spoke of earthly things Ev 177₁₉; he speaks the true law MP III 17a₁₅ — he tells the plain truth RP 286²¹. A still freer translation is found at MP III 25₄₂: You hear from Philip the saint / whose memory we commemorate today — whose memory we commemorate today RP 293a.

Active constructions of the verb to say with an accusative correspond to passive constructions with passive participles in the nominative.

Where are all your wonders recounted BrB Ps 26, 7 (one proclaims all your wonders) The fast was proclaimed BrB Jer 36, 9 (that a fast should be proclaimed)... The abomination of desolation / which was spoken of by Daniel DP 372₂ (which is foretold W 611). Miracles... will be spoken of throughout the world KN 93₁₆. That it might be said... foretold SE 89₁₀²². Jesus... who had first been spoken of ChB Apd 3,20. The prophecy... had to be foretold MP 172, 172a₉ (was not to be foretold RP 117).

Nevertheless, phrases with an accusative denoting a more distant object are by no means predominant in old writings compared with other synonymous constructions. Phrases with the same verbs in German and Polish originals are much more often translated by constructions with the preposition about or by a prefixed verb to recount, cf. he spoke about the true coming MP 207a — she spoke about the true coming RP 142₂₃, The angels told of the peace DP 353₂ — they proclaimed peace W 581₂₄. Sometimes a prefixed and an unprefixed verb occur even in the same sentence, for example: The wicked, that I might speak of their deliverance... the righteous, that I might recount their secrets MP III 24a₃₄–₃₆ — that I might tell of their deliverance... recount the mysteries RP 293.

Accusative constructions of similar meaning with other verbs of saying, e.g. to speak, talk, call out are much rarer and less evident, since they almost everywhere translate corresponding expressions in the originals.

Christus... let there be peace in the heart DP 433₂₉ (peace speaks in the heart W 707)²⁵. I shall send disaster, which I have spoken against them BrB Jer 19, 15 (that I have spoken against them). And we speak wisdom among the perfect ChB I Kor 2, 6. We speak him C Hristusiep ChB II Kor 2, 17. For your servants... will cry out against you BrB Jer 12, 6 (and cry out against you).

Just as the verb to say means “to recount,” so too to write can mean “to describe” and take an accusative without a preposition. In M. Daukša’s Postil, such phrases most often correspond to the Polish original’s opisywać + acc.

Mark and Luke write this holy History DP 368₃₈ (they describe W 606). the Evangelists write of the other three human miseries DP 175₄₄ (they describe... the wrongs W 301). they write of the great wickedness and fierce heart of the Jews DP 154₃₉ (they describe... great anger W₂ 17).

²¹ Similarly MP II 386₈₂₂ — RP 261.

²² Similarly MP 168₃₁₁, 172₁₀.

²³ Cf. also DP 48₂₁, SP I 303₁₁, MP 169₄₈, 173₁, 181₁₇, 184₁₅, 187₆, 212₄₃, ChB Apd 5, 42; 8,5; I Korint 1,23; Efez 2, 17 et al.

²⁴ Cf. MP 181₂₀: he speaks the truth — he will tell the truth, but he recounts peace — they proclaim peace RP 123.

²⁵ Cf. corresponding passive constructions: now we are said to live MP 138 (we are said to live RP 91a).

198

In some cases it is translated even more freely, cf.: Todrinąg ir du Tėwu pirmiaus raszō DP 503₈ (dwu oycu naprzod kladźie W 856).

Isolated cases of the verbs mąstyti and dūmoti used with the accusative, observed in the writings of M. Daukša and K. Sirvydas, also differ from the Polish texts, in which the corresponding forms of transitive verbs are preceded by a preposition.

You should learn for yourself | who taught such severity (Ważącyę tego) SP II 196₁₆. So that we may ponder the shortness of our life | and the unknown time of death DP 552₂₇ (abysmy obmysłali krotkość żywota W 941).

Meanwhile, M. Daukša and J. Wujek use the verb skaityti (czytać ‘to read’) with the accusative in the same way, cf.: Tris tad’ apreiškimus Wieszpatiės skaitome DP 424₁₄ (troie tedy ziawienie Pańskie czytamy W 693)²⁶.

The uses of the verb žinoti and Polish wiedzieć with the accusative also correspond in the catechism of M. Petkevičius and the anonymous 1600 postil.

Therefore, knowing God's grace toward us, we accept our share with him (znając k sobie Bożą łaskę) PK 139. We know many good deeds MP 208ᵃ₁₃ (wiemy… dobrodziejstwa RP 142ᵃ₅). We also know the Son's kindness MP 208ᵃ₁₉ (wiemy… dobrodziejstwa RP 142ᵃ₁₀). The Lord… knowing all that terrible torment of his MP 155ᵃ₃₅ (wiedząc… okrutną mękę swoję RP 107₅₂).

Among verbs of perception, the verb girdėti (‘to hear’) is worth mentioning. In the old writings, as in their German and Polish originals, it could be construed with the accusative expressing the content of information experienced through the senses.

So that you may follow him in noble deeds | hearing of his life and works DK 67₈ (DK¹ 63). When John heard of the works of Christ DP 17₁ (Vsłyszawszy… uczynki Christusowe W 14)²⁷. We heard the time | place | and manner DP 594₄₃ (Słyszelismy czas | mieysce | y sposob W₃ 199). We heard of three miraculous deeds BrP I 231. Alexander… while hearing of glorious deeds… speaks BrB I Mak 10, 15 (hóret die löblichen Thaten). And he says, tomorrow you will hear it ChB Apd 25, 22.

The correspondence with German and Polish constructions in these cases (as in the aforementioned combinations with skaityti, žinoti, byloti, kalbėti) reflects typologically and genetically motivated parallelism rather than borrowing. During the period in question, German and Polish also used prepositional accusatives to express a remote object much more widely than today; in general, a tendency to replace many constructions of this meaning with prepositional constructions can be observed in the development of many Indo-European languages (cf. 2. 1. 7.). The examples with verbs such as sakyti, mąstyti, rašyti clearly show that this accusative expressing a remote object was an autochthonous feature of Old Lithuanian²⁸. However, as early as the 16th–17th centuries it was quite rare compared with the corresponding prepositional constructions. Most combinations of the type in question had probably become archaisms by then. They are important because they help explain the origin of the double accusative and identify the original syntactic relations between the members of this construction.

²⁶ Cf. also DP 562₂₈ — W 957.

²⁷ Ibid. VEE 13, 16.

²⁸ Another matter is the constructions of the verb liepti with the accusative, found in J. Bretkūnas’s Bible and apparently arising under the direct influence of the German text, cf.: liepe mane eiti — hieß mich gehen Ezech 46, 21; liepk ghi passikelti — heis zu auffstehen II Kar 9, 2; PoNas ghi liepe — der HERR hats jn geheissen II Sam 16, 10; likewise Nehem 12, 31, Est 1, 17, etc.

199

1. 3. With the same verbs as those discussed here for word combinations, double-accusative constructions are also formed.

Since the double accusative with nouns and adjectives is somewhat more frequent in old writings than word combinations with a single accusative of the more distant object, verbs taking a double accusative are also more numerous and varied. In both cases, however, the same semantic groups and principal lexemes can be distinguished. Constructions with nouns (a) and with adjectives or pronouns (b) likewise do not essentially differ from one another in the meaning of the verb; adjectives, however, occur as the second accusative at least twice as often. Comparison of translated texts with their originals, and the remnants of the corresponding constructions in folklore and vernacular speech, leave no doubt as to the autochthonous character of these constructions. In old writings, the following verbs most often serve as their main element:

1. Verbs of perception: see, perceive, catch sight of, hear, hear (of), feel, find.

a) And her representatives reported… that she saw the lord BrP I 401. And seeing that his son John was a capable man, he made him captain BrB I Mak 13, 54 (sahe | das sein Johannes ein tüchtiger Man war) But if anyone sees a strange woman, | a drunkard, | vainglorious and haughty DP 32₄₇ (iesli ktorego wyźrzy cudzolożnika pijánice | chćiwégo y pysznego W 51). The Lord, seeing him very talkative, | took offence (widząc go wielce mownego SP II 73₂₁. Who is Saint John… felt his own Lord Mž 179₂₉. I did not find your daughter a maiden BrB V Moz₂₂ (Ich habe deine Tochter nicht jungfraw funden)³⁰.

b) They did not let Mary in, | seeing that the matter was difficult BrP I 60. That you see them poor DP 484₁₉ (Gdy ie widzą ubogimi) For I saw you as righteous BrB I Moz 7, 1 (dich hab ich Gerecht ersehen)³¹. And Aaron… saw that she was leprous BrB IV Moz 12 (wird gewar | das sie aussätzig ist). He saw them unhappy BrB I Moz 40 (sahe | das sie traurig waren) I saw them stronger MP 120³². Some are wise BrB Sir 37, 25 (Mancher ist weise) For I found your works incomplete BrB Apr 3, 2 (ich habe deine Werck nicht voellig erfunden). That he finds the house empty (znáydzie Dō prożny) SP II 73, 1. He found Jesus (tykt) alone ChB Luk 9, 36³³.

2. Verbs of mental perception and activity: recognize ‘acknowledge’, know, understand, imagine, remember, consider, believe, regard.

That we… might acknowledge you as our Lord (ciebie Panem swoim wyznawali) PK 184. The Virgin… greets Jesus, recognizing him as the eternal God SIG 45₁₁. They recognize you… as the Church’s Father, whose Majesty is unshaken BrGD XXXVI (50)³⁴. What can one understand by the seven gifts of the Holy Spirit MP 271₃₆. With-

²⁹See also BrGD Rankr 5; GD II.

³⁰In such cases the second accusative sometimes occurs without the first as well: You hear the maiden, | you hear her humbled too DP 598₃₀ (Słyszysz pannę W 723). Pilate, having found him innocent, sends him to Herod. Mž 442.

³¹Similarly BrB Iz 58, 7; Pat 22, 29; I Mak 12, 28; 4, 21; Mat 25, 37; Apr 17, 6; DP 336₁₆; SP I 165 – 166; MP 79₄; SIG I 47₁₅, 52₉, 78₁₁, 123₅, etc.

³²Similarly BrP II 33, MP 69₈, III 15₂₅.

³³Similarly with find BrB Išm 8, 11; Sir 37, 21; Zuz 14; Dan 1, 20; Luk 7, 10; II Kor 9, 4; 1, 13; Mž 442; BrGD LXIII 103; DP 289₃₇; ChB II Kor 9, 4; Luk 7, 10; MP II 365₁₈, 366 – 367; SIG II 28₂₄, 83₁₃, 98₁₈, etc.

³⁴Similarly PK 27; Mž 47, 51, 82; BrGD XII (= RG 133), ChB Jon 17, 3; MP 104₁₄, 107, 146₈₂₄, 167₈₇, 186₁₀, 265₈₃₀, 283₄₁₀, II 347₈₂₄, 351₉₉, etc.

200

understood to be idols, which... the kings know, not gods, but the work of human hands BrB Bar 6,550 (not gods | but made by human hands). We fools thought... his end a disgrace BrB Išm 5, 4 (thought... his end a disgrace). Yet they regard it as foreign teaching BrB Hoz 8, 12 (Thus it is considered a foreign teaching). What then will they believe them to be, or call them? BrB Bar 6, 55 (How, then, should one regard or call them gods? LB) He... gave them to their makers and keeps them (them) by the ears (underlined) as his own masters BrB Sir 3, 8 (he held them as his masters).

b) The apostles knew the living Jesus... Mž 280. Thus we shall know God before us SE 232₂₃... And although they know them (them) to be lifeless, yet they follow after them BrB Bar 6, 41 (And although they know | that there is no life in them). They... know the sick woman and the illness that clings to her BrP II 387. Knowing the Lord to be good (Knowing the gracious Lord) PK 179. Remembering that the Lord is good | and merciful to all people (Seeing how good the Lord is | and how merciful to everyone) PK 83. Then Eli said that she was drunk BrB I Sam 1, 14 (Then Eli thought that she was drunk). And I considered her more precious than a kingdom BrB Išm 7, 8 (And I held her dearer | than a kingdom). And the Jews whom he had previously considered unworthy BrB II Mak 91, 15 (whom he had previously not considered worthy).

3. Verbs of information transmission: say, proclaim, state, reveal, confess, testify, show.

a) Why did you say that she was your sister? BrB I Moz 12, 19 (Why did you say that | she was your sister?). He had to confess him to be God BrB Išm 12, 27 (and had to acknowledge him as a god). Who, sending his Son, revealed the eternal Father | the Son | and the Holy Spirit MT 187₁₆. Christ shows us the perfect Savior SE 152.

b) Therefore all generations will call me blessed BrP II 259. For all the Gentiles will proclaim you blessed BrB Mal 3, 12 (So that all the heathen shall call you blessed). This writing... states faithfully | and is worthy of all acceptance DP 504₁₁ (Faithful | true | word | and... worthy W 857) Then let us confess also that God justifies us as righteous MT 185₁₁. I will confess you to be righteous SIG II 36₁₅. She... testifies him guilty of death RG 85₆. Then judge him to be unclean BrB III Moz 13 (he shall judge him... unclean). I will not hide my face from you (I will not turn away) BrB Jer 3, 12 (So will I not hide my face from you). And that he might show the riches of his glory ChB Rom 9, 23.

1.4. Besides verbs from the semantic groups mentioned above, the double accusative is also formed with so-called factitive verbs, denoting the creation or maintenance of a particular quality or state, e.g.: make, raise, call, let, consider, have, or simply an action performed on an object in a particular state, e.g.: accept, lead, leave, gather, give, eat, drink, and so on.

The second accusative of the noun is used in such constructions alongside the steadily spreading instrumental³⁵.

For you make the city a heap of stones BrB Iz 25,2 (For you make the city into a heap of stones). But I will make you a wilderness BrB Jer 22,6 (I will make you into a wilderness...). Make us a worthy vessel (make us into a worthy vessel)

³⁵ For further discussion of the relationship between the accusative and the so-called predicative instrumental in such constructions, see E. Fraenkel, Der prädikative Instrumental im Slavischen und Baltischen und seine syntaktischen Grundlagen, AfSlPh 40 (1926), p. 85 ff.; Kas., p. 152 ff., 199–200.

201

PK XXXI. These are my own children, whom I can raise BrB II 275. And do not call the burden a burden any longer BrB Jer 23,36 (Vnd nennets nicht mehr Last). And you will be called a visited and not forsaken city BrB Iz 62,12 (vnd dich wird man heissen die besuchte vnd un verlassene Stadt). I thought that I would take her as a companion BrB Išm 8,9 (Ich habs beschlossen | mir sie zum Gespielien zu nemen). He made... a house for Pharaoh’s daughter, whom Solomon had taken as his wife BrB I Kar 7,8 (die Salomo zum Weibe genommen hatte). So that he might now lead us astray and gather a people for himself, his chosen ones MP 207₂ (aby nas... obral lud sobie wybrány RP 141).

Among constructions with a second accusative—an adjective—those dependent on the verb daryti are especially productive in Old Lithuanian texts. They are often used to translate expressions in the originals that have quite different structures.

We make you brave Mž 564, Br Gd XXXV. I will make their churches unholy BrB Ezech 7,24 (Ich wil... ire Kirchen entheiligen). So that you might make her detestable MP 196₈ (aby cie omierził RP 133₁₃). Making them unworthy MP 257₈ (czyniąc ie niegodnemi RP 175₄₆). Promising to make you happy in them MP 375ₐ (was w nich szczęścić RP 253₄₇).

Adjectives are used much more frequently than nouns in constructions with other verbs of the meanings mentioned above.

The Gospel calls the poor blessed (zowie błogosławionymi ubogie) SP I 303₁₆. He calls every such person accursed MP 100₂₀ (przeklętym zowie RP 64ₐ₇). He... leaves the rich empty VEE 159₃₆ (Vnd lesst die Reichen leer). He makes a man good BrB Pam 7,29 (hat auffrichtig gemacht). And Joab leaves alive those who remain in the city BrB I Kron 12,8 (lies leben die übrigen)³⁷. We will capture them alive BrB II Kar 7,12 (wollen wir sie lebendig greiffen). And I will gather all of Jacob for you BrB Mich 2,12 (versamlen gantz). You led your people through them dry Vln 60–61 (trucken hindurch gefuehrt Lut. Cat. m. 132). He brought me back and forth in good health BrB Tob 12,3 (hat mich gesund hin vnd wider bracht). And fill her after she has drawn ChB Jon 2,7. I will drink it new in the kingdom of God ChB Mork 14,25. We will devour them alive BrB Pat 1,12 (wir wollen sie lebendig verschlingen). Give the child alive BrB I Kar 3,27 (Gebet jr das Kind lebendig). The wounds... which he bears open WP 9. He attends to my health PK 50. Simon... left them alive BrB I Mak 13,50 (lies sie leben).

1.5. Comparing the double accusative with the verbal word combinations discussed in 1.2, it is clear that the syntactic relations between the verb (V) and the first nominal or pronominal accusative (N) are the same in both cases. In the phrase he certainly says while carrying us DP 409₄₄, the pronominal accusative him, referring to the person being spoken about, is directly subordinated to the verb. The same syntactic relation between V and N is also seen in the double-accusative construction Why did you tell them that you were your sister BrB I Moz 12,19. In this construction, however, another nominal accusative (N′) is also associated with N, relating to it in a way similar to that of a nominal predicate to its subject; compare she is your sister. In both cases, the relation between V and N is direct, immediate and, of course, primary. The immediate V–N relation is also confirmed by the transformational interpretation in GD (see 3.3.2). By contrast, the relation between V and N′ is mediated, based

³⁶ Cf. also BrP II 259.

³⁷ Similarly with leave BrB II Moz 1,17; I Kar 20,31; II Kar 25,22; Ezech 23,29; 29,12; 16, 39; Luk 1,53.

202

N and N’ in a semi-predicative relation. It apparently acquired a decisive role in the structure of the construction at a later stage.

Thus, in constructions with verbs that could once take the accusative of a more distant object (e.g., know / say those dead), the same primary syntactic relations can be identified as with verbs requiring a direct object (e.g., see them pure). For the double accusative of both types, the same model of primary syntactic relations can be reconstructed:

(1) V ⟶ N ←╌→ N’

This pattern presupposes the existence of a sentence type in which nominal predicates could follow case forms subordinated to the verb. This is evident first of all from the structural analogy between the double accusative and constructions with other cases. Of these, the double genitive is closest to the GD; relics of it are found in old writings and folklore, cf.: Do not look at me so dark, for the sun has so scorched me BrB GG 1,6 (Look not upon me, because I am so black; quod fusca sim); I did not find you, for your work was complete ChB Apr 3,2. A postpositional dative is likewise joined to nouns and pronouns, cf. I wrote=(rašiau) to you, Fathers BrB I Jon 2, 13 (euch Vetern); I speak to you, friends ChB Luk 12,4; corresponding constructions with a second dative—a participle (e.g., so that… everyone would believe him who speaks the truth without any doubt SP II 138₂₂–his narrative) probably once laid the groundwork for the dative absolute³⁸.

1.6. Constructions with double cases, with a direct subordinating relation between V and N and a semi-predicative relation between N and N’, are characteristic of many other Indo-European languages as well, in their earliest periods of development. Latvian parallels are particularly noteworthy; cf. pazinu vedējus devīgus cilvėkus; saka mani dzērājinu and the like.³⁹ In Old Church Slavonic, Old Russian, Old Czech, and other Slavic languages,⁴⁰ the double accusative with nouns and adjectives follows verbs belonging to the same semantic groups and can be associated with verba percipiendi, cogitandi, dicendi, and the like; cf. Old Czech mniše jeho posla spravedlivého Ote A 347⁴¹, Old Church Slavonic gl(agol)ja vašъ rabъ Mar Jon 15, 15⁴². Constructions of the same type are attested in early Germanic-language monuments as well; cf. Gothic kunnands ina wair geraihtana (εἰδὼς αὐτὸν ἄνδρα

³⁸ For further discussion, see the author, LKK V, p. 9 ff.

³⁹ J. Endzelīns, Latviešu valodas gramatika, Rīgā, 1951, p. 577—578.

⁴⁰ For Old Church Slavonic and Old Russian, see А. А. Потебня, op. cit. 296 ff., 305; J. Łoś, Gramatyka starosłowiańska, Lwów — Warszawa — Kraków 1922, p. 166; T. П. Ломтев, Очерки по историческому синтаксису русского языка, М., 1956, p. 212—216, 220 ff.; В. И. Борковский, П. С. Кузнецов, Историческая грамматика русского языка, М., 1965, 392 ff., 472; Сравнительно-исторический синтаксис восточнославянских языков, Члены предложения, М., 1968, 112 ff.; В. И. Чагишева, О втором винительном падеже имени прилагательного и причастия в истории русского языка, Уч. зап. ЛГПИ, т. 373, 1968, p. 215; for Old Czech, see J. Gebauer, Historická mluvnice jazyka českého, IV, Skladba, Praha, 1929, p. 25—26, 319 ff.; F. Trávníček, Historická mluvnice česká, 3, Skladba, Praha, 1956, p. 123—124; for Old Belarusian, see Е. Ф. Карский, Белорусы, Язык белорусского народа, 2—3, p. 397—398; for other Slavic languages, see F. Miklosich, Vergleichende Grammatik der slavischen Sprachen, IV, Wien, 1868—1874, p. 388—389; W. Vondrák, Vergleichende slavische Grammatik, ²II, Göttingen, 1928, p. 273.

⁴¹ J. Gebauer, op. cit. 321.

⁴² B. Havránek, Voice (genera verbi) in Old Church Slavonic, “Studies in the Syntax of Old Church Slavonic,” Prague, 1963, p. 51.

203

δικαιον)⁴³, Anglo-Saxon wiste hine scyldigne “knew him to be guilty”⁴⁴. In terms of internal structure, the Lithuanian double accusative corresponds exactly to the Latin acc. duplex with the verbs videre, intelligere, credere, sperare, iudicare, nuntiare, praedicare, docere, e.g.: me… tristem credas Pl. As. 842; ille navem salvam nuntiat Amph. 988⁴⁵ malitiam sapientiam iudicant⁴⁶ and to the Greek acc. dupl. with verbs such as νομίζω, ἡγέομαι, ἀποδείκνυμι, which in this construction mean “believe, acknowledge, consider someone/something as something,” or ὀνομάζω, λέγω “call, say”⁴⁷; cf. also Old Indic ná vai hatám vírtām vidmá ná jīvám⁴⁸ “we do not know whether Vṛtra is dead or alive,” and corresponding constructions with verba percipiendi and dicendi in Hittite texts, e.g.: ŠĖŠ–YA GIG-an us-ki-ir “I heard my brother to be ill”⁴⁹. The structural and semantic parallelism of these constructions convincingly shows that already in the very distant past of the Indo-European languages, noun phrases dependent on verbs of perception and information expressed predicative relations, which later came to be expressed by subordinate clauses. The same relations are also found in constructions with other double cases.

Comparison with other Indo-European languages also confirms the original relation of direct subordination between V and N. In constructions with verbs of physical perception, this is obvious, since N is the direct object of V. A. Potebnia⁵⁰ and later scholars⁵¹ convincingly demonstrated that in the oldest monuments of the Slavic languages, verbs of mental perception and information (e.g., Old Russian vědati, mnĕti, znati, dumati, glagolati, povědati, ispovědati, skazati…) were likewise followed by the accusative of a further object, denoting the content of perception or speech, just as with vidėti and other verbs of physical perception; cf. Old Russian исповѣда... пакость. К.-П. Плат. 153, помышляю жену ПСРЛ I 179, сказа явление К.-П. Плат. 180⁵², оувѣдаша... погоню I Новг. лет.⁵³. The accusative after verbs of corresponding semantics is also found in other Indo-European languages (usually in constructions of a relic character); cf. Latin iuravit morbum “regretted the illness” and the fossilized formula Iovem lapidem iurare

⁴³ W. Vondrak, Vergleichende slavische Grammatik, ² II, 273.

⁴⁴ А. А. Мухин, Функциональный анализ синтаксических элементов, М.—Л., 1964, р. 119, 189; plg. dar J. Macháček, Complementation of the English Verb by the Accusative—with—Infinitive and the Content Clause, Praha, 1965, p. 64, 66, 68.

⁴⁵ R. Kühner, C. Stegmann, Ausführliche Grammatik der lateinischen Sprache, II, 1, Hannover, 1912 (hereafter — Lat. Gr.), pp. 294—295; O. Riemann, Syntaxe latine d’après les principes de la grammaire historique, Paris, 1927, pp. 74—75.

⁴⁶ A. Ernout, F. Thomas, Syntaxe latine, Paris, ²1953, p. 35.

⁴⁷ R. Kühner, Ausführliche Grammatik der griechischen Sprache, ²II, 1., Hannover, 1870, p. 273; С. И. Соболевский, Древне-греческий язык, М., 1948, р. 255.

⁴⁸ B. Delbrück, Altindische Syntax, Halle, 1888, p. 179.

⁴⁹ A. Hahn, Genesis of the Infinitive with Subject Accusative, Transactions and Proceedings of the American Philological Association (=TAPA), v. 81, 1950, p. 128.

⁵⁰ А. А. Потебня, Из зап., I—II, 295—298.

⁵¹ Т. П. Ломтев, ор. cit., р. 257—260; Д. Станишева, Винительный падеж в восточно-славянских языках, София, 1966, р. 34—40 with literature; ibid. for the other East Slavic languages, pp. 142 ff., 170 ff.; В. И. Борковский, П. С. Кузнецов, ор. cit., р. 472; Вступ до порівняльно-історичного вивчення слов’янських мов, ред. О. С. Мельничук, Київ, 1966, р. 466. For Old Czech, see also Д. Станишева, К проблеме лексических ограничений синтаксических моделей, Славистичен сборник, София, 1968, р. 33—34; for Old Polish, ibid., р. 39—40.

⁵² Д. Станишева, ор. cit., р. 37.

⁵³ В. И. Борковский, П. С. Кузнецов, ор. cit., 472, cf. the double accusative Князя своего живого вѣдяще (Ип. лет.). А. А. Потебня, Из зап., I—II, р. 305.

204

Fam. 7, 12, 2 (instead of the usual per Iovem); Catilinam atque illa portenta loquebantur C. Mil. 63⁵⁴; similarly, in Old Indian after verbs meaning “to speak” (e.g., yām attha “what are you talking about?”, i.e., a verb concerning a person)⁵⁵, “to think” (about something, alongside the acc. duplex and subordinate clauses), etc. Such constructions are motivated by the weakness and undifferentiated nature of syntactic government characteristic of the ancient Indo-European sentence, owing to which the accusative, like other cases, could be associated with a verb not only as a direct object but also as a more remote object, and with various adverbial meanings. On the basis of the constructions mentioned here, historically it is possible to distinguish, in double accusative constructions, a noun dependent on the verb and a predicative appositive (e.g., Lat. volo te + consulem⁵⁶).

The extensive use of nominal appositives is one of the most striking features of the sentence that catches the eye in monuments of the ancient Indo-European languages. Postpositive nominals could here be freely associated with various cases, expressing relations that we now convey by nominal or verbal predicates⁵⁷. In this way they formed secondary predicative centers in the sentence, semantically dependent only on the main verb. In form, such appositional nominals did not differ in any way from attributive ones, and in ancient texts their semi-predicative function can often be inferred only from position and context⁵⁸. (Lithuanian has preserved quite well the identity of ancient predicative and attributive nominal forms.) It is precisely this type of semi-predicative relation between nominal forms (or pronouns replacing them) that can be established in constructions with double cases. In the initial stages of their development, the subordination of the first nominal (N) to the verb was undoubtedly much weaker than in historical times and was primarily semantic in nature.

1.7. In constructions with double cases, participles occupy the syntactic position of the second nominal’s case (N’). Such participial constructions are, by origin, morphological variants of double-case constructions. The use of the participle in them is grounded in its nominal nature.

Historically speaking, GD is a separate instance of the double accusative. Since Lithuanian participles, like the corresponding forms in other related languages, are derived from

⁵⁴ R. Kühner, C. Stegmann, Lat. Gr., II, 1, p. 264; on the connection between such accusatives and the acc. duplex, cf. J. B. Hofmann, Syntaktische Gliederungsverschiebungen infolge Erstarrung ursprünglich appositioneller Verhältnisse, IF, 42 (1924), p. 79.

⁵⁵ B. Delbrück, Altind. Synt., p. 173; L. Renou, Grammaire sanscrite, vols. 1–2, Paris, ²1961, p. 288.

⁵⁶ Cf. J. B. Hofmann, op. cit.,79; A. Meillet, J. Vendries, Traité de grammaire comparée des langues classiques, Paris, 1924, p. 519 ff.

⁵⁷ Cf. А. Потебня, op. cit., p. 109–110, 121 ff., 516; H. Hirt, Indogermanische Grammatik, VI, p. 196–197; VII, p. 28; A. Meillet, Introduction à l’étude comparative des langues indo-européennes, Paris, ⁷1934, p. 359 ff.; P. Chantraine, Grammaire Homérique, II, Syntaxe, Paris, 1953, p. 8,16–17, 232, 319 ff.; Г. С. Кнабе, Еще раз о двух путях развития сложного предложения, ВЯ, 1955, № 1, p. 116; К происхождению абсолютных причастных оборотов в древнегреческом языке, сб.: Вопросы античной литературы и классической филологии, М., 1966, p. 430–431.

⁵⁸ On the lack of formal distinction between nominal predication and determination in ancient languages, cf. B. Havránek, in: Otázky slovanské syntaxe, II, Brno, 1968 (hereafter OSS II), p.410; V. Hrabě, K vývoji jednočlenných infinitivních vět v ruštině a češtině, OSS II, p. 353; R. Večerka, K problematice vět s jmenným a sponově jmenným přísudkem v staroslověnštině, OSS, II, p. 373; R. Mrázek, Nominální věty neslovesné v staré ruštině, Studie ze slovanské jazykovědy, Praha 1958, p. 33, 40, 44–45; Y. Millet, Les tours au gérondif et la fonction d’épithète, OSS, II, p. 399.

205

The deverbal adjectives, GD, and the syntactic relations among the nominal constituents of the double accusative must have been identical in origin. The structural similarity of these constructions is clearly perceptible even in the period of the written tradition. In sixteenth- and seventeenth-century texts, the second accusative of a nominal and a participle is often used alongside the same verbs; cf.:

We considered his life insane and his end shameful BrB Wis 5:4 (hielten sein Leben für unsinnig | vnd sein ende für ein schande). And whoever has enough to help his brother, and sees him in need BrB 1 John 3:17 (sihet seinen Bruder darben) When you see the righteous | then you rejoice and are glad: but when you see those doing evil, then you are not at all grieved or distressed DP 484₁₉ (Gdy ie widzą ubogimi... gdy ie widzą grzeszącymi W 183).

The same closeness between the double accusative with nominals and participles is also found in other languages—Slavic, Germanic, Indo-Iranian, Greek, and Latin—especially in their earlier historical periods⁵⁹. A characteristic example in this respect is J. Bretkūnas’s translation of Matt. 25:38–39 from German, in which the double accusative with a noun, an adjective, and two participles corresponds to the nominal constructions of the Latin and Greek acc. duplex:

When did we see you as a stranger, and give you food? Or⁶⁰ naked, and clothe you? When did we see you unable to be clothed⁶⁰; cf.: Quando autem te vidimus hospitem… aut nudum… aut te vidimus infirmum, aut in carcere.

Thus, the basis of GD consisted of constructions in which participial forms in the corresponding case, originally having the meaning of deverbal adjectives, were linked in a semi-predicative relation to nominals or pronouns subordinated to the verb. The mutual relations among the constituents of the double accusative make it possible to establish the original model of Lithuanian GD:

(2) V ⟶ N ←→ P

The same original syntactic relations can also be reconstructed for corresponding constructions in the ancient Slavic, Germanic, Indo-Iranian, Greek, and Latin languages. The identical structure of the double accusative with nominals and participles, its distribution, and its archaic character in many related languages suggest that this construction was inherited from the period of the Indo-European linguistic community.

1.8. Various other views have also been advanced concerning the formation and original structure of GD in the Indo-European languages.

Semi-predicative relations between N and P (respectively, N′), which many authors simply call predicative⁶¹, are sometimes derived from attributive relations.

⁵⁹ On the closer relationship of these constructions in early Latin, cf. J. B. Hofmann, op. cit., p. 79, 83; F. Stolz, J. H. Schmalz, Lateinische Grammatik (hereafter — Lat. Gr.), München, 1928, p. 384.

⁶⁰ Underlined; above: alba.

⁶¹ For Lithuanian, cf. E. Fraenkel, Kas., 152 ff.; E. Tangl, Acc., 39 ff.; for Latvian, J. Endzelīns, Latviešu valodas gramatika, Rīgā, 1951, p. 577; for Slavic languages, cf. А. А. Потебня, Из зап., I—II, p. 491, В. И. Борковский, П. С. Кузнецов, op. cit. p. 392—393; В. Л. Георгиева, История синтаксических явлений русского языка, М., 1968, p. 91; J. Loś emphasizes that the short participial forms clearly indicate the predicative function of the participle here, Gramatyka starosłowiańska, 1922, p. 159—160; for Old Indic, cf. B. Delbrück, Altindische Syntax, p. 178; for Germanic and Romance languages, cf. В. И. Ярцева, О взаимосвязях синтаксических отношений, сб.: Проблемы сравнительной филологии, М.-Л., 1964, p. 30—31; Исторический синтаксис английского языка, М.-Л., 1961, p. 228; Н. В. Соколова, Трехчленные глагольные сочетания в современном французском языке, сб.: Грамматика, фонетика и стилистика романских языков, Л., 1971, p. 256 ff. with references; В. Е. Коваленко, Именные средства выражения предикации, Львов, 1969, p. 48 ff. with references.

206

The view of semi-predicative (or appositional) use as a later stage in the development of participles was formulated by J. Jolly as early as the second half of the previous century⁶². On the basis of the Slavic languages, O. S. Melničuk explained the transformation of attributive relations into predicative ones by the expansion of the participle with dependent words⁶³, which resulted in the formation of a secondary predicative center in the sentence. Given that in ancient times predicative and attributive nominal words were not formally differentiated, the difference of opinion on this question could be regarded as merely verbal, and the N—P/N′ semi-predicative relations as having arisen from primary, undifferentiated attributive-predicative relations⁶⁴. That they could have developed from attributive relations is highly doubtful, since this direction of development is not characteristic in other cases either. The expansion of the participle cannot explain the formation of semi-predicative relations either, since in the oldest monuments of Indo-European languages, semi-predicative constructions are most often formed by unexpanded participles⁶⁵. On the other hand, the freer use observed in many languages in ancient times of nominal adjuncts close to the sentence, and their subsequent development into syntactically subordinated attributes, adverbials, or parts of the verbal predicate, points to the opposite direction of development. This is also confirmed by the transformation: the attributive function of nominal words is derived from the predicative, but not vice versa.

For the same reasons, the view that GD (or acc. duplex) originated from direct verbal complements, between which (semi-)predicative relations subsequently developed, also seems to require clarification.⁶⁶ Such combinations, in which a verb governs two nominal words in the same case form, are undoubtedly archaic. Since these nominal words could be related to one another by part-whole and other relations of semantic subordination⁶⁷, there is no reason to suppose that their semi-predicative relations, already characteristic of the oldest stages of Indo-European language development, arose later. Their later emergence would run counter to the main direction of development of the Indo-European sentence, namely, the strengthening of its internal cohesion and of subordinating relations. However, the more distinct differentiation of GD and other semi-predicative constructions from attributive constructions is undoubtedly of later origin, as is their subsequent formal separation, already observable in historical periods of Indo-European language development.

Quite different is the hypothesis that GD originated from nominal sentences, which, drawing on J. Zubatý’s views⁶⁸, F. Travníček developed on the basis of the Slavic languages⁶⁹, and which was later supported by some other Czech linguists⁷⁰. According to F. Travníček, Czech GD constructions such as vizu tě ležiec ‘I see you lying’ developed from two sentences—a verbal and a nominal one: vizu tě, leže ‘I see you, (you) are lying’, while the meaning ‘I see you lying’ was later and secondary. Initially, the participle had the function of a nominal sentence, but later became attached to the verbal sentence (vizu tě) and began to agree with the accusative complement (te⁷¹). Since GD, by virtue of its internal structure, is a transformation of two more elementary structures (cf. 3.3), this assumption could be applied only to very remote periods in the history of language,

⁶² J. Jolly. Zur Lehre vom Particip, kn.: Sprachwissenschaftliche Abhandlungen, Leipzig, 1874, p. 89—92.

⁶³ Plg. О. С. Мельничук, Розвиток структури слов’янського речення, Київ, 1966, p. 164 tt., 169—170.

⁶⁴ Plg. B. Havránek, OSS, II, p. 410.

⁶⁵ Plg. A. Hahn, Genesis of the Infinitive with Subject-Accusative, TAPA, 81 (1950), p. 117 tt.; Verbal Nouns and Adjectives in some Ancient Languages, „Language“, 42 (1966), № 2, p. 380—381.

⁶⁶ G. Curme, A Grammar of the English Language, v. 3, 32 tt.

⁶⁷ A. Hahn discussed extensively such cases of so-called partitive opposition in Old Indian, Hittite, Latin, and Greek: Vestiges of Partitive Opposition in Latin Syntax, TAPA, 84 (1953), p. 92—123; Partitive Opposition in Homer and the Greek Accusative, TAPA, 85 (1954), p. 197—289; Verbal Nouns and Adjectives in some Ancient Languages, „Language“, 42(1966), № 2, p. 388 ff.; see also F. Stolz, J. H. Schmalz, Lat. Gr., p. 630; for Old Russian, see В. И. Борковский, П. С. Кузнецов, op. cit., p. 452.

⁶⁸ Plg. J. Zubatý, Studie a článki, II, Praha, 1954, p. 107.

⁶⁹ F. Trávníček, Neslovesné věty v češtině, II, Brno, 1931, p. 157; Historická mluvnice česká, 3, Skladba, Praha, 1956, p.186, tt.; plg. dar Slavia 7 (1928—29), p. 818.

⁷⁰ Cf. R. Mrázek, Non-verbal Nominal Sentences in Old Russian, Studie ze slovanské jazyko vědy, Prague, 1958, p. 44; Deduction and Empiricism in the Comparative Typology of the Slavic Sentence, OSS II, p. 192. R. Večerka, On the Genesis of Slavic Constructions with the Participles Praes. Act. and Praet. Act. I, ”Sborník prací filos. fak. Brněnské univ.” A, vol. 7, Brno, 1959, p. 43 ff.

⁷¹ F. Trávníček, Hist. ml. č., 3, p. 187, 189.

207

when the structure of the simple sentence was entirely primitive⁷². However, syntactic reconstruction by no means indicates that the structure of the simple sentence was primitive in the Indo-European period. Within the framework of comparative Indo-European syntax, this assumption cannot be proved, since neither in the Slavic languages nor in other languages do we find any traces from which one could infer the transformation of the nominative of a nominal predicate into the accusative⁷³. On the other hand, the double accusative with nouns and participles, especially after verbs of perception, has been attested since ancient times not only in Indo-European but also in other language families (e.g., Semitic)⁷⁴. Sentences joined paratactically, such as

I see: you can (are able) may have been used alongside the GD since very ancient times I see you lying. Thus, the considerations raised here do not conflict with J. Zubatý’s ideas concerning the function of the second, nominal, predicate characteristic of participles, which are of fundamental importance for reconstructing the structure of the Indo-European simple sentence.

G. Jacobsson likewise proposed treating the Baltic and Slavic GD constructions with inflected participles as later, secondary developments⁷⁵. In his view, these constructions originated in an accusative phrase with the locative of a verbal noun (nomen agentis), which eventually developed into a gerund (converb), while constructions with an accusative participle were later and, at least in the Slavic languages, became established under the influence of other languages. This hypothesis, however, is strongly contradicted by the historically attested spread of indeclinable participial forms in the Baltic and Slavic languages’ GD, which does not allow accusative participle forms to be regarded as younger⁷⁶.

1.9. Although evidence from sixteenth- and seventeenth-century Lithuanian texts makes it possible to establish the ancient structure of the GD, most examples of this construction found in early writings have nevertheless moved considerably away from its original state.

From the very beginning of the written tradition, the GD differs in its use from the double accusative with nouns and adjectives. These differences become increasingly pronounced over time, creating the conditions for the GD’s distinctive development during the period of the written tradition.

First, the GD differs markedly in the extent of its use. In the sixteenth and seventeenth centuries, the double accusative with nouns was already a rare, relic construction; later it disappears entirely and is not used in the modern language (apart from isolated fossilized cases). The double accusative with adjectives is considerably more frequent in early writings, but is common and productive only after factitive verbs and other verbs of similar meaning (see 1.4). It is also unproductive with verbs of perception, speech, and mental activity; it gradually recedes from use and becomes a stylistically marked archaism.

The GD occupies an entirely different place among syntactic devices. In early writings it is highly productive and at least ten times more frequent than all the other double-accusative constructions combined. Its extensive use, grounded in the living language of the time⁷⁷, is one of the most striking features of sixteenth- and seventeenth-century Lithuanian syntax. The GD also differs from other instances of the double accusative in the abundance of morphological and semantic groups of forms that make it up. All this shows that in the period under discussion, the GD inflec-

⁷² On the possible origin of the GD through the contraction of two more elementary structures, see also J. P. Hughes, The Included Sentence Structure in Function Analysis, ”Word“, 23, 1—2—3 (1967), p. 336—337.

⁷³ Cf. the criticism of F. Trávníček’s hypothesis on the basis of the Slavic languages in B. T. Коломиец, А. С. Мельничук, ВЯ, 1957, № 5, p. 150; V. Barnet, Vývoj systému participií aktivních v ruštině, Praha, 1965, p. 42—43.

⁷⁴ M. Johannessohn, The Perception Sentence with Verbs of Seeing in the Hebrew and Greek Bible, KZ, 64 (1937), p. 146—147, 253—254.

⁷⁵ G. Jacobsson, The Slavic Active Participles. Original Structure and Interference, ”Scandinavica“, IX, Copenhagen, 1963, p. 132 tt; Славянска Филология, VII, София, 1965, p.49—50.

⁷⁶ Cf. the discussion of this hypothesis at the Fifth Congress of Slavists: О. С. Мельничук, in: Славянска Филология, VII, София, 1965, p. 56; it is also assessed critically in V. Barnet, op. cit., p. 43.

⁷⁷ On the autochthonous character and spread of the GD in the sixteenth and seventeenth centuries, see the author, LKK, 11 (1969), p. 41—58.

208

to distinguish it from the double accusative with adjectives and nouns, and to generalize it as a specific means of expressing a secondary action. This process is clearly reflected in the grammatical form of GD.

2. The grammatical form of the accusative with participle

2.1. The verbal member

The choice of verb plays a decisive role in the GD construction. It determines not only the relations of V with N and P, but also the latter’s semi-predicative relation, which is possible only after verbs of a certain meaning and valency. Therefore, the verbal member is to be regarded as the principal member, while the nominal and participial members are dependent members of the GD.

The principal member of the GD may be expressed by various finite and non-finite forms of the verb: participles, half-participles, gerunds, and infinitives. The syntactic relations between the construction’s members do not change as a result, e.g.:

[Lithuanian historical-language example] BrP II 32. Likewise [that way] [Lithuanian historical-language example] MT 1 142₈. [Lithuanian historical-language example] DP 482₃₅ (seeing our nature so desirous W 811). [Lithuanian historical-language example] BrP II 469.

By semantic analogy with verbs, GD is sometimes also formed by certain nouns and neuter adjectives used predicatively; see 2.1.4.3–4.

GD may also comprise two, three, or more principal members (a). On the other hand, a GD with two or more dependent members often depends on a single principal member (b).

a) [Lithuanian historical-language example] BrP II 23–24.

b) [Lithuanian historical-language example] MT 39₈. [Lithuanian historical-language example] BrP II 36.

Since the lexical and syntactic properties of the principal member determine the structure of the syntactic relations in the construction as a whole, it is first necessary to establish which verbs can serve as the principal member of GD.

When examining corresponding constructions in other languages, verbal members are usually classified according to their lexical meanings. This classification is based on the connection between a verb’s meaning and the internal structure of the construction as a whole, observed as early as antiquity. However, it is rather subjective, since the scope, number, and boundaries of individual semantic groups are determined at the researcher’s discretion and vary considerably even among descriptions of the same language. For example, authors classify English verbs forming corresponding constructions with participles and infinitives in different ways: some divide them into (a) factitive verbs and (b) verbs of perception, mental activity, desire, and the like¹; others into (a) verbs of emotion and sensory perception, (b) verbs associated with a durative state, and (c) verbs of mental state or activity²; still others again into (a) verbs of perception, emotion, and mental activity,

¹ H. Poutsma, A Grammar of Late Modern English, I, Groningen, 1928–1929, p. 340 tt.

² K. Yamakawa, The two Constructions: „Accusative and participle“ and „Genet. and gerund“, „Annals of the Hitotsubashi academy“, Tokyo, 1957, V. 8, № 1, 58 tt.

14. 520

209

b) causative, c) desire, d) other meanings³. In such circumstances, the same verbs (e.g., want, like, hate) are assigned to different groups by different researchers—by some to volition and emotion, by others to mental state or activity. Apparently for this reason, some authors dispense altogether with classifying these verbs and instead present them in a list. There have also been attempts to divide and define verb groups according to the distribution of N and P morphological forms within the construction itself⁵. This classification is more objective, but it would be of little use in studying Lithuanian-language data, since the same verbs here can form GD with various N and P forms, which differ only in their degree of productivity. Thus, it would be impossible to interpret semantically the groups of verbs distinguished according to their distribution within the GD construction⁶. Moreover, in that case it would be difficult to compare Lithuanian GD with corresponding constructions in related languages, where verbs are discussed in terms of semantics. In our view, the most useful approach is to discuss the words functioning as the main constituent of GD according to their shared valency properties, since this makes it possible to link distributional analysis with semantic analysis. The fact that the principal meanings of verbs are generally reflected in their syntactic environment⁷ makes it possible to motivate their classification formally.

Verbs occurring with GD are first classified according to the shared valency properties relevant to the nature of the construction under consideration; individual verb meanings are then identified and, in turn, defined in terms of their syntactic environment⁸.

The syntactic relations between the constituents of GD depend greatly on whether the main constituent permits sentential complementation⁹, that is, whether it can take a subordinate clause functioning as a complement and direct speech, e.g.: I know (it is known) that you will come; I said: you will come, and so on. Verbs that permit sentential complementation, called commentative verbs¹⁰ (e.g., see, hear, understand, feel, know, say), must first be distinguished from those that do not—

³ Л. С. Бархударов, Д. А. Штелинг, Грамматика английского языка, М., 1965, р. 380—383.

⁴ Л. С. Бархударов, Структура простого предложения современного английского языка, М., 1966, р. 104; similarly in Old Church Slavonic, R. Večerka, Syntax aktivních participií v staroslověnštině, Praha, 1961, 57 ff.

⁵ Plg. E. A. Nida, A Synopsis of English Syntax, London, The Hague, Paris, 1966, p. 155—156; Д. П. Новиков, Диахронический анализ нефинитных объектно-предикативных сочетаний, авт. канд. дисс., Л., 1969, р. 15 tt.

⁶ Attempts to define semantically verb groups distinguished according to the distribution of dependent constituents in corresponding English constructions also proved unproductive; see Д. П. Новиков, op. cit., р. 17.

⁷ Plg. Р. Мразек, Синтаксическая дистрибуция глаголов и их классы, ВЯ, 1964, № 3, р. 50 tt.; Ю. Д. Апресян, Опыт описания значений глаголов по их синтаксическим признакам (типам управления), ВЯ, 1965, № 5, р. 51 tt.

⁸ Thus, a semantic class is regarded as a set of semantic features incorporated into a syntactic rule; cf. R. T. Lakoff, Abstract Syntax and Latin Complementation, Cambridge — Massachusetts — London, 1968, p. 165, 213.

⁹ On sentential complementation, see Р. Мразек, op. cit. 56; the term sentential valency is also used for this purpose, cf. Н. В. Соколова, Трехчленные глагольные сочетания в современном французском языке, сб.: Грамматика, фонетика и стилистика романских языков, Л., 1971, 255 ff. On the possibility of taking a subordinate clause functioning as a complement as a valency property of the verb, see also W. D. Klimonow, Die Valenzen der russischen Verblexeme, die einen Objektsatz mit der Konjunktion что regieren, ZfSl, 13 (1968), H. 4, p. 632—639. G. Helbig and W. Schenkel regard such subordinate clauses, like many prepositional constructions, as category I constituents defining the valency of the verb; see G. Helbig, W. Schenkel, Wörterbuch zur Valenz und Distribution deutscher Verben, Leipzig, 1969, p. 36 et seq.

¹⁰ For this term, proposed by N. D. Andreyev, see Л. Ф. Ткаченко, О лексико-синтаксической валентности коммента́бельных глаголов во французском языке, сб.: Исследования по структурно-вероятностному анализу, Горький, 1971, р. 130 ff.

210

of verbs that take them, i.e. non-commentable verbs (e.g., to work, to do, to carry, to transport, to calculate, to call). The latter are generally factitive in meaning or close to it and, being transitive, denote the making, changing, naming, etc. of particular objects. Thus, the classification of verbs used in GD as commentable and non-commentable coincides in many cases with the traditional classification into non-factive and factive verbs¹¹, but rests on a clearer grammatical basis.

Constructions with commentable verbs are the most widespread and are characterized by distinctive usage and specific developmental tendencies; it is precisely these constructions to which the term accusativus cum participio has been applied since the time of ancient grammar. Among the general valency properties of commentable verbs, the relations most important for the structure of GD are their relations with the accusative of an object with concrete or abstract meaning and with the nominative of the subject.

On the basis of these relations, we distinguish four groups of commentable verbs, or other words that occur in their position.

Table 1

Groups of words forming GD

Group | Nominative of the subject | Accusative of words with concrete meaning | Accusative of words with abstract meaning | 1: + | + | + | 2: + | − | + | 3: + | − | − | 4: − | − | −

Group 1

2.1.1. Verbs associated with the nominative of the subject and capable of governing the accusative of words with concrete and abstract meaning

Most GD constructions are formed with Group 1 verbs (almost half of all examples found in 16th–17th-century writings), yet the number of lexemes assigned to this group is not large: only 23 verbs, counting prefixed forms as separate lexemes. They can be divided into smaller subgroups, namely: verbs of physical perception, verbs of showing, and several verbs of mental perception that occupy an intermediate position between Groups 1 and 2.

2.1.1.1. Verbs of physical perception are used most widely and form the basis not only of Group 1 but of GD itself. GD with verbs of this meaning was used in nearly all Indo-European languages; corresponding constructions are also characteristic of languages in other families from the earliest periods of their development¹².

Among verbs of physical perception, the following are used in GD in 16th–17th-century writings: to see, to catch sight of, to behold, to look, to catch sight of; to hear, to hear, to listen, to listen to; to feel,

¹¹ Plg. M. Deutschebein, System der neuenglischen Syntax, Leipzig, 1926, p. 272–273.

¹² Pvz., hebrajų, arabų, etiopų, sirų ir kt. semitų kalboms, žr. M. Johannesohn, Der Wahrnehmungssatz bei den Verben des Sehens in der hebräischen und griechischen Bibel, KZ, 64 (1937), H. 3–4, p. 146–147, 253 tt.

14*

211

to perceive; to find, discover, come across. GD is most often formed with to see (519 examples—the most popular verb in this construction), to hear, hear (303), to behold (207), to find (160), see Table 2. When taking an accusative with a concrete or abstract meaning, these verbs are not associated with an object accusative. GD with them is of two types:

a) N may be considered the direct object of V, while P is associated with both N and V. For example, in the sentence ghis… ischwida du laiwu stowincziu pas ezera VEE 91₁₂, there is a direct-object relation between ischwida and du laiwu, but the participle stowincziu is also semantically associated with V (as is confirmed by the original: sahe zwey schiff am see stehen). Given the nature of the relation between V and N, such constructions are hereafter called object constructions.

b) N cannot be considered the direct object of V and depends on it only together with P. In the sentence O ischw idens, Szmones bessiba idanczias ing Wandeni eiti… pereija per Wandeni BrB I Mak 16, 6 (sahe | das das Volck eine schew hatte), Szmones is not the direct object of V, since the sentence concerns not the observation of the people, but of their fear. In this case, the verb išvysti has a different meaning from that in example (a) and is close in meaning to suvokti, suprasti. Such constructions are hereafter called non-object constructions.

Since the distinction between cases of types (a) and (b) emerges only from context, situation, and the possible substitution of V with verbs from group 2, it is not always obvious. Sixteenth- and seventeenth-century texts contain many examples that may be considered intermediate and assigned to either type. However, the differentiation of these types played a major role in the subsequent development of the GD.

to see, to catch sight of, to see, to behold, to glimpse

a) And as she went into the grave | she saw a young man sitting at the right side DP 184₃₇ (wyžrząly młodzieńca siedzącego W 315). And when we shall see this great fire placed in the world and in the ashes RG 514₁₅. Every day we see people dying and being buried BrP II 6. And how can I behold my kindred perishing BiB Est 8, 6 (zusehen | das mein Geschlecht umbkome?) Sztay izmiwisk baymey mirsztanti KN 76₂₀.

b) When they see all the things of this world abandoned around them DP 554₃₂ (gdy wyżrzą nãd sobą… wsytkie rzeczy… opuszczone W 944) From this image we see that many prayers can be said BrP II 99. Then he sees her, a poor woman, who truly has no other means SE 75₂₃. And if the priest sees that the skin has not changed before his eyes, nor become more extensive, then it is unclean BrB III Moz. 13, 55 (sehen wird… das das mal nicht verwandelt ist). Having seen this, Mary, frightened by his greeting, is comforted by him BrP I 336.

to hear, to hear, to listen, to hear out

a) I hear you weeping, speaking to yourself SIG I 97₁₆. As a blind man, he heard the people passing by | (having heard the crowd passing by) SP I 365₁₁. The eternal oneness of God | listen to us who pray RG 253₇. May the king hear us who call BrB Ps 20, 10 (hear us | when we call).

b) So that he may hear every person who has recovered life RG 563₁₄. And in the spring, when Vespasian heard that Nero had died, he was in Caesarea BrP II 341.

212

to feel, to feel

a) Those who listen to him… feel how some kind of anxiety fills their hearts (dożnawáia ze iako ošći iake kola serce ich) SP II 83₁₈. So that we too may feel the raindrops falling to the ground DP 616₂₁ (ábbysm… vczuli krople na ziemię padające W 832).

b) The little darling is alive now | after the feast you will feel her dead and buried BrP II 516. And when you feel them¹³ grieving and fleeing, then you will chase them away BrB Judith 14, 5 (mercket | das sie verzagt sein werden vnd die flucht geben).

Among the GD examples with perception verbs, object constructions are somewhat more frequent than non-object constructions, although individual verbs here have their own tendencies. For example, the verb to see is most often used in object constructions, i.e. type (a) constructions; object constructions with to behold have a smaller advantage; by contrast, the verbs to hear, to hear occur more often in non-object, i.e. type (b), constructions (see Table 2).

The verbs used in constructions of both types differ not only in their syntactic relations with the accusative, but also in their lexical meaning. The verbs to see, to behold, to catch sight of and the like in type (a) examples mean “to experience through the eyes, by sight.” In type (b) examples, however, they no longer have this specific “seeing with the eyes” meaning, and mean “to understand, to comprehend” (cf. the fifth meaning of to see in LKŽ VII 916). The meanings of to hear and its synonyms differ in a similar way: in type (a) examples they usually mean “to perceive sounds with the ears,” while in type (b) examples they mean “to know, to learn, to experience” (cf. LKŽ III 330—331). Only in type (a) examples does the verb to feel also have a physical-perception meaning; in type (b) examples it already denotes mental experience rather than physical sensation.

The meanings of the verb to find, which is conditionally classified with perception verbs, and its synonyms are similarly differentiated in examples of both types: in object constructions (a), this verb means “to discover while searching or observing,” whereas in non-object constructions (b), its meaning approaches that of the verbs to notice, to feel (unawares), to understand.

to find, to discover, to stumble upon

a) He found him sitting in the church among the Doctors, listening to and questioning them (núležli… šiedzacego) SP I 195¹⁴. And she found him already dead BrP I 393. And he found… the child lying in the manger VEE 10 (und funden… das Kind in der krippen liegen.). The bridegroom came | and where he found them making merry | he took them for himself DP 57025 (ktore zostały gotowe W 969¹⁵).

b) We shall find another to be our great enemy BrP II 248. The spirit… speaks, returns home… And when it comes back, it finds them swept clean and adorned. ChB Luke 11, 25¹⁶. And when he returned, he found them sleeping SE 250³².

A meaning similar to that in type (b) examples can also be expressed by to find, to discover when they are associated with subordinate clauses of explanation, e.g.: He found that my bone was broken Rm. Rado [vaikas], ka barbaras rūgštus i grauż Krš. Atrádo lenkai,

¹³ Underlined, above: ios.

¹⁴ Similarly SE 25₁₆, Ev 22—23.

¹⁵ Similarly DP 570₈.

¹⁶ Similarly ChB Matthew 12, 44; VEE 56₄; SP II 44₁; SE 64₆.

213

what the colony now needs to have done. Pls. Nevertheless, this meaning of the verbs of speech is much rarer than the one found in “non-object” constructions. In forming GD, the verbs find, discover, more often than other verbs of perception, retain direct government relations with N and the literal meaning of finding a particular object.

2.1.1.2. Verbs of showing rodyti, parodyti, ženklinti ‘show, indicate, mark’ may occur with either an object in the accusative or a dative object. They also form both types of GD constructions. In object constructions they mean ‘cause to see’, while in non-object constructions they mean ‘make known, make understood’. Object constructions (a) are close to attributive constructions.

rodyti, parodyti, ženklinti

a) Show us the heavenly Father… and you, the Holy Spirit, dwelling in one Godhead PK 175. And he showed me the great city, holy Jerusalem, descending from Heaven from God BrB Apr 21, 10 (show me the great city… descending). The bush there… signified… only Mary giving birth and feeling no pain DP 398₁₉ (signified… Mary giving birth and feeling no pain W 654).

b) This bottomless depth… shows God to possess within himself all good (wyraža że Bog ma) SP I 92₁₉. Through your prayers God… shows them blessed SIG I 114₁₀. We will show the Lord… truly risen from the dead and alive BrP II 6. So that it might signify that these people love themselves | who are baptized | and themselves wish to be friends of such baptized people BrP I 144.

GD constructions in which rodyti has the meaning characteristic of examples of type (b) are related to constructions with išrodyti ‘show/prove’, which usually occurs only with abstract objects.

These five things… will show your one religion to be steadfast MT LŽ 7₁₃.

2.1.1.3. The verbs pažinti, atmintti ir laikyti ‘recognize, remember, and consider’ merit separate mention: they have a meaning of mental perception, but may occur with accusative nouns denoting either concrete or abstract entities.

pažinti

a) Tikėjimas… Pazista Sūnų Dié wistikelėse gūlinti | pazista ant’ médžio káranti pazistá mirstanti DP 426₅. Tawe… sweykiname | Ir su priewu gimusi pažistame SIG I 58₂₇.

b) Iau nu paszistam | tawe wissus daiktus szinanti VEE 79–80 (Nu wissen wir | das du alle ding wissest). Ir iżg cze iau pigái pažįsi moteriškę wirą apgóbiąnčią DP 399₂₉ (pozna niewiastę mężą ogarniającą W 656).

In object constructions, pažinti means ‘find out what someone or something is’, whereas in non-object constructions it means ‘understand, know’. This second meaning of pažinti, which is considerably removed from the first, is rare in contemporary Lithuanian and probably should be regarded as a Slavicism; cf. the meaning of Polish poznać, ‘to penetrate with the mind; to become acquainted with something; to learn something’¹⁷. (In the non-object constructions of M. Daukša’s Postil, the verb pažinti is translated by poznać and znać.)

The meanings of the verb at(si)minti in object and non-object constructions are less distinct from one another. In the former case, the reference is to keeping concrete things and people in memory; in the latter, to keeping actions, states, and concepts in mind. In object constructions, P most often has an attributive function and is associated only with N, not with V, and thus does not belong among the GD constructions discussed here at all. However, rem-

¹⁷ Słownik języka polskiego, VI, Warszawa, 1964, p. 1323.

214

also applies to cases in which N can still be regarded as the object of V, although semantically V is associated with the entire NP phrase.

remember

a) Let us remember, dear souls, our enemy, who takes God’s word away from many people BrP I 249. This multitude of holy angels remember great suffering… And God, who promised the world out of grace | to save BrP I 88.

b) Remember my life, which is but a breath, and how¹⁸ my eyes will not return to see Gero BrB Job 7, 7 (gedenck das mein Leben ein wind ist). Wieschpatie remember me as having lived well | with a heart full before your eyes | and having done what is pleasing to you BrP II 462. From these sorrows let us remember Christ, who truly rose from the dead on the third day BrP I 19. So that everyone may remember us, who have been put to death BrP II 406.

The verb laikyti ‘hold, not allowing to fall or escape’ in its primary meaning is non-commentable and is used only with nouns or pronouns of concrete meaning. With the meaning of mental perception, ‘consider to be something or to have certain qualities’, it is always trivalent, i.e. it requires, in addition to a nominative, two accusatives or (most often) an accusative with an instrumental; GD is formed with it comparatively rarely.

Mana Scheimatis ir mana Tarnaiteś laika¹⁹ mane Swetim²⁰ santi BrB Job 19, 15 (achten mich fur frembde). For they regard²¹ ghe prietelumi esanczę BrB Exod. 1, 16 (sie halten jhn fur Freund).

Such constructions are very close to GD with the group 2 verbs of mental perception (cf. 2.1.2.1.); this is clearly shown by the synonymy of laikyti and tarti in J. Bretkūnas’s Bible.

2.1.1.4. A comparison of the GD of types (a) and (b) formed by group 1 verbs shows that the syntactic valency of V differs between them. When these verbs have meanings characteristic of non-object constructions, i.e. those of type (b), they can no longer be associated with a single accusative of a concrete noun and thus move beyond the boundaries of group 1. Their mental-experience meaning can be realized only in phrases with abstract nouns or in three-member constructions. Thus, already in group 1, a gradual transition becomes apparent from verbs taking concrete objects to those that can occur only with abstract objects. In all the cases discussed here, the meaning of physical perception or of indicating concrete things is primary, whereas that of mental experience and indicating abstract things is secondary, having developed on the basis of the former.

This tendency to use verbs of physical perception with a secondary meaning of mental perception, apparently conditioned by general patterns of semantic development, is observable in many languages. For example, Latin video, audio, invenio (like modern French voir, entendre, trouver), German sehen, hören, finden, and English see, hear, find correspond to Lithuanian matyti, girdėti, rasti not only in their secondary meanings ‘understand, notice, experience, find out’, but also in that these meanings are generally realized in three-member constructions with a predicative constituent; cf. Latin inveniebat ex captivis Sabim flumen milia pas-

¹⁸ Underlined, above: iog.

¹⁹ Underlined, above: tar.

²⁰ Corrected to Swetim.

²¹ Underlined, above: taria.

215

suum X abesse Caes. B. G. 2, 16, 1, ²², cf. Old English Θu… fundest him treowe²³, Modern English He finds him guilty, or with a subordinate clause, cf. German die Betriebsleitung findet, daß sich die Arbeit verbessern muß²⁴. It is noteworthy that verbs of perception, precisely when they retain their basic, oldest meaning, form constructions corresponding to the original GD model (in which N functions as the direct object of V), whereas the abstraction of the lexical meaning of the verb is associated with the subsequent development of the internal structure of GD.

Group 2

2.1.2. Verbs associated with a nominative subject and capable of governing an accusative of abstract meaning

In 16th–17th-century writings, GD with Group 2 verbs is formed only slightly less frequently than with Group 1 verbs. Group 2, however, is the largest in terms of the number of lexemes: it includes as many as 66 verbs, of which the most frequent in forming GD are to know (264), to say with various prefixes (229), and to believe (118). We divide the verbs in this group into three subgroups: verbs of mental perception and activity, verbs of judgment and promising, and verbs of information.

2.1.2.1. Verbs of mental perception and activity, when used with the same meaning as in GD constructions, can govern the accusative of abstract declinable words, but do not govern the dative of the object. These include the verbs to know, understand, infer, fathom, comprehend, invent, misuse; investigate, examine, test; think, ponder, reflect; say “to think,” to imagine, dream “to think unrealistically,” to opine, imagine; believe; undertake, agree, seem, determine, decide. In meaning, verbs of group 1 used in the sense of mental perception in non-object (type b) constructions are also related to these.

The most common GDs with the verb žinoti account for almost 24% of the examples in this group and are used by various authors. By contrast, GDs with the verb sužinoti could not be found at all in 16th–17th-century texts; verbs usually used with a meaning close to this verb are ištirti, pertirti. In J. Bretkūnas’s writings, constructions with suprasti are often replaced by the synonymous verbs permanyti, numanyti, išmanyti; in SE, nuvartoti is used once with this meaning²⁵. With the meaning ‘to think’, J. Bretkūnas generally uses tarti in GD constructions; in the writings of other authors, we find isolated instances with the verbs dingoti, sapnuoti ‘to think unrealistically’, and nuomoniauti. The widespread use of GDs with tikėti is partly conditioned by the content of 16th–17th-century writings.

²² R. Kühner, C. Stegmann, Ausführliche Grammatik der Lateinischen Sprache, II, 1, Hannover, 1912, p. 697.

²³ J. Macháček, Complementation of the English Verb by the Accusative—with—Infinitive and the Content Clause, Praha, 1965, p. 70.

²⁴ G. Helbig, W. Schenkel, Wörterbuch zur Valenz und Distribution deutscher Verben, Leipzig, 1969, p. 208. The dictionary notes that finden has the meaning ‘meinen, feststellen’ only when it occurs with a subordinate clause. Similarly, phrasal valency is also associated with the secondary meaning of zeigen, pp. 204–205. However, the mental-perception meaning of schauen, sehen, likewise associated with phrasal valency, is not distinguished (cf. p. 105). H. B. С о к о л о в а, op. cit., pp. 253–255, also points to the less pronounced differentiation of the secondary meanings of corresponding French verbs voir, entendre, etc., compared with trouver. On the distinction between the primary and secondary meanings of the corresponding Old and Modern English verbs (see, hear, feel, etc.), as indicated by phrasal valency and the ability to occur with acc. c. inf., see J. Macháček, op. cit., pp. 43–44, 103 ff.

216

to know, to understand

But your Son … whom we know to stand up for us before your face KM 160¹⁴. And we know for certain that we shall all one day die BrP II 6. For you know that we are mocked because of you BrB Jer 15, 15 (du weissest | das wir umb deinet Willen geschmächt werden). For you already know that our grain is gathered from the field BrP II 395. I understood the listeners wishing to have the word of God in song Mž 81. No sooner had the Savior been born than, as everyone understood, she left home and set out BrP I 155. Then we shall understand that you are gods BrB Iz 41, 23 (so wollen wir mercken daß ihr Götter seyd).

to surmise, to perceive, to understand; to invent, to interpret as “understand”

He surmised that the new King of the Jews had been born BrP I 152. We can perceive that the Jews… had misunderstood the commandment BrP II 288. For some understood | that he had great need | to speak to the Jews IEzusa SE 248¹⁸. [The Stoics] invent God sitting in heaven | and writing decrees that bring destruction or ruin MT 177¹⁶. By interpreting Jesus the Son as… often before their disciples he remembered and said concerning the suffering… we can see SE 50¹⁶.

to investigate, to examine, to test

And you shall discover that your dwelling is at peace BrB Job 5, 24 (wirst erfahren | das deine Hütten friede hat). The Lord’s messenger… the first to proclaim Christ risen from the dead, finds out SE 79²⁴. We send… a Brother whom we have often tested (above: tried) in many matters, who is very diligent, and now even more diligent (in the margin: zealous) BrB II Kor 8, 22 (den wir oft gespüret haben in vielen stücken | das er vleissig sey | nu aber viel vleissiger). But the king, having tested the Jews as being so courageous, rejoices BrB II Mak 13, 18 (als er versucht hatte, das die Juden so freidig weren),

to think, to ponder, to conceive

Begin to imagine your Lord Jesus Christ stretched out and extended on the cross (imagine the Lord… stretched out) SP II 58⁴. Those who in their ears think themselves to be serving God… that service… is useless VEE 70²¹ (sich lesset düncken | er diene Gott). Human reason establishes | or imagines | that (235) this world will remain forever MT 235 – 235ᵃ. He conceived in his heart | that he would fulfill God’s command | and had now sent his son to earth BrP I 119

to utter ‘to think’, to imagine, to dream ‘to think unrealistically’, to suppose

For he will say that we are fleeing from them BrB I Joz 8, 6 (sie werden gedencken | wir fliehen für jnen). But do you imagine²⁶ that I want to eat beef PsD 93 (Meinstu das ich Ochsen–fleisch essen wolle). [The Jews] dreamed of a kingdom lasting on earth for a thousand years and possessing worldly power and glory MT 257⁷. The papacy… lays its trust in the notion that Christ’s body is falsely changed into bread… MT (PM) 30, 12.

to believe

And who believes you to be so very angry? PsD 164 (Wer glaubts aber | das du so zeer zörnest?). He… believes that forgiveness of sins is given to others, but not to himself MT 138¹⁸.

²⁵ Cf. D. Sutkevičius’s 1848 “Dictionary”: Nuwartoją intelligo, – Rozumiem.

²⁶ BrB Ps 50, 13: he says.

217

2.1.2.2. Verbs denoting decision and promise—undertake, agree, come to an understanding, determine, decide ‘decide’; swear, pledge, promise—govern a single object accusative relatively rarely; they are most often associated with an infinitive or with subordinate clauses of explanation. Examples of GD with these verbs are found only in J. Bretkūnas’s Bible.

undertake, agree, come to an understanding, determine, decide

And the Jews agreed to this and undertook upon themselves… to observe these two days, intending to celebrate them BrB Est 9, 27 (namens auff sich… das sie nicht vbergehen wolten | zu halten diese zween tage). They gathered together and agreed, intending to save their people BrB I Mak 3, 42 (vereinigeten sich | Sie wolten jr Volck retten). He conferred with them. Intending to march against him BrB II Mak 13, 13 (beratschlagt er mit jnen | Er wolte… jm entgegen ziehen). The priest must determine whether it is good or bad. Then he must decide that it is unclean BrB III Moz 13, 25 (Darumb sol jn der Priester vnrein vrteilen).

swear, pledge, promise

King Solomon, swear to me today that you will not kill your servant, who wishes it BrB I Kar 1, 51 (schwere mir heute das er… nicht tödte). And when they pledged strongly, not knowing where he might be²⁷would be BrB II Mak 14, 32 (sie schwuren | sie wusten nicht). He… promised them likewise, intending to persuade the king BrB II Mak 11, 14 (werhies jnen daneben | er wolte den König dahin vermugen).

2.1.2.3. Verbs of information can govern the accusative of abstract declinable words as well as an object dative, but are generally not associated with an infinitive. Of these, the following are used in GD in the 16th–17th centuries: say, tell, narrate, answer; speak, talk, persuade, state, mention, remind, announce, report, proclaim, publish, reveal, confess, acknowledge; testify, accuse, call out, defend ‘demonstrate’, lead, assert ‘affirm’, dispute, assert ‘state’, murmur, grumble ‘murmur’; write, record, read; partly related to these are also worship, praise, used in a sense close to the verb proclaim. GD with say is the most common; various authors also use GD with proclaim, testify, confess, state. On the other hand, some information verbs form GD only in the writings of one or another author and are not attested elsewhere; for example, speak is used this way only by J. Bretkūnas and “Suma Evangelijų,” write only by J. Bretkūnas, defend, assert only by S. Vaišnoras. Finally, there are verbs with which only one or another GD example has been found altogether, e.g.: argue, dispute, assert (in S. Vaišnoras’s writings), call out, pledge, announce (in J. Bretkūnas’s Bible).

say, tell, narrate, express, answer

Thus we say that they have fallen away from eternal salvation MT LĮž 6, 4. He says that we were raised from the dead on the third day SP II 252ₐ. And David’s servants feared to tell him that the child had died BrB II Sam 12, 18 (furchten sich jm anzusagen | das das Kind tod were) The Liepe, that the Hetmonga… was coming upon the earth, might be told to them KN 139₁₄. The letter… states that the Lord… sits at the right hand of God | and that he is the eternal high priest MT (PM) 18ₐ, The holy Angel first recounted all this to the delivered, born father (79) BrP I 79–80. They answered him that there was a very large crowd

27 ghis is crossed out.

218

to speak, to talk, to talk over, to testify, to mention, to remind

To the person suffering from bone disease | we speak of him whose bones… [are] very sore BrP I 188. And Simon… spoke ill of Onias, saying he had done such a thing that Heliodorus did not…give in BrB II Mak 4, 1. (redet dem Onia vbel nach | wie er solch vnglück gestiftet hette). Szietonas… Nor… talk over | the Lord God concerning them in their hardships and troubles, taking no care of anything SE 55₂₀. And Irenæus also speaks, lib. 4 cap. 34. Eucharist, that is | Supper, Christ… to the two things MT 153₁₆. And those Jews are much more righteous, whom Ambrose S. mentions as speaking thus DP 456₁₇ (wspomina tak mowiące W 761). But having mentioned God calling us through hardships to repentance | let us… BrP II 469.

to report, to make known, to prophesy, to announce, to proclaim, to reveal, to confess, to acknowledge; to worship, to praise

I announce to you that the LORD intends to build you a house BrB I Kron 18, 10 verkünde dir | daß der HERR dir ein haus bawen wil). Then God the Father himself… announced to the world | declared and revealed | this Child, Christ, | born BrP I 78. I spoke through… the Prophets… who at that time prophesied that I intended to send you against them BrB Ezech 38,17 (weissagten | daß ich dich über sie kommen lassen wolt). Her sons rise up and proclaim her blessed VEE 169₃ (preisen sie selig). They proclaimed you as having recovered and having a healthy body BrP I 194. So that he might reveal to you, man, that you do not live by bread alone BrB V Moz 8, 3 (das er dir kund thet | Das der Mensch nicht lebet vom Brot allein). The papists confess that the sacrifice is not bloody MT (PM) 8ₐ₁₆. The centurion acknowledges Christ to have soldiers beside him (przyznawa | że Chrystus ma…) SP I 258₁₉. When they saw the daughters, they praised them (marginal note: proclaimed them) as blessed (blessed). BrB GG 6, 9. And prais (above: proclaim) her sav (above: blessed) BrB Pat 31, 28 (preisen sie selig).

to witness, to testify, to call out, to defend, “to prove,” to bring forward, to affirm, “to assert,” to dispute, to posit, “to state,” to lie, to lament, to murmur

I testify | that I accept you MT 121ₐ₁₈. We testify against them on this day… we (that is, us) who asked them, to hand over the city to Holofernes BrB Judit 7, 17 (Wir bewegen… das wir euch gebeten haben). So that the people would not be able to cry out against him, Jonathan had therefore perished BrB I Mak 13, 18 (das das Volck nicht vberjn klagete | Ionathas hette derhalben müssen vmbkomen). The impious defend the holy one against God MT 10. By what arguments do you prove that all things will indeed be judged? MT 246ₐ₁₆. (Aristotle) He asserts that the world will remain forever MT 46₁₉. Origen… disputes that spiritual bodies will be resurrected MT 246₁. (The mockers) nevertheless assert that all the sacraments are of one kind MT 120₂₄. Who is a liar, if not the one who lies, saying that Jesus is Christ? BrB I Jon 2, 22 (der da leugnet | das Ihesus der Christ sey?). Prades — to lament such goodness of the Lord God… WP 244₃₇. Those who murmur against the natural order | and desire | themselves, by themselves, are not sinful MT 61ₐ.

to write, to copy out, to read

Thus the holy Luke writes that Gabriel spoke to Mary BrP I 135. For thus the holy Paul writes not only that Christ… gathered the Apostles around his table BrP I 64. And we do not read that there were many more

219

as many as four times, having spoken the Gospel S. DP 474₅₄ (we do not read *that… spoke* W 795).

2.1.2.4. Verbs assigned to group 2 normally govern only abstract objects. In GD, by contrast, they most often occur with accusative forms of nouns and pronouns with concrete meanings. This fact alone shows that in these constructions N, taken separately, is no longer the direct object of V. However, the connection with object constructions had not yet disappeared in the 16th–17th centuries. As with constructions containing verbs of physical perception, here too we observe transitional stages between object and non-object GD.

For example, such a gradual transition is evident between constructions with believe, where V is primarily associated with the concept denoted by N, e.g.: We believe Jesus Christ to have risen and opened heaven Mž 90, and those in which the connection of V with P or with the entire N-and-P phrase already predominates, e.g.: They believe only this life to be DP 584₃₁ (they believe only this life to be W 995); here the two types can sometimes be distinguished only on the basis of the broader context or the source used by the translator.

Since in the 16th–17th centuries verbs of mental perception were sometimes still associated with the accusative of a more remote object (cf. 1. 2.), the same relation can also be discerned in GD in which N denotes the content of perception. Thus, for example, the construction Knowing the Welikas to be there in two days BrP I 356 may still be considered object-like. The same applies to GD with verbs of information. In the sentence A host of angels… told them, the Child was beautiful | able to [be] on the field of shepherds BrGD VI the accusative Child may be regarded as the object of told (cf. Saul… said Christ to be chosen VEE 144₃). In the sentence describing the same situation, The angel recounts Christ to have risen BrP II 432, the object of V is the entire NP phrase. From here, it is only a short step to constructions in which a direct government relation between V and N can no longer be established, e.g.: tell them, his knighthood to have ended BrB Is 40, 2 (preach and | that their knighthood has come to an end).

That GD with verbs of information were inherently object-type is also shown by the connection between corresponding folk constructions, e.g., they say of me, the daughter-in-law, that I do not know my work (Žž), and relic word combinations preserved in folksongs: everyone says of me BM 463; the two of us, little people, speak, but not in the same words JV 61, and the like (cf. LKŽ V 114). Synonymous expressions with subordinate clauses, cf. my mother says of me that I do not know my work (Mrk), show that N’s dependence on V did not prevent the NP from being understood as a semi-predicative phrase.

In terms of valency properties, verbs of groups 1 and 2 are likewise not strictly distinct. With a secondary (figurative) meaning, group 1 verbs can occur with abstract objects. In certain cases, group 2 verbs may also still take concrete objects. However, combinations of group 1 verbs with concrete objects are predominant and productive, whereas corresponding combinations with group 2 verbs were already disappearing archaisms at the very beginning of written tradition.

The distinction between GD with group 1 and group 2 verbs is relative and historical. It reflects different stages in the development of GD and the process of restructuring internal relations within the construction. The structure of GD in groups 3 and 4 is already directly determined by the consequences of that restructuring.

220

3 grupė

2.1.3. Verbs associated with a subject in the nominative but incapable of governing the accusative

GD with verbs of this group are considerably less frequent and account for only about 6% of all examples in 16th–17th-century texts. They are nevertheless very important for the development of GD, because the intransitivity of the verbs makes changes in the construction’s internal structure particularly evident. Within this group, six subgroups can be distinguished, with different meanings and valency properties: reflexive verbs of mental perception, expectation, teaching/learning, mental and emotional states, and wanting.

2.1.3.1. Reflexive verbs of mental perception

have meanings similar to those of the non-reflexive verbs of group 2 (or those with a reflexive particle having dative meaning, e.g., vaizduotis ‘imagine oneself’), but differ from them in that they cannot govern the accusative. They generally do not govern other cases either. Besides GD, they may also be associated with subordinate clauses of explanatory sentences.

tartis ‘think’; išsimanyti, išsireikšti, ženklintis ‘mean’

The wicked will increase still more | who will call it a sin BrGD XLII (p. 72). Therefore he imagines us all to be very guilty BrP II 489. That salvation… might be expressed, the Church… the divine power being proclaimed | defended and preserved MT 182, 3. And when Pharaoh dreamed this a second time, he recognized God as truly showing him what was to come BrB I Moz 41, 32 (bedeut | Das solchs… thun wird).

The use of GD with these verbs is readily explained by their semantic similarity to verbs of group 2, whose governance of the accusative is already very restricted. For example, tartis here has the same meaning as tarti ‘think’, and ženklintis means the same as ženklinti (cf. 2. 1. 1. 2. and 2. 1. 2. 1.). Since GD is generally non-objective even with verbs of group 2, the semantic analogy extends further: following constructions with verbs, GD also comes to be associated with verb phrases of corresponding meaning that cannot govern the accusative. For example, following ženklinti(s), J. Bretkūnas associates GD with duoti ženklą, turėtis ženklą, gauti ženklą.

And he gave them a sign, God being their deliverer BrB II Mak 8, 23 (gab jnen ein Zeichen, das jnen Gott helffen würde) You shall have this sign from the LORD, the LORD being the one to do it BrB Iz 38, 7 (habe dir das zum Zeichen… das der HErr solchs thun wird). This sign you will receive from the LORD, the LORD being the one to do it BrB II Kar 20, 9 (Das zeichen wirstu haben… das der HERR thun… hat).

Similarly, following constructions with paskelbti, GD is used with padaryti garsą, following liudyti—with liudymą duoti, and following susitarti—with padaryti ryšį.

The whole congregation murmured, because they were making the land infamous BrB IV Moz 14 (they made an outcry | that it was evil). All these things, I say, provide strong testimony that God is the great Lord BrP II 177²⁸. And I… an eternal covenant29I will make with them, being their God, and they being my people BrB Bar 2, 35 (And will establish an everlasting covenant… | that I will be their God).

²⁸ Similarly, BrP II 88, 238.

²⁹ Underlined, in the margin: Siderejima.

221

2.1.3.2. Verbs of expectation tikėtis, daboti(s) ‘to watch, heed’, žvalgytis ‘to look around’ may be associated with the genitive³⁰, verb infinitives, or subordinate clauses of explanation. Like other reflexive verbs of mental perception, they are close in meaning to group 2, but differ in syntactic valency. The GD with tikėtis and dabotis is attested quite extensively in the writings of various authors (J. Bretkūnas, M. Daukša, S. Vaišnoras, B. Vilentas, and others).

to expect, to watch over, to look around

There are also many who expect the greatest bliss to be bestowed upon them through the grace and love of great lords DP 596₃₂ (who… think that the greatest happiness has been laid up W 202) Ames we hoped | that he would redeem Israel VEE 59₂₉. Look up here to the Lord, for he cares greatly for those who believe SE 68₂₃. It is worth looking: Sacraments are not suitable for anyone MT 117₈. Look at the heavens, and gaze up at the clouds, how (underlined, above them) they are higher than you BrB Job 35, 5 (look at the clouds | that they are too high for you).

2.1.3.3. Verbs of teaching and learning mokyti(s), pamokyti, išmokti, primokti may be associated with an object in the genitive and verb infinitives. They relate to the information verbs of group 2 in a way similar to that in which tikėtis, dabotis relate to the mental-perception verbs of group 2. Although mokyti, pamokyti can govern an accusative denoting the person being taught, the GD never occupies the syntactic position of this accusative. Meanwhile, the accusative denoting the object of teaching or learning³¹ is used relatively rarely. In 16th–17th-century GD constructions with verbs of teaching and learning, N can nowhere be regarded as the object of these verbs. These constructions are widely used in the writings of J. Bretkūnas and S. Vaišnoras and display variety in both structure and lexical expression.

to teach, to instruct, to teach; to learn, to study, to learn, to teach additionally

Schwenkfeldius teaches the human nature clothed in the divine MT 31₁₈ The Lord… teaches that little children have holy angels BrP II 430. The Lord… wishes to mark and instruct us all: this Child whom the Virgin Mary bore (226) being the true firstborn BrP I 226 – 227. Therefore, during these Christmas days, let us learn of Christ, born for us people BrP I 72. His body lay… until he learned that he had the highest power of God over the kingdom of men BrB Dan 5, 21 (until he learned that God… has power). We heard and learned that we are all greatly afflicted because of sin BrP II 246.

2.1.3.4. Verbs of mental and emotional states and their expression abejoti, nusistebėti, nusitikrinti, girtis, skųstis, linksmintis are associated with an object in the instrumental or with subordinate clauses of explanation. The GD is used infrequently with these verbs. Apart from somewhat more common constructions with abejoti, skųstis, only one or two examples of each are found (mostly in J. Bretkūnas’s writings).

to doubt, to be surprised, to check oneself

There is no reason to doubt that she, having received her own body and the blood of Christ with other learned mouths MT 156₅. And Pilate was astonished that Jesus had already died BrP I 394.

³⁰ For the use of dabotis with the genitive in the corresponding sense, cf. Look at the birds of the air, how they neither sow nor reap ChB Mat 6, 26.

³¹ See E. Fraenkel, Kas., p. 76.

222

Through faith… we assure ourselves that God is and will be present | to us sinners, a comfort SE 231–232.

to boast, complain, take comfort (“console oneself”), rejoice

He seeks a reason, how he might (underlined) boast. of himself that he is such as we are BrB II Kor 11, 12 (that they might boast | that they are like us) Alcimus… complained against Nicanor, saying that he was unfaithful BrB II Mak 14, 26 (accused… that he would become disloyal) We comfort ourselves, imagining ourselves living among the Gentiles as his followers BrB JerR 4, 20 (we wished… to live). I rejoice that you have returned safely to your home BŽ 496.

As in the cases mentioned in 2. 1. 3. 1., by analogy with verbs, GD is occasionally also associated with word combinations of corresponding meaning. For example, alongside GD formed with to assure oneself, a corresponding construction is also formed with to be certain: it is pleasant | being certain | that I am strong | for the Lord God | and being dear… serves SE 30, 10. Examples from J. Bretkūnas’s Bible illustrate the analogical formation of such GD: alongside a verb, a word combination of corresponding meaning is also recorded as a possible alternative: The man… spoke, believing (reckoning us to be spies) us to be spies of the land³² Szemes I Moz 42, 30 (and held us for spies of the land) The verb itself, to believe, although it can govern the instrumental, is closer to verbs governing the accusative of abstract nouns (see 2. 1. 2. 1.).

2.1.3.5. A separate subgroup, more distinct from the others, consists of GD with the verbs of wishing to want and to wish. These verbs can govern the genitive or a verb infinitive, but are associated only with subordinate clauses of explanatory sentences whose predicates are in the subjunctive (cf.: he sees that the guest is coming and wants the guest to come). These verbs also differ in meaning from those discussed so far. GD with them are more frequent in the writings of J. Bretkūnas and S. Vaišnoras; isolated examples also occur in the Wolfenbüttel Postil and Slavocinskis’s hymnal.

to want, to wish

Therefore she is also worthy of compassion WP 284. For who would wish the strongest to be weak MT 42₁₂ (Peter) strengthens the resolve of those | who wished to keep the Mosaic laws MT 208₁₄. Those whom he did not wish to see living in such a state, afflicted there… he wished to comfort MT L[ž 17₂₀ We want the Holy Spirit dwelling among us BrP II 142. The drunkard would wish that his neck be filled, not like me, fleeing, but struggling at this moment SIG I 61. I would wish that those coins had never existed BrB Tob 5, 24 (I would wish | that the money had never existed) Ah, I would wish my lord to be with the prophet in Samaria BrB II Kar 5, 3 (Ah! that my lord were with the prophet in Samaria).

As the examples show, GD with to want, to wish include constructions of highly varied structure, with participles of different kinds and tenses. In translated texts whose originals are available, no direct influence of the latter on this use of GD is apparent; on the contrary, in some cases the translation is particularly free. Only Latin could have exerted an indirect prompting influence here: after verba voluntatis, the accusativus cum infinitivo construction customary in Latin may have been associated by authors of early writings with Lithuanian GD. While revising the Bible manuscript, J. Bretkūnas in one place changed the subordinate clause of an explanatory sentence—

³² Crossed out; in the margin — Spiegoreis.

223

GD replaces the accusative after norėti ‘to want’: Bet jei noriu, kaip pasilikčiau (underlined, above: pasiliekanti) iki ataimi John 21, 23 (So ich wil das er bleibe);³³ in the Greek original and the Vulgate there is an accusative with infinitive, which may have suggested this usage (cf. ἐὰν θέλω μένειν, eum volo manere). However, in another revised passage of the translation, GD after norėti (Er nori scheschūli! deschimtį trapų toliaus aisenti II Kings 20, 9), used without any influence from the original, is replaced by a subordinate clause of explanation: Er nori iog scheschūlis … eitu (Sol der Schatten zehen stuffen forder gehen, Vis ut ascendat umbra). We have no evidence that would allow us to determine whether such constructions were borrowed; on the contrary, their rather varied and free use in early writings suggests that they were not alien to spoken Lithuanian in the 16th century.

The genitive government of norėti ‘to want’ (like that of other verbs expressing striving or purpose) is an archaic feature inherited from the period of Indo-European unity, characteristic of many related languages³⁴. The accusative with this verb is rather rare, but when the desired object is designated as a whole, it was also familiar in Baltic, Slavic, Germanic, and other Indo-European languages. Moreover, in many of them we find GD with verbs of the same meaning as Lithuanian norėti; cf. Latvian gribēj mani drīz augam BW 1923³⁵, Latin te volo monitum; patriam exstinctam cupit³⁶, German ich möcht ihn gern getödet³⁷; corresponding infinitival constructions are even more widespread³⁸. Under such circumstances, it is not impossible that such GD constructions once arose on the basis of old object constructions and gradually disappeared as the verb norėti ceased to be associated with the accusative.

Group 4

2.1.4. Verbs and other words and word combinations that are not associated with a nominative subject and cannot govern the accusative

With words from Group 4, GD is formed least frequently in the early writings—only about 2.7%. All examples assigned to this group are from the works of J. Bretkūnas and S. Vaišnoras, that is, from the sources in which GD is used most extensively and variously. Moreover, the works of J. Bretkūnas alone account for approximately 40% of all the 16th–17th-century texts examined.

In addition to finite and non-finite verb forms, Group 4 also includes certain modal words and nouns that can take a subordinate clause of explanation (i.e., that share with verbs the property of phrasal government). When associated with words in this group, the N and P construction retains its accusative form but occupies a syntactic position that is no longer accusative, but nominative.

³³ This construction was subsequently carried over into the Postil as well; cf.: Iei noriu ana pasiliekanti iki ataimi BrP I 411.

³⁴ B. Delbrück, Vergl. Synt., I, p. 324—327.

³⁵ J. Endzelīns, Latviešu valodas gramatika, Rīgā, 1951, p. 1001.

³⁶ A. Ernout, F. Thomas, Syntaxe latine, Paris, ²1953, p. 282; R. T. Lakoff, Abstract Syntax and Latin, Complementation, Cambridge – Massachusetts – London, 1968, p. 228—229.

³⁷ O. Behaghel, Deutsche Syntax, II, Heidelberg, 1924, p. 417.

³⁸ For Greek, cf. R. Kühner, Ausführliche Grammatik der griechischen Sprache, Hannover, 1870, p. 593 ff.; С. И. Соболевский, Древнегреческий язык, М., 1948, p. 334—335; for Latin, R. Kühner, C. Stegmann, Lat. Gr., II, 693—694; for Anglo-Saxon, J. Macháček, op. cit., 40 ff.

³⁹ Corrected above in Z. Blotno’s hand: Regetis jums.

224

Group 4 includes impersonal reflexive verbs, neuter forms of passive participles, constructions with modal meaning, and constructions with verbal and other nouns.

The few examples of Group 4 are significant because they show the extreme limit of the expansion of GD usage and, in this respect, are unique among participial constructions of corresponding structure in the Indo-European languages.

2.1.4.1. Impersonal reflexive verbs regisi, rodosi, įsirodo have meanings similar to those of the corresponding non-reflexive verbs, which govern the accusative and often form GD; also belonging here is sekasi, used twice by S. Vaišnoras in the sense ‘it works out, it seems’. Since only 8 such constructions have been found in all the texts examined, they are presented in full.

regėtis, rastis, įsirodyti, sektis

Schittap regesi mums | anus išch gerros schirdies klaususius BrP I 44. Regissi mane walgantį ir gerinti su jumis BrB Tob 12, 19 (Es schein wol | als esse ynd trüncke ich mit euch). Regissi newerta santi Baszniczo Chrikscronischko tą abijoienti klausimą MT 258⁸. Iei pirmaghi uschrakinima kalbos galli priewesti | tad regissi nepau-dinamą gruntą santi? MT (PM)8^8–10. Pirm ne ie pardawė, radosi ghę nieschzę sanczę BrB Mat 1, 18⁴⁰ (Erfand sichs | das sie schwanger war); Vszmokiek tiemus, kurie tawes lauk⁴¹, idant issiroditų tawa Prarakus tiesą kalbeiusius BrB Sir 36,18 (das deine Propheten wahrhafftig erfunden werden). Ischtogi sekassi Moyšiechaus rądime didę sandorą pirm Patapą su tewais Schwentaisėis buwsse MT LĮŽ 14⁵. Tad sekassi neatkalbėtinai | tą Mische sanczę ne nūg Christaus įstatita | bet… nūg wissų welmu isch Peklas atneschtą ir jwestą. MT (PM) 2¹⁵.

2.1.4.2. Neuter forms of passive participles, which also have an impersonal meaning, form GD much more frequently. They are derived from transitive verbs in Groups 1 and 2.

Neuter forms of present-tense passive participles are mostly formed from the verbs regėti, žinoti, kalbėti; in isolated cases, from rasti, perman yti, sakyti, rašyti, tikėti, suprasti, giedoti.

regėti, žinoti, kalbėti, rasti, perman yti, rašyti, sakyti, giedoti, tikėti, suprasti

It is seen from this, that they are⁴² not gods BrB Bar 6, 14 (daran ir ja mercken könnet | das es nicht Götter sind). And it is known that the Son of God was sent not in order that… we should perish MT 137³. This is certainly known and believed: that the immortal shall not die MT 258⁸. Thus it is commonly said: in the Godhead there is one spiritual essence MT 7¹¹. It is also found in the writings that Jeremiah (above: -schu) had instructed the prophet BrB II Mak 2, 1 (Man findet auch in den Schrifften | das Jeremias… geheissen habe). So that forgiveness of sins might be possible and reconciliation with God might be compatible with the grace bestowed upon us MT 79¹³. It is written | said | believed | and sung… that he himself rose from among the dead on the third day BrP II 6. Therefore now

⁴⁰Corrected in Bretkūnas’s own hand from ana nieschęze buwa. In the NT, a subordinate clause with jog is used here; in the 1823 Bible, with kad; the GNT, following the Latin text, has ira nesczia (inventa est in utero habens).

⁴¹ -tą deleted.

⁴² Underlined, above: ios.

15. 520

225

is understood to be | willing BrB Bar 6, 50 (Darumb kan man mu fort wol erkennen | das es triegerey ist).

Thus, in impersonal constructions with neuter participles, the lexemes to see, to say, and to know predominate; these most often form an accusative-with-participle construction in general. This confirms the analogical origin of these constructions. Only after, in constructions of the type I know my brother to have returned, the government relations between I know and my brother had faded, could the phrase my brother having returned, understood as equivalent to that my brother returned, be associated with it is known in the following way:

I know (that) my brother returned: it is known (that) my brother returned = I know my brother to have returned: x; x = it is known that my brother returned

The neuter forms of past-tense neuter participles are mostly formed from to say, to tell, to answer, and in isolated cases from to write, to write down, to suppose, to know ‘to recognize, to learn’.

to say, to tell, to answer, to write, to write down, to suppose, to know

It was first said | that the old Church had been this way MT 145ᵃ₂. On the third day it was told to Laban that Jacob was fleeing BrB I Moz 31, 22 (wards… angesagt | das Iakob flöhe). But it was answered to them that such a thing was impossible MT 266ᵃ₁₈. Likewise, it was written that the commandment was given to all the lands BrB Est 3, 14 (Also war der inhalt der schrift | Das ein Gebot gegeben were). It was written | that the king… speaks thus BrP II 472 But in Moses’ books it is written that God hardened Pharaoh’s heart? MT 174₁₉. It may also be inferred from this that they are not gods BrB Bar 6, 71 (Auch kan man es daran mercken | das sie nicht Götter sind). For what should have been known from this? That I and your people had found grace BrB II Moz 33, 16 (sol… erkannt werden | das ich vnd dein Volck… gnade funden haben).

Neuter forms that form an accusative-with-participle construction can themselves be dependent constituents of another such construction. Three instances of this kind have been observed in all:

I have heard it said of you that you have the spirit of the holy gods, and that enlightenment, understanding, and great wisdom are found in you. BrB Dan 5, 14 (Ich habe… hören sagen | das du den Geist… habest | vnd Erleuchtung… bey dir funden sey). The people of Abas heard it said that Zimri had made a conspiracy and killed the king BrB I Kar 16, 16 (höret sagen | das Simri einen Bund gemacht vnd auch den König erschlagen hette). And they found written in the Law which the Lord had commanded through Moses BrB Neh 8, 14 (Sie funden geschrieben im Gesetz | das der Herr… geboten hatte).

In one passage of J. Bretkūnas’ Bible, an accusative-with-participle construction is formed in accordance with constructions with neuter forms, but with a masculine form of a past-tense neuter participle:

Thus I will trample it upon the earth, as its foundations will be seen, it lying there Ezech 13, 14 (das man jren Grund sehen sol | das sie da lige).

The neuter form of the necessity participle serving as the main constituent of an accusative-with-participle construction was observed in only one example: Thus it is to be seen | that there is no greater sin before God than being terribly wicked MT 196ᵃ₁₄.

Constructions dependent on neuter forms of neuter participles are not a feature specific only to the language of J. Bretkūnas and S. Vaišnoras. J. Rėza, who edited volume VI of J. Bretkūnas’ Bible, carried over into his “Psalter of David” a construction associated with it would be seen: the true God being in Zion (Ps 84, 8), which in the manuscript replaced the explanatory—

226

the subordinate clause of the main sentence that the true God is in Zion (cf. PsD 152; LB: Das man sehen mus | der rechte Gott sey zu Zion). Corresponding constructions (but now with gerunds) are also found in contemporary Lithuanian, cf.: It is not for nothing that people say that the fox is cunning Vaižg RR II 173.

2.1.4.3. In several passages of S. Vaišnoras’s „Žemčiūgos“, GDs are formed by constructions with the modal meanings it is certain and it is true, which in meaning are close to predicates with impersonal neuter forms of participles.

it is certain

It is certain | that the Son of God will come to judge people now on the earth 243ᵃ. It is certain then that the Son of God was sent by the Father for this purpose (106ᵃ), so that he might rule over the world 106ᵃ–107.

it is true

And it is true | that the genuine and everlasting coming … cannot be excluded from the most precious possession 155ᵃ⁻¹⁰. For it is true | that unbelievers deny the promise of grace by their faithlessness 168¹.

2.1.4.4. Nouns, usually deverbal, sometimes serve as the principal member of a GD, and a subordinate explanatory clause may be associated with them. These nouns are most often derived from verbs used in GDs (or from verbs of related meaning). In nearly all cases, syntactic relations of analogy with constructions in groups 2 and 3 can be discerned here.

For example, following GDs with to testify, to give testimony, constructions are formed with the nouns testimony, witness.

testimony, witness

God established that union … so that it might be a clear testimony that the bride is living a holy life MT 221₈. So that the suffering and death of the saints might be testimony about the teaching | that the true Gospel teaching is from the heart MT 181₈. You are witnesses this day that I have bought all that belonged to Elimelech BrB Rut 4, 9 (and seid… Zeugen | das ich alles gekauft habe).

Similarly, the noun smoke is associated with to smoke; likewise to trust — trust, to mark — mark, to marvel — wonder.

smoke, trust, mark, wonder

That most wicked smoke takes in | the teaching given by GOD MT 3ᵃ₉. Faith is strong trust in the Lord God | who has let us humans | love | cherish | cling to all things | wishing to save BrP I 286. What is the sign that I shall go up to the house of the LORD BrB Iz 38, 22 (Welch ein Zeichen ist das | das ich hinauff… sol gehen?) This is the miracle of the Lord speaking thus BrB I Kar 13, 3 (Das ist das Wunder das solches der HERR geredet hat).

Other nouns are related to the verbs occurring in GDs more by meaning than by derivation. For example, the GD with sound, ‘news’ is associated with corresponding constructions dependent on verbs of informing, such as to announce, to report, to call out and so on (see 2.1.2.3.); similarly, the noun sentence is related to to teach, to say; ‘gramotos’ with ‘to write’.

15*

227

sound, sentence, thunderclaps

And there arose a false (above: unjust) report that Antiochus had died BrB II Mak 5, 5 (Aber es kam ein erlogen geschrey aus / wie Antiochus solt tod sein). This sentiment must be accepted from the bottom of the heart: that God is no adversary of sin, nor wishes it, nor appoints it... MT 54a. Then letters were sent... with the command to the people that each man should be master in his own house BrB Est 1, 22 (Da wurden Brieue ausgesandt... das ein jeglich Man der Oberherr... sey).

That such constructions were familiar to the authors is evident from BrB II Sam 13, 30: a report comes to David that Absalom has slain all the king’s children, where the subordinate clause of explanation that Absalom had slain all the king’s childrens is replaced by an GD, when correcting the manuscript. Neither the German nor the Latin text could have prompted this correction, cf. kam das gerüchte für David | das Absalom hette... erschlagen; fama pervenit ad David dicens: Percussit Absalom...; the influence of these languages is likewise not apparent in the other cases mentioned. Some of these constructions are close to GDs dependent on the neuter forms of impersonal participles, cf. the cases discussed in 2.1.4.2 with it was said and this example: And when that report was told in Pharaoh’s house, that Joseph’s brothers had come, it pleased Pharaoh. BrB I Moz 45, 16. (Vnd da das geschrey kam... das Iosephs brüder komen weren). Similarly, it was written (see 2.1.4.2.) corresponds to it was in the writing in this example: And this too was in that writing: the prophet had taken it... and commanded them... to take the casket along. BrB II Mak 2,4 (So stund das... in der selben Schrifft / Das der Prophet... sie geheissen habe)

Thus, the nouns forming the GD in group 4 are closely related in meaning, and often also in derivation, to the verbs of groups 2 and 3. However, the syntactic independence of the construction is even greater in this group: the use of NP in the subject nominative position shows that not only the genitive, but also the function of an object case in general, has been entirely lost here.

It is noteworthy that such independent usage is characteristic precisely of those authors—J. Bretkūnas and S. Vaišnoras—in whose writings the GD is most widespread. The construction’s greater productivity apparently stimulated its development toward syntactic independence.

2.1.4.5. The independence of the GD was so strongly felt by J. Bretkūnas that in some places he even tried to use it with the meaning of a separate sentence or a coordinate clause. In vol. I of the Postil we find the heading: On the First. We all being servants (487), although there is no verb or other word anywhere nearby with which it could be connected; only from the sense can we infer the implied we must know (remember) or something similar.⁴² A coordinate clause is rendered by a GD in this example: And he (Ptolemy) sent messengers to Demetrius, that he might come to him, to make an agreement with him; then he would give him his daughter. BrB I Mak 11,9 (Vnd schickt Boten zu Demetrio | das er zu jm komen solt | einen Bund mit jm zu machen | So wolt er jm seine Tochter geben); the Latin text offers no prompt for using a GD here either (cf.: Veni, componamus inter nos pactum; et dabo tibi filiam meam). That NP was understood as a separate predicative unit is also evident from the examples,

⁴² NP is used entirely without V also in BrB Job 36, 19: But what profit would your wealth or gold bring you, but here tari was most likely omitted through oversight, you think or something similar (cf.: Meinſtu dus er deine gewalt achte).

228

where a whole chain of subordinate clauses intervenes between it and V. For example, in BrB Neb 10, 30, after the main clause And they came to make an oath and a sworn promise to walk in the law of God comes a restrictive subordinate clause linked to the law, then two long subordinate clauses of explanation linked to came (so independent that both M. Luther and J. Bretkūnas set them off with full stops), followed by another second-degree clause of explanation, and only then the GD:... And when the people brought provisions and all kinds of merchandise to sell on the Sabbath, we would not take any of them. (das wirs nicht von ihnen nemen wolten).

2.1.4.6. A case in which an interjection serves as the main member of the GD deserves separate mention: Alas, that I was born to see the ruin of my people and of the holy city, and must remain silent BrB I Mak 2,7. In the German original we find a subordinate clause with daß (Ah | das ich dazu geboren bin), and in the Latin one with ut (ut quid natus sum); thus, the GD may have been used here as the equivalent of a subordinate clause, as in other examples of group 4. However, sporadic instances of the accusative after interjections in old writings (cf. Sapūnas and Šulcas’s Ay Budą szio Čieso — O mores hujus temporis... Ak manę Biedną. O me miserum 94⁴³), its use in Latvian folklore (cf.: ai tu manu lētu dabu!⁴⁴), and abundant parallels in Slavic⁴⁵, Latin⁴⁶, and other languages allow us to suppose that the accusative after interjections was once not alien to Lithuanian either; in that case the cited example could be considered a relic of object constructions with a dependent accusative, structurally corresponding in its original form to a GD with say and similar verbs (cf. 2.1.2.3.).

2.1.5. A review of GDs according to their main (usually verbal) member shows that most examples found in sixteenth- and seventeenth-century writings no longer conform to the original model of syntactic relations in this construction. In the groups distinguished by the general valency properties of the verbs, the relation between V and NP varies. Direct government relations between V and N are clearly evident only in the object (type a) constructions of group 1. Such relations can sometimes also be discerned in constructions of groups 1b and 2, but here they are no longer active or productive; rather, they are reconstructed as a legacy of earlier periods. In most group 1b and 2 constructions, NP already depends on V as a separate syntactic unit. NP is even more independent in group 3 examples, where the government of an object N can no longer be established at all. Finally, in group 4 NP is used even in the nominative case position. Thus, moving from group 1 to group 4, we see a gradual change in the syntactic relations between NP and V, which is more likely to reflect the actual development of the GD. Its essence is that case (accusative) government, characteristic of the original GD model, gradually turns into phrasal government of the entire NP phrase, indifferent to its case form.

The change in the government relation between V and N is also indicated by the fact that after verbs with negation, the accusative N here most often does not become a genitive of negation, as in other cases of direct government, e.g.:

⁴³ Žr. A. Bezzenberger, Beiträge zur Geschichte der litauischen Sprache, Göttingen, 1877, p. 239; E. Fraenkel, Kas., 151.

⁴⁴ J. Endzelīns, Latviešu valodas gramatika, Rīgā, 1951, p. 580 (§ 436).

⁴⁵ Fr. Miklosich, Vergleichende Syntax der slavischen Sprachen, Wien, IV, 1868–1874, p. 392; W. Vondrák, Vergleichende slavische Grammatik, II, Göttingen, 1928, p. 275; F. Trávníček, Historická mluvnice česká, III, Skladba, Praha, 1956, p. 9.

⁴⁶ F. Stolz. I. H. Schmalz, Lat. Gr., p. 385–386.

229

Why did you not tell me that your wife was your sister? BrB I Moz 12,18 (Warum sagestu mirs nicht | das dein Weib were?). You did not send Christ risen from the dead as a messenger BrP II 21. So that you would not call him dead’ RG 539b. But among the Lithuanians, such a thing was said | I do not know BrP II 517. If you do not want to send all my troubles upon yourself this time BrB II Moz 9,13 (Ich wil anders dis mal alle meine Plage über dich selbs senden)

The genitive form with negative verbs is more characteristic only of object constructions (after verbs of perception and the like), in which the governing relation between V and N is still clearly felt, e.g.: I have never seen a righteous man abandoned MP 109²⁷. In non-object constructions, the genitive of negation occurs only in isolated instances.

The Creator does not want a single sinful person to die DP 403₂₃ (niekche śmierči żadnego grzesznego W 661). How not to know (=not to know) | what kind of women they are | if you have once touched them⁷ DP 479₄₄ (iākoby niewiedział co to żá niewiasti W 808). Yet he did not know that the things from his hands were given | for his own betterment DK 124—125 (ich nie poznał być z rąk iego dāne ku swemu lepszemu DK₁ 117).

Such constructions, grounded in the original relations between their constituents, had already become syntactic archaisms by the sixteenth and seventeenth centuries. Although N here has the form of a governed case, semantically it can no longer be regarded as the object of V, just as in many corresponding constructions with the accusative. For example, in the expression he did not know that they were… the object of V is no longer them, but their being (=that they are…), just as the meaning of GD I know that you remember this BrB Job 10,13 allows only the NP you remembering (=that you remember) to be regarded as the object. The change in the type of government is quite obvious here.

When the type of government changes, the internal structure of GD itself is also reorganized. Non-object constructions represent not the original, but a reorganized model of GD syntactic relations, which can be represented by the following schema:

(3) V ⟶ (N ← - - - → P)

Constructions formed according to this model are three-member word groups, in which the syntactic binomes (V and NP) can be identified only at different levels of analysis⁴⁷.

In sixteenth- and seventeenth-century object GD constructions (e.g., to see him coming Mž 299), syntactic relations likewise can no longer be considered identical to those reconstructed for prehistoric stages in the development of Lithuanian (see 1, 4.). The original model of GD syntactic relations (2) fully corresponded to the unstrained-structure sentence characteristic of the ancient Indo-European languages⁴⁸, which was formed gradually—

⁴⁷ Cf. E. Nida’s double-object model for English case forms, which marks the dependence of both object cases in the combination on V, while at a lower level of analysis it indicates the exocentric relations between the members of that combination; see E. A. Nida, A. Synopsis of English Syntax, London — The Hague — Paris, 1966, p. 30, 155—156; constructions with the -ing form and the infinitive are treated similarly, pp. 42, 111—112. Many other scholars also indicate the syntactic relation of the whole NP constituent, rather than any one of its members, to V; for English, cf. also O. Jespersen, A Modern English Grammar on Historical Principles, II, 1914, 1. 67; The Philosophy of Grammar, London, 1929, p. 115 ff.; Z. S. Harris, Co-occurrence and Transformation in Linguistic Structure, in The Structure of Language, Englewood Cliffs, New Jersey, 1964, p. 178; Ch. F. Hockett, A Course in Modern Linguistics, New York, 1969, p. 195—196; J. Machaček, Complementation of the English Verb by the Accusative — with — Infinitive and the Content Clause, Praha, 1965, p. 8; for French, cf. H. Nilsson-Ehle, L’attribut de l’objet en français, „Studia Neophilologica“, 25 (1953), № 3, p. 110.

230

as one moves from one sentence constituent to another and defines the meaning of each by the form of the word itself, without any prior projection of the whole sentence determined by verb valency. As sentence centralization proceeded—a development already clearly perceptible in sixteenth-century Lithuanian—the object constructions inevitably had to reveal a contradiction between the reciprocal relation of N and P and the one-sided relation of N to V. This contradiction was eliminated by linking the second member of the semi-predicative NP construction to V as well. In this way, the valency of the verb forming the GD increased by one position, and the construction acquired the following form:

(4)

Thus, an object-type GD forms a three-member word combination that cannot be divided binarily at the level of syntactic form⁴⁹. Attempts to isolate binary relations in corresponding constructions in other languages proved one-sided (this question was examined especially intensively on the basis of English, which widely uses constructions with a double-object case, an object case and an infinitive, or an -ing form corresponding to the Lithuanian participle). If N and P (or N¹) are treated as separately dependent on the verb⁵⁰, thus isolating two binomes, V→ N and V → P, the unequivocal relation between N and P remains unspecified—a relation clearly indicated in Lithuanian, as in other inflectional languages, by agreement. On the other hand, if only the first nominal⁵¹ (N) is considered dependent on V, the connection between V and P is not captured, as shown by the fact that P cannot be eliminated without changing the syntactic meaning of the construction⁵². Such complex syntactic relations are explained by the fact that, synchronically, the GD is not a primary but a derived construction, which can be regarded as the result of transforming simpler constructions and can be divided binarily at the level of its internal structure (see 3.3.2).

2.1.6. In terms of the relations between their members, object-type GDs diverged from non-object types and remained close to constructions with verbs that do not take phrasal complements, i.e. non-commentable verbs. Of these, those that also form corresponding double-accusative constructions with nouns and adjectives are generally used with a second accusative—the participle (cf. 1.4), e.g.:

make, call, nickname

Kurtinius padarė girdi̇nčius ir nebylius kalbančius VEE 113⁴. Wissus daiktus sau padėtus dara VEE 125¹⁸. Schicze girdim Apaschtala numirusius wadinnanti

⁴⁸ Cf. Б. Г. Адмони, Развитие структуры простого предложения в индоевропейских языках, ВЯ, 1960, № 1, 24 ff.

⁴⁹ Plg. А. М. Мухин, Структура предложения и их модели, Л., 1968, p. 122; Л. С. Бархударов, Д. А. Штеллинг, Грамматика английского языка, М. 1965, p. 379, 385—386; H. Nilsson-Ehle, op. cit., p. 109; Н. В. Соколова, Трехчленные глагольные сочетания в современном французском языке, сб.: Грамматика, фонетика и стилистика романских языков, Л., 1971, p. 251 tt.; В. И. Чагишева, История простых предложений с предикативными падежами имени, авт. докт. дисс., Л., 1971, p. 6—7.

⁵⁰ See Л. С. Бархударов, Структура простого предложения современного английского языка, М., 1966, p. 104 ff. with lit.

⁵¹Such relations are indicated in A. Mukhin’s diagram depicting the structure of corresponding English constructions (e.g., I saw him cross the road); see А. М. Мухин, Функциональный анализ синтаксических элементов, М.-Л., 1964, p. 42 ff., 104 ff.; Структура предложений и их модели, Л., 1968, p. 91 ff., 133.

⁵² Cf. Z. S. Harris, op. cit., p. 178—179; the impossibility of removing either one of the members linked by a semi-predicative (=sociative) relation is also noted by A. Mukhin; see А. М. Мухин, Структура предложений..., p. 94—95.

231

Mieganczius BrP II 519. You will no longer be called Abandoned, nor will your land be called Desolate BrB Isa 62, 4 (Man sol dich nicht mehr die Verlassene... heissen). And therefore calls her blessed DP 506⁴ (ia blogoslówiona zowa W 861)⁵³. The Lord, through Pránaszę prámine nopipiaustitivus MP 174ᵃ₃₅ (przewal / nie obrzezáne RP 118ᵃ).

to guard, protect, have, keep

Those... guarding the sleeping RG 224₁₃. The Angel guarded me, going out from there / and living there / and returning to us from captivity DP 522₃₁ (guarded me... going out... living... returning W 892) I have a great task to do; I cannot come BrB Neh 6, 3 (I have a great task to accomplish)⁵⁴. Let the others drive in the cart, filled up SIG II 3₄⁵⁵. I consider Jonathan... captured BrB I Mak 13, 15⁵⁶ (I have kept Jonathan...)⁵⁷.

to lead, accompany, send, let, leave, meet, receive, strike, frighten, revive...

He came here so that you would find them VEE 152₈⁵⁸. He led you, the accused, to death BrB II Kar 23, 30 (led him to death). We... having blessed you, we led DP 617₄₀ (leading... they brought W 834). Paul and Barnabas... sent the consecrated DP 619ₐ₆ (sent the consecrated W 849). Having gone out, they left the house VEE 103₅⁹. You... have left all things behind MP 185ₐ₁₉⁶⁰. When you meet your enemy’s ox or donkey straying, you will encounter it / that it may stray BrB II Moz 23, 4. Then on that day I bring her into the city DP 617₂ (receives the one entering W 833). The young woman... comforted her as they ran BrB Judit 16, 14 (they struck them as they fled). The people... frightened them as they worked BrB Ezr 4, 4 (they discouraged them from building). Here W. Christus raised the dead DP 336₂₆ (he... revived the dead W 556).

In such constructions, the accusatives of the noun (pronoun) and participle are subordinated to the verb just as in object GDs after verbs of saying. The participle’s direct dependence on the verb is also indicated by the fact that it sometimes takes the instrumental form, that is, the form of a semantic case directly associated with the verb, e.g.: The Apostles... only Jesus Christ truly risen from the dead BrP II 38; keep yourselves from being deceived MP 213ₐ₆, and so on.⁶¹

The instrumental forms of nouns in the corresponding constructions are entirely regular and tended to spread rapidly during the period of the written tradition; compare: I want you

⁵³ Similarly DP 471₁₄, 473₄₀, 475₁₉, SE 280₂, MP III 5ᵃ₂₉, 35₁₅; the construction is based on Luk 1, 48: beatam me dicent. ⁵⁴ Similarly BrB Ps 96, 10; Mat 6, 16; Luk 14, 18; Apr 14, 1; 19, 16. ⁵⁵ The frequent use of constructions with have in old writings was stimulated by corresponding expressions in German, Polish, and Latin originals; compare: Everyone had his head covered BrB II Sam 15, 30 — hatte ein ieglicher sein Heubt verhuellet; I have her living DP 606₃₉ — mnie ma mieszkającego W 742; I have excused myself BrB Luk 14, 19 — habe me excusatum. ⁵⁶ Similarly BrB II Kron 30, 9; I Kar 8, 50; Ps 137, 3; Wisdom 18, 4; Rom 7, 6. ⁵⁷ On GDs with consider, meaning ‘evaluate’ or ‘understand,’ see 2. 1. 1. 3. ⁵⁸ Similarly MP III 14₃; BrB II Kron 28, 55; I Kar 8, 46; II Kar 24, 15; Jer 43, 11; 48, 46; Hoz 4, 19; Ob 1, 11; I Mak 13, 12; Apd 22, 5; compare also bound in chains VEE 203₆; ChB Jon 18₂₄. ⁵⁹ Similarly MP 266₆, 266ᵃ₆. ⁶⁰ See also BrB I Kar 7, 47; Jer 46, 28; Apd 24, 27; II Timot 4, 20. ⁶¹ For more examples see E. Fraenkel, Der prädikative Instrumental im Slavischen und Baltischen und seine syntaktischen Grundlagen, AfSlPh, 40, p. 86; Kas., p. 200, 204, where the influence of Polish on this use of the instrumental is rightly noted.

232

to make people fishermen VEE 137¹³; calling you lords VEE 23, 31, we saw you coming ChB Mat 25, 38; they believed them to be gods BrB Išm 13,3; they found them praying ChB Apr 2,2, and so on.⁶²; in many cases the instrumental here eventually completely displaces the second accusative of the noun. The instrumental of adjectives, supported by the influence of Slavic languages, is also penetrating these constructions very actively, e.g.: he made us worthy VEE 135; makes you ready^⁶³ BrB Hebr 13, 21; they brought him back healthy ChB Apd 23, 24; even with verbs of seeing and saying: we saw you hungry ChB Mat 25, 37; they called him good ChB Jon 7, 12; he must be declared clean by the priest BrB III Moz 13; with the verb to consider the instrumental of adjectives has now become predominant and normative.⁶⁴ All this shows that the verb’s syntactic connection with the second dependent case gradually grows stronger, even changing the latter’s morphological form.

Direct subordination of the second of the two cases to the verb is a general phenomenon, also observed in other constructions with double case marking. In the dative-with-infinitive construction, too, this is shown by the gradual establishment of the instrumental in place of the second dative, cf. you will have to be a leader instead of the older a leader (dative); the instrumental of adjectives is also penetrating such constructions very actively. Slavic languages have gone still further along this path. For example, in modern Russian, Czech, and Polish, the second accusative and dative have already yielded completely to the instrumental associated with the verb. Thus cases formerly connected to nouns in a semi-predicative relation gradually became dependent on the verb, the center of the sentence’s syntactic relations.

2.1.7. The development of the GD construction was favored by the general direction of change in subordination relations: in both Lithuanian and other Indo-European languages, the strengthening and differentiation of verbal government can be observed. In old Lithuanian texts and folklore, the use of cases is considerably freer than in the modern literary language. Even the accusative—the principal object case—is quite often used here with various adverbial meanings of place, direction, time, and means, which are also expressed by forms of other cases. Its primary meaning of direct object has not yet been clearly distinguished from these secondary (adverbial) meanings⁶⁵; in these uses, the accusative functions as a semantic case. Compare such combinations as grit one’s teeth, tire one’s head, ride a horse, sail the seas, walk the fields, and so on.⁶⁶ The semantic rather than syntactic principle of case use emerges even more clearly in languages with a very old written tradition. In the Vedas, the same accusative case is not only freely used with a great variety of circumstantial meanings—

^⁶² See also E. Fraenkel, Kas., pp. 199—200.

^⁶³ Corrected from gatewus; such corrections in the Bible manuscript, where the accusative with the instrumental after to make is used extensively, are quite frequent.

^⁶⁴ The gradual spread of the instrumental in place of earlier accusative forms confirms the antiquity of the double accusative and clearly contradicts the hypothesis that the predicative instrumental is primary (A. Valeika, A Few Observations on the Structure and Semantics of Predicate Nominal Constructions, LKK, 14 (1973), p. 237).

^⁶⁵ For primary and secondary meanings of cases, see J. Kurylowicz, Le problème du classement des cas, in: Esquisses linguistiques, Wrocław — Kraków, 1960, p. 137 ff.

^⁶⁶ See J. Jablonskis, Rinktiniai raštai, I, 240, 611 ff.; E. Fraenkel gives very similar examples both under the object accusative and under the accusative of space, time, and direction; cf. E. Fraenkel, Kas., pp. 137 ff. and 161 ff.; for Latvian equivalents, see J. Endzelins, Latviešu valodas gramatika. Rīgā, 1951, pp. 578—579.

233

…but in fact can occur with almost any verb, differing from other cases only in a certain nuance of meaning⁶⁷. The further back we go into the past, the more independent a role the individual case forms have, and the weaker their syntactic dependence on the verb. Monuments of the oldest Indo-European languages allow us to glimpse a stage in the development of these languages when government relations were still very weak, and the use of autonomous cases in the sentence was determined primarily by their semantics⁶⁸. It was precisely this stage in the development of word relations that conditioned the original model of the GD structure. Since constructions with double cases were likely formed under the conditions of the Indo-European linguistic community (cf. 1.6.), the subordination relation between V and N must initially have been not as strong and syntactic as present-day government; apparently, its nature was essentially semantic. Therefore, in constructions with the double accusative, only the first accusative could be directly subordinated to the verb.

In the subsequent development of many Indo-European languages, the subordination of cases to the verb grew stronger, while the semantic and syntactic functions of cases became more differentiated. In Lithuanian, this process is associated with the gradual establishment of prepositional constructions in place of semantic cases. During the period of the written tradition, the accusative, genitive, and dative became increasingly specialized for certain object relations, while prepositional constructions spread in their place for adverbial meanings. Over time, they entirely displaced from the literary language the so-called accusative of direction (cf. keliauti į svečią šalį, lipti į viršūnę, tekėti į ežerėlį instead of the older keliauti svečią šalelę, lipti viršūnę, tekėti ežerėlį), the accusative of path (plaukti per jūras, eiti per lauką instead of plaukti jūras, eiti lauką), the genitive of separation and time (gelbėtis mirties, joti nakties), and the dative after the verbs susimilti, juoktis, pabėgti, etc.⁶⁹. Government relations were also strengthened by the formal differentiation between direct and indirect objects discussed in 1.2. Once the direct object, expressed by the accusative, was distinguished from the indirect object, which came to be expressed by a prepositional construction, the use of the accusative became even more clearly defined. In present-day Lithuanian, the accusative after transitive verbs is a syntactic case with the function of a direct object. Phrases in the accusative that were formerly associated with the verb only semantically are now syntactically subordinated to it. Thus, the revised structure of the GD, in which N and P or the entire NP phrase are syntactically dependent on V, fully corresponds to Lithuanian case usage during the period of the historical tradition.

2.2. The participial constituent

In the GD construction, the participle (P) agrees with a noun or pronoun (N) and is thereby grammatically dependent on it. However, by virtue of its function in the NP phrase, it is the central constituent and cannot be eliminated without changing the construction

⁶⁷ W. D. Whitney, A Sanskrit Grammar, Leipzig, ²1889; p. 90 ff.; L. Renou, Grammaire de la langue Védique, Lyon — Paris, 1952, p. 355—356; Remarques générales sur la phrase Védique, in: Symbolae linguisticae in honorem G. Kuryłowicz, Wrocław—Warszawa—Kraków, 1965, p. 231—232.

⁶⁸ Cf. A. Meillet, J. Vendries, Traité de grammaire comparée des langues classiques, Paris, 1924, p. 519 ff.; A. Meillet, Introduction à l’étude comparative des langues indo-européennes, Paris, ¹1934, p. 359 ff.; P. Chantraine, Grammaire Homérique, 2, Syntaxe, p. 8, 16—17, 232, 319 ff.

⁶⁹ Plg. E. Fraenkel, Kas., p. 62, 83, 113, 137 tt., 140, 164.

234

2 lentelė

Classification of the accusative with participle in 16th–17th-century texts according to the main constituent*

Main constituent | Accusative with participle | pres. act. | past act. | fut. act. | all act. | pres. pass. | past pass. | fut. pass. | all pass. | Total number | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | Group 1 | (pa)matyti a | 13 | 13 | 2 | 2 | 15 | b | 2 | 2 | 1 | 1 | 3 | regėti a | 228 | 29 | 1 | 258 | 8 | 50 | 58 | 316 | b | 123 | 70 | 3 | 196 | 4 | 3 | 7 | 203 | veizdėti a | 2 | 2 | 2 | b | 2 | 2 | 2 | išvysti a | 160 | 8 | 168 | 1 | 19 | 20 | 188 | b | 13 | 3 | 16 | 3 | 3 | 19 | girdėti, išgirsti a | 92 | 92 | 1 | 3 | 4 | 96 | b | 85 | 83 | 6 | 174 | 9 | 24 | 33 | 207 | (iš)klausyti a | 6 | 6 | 6 | jausti, (pa)justi a | 4 | 4 | 2 | 2 | 6 | b | 12 | 4 | 16 | 1 | 1 | 2 | 18 | (at)rasti a | 86 | 28 | 114 | 36 | 36 | 150 | b | 7 | 2 | 9 | 1 | 1 | 10 | užtrop­ti a | 2 | 2 | 2 | (iš) (pa)rodyti a | 2 | 2 | 1 | 1 | 3 | b | 22 | 16 | 2 | 40 | 1 | 1 | 41 | (pa)ženklinti a | 1 | 2 | 3 | 3 | b | 3 | 3 | 3 | pažinti a | 14 | 2 | 16 | 2 | 1 | 17 | b | 42 | 5 | 47 | 2 | 2 | 4 | 51 | at(si)minti a | 2 | 1 | 3 | 3 | b | 4 | 18 | 1 | 23 | 1 | 2 | 3 | 26 | laikyti a | 2 | 2 | 2 | All group 1 | 612a | 70a | 3a | 685a | 10a | 114a | 124a | 809a | 313b | 203b | 12b | 528b | 18b | 37b | 55b | 583b | 952ab | 273ab | 15ab | 1213ab | 28ab | 151ab | 179ab | 1392ab | Group 2 | žinoti | 166 | 59 | 30 | 255 | 4 | 5 | 9 | 264 | suprasti | 34 | 13 | 1 | 48 | 5 | 5 | 53 | (iš) (nu) (per) (pra)manyti | 56 | 15 | 2 | 73 | 2 | 3 | 5 | 78 | nuvaruoti | 1 | 1 | 1 | (iš) (per)tirti | 70 | 19 | 89 | 1 | 4 | 5 | 94 | išmėginti | 1 | 1 | 1 | mąstyti | 1 | 1 | 1 | dimoti | 11 | 3 | 5 | 19 | 1 | 2 | 3 | 22

235

2 lentelė (tęsinys)

1 2 3 4 5 6 7 8 9 10 be confused 1 1 1 say “think” 44 12 7 63 1 1 2 65 suppose 1 1 1 dream 2 1 3 3 opine 1 1 1 imagine 1 1 1 reckon 2 2 2 believe 64 38 11 113 4 1 5 118 undertake 1 1 1 agree 2 2 2 seem to one another 1 1 1 determine, decide 14 14 14 swear, make a vow 3 1 4 1 1 5 promise 3 1 4 4 (pa) (ap) (api) (at) (iš)say 142 52 12 206 3 20 23 229 (pa) (per)speak 4 2 1 7 7 testify 6 2 8 8 mention 3 3 3 remind 3 3 3 announce 4 6 10 10 make an announcement 1 1 1 (pa)publish 2 1 3 1 1 4 reveal 3 2 5 5 confess 15 2 17 17 acknowledge 1 1 1 worship 1 1 1 praise 1 1 1 accuse 1 1 1 (ap)testify 11 24 35 1 1 36 call 1 1 1 defend 6 6 1 1 7 lead to 2 2 2 astonish 3 3 3 dispute 1 1 1 assert 2 1 3 2 1 3 6 tell lies 1 1 1 lament 1 1 restrain 1 1 (iš)write 4 18 22 1 1 23 read 1 12 13 1 1 14 Total group 2 691 279 81 1051 19 49 1 69 1120

236

2 lentelė (tęsinys)

1 2 3 4 5 6 7 8 9 10

Group 3

confer 6 2 8 2 2 10

know thoroughly 5 5 5

express oneself 1 1 1

signify “mean” 1 1 1

know 1 1 1 1 2

expect 4 9 13 1 2 1 4 17

observe 8 2 10 10

look around 1 1 1

(not)pay attention 2 2 2

(teach) 22 5 9 36 1 1 1 3 39

(teach) 15 1 16 1 1 17

doubt 2 1 1 4 4

be astonished 2 2 2

check oneself 1 1 1

to comfort oneself 1 1 1

to boast 1 1 1

to complain 2 1 3 1 1 4

to rejoice 1 1 1 2

to want 10 2 1 13 6 6 19

to wish 5 5 5

to give, have a testimony 4 1 5 5

to give, have, receive a sign 1 4 5 5

to make a sound 1 1 1

to make a connection 1 1 1

to lay to heart 1 1 1

to bring before one’s eyes 1 1 1

not to impute sin 1 1 1

to be certain 1 1 1

to send a letter 1 1 1

to be a witness 1 1 1

All of group 3 95 21 28 144 9 7 2 18 162

Group 4

to be seen, appear, show up 4 1 2 7 1 1 8

to occur 1 1 1 1

to seem 1 1 1 1

to succeed, “to turn out” 1 1 2 2

it is seen, known... (is, was) 19 12 31 1 1 2 33

237

Table 2 (continued)

1 2 3 4 5 6 7 8 9 10 said, written … (is, was) 4 4 8 1 1 9 it is to be seen 1 1 1 certainly is 1 1 2 2 truth, testimony, sign … (is, was) 7 4 11 11 Ah! 1 1 1 Without a principal constituent 4 1 5 5 All of group 4 43 24 3 70 1 3 4 74 Total 1754 597 127 2478 57 210 3 270 2748

syntactic meanings¹. Expressing a secondary action (state), in an NP construction it plays a role similar to that of a finite verb form in a sentence. This role of P in the construction, conditioned by the regrouping of syntactic relations already discussed, is also reflected in its grammatical form. The participial constituent of GD has many of the morphological and syntactic properties characteristic of a sentence predicate.

2.2.1. GD with active participles

Active participles are used in the GD of Old Lithuanian writings almost ten times more often than passive participles (ratio 2478: 270; see Table 2). This is explained first and foremost by the generally greater prevalence of active participles, evident both in old writings and in present-day dialects. More importantly, active participles are more fully integrated into the system of verbal categories, have more verbal features, and are therefore better able to perform the so-called secondary predicate function, which is precisely the function characteristic of them in the GD construction.

2.2.1.1. Present active participles are the most widespread in the GD: nearly two thirds of all constructions found in old writings are formed with them. Their predominance is especially pronounced in Group I object constructions (type a), in which 612 of the 685 active participles are present-tense forms (the relative simultaneity characteristic of these participles corresponds here to the semantics of verbs of physical perception: only an action taking place at the time in question can be observed directly). Present-tense forms likewise predominate in other semi-predicative constructions with oblique participial cases, for example, the absolute dative (cf. LKK V 53, 57), where they account for about 75%.

* Constructions with attributive accusatives of participles (often having dependent words) are not included in the statistics.

¹ Such reciprocal relations, in which the same constituent is both central (constitutive) and grammatically dependent, are generally characteristic of words linked by a predicative (or semi-predicative) relation; see J. Kurylowicz, Esquisses linguistiques, Wrocław — Kraków, 1960, p. 12, 40; B. Trnka, On the Basic Categories of Syntagmatic Morphology, “Travaux linguistiques de Prague”, 1 1966, p. 167.

238

Most present-tense participles are formed without prefixes, though prefixed forms are also common and usually have a specific lexical meaning. The lexical range of present-tense participles in the GD is unrestricted: it includes participles from verbs of various semantic groups, denoting both actions and states.

2.2.1.2. Past simple active participles rank second in frequency of use in the GD

by frequency. Nevertheless, in old writings they are more than three times less frequent than present-tense active participles. Unlike the latter, past simple active participles are more characteristic of non-object constructions. For example, in Group I, non-object constructions with past simple active participles are much less numerous than object constructions (203 b: 70 a), whereas the ratio for corresponding constructions with present-tense active participles is the reverse (313 b: 612 a). After the verbs girdėti, išgirsti, object constructions with past simple active participles do not occur at all, while non-object constructions are as widespread as the corresponding constructions with present-tense active participles (83: 85). However, past simple active participles are more frequent than present-tense active participles in constructions with the verbs (ap)liudyti (24: 11), (iš)rašyti (18: 4), and skaityti (12: 1), found mostly in the writings of J. Bretkūnas.

Both prefixed (somewhat more common) and unprefixed past simple active participles are used in the GD. Most of them have perfective aspect and denote an action completed earlier (than the verb). However, constructions with participles denoting an ongoing action are also fairly common, though less frequent, for example:

We know that she suffered many hardships BrP I 259. As we read, she herself taught Christ and his apostles the world BrP II 328. For some thought that Judas had a purse, when he spoke to Jesus SE 248₁₈. Remembering the Lord… who had died and spoken of his suffering and death… we can see SE 50₁₆.

The meaning of an ongoing action is also found in many past simple active participles formed from the verb būti ‘to be’ and functioning as a copula (see 2.2.3.). Thus, compared with appositional constructions (see LKK VII 52 ff.) and the absolute dative (see LKK V 58–59, 64), the aspectual opposition between present and past simple active participles in the GD is much less pronounced.

2.2.1.3. Future active participles are used relatively rarely but regularly in GD constructions in old writings (unlike in present-day Lithuanian and many other Indo-European languages), accounting for about 4.6% of all examples. They occur in the writings of J. Bretkūnas, S. Vaišnoras, M. Daukša, K. Sirvydas, J. Rėza, and others. With a few isolated exceptions, future active participles occur only in non-object constructions. With Group 1 verbs they account for only about 1%, with Group 2 verbs, 8%, and with Group 3 verbs, as much as 20% of all active participles found in the GD; in the last group, their number even exceeds that of past simple active participles. Thus, future active participles are most often found where an attributive function is impossible not only semantically but also structurally. This distribution of these participles accords with the general direction of their development: attributive use of future active participles had already declined considerably in the 16th–17th centuries and is entirely uncharacteristic of present-day Lithuanian, while predicative use remains alive in many dialects even today (especially in forming the so-called indirect mood).

In terms of lexical expression and aspect, future active participles in the GD are unrestricted, although prefixed forms with perfective meaning are used more often.

Tikrai szinwam. Newiena Szmogu amszinai ne passilikſenti szemeie BrP II 520. Nesa mussitikik, ghin man ios (sanarius) attadusenti BrB II Mak 7,11

239

(I hope | he will give it back to me). Aristotle… the world … will remain forever without end in thought MT 245₁₅. Whoever loves someone does only this: believes in him who is to come SE 124₇. Knowing that death will come suddenly, as winter, prepare yourself SE 205₁₄. How then do you say to my soul, if it is to fly away PsD 28 (sagt… sie sol fliegen)². The Virgin Mary said that she would become a human being (was to be incarnated in her) PK 152. (A person) does not know about his own… rejection of eternal punishment for sins | in which he sees God dying for them without lament SP I 116₁₈. Then those who had despaired, and had not trusted God, seeing him about to punish, turn away and flee BrB II Mak 8,12—13 (hatten das vertrauwen nicht… das er straffen würde). The lord… announced the wretches, coming out of prison, and spoke BrP II 77. Knowing a person to listen so attentively | then he says such things BrP I 243. And when Pharaoh dreamed thus a second time, he understood God to be truly and soon to act BrB I Moz 41,32 (bedeut | Das solchs… eilend thun wird).

For GD with the future active participle forms būsiąs, -antis, see 2.2.3.

2.2.2. GD with passive participles

GD with passive participles are relatively uncommon. Being closer to adjectives, passive participles mark a secondary action (state) in a sentence much less often than active participles; this is also characteristic of other participial constructions³. For the same reason, passive participles, especially those of the past tense, are much more likely to form object-oriented (a) than non-object-oriented (b) GD. Among type-a constructions with passive participles, they account for about 15%, whereas among type-b constructions they account for only about 7.7%—almost half as many.

2.2.2.1. Present-tense passive participles make up a minority (about 22%) of the passive participles used in GD. Almost all GD containing them are found in the writings of J. Bretkūnas and S. Vaišnoras, apart from isolated constructions in M. Daukša’s “Postilė,” K. Sirvydas’s “Punktai,” and “Suma Evangelijų.”

Both non-prefixed (1) and prefixed (2) forms of present-tense passive participles in GD generally have an imperfective meaning; prefixes here modify the participle lexically.

1) We hear you being praised as a glorious man BrB I Mak 10,19 (Wir hören dich preisen). Here we hear the false prophets… sending us away and coming BrP II 312. That one… believes in the forgiveness of sins | being given to others MT 138₁₈. Foolish people… are governed as Christians by these very [commandments] MT 204₁₂. The text… stepin: the Egyptians and Canaanites were trodden down because of their shame MT 223₈₁₄.

2) The faithful… hear their sins being forgiven SE 97₂₃. He hears that forbearance and wrath are attributed to God SE 97₂₅. They say that you are received by the Church for the sake of the Son Redeemer | and by the Holy Spirit MT 122₄. Nor is it necessary (they say) to reveal this: the souls of the dead are sanctified, and the living are made known MT 265₉₁₆. The Sophists… prove | that a person is justified by faith, formed through love MT 88ₐ₁.

² Cf.: Sakat duschei mana. Jei paleksi BrB Ps 11,1.

³ Cf. constructions with the dative in LKK, V, pp. 56—57; with connectives in LKK, IX (1967), pp. 47, 55; with the conjunction and in “Baltistica,” I (1) 1965, p. 62; appositional participles in LKK, VII (1964), p. 61.

240

Only occasionally does a finite action also get expressed by a present passive participle.

For we saw the writing on the wall: the enemy had seized our houses and was leading the rioters to the gallows MT 216₃ Then a great storm rose (variant: storm) against the ship, as if it were about to be broken to pieces BrB Jonah 1,4 (that they thought the ship would break apart)

In the GD construction, as in other contexts, present passive participles sometimes have an adjectival modal meaning of possibility.

The Albanians say: let us surrender the city to those who intend to take it, since it cannot be saved BrB Judith 7,9 (which they thought | could not be conquered). As he... saw that the Jewish people could not be overcome... BrB II Macc 11,13 (saw | that the Jewish people were invincible). For it shows that his teaching is easily fulfilled and accomplished SP II 136₂₁

The GD may also contain an adjectival present passive participle that does not have a passive meaning, e.g.: Isch schu szodziu paties Kristaus girdim io dangaus szengima didei reikema BrP II 84.

Lexically, present passive participles occurring in the GD are not subject to any notable restrictions.

2.2.2.2. Past passive participles The GD is more widespread with these than with any other passive participles, and they are used in most sixteenth- and seventeenth-century texts. Unlike present passive participles, they occur most often in object constructions of Group 1: there are more instances of them here than of all other groups in non-object constructions (114a: 96 b), especially after the verbs regėti, išvysti, rasti.

And I saw heaven opened ChB Rev 19,11. Without doubt we will see her laid in the manger DP 35₁₇ (we shall endure him lying in the manger W 55). There you will find him laid Mž 185.

Most past passive participles in the GD are prefixed and have perfective meaning. Finite action is usually also denoted by the rare unprefixed forms, e.g.: Esch isch PONO girdeiau, paslus Pagontump siųstus BrB Jer 49,14 (Ich hab gehört... daß eine Botschafft... gesandt sey). The meaning of an ongoing action is expressed by past tense passive participles in the GD only in isolated instances, e.g.: Gir déiot taipiaeg Szidus keturaku budu nūdem sunkei vargintus BrP II 347

The GD also occurs with past passive participles of qualitative meaning, semantically very close to adjectives.

The pagan considers you accursed (above: an accursed state) BrB Ezech 22,15 (you... must be considered accursed). And they praise (above: proclaim) her as blessed⁴ BrB Pat 31,28 (they praise her as blessed)⁵.

Meanwhile, in non-object constructions, the number of past passive participles is only twice that of present passive participles. They occur most frequently here after the verbs girdėti, išgirsti, sakyti (with various prefixes).

We heard that he was being glorified... spoken of with great praise BrP I 219.

And when the Philistines heard that David had been anointed king over Israel, they all came BrB II Sam 5,17 (heard | that... had been anointed). They say that on the third day we were raised from the dead (announces | that we... were raised from the dead) _{SP II 252₆}

⁴ Above: blessed, praised.

⁵ Here one may also include the construction, frequent in old texts, they call (will call) me blessed, they proclaim me Mž 488, VEE 158ₐ, 161ₐ, DP 121ₐ, RG 150, etc., whose basis is Greek μακαρισουσιν με πάσαι and Latin beatam me dicent Luke 1, 48.

16. 520

241

Much more rarely, passive participles of the present tense are used with verbs that do not govern the accusative, e.g.: They proclaim him before all the people of the Jews, mourning that innocent Onias had been murdered BrB II Mak 4,36 (und klagten jm | Das Onias vnschüldiglich ermordet were), see 2.1.3–4.

2.2.2.3. Future passive participles have been found only in the following three examples in S. Vaišnoras’s “Žemčiūga” GD.

Knowledge about predestination… the elect | all saved by the Son of God | and for his obedience… holy chosen ones 175₁₉. Origen… discussed raising the dead spiritual bodies 246₁. The wicked… having endured the pains and torments of the eyes in their soul and body, put to death, and turning to nothing | without hope 254₂₀.

All three constructions are non-object constructions and occur with verbs that are not normally associated with an object accusative of the corresponding meaning.

2.2.3. GD with participial phrases

About 20% of all GD found in sixteenth- and seventeenth-century texts are formed with participles of the verb būti ‘to be’, which may occur alone or together with forms of other participles, nouns, adjectives, and adverbs. GD uses participles of the verb būti more extensively than those of all other verbs.

The history of many Indo-European languages clearly shows that verbs with the roots *es-, *bhū-, from which the copulas eventually developed, originally had a concrete lexical meaning of existence⁶. In the GD construction, participles of the verb būti have retained this meaning in many instances. It is most clearly felt in cases where the participle with N occurs alone or is expanded only by an adverb of modal meaning.

a) (Epicureans) the world … thus believed to be and to perish MT 46₆. Such a shining light appeared | as each one said | to be day BrP II 337. God’s being witness | the grievous torments of the wicked world BrP II 178. The Scripture… bears witness and teaches (177) that the Lord God is truly and eternally alive BrP II 177–178. That … you may understand that I am BrB Iz 43,10 (verstehet | das Ichs bin)

b) There have likewise remained some masonry buildings and towers on the site | where the town had been BrP II 346. And the greatest sins that appeared… had been BrP II 196.

c) Let us then learn to know and consider | that my end is to come MT Mld 1₁₃. They believe only this life to be DP 584₃₁ (oni tylko ten żywot być wierzą W 995). The people, hearing that the day of Judgment would come… thought BrP I 18. The Jews rejoiced greatly and were delighted to be BrP II 339₇.

⁶ K. Brugmann, Die Syntax des einfachen Satzes im Indogermanischen, Berlin und Leipzig, 1925, p. 72 ff.; E. Benveniste, La phrase nominale en indo-européen, BSL, XLVI, f. 1, 1950, p. 19 ff.; „Etre“ et „avoir“ dans leur fonctions linguistiques, BSL LVI, f. 1, 1960, p. 114 ff. It should be noted that the old meaning of the verb *es- ‘to really exist’ is reconstructed above all precisely on the basis of participial forms, which have best preserved it; cf. Hittite ašant-, Old Indian sant- ‘existing, true, reliable’, Old Icelandic sannr, Russian сущий ‘true’, and so on; see Вяч. Вс. Иванов, Общеиндоевропейская, праславянская и анатолийская языковые системы, М., 1965, p. 266.

⁷ See also BrB Jer 14, 15; 14, 19; Sir 48, 7.

242

The specific lexical meaning of the participle in such cases is sometimes emphasized by accompanying synonyms, e.g.: Bei isch to permanik Dwase Schwenta ir tawip giwenancze alba esancze BrP II 125. It is also clearly perceptible in constructions with relative adverbs where, how, which at the same time function as connectives.

For what would it profit a person to know where something is, but not know how to obtain it BrP I 76. The truth came to me: the men were with me, but I did not know where they were from BrB I Joz 2,4 (ich wuste nicht von wannen sie waren). It is necessary to care for one's own children… so that they would know each year | where they are DP 65₄₁ (aby wiedzieli… gdzie się obracają W₁ 65) It is good and from afar it is clear | what each thing will be BrP II 335.

With locative and temporal adverbs, the lexical meaning of būti ‘to be’ in the participle has already faded considerably. In such cases, it is used primarily as an indicator of grammatical tense and modality, i.e. it performs a copula-like function, although it can still be replaced by participles of verbs such as to live, to exist, and the like; cf.:

They heard that they were near them and living among them BrB I Joz 9, 16 (das jene nahe bey jnen waren | vnd würden vnter jnen wohnen). They heard that Christ was near the city BrP I 12. John… recognizes that the Messiah is present here MT 127₁₆. Seeing you to be present DP 177₁₅ (widząc cię być obecną W 303), (the devil) imagines that Judgment Day is near BrP II 437.

When occurring with locative forms of nouns and with various prepositional constructions expressing circumstances, the participle of the verb būti ‘to be’ performs the same copula-like function as with locative adverbs (except for the aforementioned relative adverb kur ‘where’).

We see and understand that the whole world is in great misery or is living BrP II 148. Say: You firmly believe | that the body is in the grave RG 305₁₁. And thinking that he was in their company, they went a day's journey (A mniemając że on był w towarzystwie) SP I 195₈. We see | the LORD with you BrB I Moz 26,28 (wir sehen… das der HERR mit dir ist). They must know that a prophet is among them BrB Ezech 2,5 (das ein Prophet vnter jnen ist). How he might understand that he is among them MP III 31₁₋₂₀ (go być v siebie rozumieli RP 300ₐ₁₉). Sirata, seeing the winter nearing, began to grieve BrP II 395.

In the same kinds of constructions, though much less frequently, the past simple and future forms of the active participle of būti ‘to be’ are also used.

Aristotle… the world’s eternal ages, which were to come MT 245₉₁₅. Such things must be known about these wise men | who were not of the tribe of the Jews BrP I 153. The third says that the venerable ones will be in the waters WP 6. So that Christ might know… that more would not be among the people BrP II 115. I know that they (underlined) will be in their hands BrB Judith 10, 12 (ich weis | das sie euch in die hende komen werden)

The participial forms of būti ‘to be’ have a fully grammaticalized copular function when used with cases and prepositional constructions expressing attributive (qualitative, possessive, etc.) meanings. Here they have become abstract grammatical indicators of tense and mood and cannot be replaced by forms of other verbs expressing existence.

⁸ Similarly VEE 24₁, Ev 22₁₆, SE 25₁₃ (Luk 2, 44).

⁹ Similarly MT 46₁₇; cf. also BrP I 152—153, 334

16*

243

But I know you to be of no small virtue KN 265₁₄. That… teach his disciples the treasures of the world, not falsely ours to be BrP II 321. We consider these two not only friends, but also members of one family and former relatives BrP II 262. And when he recognized him to be under Herod’s authority, he sent him to Herod DP 164₂₃. You know my wife to be without fault RG 47₆. Ius regit tha Wira be proto santi BrB I Sam 21, 15 (sehet | das der Mann unsinnig ist). Dumoket teipaieg, mane isch iusų kaulų ir mėsos santi BrB Teis 9, 2–3 (gedencket… das ich ewr Bein und Fleisch bin)

A function close to that of a copula is also served by the accusatives of participles of the verbs tapti, pastoti ‘to become’, which are occasionally used in GD, cf.:

You will investigate your people … having become difficult BrP II 265. He speaks disputatiously | accusing a person of being righteous by his own strength before God MT 61₁₅.

The accusative of the participle of the verb būti functioning as a copula is most often associated with the accusative of other participles and nouns. Such combinations of a būti participle with the forms mentioned correspond functionally to compound tense forms (with participles) or a compound predicate (with nouns and pronouns).

2.2.3.1. Constructions with the accusative of participles

Active participles in GD generally display clear verbal properties and do not require a copula. A participial copula is occasionally used only with some verbs that have an adjectival meaning.

It is true that she confesses … God to be truly almighty, present and helpful MT 196₉. And when Jesus came, he saw him already dead DP 179₂₅ (vyžtavvšy go už vmárlego W 307). That… in our affliction… this faith, being not dead, but living and active in our hearts MT 92₁₅. The Pharisee judged the woman to be still sinful DP 481₃ (osądzil niewiastę być ieszcze grzeszną W 808). I have not seen this first coming, so weary was the holy prophet … who had foretold it BrP I 9₁₀.

Like an adjective (cf. 2.1.6.), such an active participle may also occur in the instrumental, e.g.: we considered that wage to be that of a sinner DP 348₁₆ (ktorego grze-sznym być wyznawa W 574).

Passive participles, especially those of the past tense, are more frequently used with participial copulas. However, such constructions make up only a small proportion (about 15%) of all GD examples with passive participles. Alongside nouns and adjectives occurring with copulas, we often also find participles without a copula. In J. Bretkūnas’s Bible manuscript there are examples where the copula before a participle has been crossed out in such a case, e.g.: You consider me cursed (crossed out to be). And he will find me to be the Lord Ezech 22, 16 (du… must verflucht geacht werden). Meanwhile, Z. Blotnas, while revising J. Bretkūnas’s translation of the Book of Tobit, already proposes adding a participial copula, cf.: Tobias considered his prayer heard BrB Tob 4,1; in the margin; he hopes his prayer to be heard (gedacht | das sein gebet also erhöret werde). Nevertheless, in 16th–17th-century texts, most passive participles, like active ones, occur in GD without a copula. In this respect, Old Lithuanian differs markedly from present-day Lithuanian, where a copula is obligatory in non-object GD constructions with passive participles.

¹⁰ Similarly BrP II 220, 222, 223.

244

In constructions with present passive participles, a participial copula in the present tense is characteristic only of examples from S. Vaišnoras’s writings. In some of them, present passive participles are close to adjectives; in such cases, the participle with the copula corresponds to a compound predicate, e.g.: Atgimeghie... tikis... tas biauribes sanczes apdengemas bei malone sancze silingesne MT 59ᵃ₁₉. However, most present passive participles with būti in participial form function as equivalents of compound passive forms.

Powilas... teaches that the promise of faith is received MT 138₁₆. Oieigi ghie sakitu Missche sancze wadinama affiera ijednojma MT (PM) 9₁. I... through those works... I pray... you... are washed in the baptism of regeneration, received in eternal life | and in the grace of the true God MT 121–121ᵃ.

In constructions with past passive participles, the copula is used not only by S. Vaišnoras but occasionally also by other authors—M. Daukša, J. Bretkūnas, B. Vilentas, and K. Sirvydas. Combinations of past passive participles of various lexical meanings with būti in participial form likewise generally function as equivalents of compound passive forms.

And that the Gentiles might know that the houses of Israel were carried away captive for their sins BrB Ezech 39,23 (learned | how the house of Israel... was carried away). We see that the resurrection of our Lord... from the dead | was first proclaimed or announced BrP II 8–9. The prophet said that his sins had been forgiven (65) SP II 65–66. John... says | that death and hell were cast into the lake of fire SP II 42₁₅. From what can we know | that we are chosen for eternal life? MT 171. The Church... knows | that this world was created at the appointed time | and will cease at the appointed time MT 235ᵃ₃.

Past and future tense forms of the participle būti are also used with past passive participles; thus, in the GD construction, equivalents of past and future tense periphrastic passive forms are formed.

Further, the Evangelist says that the Virgin Mary, descended from the family of David, had been betrothed to a man BrP I 334.

How dare they proclaim to us that Israel, being protected by its God, will be? BrB Judit 612 (prophesy that the people... should... have protection). They... thought that the Kingdom of God would appear immediately BrB Luk 19,11 (they supposed that the kingdom of God would immediately be manifested). Christ says and also teaches | that the Jews, being rejected, | and many Gentiles are received into eternal life BrP I 186. That he might teach | himself thus appointed for death BrP I 144.

2.2.3.2. Constructions with the accusative of nouns and pronouns

A participial copula is used with nouns, adjectives, and adjectival pronouns much more frequently than with participles. Although nouns can themselves serve as the second member of a double accusative, in old writings they are most often already linked with accusatives of active participles of the verb būti or with gerundial forms¹¹. These participial constructions with accusatives of nouns (pronouns) relate in exactly the same way as, in verbal clauses, finite forms of būti relate to the nominatives of the corresponding words. GD constructions such as I know him to be good correspond in meaning and in many features of syntactic form to verbal expressions I know (that) he is good.

¹¹ For constructions with gerundial copulas, see the author, “Baltistica,” IV (2), 1968, pp. 198 ff.

245

Similarly to a verbal predicate, nominal words accompanying a participial construction often have dependents and even subordinate clauses.

In constructions with participial phrases, nouns and adjectives occur in almost equal numbers; a marked predominance of nouns emerges only in J. Bretkūnas’s Bible, due to the frequent repetition of certain stereotyped constructions (e.g., ischtirs mane Poną santį); cf. LKK XI 43.

Pronouns with participial phrases occur only occasionally; among numerals, only vienas. has been observed.

For she does not wish to know that it is I who give it to her. Rye, new wine, and beer BrB Hoz 2, 10 (Sie wil nicht wissen / das ischs sey der jr gibt). We understand the works of the Holy Spirit, these holy ones BrP II 138. You say that I will be like you, being what you are BrB Ps 50, 21 (meinestu / ich werde sein gleich wie du)

Nes szitė nórint’ tiki… Die<w>ą wieną sąntį DP 257₂₃ (wierża… Bogą być iednego) W 429. Kurių daiktų prieszasti viena sancze regiū Vln 53.

Accusatives of nouns (sometimes with dependents) can be linked by a participial phrase to the neuter pronoun form tai, tatai. Such constructions are characteristic of J. Bretkūnas’s writings.

Æsch reggiu tatai sawa walios Szmones sanczias BrB V Moz 9, 13 (das dis Volck ein halstarrig volck ist). Ponai piauwia ies nei tar ia tatai grieką santį BrB Zach 11,5 (und haltens fur keine sünde)¹².

The nominal member (N) of the GD may also be unexpressed and implicit. In such cases, the construction consists of a participial phrase together with a noun accusative functioning as an equivalent of the nominal part of the predicate.

Nesa Mamah neszinnoia Angelą PONO santį BrB Teis 13,16 (wuste nicht / das es ein Engel des HErrn war). Kaczei Simonas gerai suprata, wilių santį BrB I Mak 13, 17 (mercket / das eitel betrug war) (Cf. constructions of the same external form in 2. 2. 3, in which a noun accusative marks the subject of the participle’s state.)

Accusatives of nouns and adjectives most often occur with the present-tense form of the active participle of būti ‘to be’:

a) Believe that your body lies in the dust MŽ II 418. The whole city knows me, [knows] you to be a virtuous woman BrB Rut 3, 11 (weis / das du ein tugendsam Weib bist). Esther said that she was of his kin BrB Est 8,1 (saget an / wie er jr zugehöret). The church that God desired to be a virtuous fellowship MT 221₁. From these examples you can notice, O most merciful one, such love and compassion, when one loves another BrP II 247. Do you believe my forgiveness to be God’s forgiveness? Vln 31₉.

b) Do you suppose that the Lord is righteous SE 75₂₄. Your … deeds show you to be good (świadczą iżes ty Bog prawy) PK 78. Your teachers … heard my sermon to be dear PsD 246¹³. I knew my Redeemer to be alive BrP II 40. Elsbeta … acknowledges … her kinswoman to be ill BrP II 266.

Forms of the simple past are much rarer here (about 10 examples).

a) Teip girdim … io tewa buwusi Zacharioschu / io augiwe Elsbi̇eta BrP II 236. Suprantam Schwenta Iana… paschlowinta ir czesninga wira buwusi

¹² Žr. dar Jer 32, 9; II Moz 32, 9; 32, 22; Sir 29, 6

¹³ Plg.: Bus girdėta, jog nama pamokslas mielas ira (pabr., parašėi tas pat tekstas, kaip PsD) BrB Ps 141, 6 (wird… Lehre hören / das sie lieblich sey).

246

BrP II 234. Then it became apparent that the Samaritans had been the most wicked pagans BrP I 322. And as the executioner was found, he showed what foolish men they had been¹⁴ BrB Wis 1, 3 (beweiset sie / was jene für Narren gewesen sind)

b) Didžukem. So that everyone would understand us to have been joyful BrP I 237. He shows the Lord not to have been guilty | yet to repent on our behalf SP II 128_a.

Future-tense forms are also found only in isolated instances.

a) That suffering | foretold your holy Christ to be coming Vln 61₃. The pagans… considered the misfortune of the Jews to be their own good fortune BrB II Mak 14,14 (hoffeten der Jüden vnglück solt jr glück sein).

b) So that people might know that his Kingdom is not worldly but spiritual BrP I 349. For I hope that he will be gracious and good to you BrB II Mak 9,27 (He will deal graciously and kindly with you). Ianas… knowing himself to be no longer alive BrP I 31. For who knows whether he will be wise or foolish BrB Pam 2,19 (wer weis | ob er Weise oder Toll sein wird?)

In constructions with participial clauses, verbs of groups II and III are used most often; group I verbs are rarer here and usually have a meaning of mental perception, cf.: The king saw (i.e., understood) that the battle was too severe for him BrB II Kar 3,26 (sahe | das jm der streit zu starck war). All GDs with participial clauses are of the non-object (b) type. By contrast, constructions in which the participle of the verb būti ‘be’ is linked with an N directly governed by V are attributive, and the participle in them has a concrete existential meaning, cf.: (Jonas) while being in the womb, knew his Lord | likewise, in the womb, the existence of the holiest maiden DP 462₂ (poznal Páná… w żywocie… będącego W 770₂₀.)

The accusative of the participle of the verb būti ‘be’ could acquire a linking function only when N was no longer considered the object of V, and the NP construction became equivalent to a subordinate clause. By analogy with sentences containing a compound predicate or a compound tense form, entirely new syntactic relations developed within the GD between the participle of būti and the accusative of a noun, pronoun, or another participle, corresponding to the relations between a copula and a nominal predicate (cf. I know him to be a king and I know: he is a king). These new syntactic relations also affected the external structure of the construction. The noun or pronoun corresponding to the nominal part of the predicate, as in a sentence with the finite verb būti (I know: he is a king), in many cases took the instrumental form and thereby lost agreement with N and P (I know him to be a king).

2.2.3.3. Constructions with the Instrumental

In 16th–17th-century texts, these account for about 28% of all examples with participial clauses (ratio ≈ 140: 550) and about 5% of the total number of GDs. In them, the instrumental of a noun is most often linked with a present-tense copula.

We know them to be thieves and robbers BrP II 165. So that everyone might suppose | him to be a true human being BrP I 132. Rimianis heard that Simon, his brother, was high priest BrB I Mak 14,17 (höreten das Simon… Hoherpriester war). In her heart, that Pharisee judged me not to be a prophet DP 479₄₉ (osądził mnie być Prorokiem W 808). We can… boast boldly, saying that we are holy townspeople Baszniczes MT LĮŽ 3₄.

¹⁴ Nuo kakius in the margin, instead of the underlined text: kakie anie buwa Durnus.

247

The accusative and instrumental forms of nouns are sometimes found in sentences of exactly the same type and compete with one another. For example, in J. Bretkūnas’s Bible, the accusative was replaced by the instrumental in 12 examples during manuscript revision. Nevertheless, the accusative form remained more popular in the 16th–17th centuries and occurred with the copula more than twice as often.

Instrumental forms of nouns are also sometimes associated with the copula in the past and future tenses.

For remember that thou wast a servant in Egypt! BrB V Moz 24,18 (solt gedenken / das du Knecht…. gewesen bist)¹⁵. We read that they were prophets and teachers MT (PM) 16^a19. Saul the king… the writing says… elsewhere that he had been king for forty years (says / that he reigned two years) SP I 136¹². We hear that Mary and Elizabeth… had been good friends BrP II 261.

The Lord … wanted… to teach (226) … that he would be our true king and High Bishop BrP I 226—227. So that they might know that he would not be a worldly king or lord of any kind / nor that they themselves would likewise be worldly officials BrP II 71¹⁶.

The instrumental of adjectives is much rarer here than that of nouns (it is likewise rarer with finite forms of the verb to be).

They … consider him to be worthy DP 77¹⁷. (they know him to be worthy W₁ 82). People … believed that he was righteous MT 88¹⁵. Whomsoever they wish … to be equal to themselves MT LĮŽ 17²⁰.

Do you think that a dead man will live again? BrB Job 14,14 (Meinstu ein todter Mensch werde wider leben?)

Instrumentals of adjectives occur alongside accusatives in entirely analogous sentences, cf.: you must judge him to be unclean BrB III Moz 13,27 and you must judge him, the leper, to be unclean ibid., 13, 11¹⁷ (sol jn… wrein urteilen). By analogy with adjectives, the instrumental of a past passive participle is also used in one place in DP, directly translating the corresponding Polish form: the one whom they believed to have been killed DP 198²⁶ (ktorego wierzyli być zabitym W 337).

Among pronouns, instrumental forms of who, which, what kind are sometimes used.

Who do you say that I am? VEE 156²¹ (wer saget denn jr / das ich sey? Mat 16,15)¹⁸. And whom do you say that I am (465)? DP 465—466. The intriguer asked whether W. Christ was such as they had heard him to be DP 318²⁹ (iakimgo być slyszeli W 528).

Although constructions with the instrumental are used much less frequently than those with the accusative in 16th–17th-century texts (the ratio is approximately 95:230), they nevertheless reveal with particular clarity the altered syntactic relations between the participle and the other constituents of the construction. Here, the participle of the verb to be is associated primarily not with the accusative, but with the instrumental corresponding to the nominal part of the predicate. The agreement of the participial copula with the accusative has lost its syntactic motivation, since the nominal part of the predicate is otherwise not inflected for case with a copula (the verbal copula has no case form at all). Therefore, the participial copula is sometimes made to agree with the equivalent of the nominal part of the predicate—the instrumental, whose relations with the participle of to be are the closest.

¹⁵ Similarly, V Moz 24,22; 5,15; 15,15; I Sam 24, 21; I Kar 14, 2; Ism 8, 9.

¹⁶ Similarly, BrP I 9.

¹⁷ Similarly, BrB III Moz 13,6; 13,7; 13,8; 13,25.

¹⁸ Similarly, VEE 134¹⁴, 134¹⁸, 156¹⁷, DP 465⁸, Ev 181⁹.

¹⁹ See also MP III 30⁴³, 30²⁹, DP 465³³, Ev 181⁹.

248

The abbreviated form | which God... wanted to sanctify with the saints, the goodness of the Son MT 178a₅. Do not consider me to be some other kind of man, a stranger among the people BrP II 37. He healed the Man... showing himself to be a true God | and the Savior of the world BrP II 453. Nor did they believe Christ to be the true Son of God BrP II 349. Let us learn | not to regard men as false lords BrP II 325. Christ proved himself to be a true God BrP I 279. One might suppose | that he would not be any worldly king BrP I 258. For I knew well that he could not be alive after his fall BrB II Sam 1,10 (ich wuste wol das er nicht leben kundte).

That the instrumental form of the participle was determined here by its association with the instrumental form serving as the equivalent of the nominal part of the predicate is apparent from examples in which participles used alongside the participial construction with another verb retain the accusative form:

We see... those people truly believing... him to be a true God | and able to drive away all... diseases from a person BrP II 456. He showed Christ risen from the dead | and alive (9) times his appearances BrP II 9–10.

A closer connection between the participial construction and the equivalent of the nominal part of the predicate sometimes becomes apparent even when it has the accusative form: this is evident from number agreement, for example: He says Iron! s c h a u d u s s a n c z i u s (not *sančią!) / BrB Job 41,18 (er achtet eisen wie stro)

2.2.3.4. Genitive constructions

In old writings, constructions in which the connective function is performed by a negative form of the participle of the verb *būti* ‘to be’ are of two kinds.

One kind is formed according to the usual GD pattern: N, P, and another noun associated with it, serving as the equivalent of the predicate complement, are in the accusative, for example: Know, then, that they are not gods; do not fear them BrB Bar 6, 64 (wisset | das es nicht Götter sind). Then you will see that there is no other God than (above: only) one Nebuchadnezzar BrB Judit 6, 2. (wirstu sehen | das kein ander Gott ist). Ask... knowing that I am not wicked BrB Job 10, 7. (weissest | wie ich nicht Gottlos sey)²⁰.

In the other kind, the negative participle of *būti* ‘to be’ is linked to a genitive of negation and itself takes the genitive form, much like a finite verb (cf. *išvydau: tu esi* ‘I saw: you are’ and *išvydau: nėra tavęs* ‘I saw: you are not there’). In old writings (especially those of J. Bretkūnas), such constructions occur with participles in various tenses:

So that you may find out that there is none like me throughout all the lands II Moz 9,14 (innen werden solt | das meins gleichen nicht ist in allen Landen). And so that the pagans may know that there is no other one... Lord Tob 13, 4 (erkennen | das kein allmechtiger Gott ist). For you know that there is not one of us I Kar 5, 6 (du weissest das bey vns niemand ist). In addition, know this | that there is no person so good and holy SE 65₂₆.

The people of that place say that no prostitute had been there I Moz 38, 22 (sagen die Leute... es sey keine Hure da gewesen).

We hoped for good, but there will be no good Jer 8, 15 (verlasset euch drauff | Es solle keine not haben).

²⁰ See also BrB I Moz 31, 5; 44, 31; II Moz 2, 12; I Kar 22, 33; MT (PM)1, 15.

249

Such genitives of nouns and participles may occur alongside corresponding accusative constructions governed by the same verbs without negation; cf.: Er regit nu, mane vieną santi, ir nesanczio niewieno Diewo prieg manens? BrB V Moz 32, 39 (SEhet jr nu | das Ichs allein bin vnd ist kein Gott neben mir?)

The genitive form under negation may also occur in constructions with participles of other verbs, e.g.: Tūgi nu numanat, niewieno szodzio PONO ant szemes nepūlusio BrB II Kar 10, 10 (erkennet… das kein wort… ist auff die erden gefallen).

Corresponding partitive genitive constructions are formed in the same way with affirmative participial forms of būti ‘to be’, e.g.:

Turim szinnoti | tokiu kleidunu pamoksle | isch pradzio swieto buwusiu | dabar sancziu ir ikki galo Swieto busencziu BrP II 309 (cf. žinoti: buvo ir bus klaidūnu).

Here we see reciprocal dependency relations between the participle and the noun (pronoun): the negative-form participle determines the noun’s genitive and is at the same time concordant with it.

As with the partitive genitive, the participle of the verb būti ‘to be’ agrees with the genitive of nouns governed by the adverb daug ‘much/many’.

The pagan… said that there were many gods BrP II 180. We hear of many people who have lived since the beginning of the world, but all have died BrP II 6. Do you not know that afterward there may be more troubles (could there be)? BrB II Sam 2, 26 (Do you not know | that afterward there might be more sorrow?)

In such constructions, the participle has retained its lexical existential meaning. Similarly, genitives of participles with other lexical meanings are also used with the adverbs daug, mažai ‘much, little’, cf.: Janas ischwiddens daug Phariseuschu ir Sadduceuschu savo krikschtop atencziu | bilba iemus BrP II 240²¹.

The case form of the participle here is determined by its relations with the genitive noun. However, by analogy with regular GD cases, the accusative of the participle of the verb būti ‘to be’ is occasionally used alongside a noun in the genitive, e.g.: Hoitmanas Sýrios girdeia, gerūsius pas Judą sanczius, ir daug Szmonių kruwoj essanczias BrB I Mak 3, 13 (hôret… das ein gros Volck bey einander war); cf. corresponding instances with participles of other verbs: Rasche regim atspraschiusius daug schwentu szmoniu BrP II 414.

Thus, in non-object constructions in which the participle functions as a copula, as well as in the other cases discussed here, variation in the case of the participial constituent is clearly apparent. This is conditioned by newly emerging syntactic relations within the construction. As the government relation between V and N faded, the agreement between N and P weakened, and the participle came to be formally linked to the noun functioning as the equivalent of the predicate’s nominal part. The new syntactic relations between constituents destabilized the earlier grammatical structure of the GD and created favorable conditions for the ingress and spread of indeclinable participial forms—gerundives. Constructions of precisely the type discussed here, with the instrumental and genitive, in which agreement between N and P broke down most extensively, were among the first and earliest centers for the spread of the gerundive in the GD (cf. 4. 1. 1 and 4. 1. 6.).

²¹ See also BrP II 308, 346, 455; for corresponding gerundive constructions cf. the author, “Baltistica,” IV(2), 1968, pp. 197—198.

250

2.3. Nominal constituent

In the GD, the nominal constituent (N) is formed by accusative forms of nouns and pronouns, and occasionally by adjectives or numerals used substantivally.

2. 3. 1. Nouns forming N may denote concrete or abstract entities.

Among nouns with concrete meanings, those most commonly used denote persons or other living beings capable of action. They may be proper nouns (a) or common nouns (b).

a) And David saw Saul seeking him BrB I Sam 23, 14. Alexander understood that Demetrius was seeking Jonathan, his friend BrB I Mak 10, 15.

b) And I saw a woman sitting on a scarlet beast BrB Apr 17, 3. At once you will find the donkey tied up VEE 54₉. I read... the Samaritans who had slain so many thousands of people BrP I 106.

N is also formed by other names of entities, for example:

so that you may see all the things of this world abandoned around you DP 554₃. The princes... saw the fire having done no harm BrB Dan 3, 27.

Nouns with abstract meanings are used less often in the GD:

Having met | such precious gifts for us out of kindness | bestowed without deeds MT 71₉. that you may see the terrible destruction... standing in a holy place Ev 139₁₂. How great deeds we hear have been done BrB Luk 4, 23. I know a little thing having gone out from me BrB Luk 8, 46.

Several nouns can be associated with the same participle.

From where do you know Saul and his son Jonathan to be dead? BrB II Sam 1, 5.

A noun forming N may be expanded by other dependent words or even by subordinate clauses.

Judas and his companions understood that Demetrius, the son of Seleucus, had come to Tripoli BrB II Mak 14, 1. He wanted to see | his dear... little boy, and, not knowing what was wrong now, | being struck down BrP I 106.

Adjectives, numerals, and participles very rarely function as the nominal constituent, and only when they have the meaning of nouns (or when a noun is understood), for example:

In this way he says that the righteous and unrighteous, taking nothing else into the body through their mouths MT 161₁₅. And he returned and told Tobias that he had found one lying dead in the street BrB Tob 2, 3 (das einer auff der gassen tod lege). And he also says that the dead, neither willing nor able, will be constrained to rise MT 179₅.

2. 3. 2. Pronouns very frequently function as the nominal member of the GD. Personal pronouns are particularly common, of which the third-person forms are the most widespread (jis, ji, and the forms anas, ana, used with their meaning).

The spirit has no body and bones | as you see | me having BrP II 30. The Lord... teaches | us all proclaiming much to God BrP II 486. I know that you can do everything BrB Job 42, 2 Regeiot g hy enti VEE 73. We hear them speaking MP 206₂. For she saw that she herself could not save BrB Wis 13, 16,... Hearing... them reviling | and mocking BrP I 301.

Among other pronouns, this, that, and other occur more frequently as the GD, and all occasionally.

Who finds our people willing BrB Luke 23, 2. You say those who have departed for eternal salvation MT LĮŽ 6₄. He saw others standing in the marketplace SP I 317₇. We see not all being Christians BrP I 249.

251

The nominal member of the GD can also be a relative pronoun, e.g.: What we see here done bodily on the walls, that continues spiritually in the hearts DP 577³³. Ezekiel spoke about those bones | which he had seen lying in the field MP 165^a9.

In general, any of the words that can function as the subject of a verbal clause may constitute the nominal member of the GD. In accordance with the prevailing tendencies in the expression of N, the GD corresponds to the absolute dative, likewise based on a construction of an oblique case with a participle (cf. LKK V 67 ff.).

2. 3. 3. A separate group consists of GDs whose nominal member, a personal or reflexive pronoun, denotes the same participant as the clause subject (the subjects of the verb’s and participle’s actions are identical). Such GDs are characteristic of old writings, and are now very rare in the modern language. The accusative of the reflexive pronoun is most often used in this way, e.g.:

He said | himself to be the holy King of the Jews VEE 207²². Let us recognize then | ourselves as rightly punished | and deserving harsher torments MT 186^g. He told them, himself wishing to do this BrB II Mak 7, 26.

Much less frequently, N denoting the same subject occurs in the accusative forms of personal pronouns: me, you, us, e.g.:

Then I will show myself to them before the heathen, being holy BrB Ezek 28, 25. Did he know you to be born at that time? BrB Job 38, 21. So that we might believe, having forgiveness of sins Mž II 376.

These constructions are primarily related to the problem of the nominative with participle (nominativus cum participio)¹, and are therefore not discussed in greater detail here.

2. 3. 4. The nominal member of the GD may be entirely unexpressed. Cases of N ellipsis occur when the subject of the participle’s action is mentioned in the same or an adjacent clause, or is inferred from the context and situation, e.g.:

Truth came to the men to me, but I did not know where they were from BrB I Josh 2, 4. For the thieves on the cross | and the wise men in the manger recognized: those filled with wine | and these entwined with chains DP 426².

Of a different kind are constructions in which neuter forms of present and past passive participles are used impersonally, and their action is not associated with any specific participant, e.g.:

You have heard the elders say: Do not grieve BrP II 188. It is heard spoken throughout the world about the Devil… nothing lets that go BrP I 293³.

A different GD may sometimes also depend on a neuter participle, e.g.: The people of Abuse heard it said that Simri had made an agreement BrP I 1 Kings 16, 16 (see 2. 1. 4. 2). In such cases, instead of active participles, converbs have been used since the earliest writings⁴.

¹ Su nom. c. part. jas sieja jau D. Kleinas, žr. Grammatica Litvanica, 1653, p. 165; dėl šių konstrukcijų genetinio ryšio plg. A. Bielenstein, Lettische Grammatik, Heidelberg, 1923, § 817; E. Tangl, Acc., p. 50; I. Marvan, О некоторых предикативных функциях литовского причастия, „Kalbotyra“, IV, Vilnius, 1962, p. 35—36; Verbum infinitum в балтийских языках и его отношение к славянскому состоянию, Slavica Pragensia, II, 1960, p. 13 (kur kalbamo tipo konstrukcijos laikomos nepaliudytomis); kiek kitaip aut., Donum Balticum, Stockholm, 1970, p. 10.

² Similarly DP 165₁, 407, 265₉₇, WP 6, SIG I 58₂₇, II 111₅; RG 539₁₄.

³ Similarly BrB Pam 5, 7; I Joz 22, 11—12; I Mak 16, 21. For the corresponding Latvian GD, see J. Endzelīns, Latviešu valodas gramatika, Rīgā, 1951, p. 1001.

⁴ See the author, „Baltistica“, IV (2), 1968, 197 ff.; ibid. for one problematic case with a passive necessity participle, p. 198.

252

Closely related to these constructions are occasional GDs whose N is a neuter pronoun, e.g.: Atmink tatai teipo PONO likta BrB Sir 41, 5; in similar cases, the neuter form of the present active participle is also used, esą, e.g.: tikra tariu esą BrB II Petr 1, 13⁵, but impersonal participles predominate.

2. 4. Overview of the grammatical forms of the GD shows that this construction in Old Lithuanian can be formed by almost all words and word combinations that govern a subordinate explanatory clause. Meanwhile, the dependent members of the GD correspond, in their syntactic properties, to the subject and predicate of a verbal clause. Having verbal government and other verbal properties (except for the categories of person and mood forms), the participle in this construction can take various cases, with or without prepositions, and infinitives together with their dependents; it can itself serve as a copula and form equivalents of compound verb forms and compound predicates. In other words, here the participle has the same constructional possibilities as a finite verb form. For this reason, in Old Lithuanian writings GDs often develop into complex, expanded constructions that include other constructions and subordinate clauses.

Although the basic GD pattern is common to all Indo-European languages that have corresponding participial forms, the varied and complex structure of this construction found in Old Lithuanian is the result of a later, independent development. In terms of its range of use and the diversity of its syntactic development, the Old Lithuanian GD occupies a distinctive place among the corresponding constructions in Indo-European languages.

The Lithuanian GD has most in common with the Latvian GD, in which we find the same participial forms (largely fossilized in the modern language), which can occur not only after transitive verbs but also after some intransitive verbs (e.g., šķist “to seem”⁶), although not to the same extent as in Old Lithuanian. The GD is not attested in Old Prussian texts (apparently because of literal translation). Among other languages, the closest to Lithuanian is the GD in the Old Slavic languages, with corresponding participial forms in -nt-, -us-, -mo- and -to-, which have similar meanings and syntactic properties. Constructions with verbs of mental perception and speaking that do not involve an object are common to Baltic, Slavic, and Germanic⁷, and are also associated with the use of the accusative in the sense of a more distant object (see 1. 6.). Finally, in almost all Indo-European (and not only Indo-European) languages, GDs can be formed by verbs of physical perception (especially seeing and hearing)-

⁵ For more detail, see the author, op. cit., pp. 202–204.

⁶ Cf. Škīta mani neprotuotam (nepruotuoat) B W 8683, J. Endzelīns, op. cit. p. 1001.

⁷ Nevertheless, they are much rarer in Slavic languages than in Lithuanian: in Old Slavic, where about 90% of GDs are formed with verbs of physical perception, they are, according to R. Ružička, “especially rare,” R. Ružička, Das syntaktische System der altslavischen Partizipien und sein Verhältnis zum Griechischen, Berlin, 1963, pp. 240, 243; moreover, they are used here only with a few verbs of the meanings mentioned: minėti, razuměti, (u)vĕděti, glagolati, kazati, see R. Večerka, Syntax aktivních participií v staroslověnštině, Praha, 1961, pp. 57–58. Their distribution is very limited in Old Russian as well; see А. Потебня, op. cit., p. 308 ff.; АН СССР, Сравнительно-исторический синтаксис восточнославянских языков, М., 1968, p. 124; for Old Czech, see J. Gebauer, Historická mluvnice jazyka českého, Praha, 1929, p. 600; Dvořák, Vývoj přechodníkových konstrukcí ve starší češtině, Praha, 1970, p. 160 ff. A similar situation is found in the Old Germanic languages: constructions with verbs meaning “see,” “feel,” and “find” also predominate here; see O. Behaghel, Deutsche Syntax, II, Heidelberg, 1924, pp. 388–390, 417–418; А. М. Мухин, Функциональный анализ синтаксических элементов, М.-Л., 1964, pp. 104–105, 188–189.

253

verbs⁸. However, in vain do we seek parallels in other language families for the GD in Old Lithuanian with intransitive verbs (group 3), nouns, modal words, and other words (group 4). In those cases expressing semi-predicative relations where the GD became widespread in Lithuanian, many other languages generalized infinitival constructions. In terms of range of use and the meanings of the words serving as the main element, the Lithuanian GD roughly corresponds to the Latin acc. cum participio and acc. cum infinitivo taken together⁹. A similar, though much less pronounced, picture is also found in the ancient Germanic languages (Gothic, Old High German, Old Anglo-Saxon)¹⁰, where infinitival (as well as participial) constructions following verbs of perception and speech were stimulated by Greek and Latin models; in the later development of these languages, infinitival constructions became firmly established with the meaning characteristic of the non-object GD in Lithuanian. Romance languages¹¹ extended the use of infinitival constructions instead of acc. c. part. even further.

The Lithuanian GD also differs in the variety of its syntactic structure and its clause-like character. It uses forms of active and passive participles in all tenses (except the relatively recent necessitative participles), even future active and passive participles, which are not found in either the Slavic or Germanic languages. In the GD, active participles correspond to the simple forms of finite verbs, while passive participles correspond to compound passive-voice forms; in the latter case, they may occur with or without participial copulas. Although constructions with participial copulas were also formed in the Slavic¹² and Germanic¹³ languages, their range of use is by no means comparable to that of the Lithuanian GD. In this respect, Greek acc. c. part. with frequent accusative forms of the participle ὤν comes much closer to Lithuanian¹⁴, although its use, too, is restricted by constructions with infinitival copulas (ἔνιαι and the like).

⁸ They have already been attested in Hittite texts; see A. Hahn, Genesis of the Infinitive with Subject—Accusative, TAPA, 81 (1950), p. 128. For the GD in Avestan with “see,” see E. Kieckers, Historische griechische Grammatik, IV, Syntax, Berlin u. Leipzig, 1926, p. 89.

⁹ Particularly noteworthy in this respect is the use of acc. c. inf. in the position of a nominative subject with existential verbs apparet, liquet, constat, convenit, sequitur, efficitur, with expressions of modal meaning verum est, verisimile est, credibile est, certum est, and even with nominal predicate expressions fama est, spes est, testis sum (!) and the like—that is, precisely in those cases where in Old Lithuanian texts we find the GD after words of group 4. Compare A. Ernout, F. Thomas, Syntaxe latine, Paris, ²1953, p. 321; compare also the Lithuanian GD with neuter-gender passive participles sakoma, žinoma (“it is said,” “it is known”) and Latin acc. c. inf. constructions such as dicitur Gallos in Italiam transisse or adesse Romanos nuntiatur; see ibid., p. 327.

¹⁰ For Gothic, compare W. Willmanns, Deutsche Grammatik, III, 1, p. 118–121; for Old Anglo-Saxon, M. Callaway, The Infinitive in Anglo-Saxon, Washington, 1913, p. 120 ff.; B. H. Ярцева, Исторический синтаксис английского языка, М.—Л., 1961, p. 231–232; А. М. Мухин, op. cit., p. 120, 188 ff.

¹¹ For French, compare H. Nilsson—Ehle, L’attribut de l’objet en français, „Studia Neophilologica“, 25, № 3, p. 109 ff.; H. B. Соколова, Трёхчленные глагольные словосочетания в современном французском языке, in: Грамматика, фонетика и стилистика романских языков, 1971, p. 251 ff.; A. Ernout, F. Thomas, op. cit., p. 283; for other Romance languages, see M. Manoliu Manea, „Accusativus cum infinitivo“ dans les langues romanes, „Cahiers de linguistique théorique et appliquée“, VI, 1969, p. 125–136.

¹² Comparative-historical syntax of the East Slavic languages, Sentence Parts, M., 1968, p. 126–127; R. Večerka, op. cit., p. 58.

¹³ Compare Old Anglo-Saxon Ie hine cūðe niht wesende Beowulf 372 “I knew him to be a child,” K. Yamakawa, The two constructions: „Acc. and Participle“ and „Genetive and Gerund“, „Annals of the Hitotsubashi Acad.“, 8, № 1, Tokyo, 1957, p. 49.

¹⁴ See R. Kühner, Ausführliche Grammatik der Griechischen Sprache, II. 1, Hannover, ²1870, p. 611 ff.

254

The possibilities for forming GD inherited from the period of Indo-European unity were developed in Old Lithuanian to their furthest limits. The abundance and productivity of participial constructions in popular speech provided the conditions for this. Participial constructions with oblique cases, which in Slavic and Germanic languages were sustained by literary tradition and the influence of classical languages, were grounded in living usage in Lithuanian and took shape and spread on that basis. In this respect, their development is important both for historical linguistics and for general linguistic typology.

3. The accusative with a participle and the explicative clause

3. 1. The synonymy of this construction with subordinate explicative clauses was of great importance for the development of GD syntactic relations.

Alongside the double accusative with verbal adjectives and, later, with the participles that developed from them (e.g., I see my father coming), Indo-European languages had long been familiar with paratactic expressions such as I see: my father is coming. They have survived in Lithuanian to the present day¹ (especially in dialects and folklore), but in the modern standard language they are generally used only for particular stylistic purposes (to emphasize unexpectedness, give a narrative a colloquial flavor, and so on). In sixteenth- and seventeenth-century texts, we find them in circumstances similar to those in which GD occurs, but much less frequently, for example:

Schitai ghi esti taw Mote, kamgi tarrei ghi est mana sessu? BrB I Moz 26, 9. Kitti wel isch Szidu ape lana dumnoia | ghis ira prarakas koksai BrB I 42. Nes pasakia: esmi sumumi Diewo SE 261, 24.

GD and the asyndetically connected clause sometimes depend on the same verb and correspond to original constructions with precisely the same structure; cf.:

Sakikite sawa Brolims Ammi, anis ira mana Sźmones, ir iusu Sęserei, įę sanczę malonie BrB Hoz 2, 3 (Sagt ewern Brüdern (Ammi) sie sind mein Volck | vnd zu ewr Schwester (Ryhamo) Sie seyn in gnaden).

Subordinate explicative clauses developed from such paratactic expressions; used in the same syntactic environment, they eventually became syntactic synonyms of GD. At the beginning of the written tradition, they were already widespread, although in a large part of the Lithuanian-speaking area, at least in the Samogitian and western Aukštaitian dialects, their sphere of use at that time probably still did not match that of GD. This can be inferred from old texts bearing features of these dialects. Although the use of subordinate explicative clauses after verbs of perception in these texts was stimulated by German, Polish, and Latin originals, themselves in turn influenced by the original Greek and Hebrew texts², GD was still considerably more usual and popular in this context. This is especially true of the relatively rare original sources. For example, in the original Lithuanian introduction to S. Vaišnoras’s “Žemčiūgos,” alongside 10 GD and 4 nominative-with-participle (nominativus cum participio) constructions, we find only 6 synonymous clauses with jog and 1 with kad. The corresponding explicative—

¹ Plg. E. Hermann, Über die Entwicklung der litauischen Konjunktionalsätze, Jena, 1912, p. 5.

²In Greek New Testament texts, explicative clauses after verbs of perception spread under the influence of Hebrew, where they are more frequent than participial constructions of the GD type; see M. Johannesson, Der Wahrnehmungssatz bei den Verben des Sehens in der hebräischen und griechischen Bibel, KZ, 64 (1937), 3–4, p. 145–260, especially p. 157, 242. By contrast, explicative clauses were probably not long characteristic of Indo-European languages, at least after verbs of seeing; cf. ibid., p. 252–253.

255

The number of explanatory clauses in the entire translation of “Žemčiūga” (about 100, by approximate count) likewise does not match the number of GD. In pp. 50–200 of Vilentas’s “Evangelijos ir epistolos,” about 80 explanatory clauses with jog occur, so their number slightly exceeds that of GD³; however, considering that this is a translation from German, where subordinate clauses were used in all corresponding cases, and the sacred nature of the translated text, which compelled fidelity to the original, this ratio of the constructions in question undoubtedly indicates a greater abundance of GD⁴. The same can be said of J. Bretkūnas’s writings, whose author, using GD extensively, had to overcome the pull of the original in each case. It is worth noting that in many places in J. Bretkūnas’s partly original “Postilė,” explanatory clauses are no more frequent than GD.

The situation is different in old writings representing the eastern or so-called central Aukštaitian dialects. Here, explanatory clauses far surpass GD in number, in some instances tenfold or even more. For example, in just pp. 440–470 of M. Daukša’s Postilė, 85 explanatory clauses with jog are used (more than half the total number of GD found in the entire Postilė); since explanatory clauses are used no less frequently in other parts of the text, these data suggest that their ratio to GD is approximately 10:1. This is also confirmed by E. Hermann’s data on the use of explanatory clauses in excerpts from M. Daukša’s Postilė and in the Catechism. On pp. 6–33 of K. Sirvydas’s “Punktai,” 17 explanatory clauses with jog and kaip are used (according to E. Hermann’s table); if they occur at the same frequency elsewhere (as our approximate counts for SP I, pp. 34–380, suggest), then the ratio of explanatory clauses to GD here is 5:1. Explanatory clauses are used several times more frequently in the anonymous Postilė of 1600, where GD are comparatively rare overall; their widespread use is also reflected in B. Chylinski’s Bible, M. Petkevičius’s Catechism, and other texts. Although the data from these writings, given the influence of the originals and other features of translation technique⁵, by no means directly reflect the actual ratio of the constructions in question in the spoken language of the period, they nevertheless clearly show that in the sixteenth and seventeenth centuries, over a large part of the Lithuanian-speaking area, explanatory clauses with conjunctions such as jog, kad, kaip, etc., following verbs of perception, speech, and other commentative verbs, were fairly widespread and in frequency of use probably surpassed GD in some areas.

3.2. Almost all GD found in sixteenth- and seventeenth-century writings can be replaced by explanatory clauses with jog or kad without distorting the meaning of the text (but not vice versa, since the scope of use of clauses with kad and jog is considerably broader). The synonymy of GD and explanatory clauses is also confirmed by other features of their use.

3.2.1. When comparing texts translated from the same original, we see that different authors rendered the corresponding passages using GD or explanatory clauses. Cf.:

Sakot Beelzebubi mani išmetineunti welinus SP II 43₁₂. — Bilote / idant asz Beelzėbube iżtremdinėcże welinuwas DP 117₄₀. — Sakote nesang | iuog asz Beelzebubba mace izmetineiu welinus SE 63₁₃.

O Pilotas regiedams… tů daugiau didesni trunksma santi… priesudija VEE 205. — O kaip regėjo Pilotas… kaip io didesnis trunksmas stoiosi BrB Mat 27, 24⁶.

³ Cf. statistical data on the use of GD in LKK, 11 (1969), pp. 54–55.

⁴The ratio of these constructions is different in the “Enchiridion,” where alongside 7 GD there are 41 explanatory clauses with jog and jeng; see E. Hermann, op. cit., appendix.

⁵ Cf. the author, LKK, 5 (1962), pp. 50–51; LKK, 11 (1969), pp. 52 ff.

⁶Similarly, SE 259₁₇.

256

And when they looked, they saw the stone rolled away from the tomb BrP I 399. – And looking, he saw that the stone had been rolled away VEE 56⁷.

3.2.2. That GD and the explanatory clause were regarded as parallel means of expression, capable of replacing one another, is clearly evident from J. Bretkūnas’s revisions to the Bible. While revising the translation manuscript, the translator himself replaced many explanatory clauses, usually corresponding to sentences in the German original containing daß, wie, etc., with GD, for example:

So that they might understand the fire… destroying the enemies’ fruits (revised from: that the fire destroys the enemies’ fruits) Wis 16, 26 (Auf das sie inne würden | wie das fewr… der Feinde früchte verderbete).

O Lord, teach me that I must come to an end, and that my life is fleeting and that I must die (revised from: that my end will come, and that my life is fleeting, and that I must depart Ps 39,5 (lere doch mich | das ein ende mit mir haben mus… vnd ich davon mus).

And it was reported that he was in the house (revised from: that he was in the house) Mark 2,2 (vnd es ward rüchtbar | das er im Hause war; quod in domo esset⁸).

No one can complain as if he were not helping (not helping him) (underlining from “as if” to the parentheses) Sir 40 (vnd man darff vber keinen mangel klagen | an seiner Hülffe).

Such revisions are particularly numerous in the New Testament translation manuscript. For example, in the translation of the Gospel of John alone, J. Bretkūnas replaced about 20 explanatory clauses with clauses containing GD⁹; about 10 such revisions are also found in the Gospel of Matthew¹⁰ (there are more than 40 in the New Testament altogether)¹¹. On the other hand, in the Gospel of Luke, translated from Latin, only one passage has been noted where GD was used instead of an explanatory clause in revising the manuscript.

If the farmer knew at what time the thief would come (the underlined words above: at what time the thief was to come), he would certainly Luke 12,39 (si sciret… qua hora fur veniret).

Sometimes both the explanatory clause and GD are retained in the manuscript and appear side by side as two possible translation variants equivalent in meaning.

Lest anyone say that I had baptized in my own name (that I had been baptized in my own name) 1 Cor 14 -15 (Das nicht jemand sagen möge | Ich hette auff meinen namen getaufft).

So that you might know that the Most High has power over the kingdom of men (the Most High has power over the kingdom of men) and gives it to whomever he wishes. Dan 4,22 (das der Höchest Gewalt hat… vnd gibt sie).

Opposite examples, in which GD is replaced by an explanatory clause, occur very rarely, for example:

⁷The translations of this passage from the Polish original likewise vary; cf. DP 184, SP II 225₈, Ev 64₅ – GD, but SE 79,5 – explanatory clause. See also the translation of Luke 24,20 in BrB II 17–GD, SP II 235₁₈, Ev 67₃, DP 189₂₀ – explanatory clause with iog, SE – explanatory clause with kad.

⁸ See also BrB Wis 1,3; 18,13; Tob 9,1; 10,9; Bel 23; 2 Macc 2,25, etc.

⁹ Jon 1,32; 4,1; 6, 15; 7, 26; 7, 32; 8, 54; 9, 20; 9,31; 11,7; 11,13; 11,17; 12,12; 12,34; 19,10; 19,33; 21,17; 21,24.

¹⁰ Plg. Mat 5,17; 10,34; 14,5; 16,13; 16,14; 17,10; 17,13; 21,26.

¹¹ Cf. also Mark 2,8; 2,10; 11,32; 15,39; 16,4; Acts 9,27; 10,3; 14,19; Rom 2,3; 2,15; 4,21; I Cor 4,9; 6,19; I Cor 7,3; Gal 1,20; Phil 1,30; 2,22; Heb 11,14.

17. 520

257

And another one will come by, saying… that he saw Tobias lying in the street, beaten (in the margin: that one lay dead in the street) Tob 2,3 (that someone lay dead in the street).

So that the world may know what he can do (corrected from: what he is able to do) Job 37,7.

3.2.3. The best authors of the sixteenth and seventeenth centuries sought to exploit the parallelism between GD and explanatory clauses for stylistic purposes. By translating some German or Polish explanatory clauses with corresponding Lithuanian clauses, and others with GD, they avoided repeating the same expressions and creating monotony. When a subordinate clause had exactly the same structure, J. Bretkūnas translated it in one sentence with a corresponding subordinate clause introduced by that or so that, and in the next adjoining sentence with GD. Cf.:

If we say that it was from Heaven, then he will say to us, “Why did you not believe him?” But if we say that it¹² was from men, then we must fear the people Mat 21, 25—26. (Sagen wir | sie sey vom Himel gewesen… Sagen wir aber | sie sey von Menschen gewesen).

In a similar way, variation was also used when translating two subordinate clauses of the same explanatory sentence: one was translated with a subordinate clause introduced by that, so that, or the like, and the other with GD.

Make it known this day that you are God in Israel, and that I, your servant, have done all this according to your word BrB I Kings 18,36 (Las heute kund werden | das du Gott… bist… vnd das ich solchs alles… gethan habe).

One messenger met the others there, to tell the king of Babylon that his city had been captured… and that the fords had been seized, and the soldiers had lost heart. BrB Jer 51, 31—32 (dem Könige… anzusagen | Das seine Stad gewonnen sey… vnd die Kriegsleute seien blöde worden¹³).

But Pilate, seeing|that he could accomplish nothing|but that the uproar was growing greater… testified VEE 205¹⁶.¹⁴ (Da aber Pilatus sahe|das er nichts schaffet| sondern das viel ein grösser getümmel ward).

A subordinate clause of an explanatory sentence and an NP phrase, both dependent on the same verb, occur side by side in the original texts as well.

Furthermore, we have discovered that many Curonians and Lithuanians… practise the worship of gods or idols, or wear… and use images depicting various things in churches, and observe other harmful customs or superstitions | and witchcraft IF 10—20.

Learn from this graciousness of God’s Son | who for the sake of us human beings was born as a man | and that much good comes to us, the weary, through his redemption | Then… let us learn… BrP I 76.¹⁵

3.2.4. Because GD and an explanatory clause are functionally equivalent, cases of contamination between the two constructions occur in old writings: following the pattern of a subordinate clause, a subordinating conjunction, relative adverb, or pronoun is also used at the beginning of GD¹⁶, for example:

And we do not read that (Mary) spoke in the Holy Gospel only four times DP 474⁵⁴. So that they might believe firmly | that (26) through that sin

¹² Above— that.

¹³ Cf. also Ezek 20,48; Judith 6,13; I Macc 3,13; 14,21; 7,25.

¹⁴ A very similar construction in DP 167¹⁹ corresponds to the original sentence with iš.

¹⁵ Similarly, BrP I 422, II 241, 403, 456—457, 475, Vln 65⁷.

¹⁶ See also 2. 2.3. Other semi-predicative participial constructions, such as the absolute dative, similarly become conflated with subordinate clauses; see LKK, 5 (1962), p. 79 ff.

258

those who have been released to Heaven Vln 28–29¹⁷. Then release them wherever you wish, and do not make them wear out. BrB V Deut. 21, 14 (wo sie hin wil). And as the Sun set, they departed, as if unaware where they had left them BrB Nah 3, 17¹⁷ (wo sie bleiben). And the men did not know where I was going, or what I was doing BBbB Nehem 2,16 (wusten nicht | wo ich hin gieng | oder was ich machte) For he knew beforehand what they would do BrB Wis. 19,1 (Denn er wuste zuvor wol | was sie künftig thun würden)

3.2.5. When a sentence contains two or more GDs, one of them may depend on another. In this way, complex “two-storey” constructions are formed, corresponding to subordinate clauses at two levels. In them, a GD sometimes also has a subordinate clause attached.

From these words… we can understand that the Jews, having wrongly understood the fifth commandment of God | believed only this | that the supreme punishment had been proclaimed BrP II 288. They say… that they considered themselves | to be the true Messiah, who was to redeem Israel BrP II 26. We see… this woman… firmly believing | that the Lord Jesus Christ could and wished to forgive her sins BrP II 304¹⁹. We also heard many of the Jews | not believing that this Christ was the true Messiah promised by God BrP II 5²⁰. So that we might appear to know | what the true God is | whom we worship in this way MT 119⁵.

As with the subordinate clauses of an explanatory sentence, several NP phrases may depend on the same V and be linked to one another by various coordinating relations.

They… complained as they worked. Pagans from all the towns… the Galileans who had gathered from Ptolemais, Tyre and Sidon, and all Galilee filled with enemies BrB I Macc. 5,15. Let us then recognize | ourselves as rightly condemned | and deserving of harsher punishments: yet those punishments are eased for the sake of the Mediator MT 186⁶.

In the writings of J. Bretkūnas and S. Vaišnoras, we sometimes find whole series of GDs linked to one another in various ways, which to a large extent take the place of verbal expressions. Here is a characteristic example of such a period:

Saint Matthew… writes | Wise men having seen a new star in the eastern land | and having inferred from it in the land of the Jews that a new king of the world’s Savior had been born | and having therefore come from their own lands to Jerusalem | to worship him and make offerings to him | and King Herod | with all the city of Jerusalem | having been greatly frightened by this news. And he ordered these wise men to go to Bethlehem, to Christ. And when they had arrived, they found the child Jesus BrP I 105²¹.

Thus, GDs contribute significantly to the participial (i.e. essentially nominal) style characteristic of many old writings. Similar tendencies to express a sequence of mutually related actions through participial constructions, especially prominent in Classical Latin and Sanskrit, are also clearly visible in the old Germanic (e.g. Gothic) and Slavic languages²².

¹⁷ Cf. E. Hermann, op. cit., p. 25.

¹⁸Similarly, BrP II 288, II 387.

¹⁹Similarly, BrB II 456.

²⁰ Cf.: You can infer | knowing that this Christ is the true Messiah BrP I 31. Here the subordinate clause of the explanatory sentence occupies the same syntactic position as the GD in the cited example.

²¹See also BrP I 64; BrB I Macc. 5,38–39.

²² Cf. E. Hofmann, Zum Gebrauch der Partizipien in den Skeireins, in: Indogermanica, Festschr. für W. Krause, Heidelberg, 1960, pp. 28–30.

17*

259

3.2.6. The punctuation of GD in sixteenth- and seventeenth-century writings also shows that the accusative combination of a noun (pronoun) and a participle was regarded as a separate semantic, and sometimes also intonational, unit, equivalent to a subordinate clause of an explanatory sentence. Although punctuation marks are often used inconsistently and vary considerably in old writings, the prevailing tendencies clearly emerge from the large number of examples. It is noteworthy that object constructions in which N is directly governed are almost never separated by commas or other marks. On the other hand, in those sentences where V, through phrasal government, is linked to the entire NP as the equivalent of a subordinate clause of an explanatory sentence, a comma or an equivalent oblique slash (/) is very often found. Here are a few examples:

O kas tik, tawę taip didei rustaujenti? PsD 164 (Wer gläubts aber / das du so sehr zürnest?) Bei dabokessi, schitas Szmones tawa Szmones sanczes BrB II Moz 33,13 (Vnd sihe doch / das dis volck dein volck ist). Tau uu (= nu) paszistam / tawę wissus daiktus szinanti VEE 71–72 (Nu wissen wir / das du alle ding weisset). … Nepriwala tie / kurie isch schirdies tikki / Christu sedinti descheneip macės MT (PM) 32₆. Pirmiaus / žinokim / Tą ceremoniją / senoja Zokaną… pradžia sawo turinczią nuog Diewą paties SE 23₄. Bau atmeni / Bagocziu pirm lepawusi BrGD LXXIII.

The NP combination is most often set off in the writings of J. Bretkūnas and S. Vaišnoras. Non-object constructions with participles of the verb to be, following V, are so frequently set off here that the unmarked cases are in the minority (see 2.2.2.). In the writings of these authors, a period or colon is also sometimes placed between V and NP.

Saint Luke writes, the Lord having worked / having shed his holy tears / and having repented from the bottom of his heart BrP II 335. First of all, remember this from this history. Christ having been born / when the Jews had already begun to be … lawless BrP I 56. Hymenaeus and Philetus… said. The resurrection then being denied / that people in this life close their eyes to God MT 245₁₃. Christ John 3. judges: Us having been born again of water and spirit MT 127₂₀. And Augustine… says, writing clearly: the Christian reddening the earth with Christ’s blood. MT 122₂₁.

Thus, the punctuation of GD in old writings confirms the semantic independence of this construction. Treating GD as the equivalent of an explanatory sentence, the authors of old writings often applied the same punctuation principles to it.

3.3. The synonymy of GD and the explanatory sentence is based on the commonality of their internal structure.

The complex interrelations among the members of GD show that, even from a synchronic perspective, this construction is not primary but derived. Traditional grammar emphasized its similarity to the sentence and explained this by claiming that GD, like other participial constructions, is a kind of reduced sentence²³ or a sentence equivalent with a subject–predicate structure²⁴. The relationship between GD and the sentence is more precisely defined in transformational

²³ Plg. Ф. Буслаев, Историческая грамматика русского языка, М. 1959, p. 281 tt, 538–539. Remdamasis rusų kalbos gramatikomis, panašiai aiškino lietuvių kalbos dalyvines konstrukcijas J. Jablonskis, kartu pažymėdamas „sutrauktinio sakinio“ termino sąlygiškumą ir tradiciškumą, žr. J. Jablonskis. Rinktiniai raštai, I, p. 516–518. H. Poutsma taip pat laikė anglų kalbos infinityvines konstrukcijas, atitinkančias lie. GD, sakiniais, įtrauktais į kitus sakinius, žr. H. Poutsma, A. Grammar of Late Modern English, I, Groningen, 1904, p. 433 tt.

²⁴ O. Jespersen, A. Modern English Grammar on Historical Principles, IV, New York, 1956, p. 5 tt.; Философия грамматики, М., 1958. p. 133 tt., 139 tt.

260

a method that in this case can be used as a means of sentence analysis, helping to determine the internal structure of the construction under investigation through its relationship to other constructions in the same language²⁵.

From a transformational perspective, the basis of GD is a sentence embedded in another sentence according to certain formally defined rules for forming the whole complex. The rules for forming (deriving) object and non-object types of GD are different.

3.3.1. Non-object-type constructions are formed from sentences in which the first (matrix) sentence contains an empty object position (P^ob), determined by the verb’s phrasal valency and marking the place of the second (constituent) sentence²⁶, e.g.: I know + P^ob (brother returned) ⇒ I know brother returned. The object position may also be expressed by a pronoun functioning as a resumptive word: I know [that]. The accusative combination of a nominal and a participle (brother returned) here represents a noun phrase (NP), rewritten by the rule NP→N(S) or simply NP→S²⁷. This rule specifies that a noun phrase consists of a constituent sentence S, occupying the same position as a nominal N (cf. I know the story). The form of the constituent sentence is defined by transformation rules according to which its main parts are expressed in the accusative case of a nominal and a participle.

Without examining in detail the process of generating non-object constructions, it is important to note that their internal structure is exactly the same as that of subordinate complement clauses. The sentence I know that brother returned is formed from the same sentences and according to the same derivational rule as GD I know brother returned²⁸. The only difference here is in the transformation rules, which give the constituent sentence a different form. Constructions with double case in other languages have a similar relationship to complement clauses²⁹. Since the internal structure of the sentence is based on semantic

²⁵ For the view of transformations as a means of identifying syntactic relations between constructions and sentences, cf. Z. S. Harris, Co-occurrence and Transformation in Linguistic Structure, ”Language“, 33 (1957), № 3, p. 283 ff. (= The Structure of Language, Englewood Cliffs, New Jersey, 1964, p. 155 ff.); Transformational Theory, ”Language“ 41(1965), № 3, P. 1, p. 363, 370 ff.; L. G. Heller, J. Macris, Perspectives in Functionalism, ”Word“ 23 (1967), № 1 – 2 – 3, p. 291; on the significance of the transformational method beyond generative grammar, cf. О. Лешка, Место трансформационных отношений в языковой структуре, „Československá Rusistika“ 1966, XI, 2, p. 65—66. Ch. Bally had already pointed out that the study of functional substitutes of constructions is one of the most important tasks of syntax, in Syntaxe de la modalité explicite, „Cahiers F. de Saussure“, 1942, № 2, Genève, p. 6.

²⁶ For elucidating the structure of GD, it is not decisive whether we consider it to be formed by means of a generalized transformation (the structure of corresponding utterances was explained in this way earlier; cf. R. B. Leest, The Grammar of English Nominalizations, Indiana, 1960, p. 54—55; E. Bach, An Introduction to Transformational Grammars, New York — Chicago — San Francisco, 1964, p. 67; P. Roberts, English Syntax, New York—Chicago, 1964, p. 169; J. Lyons, Introduction to Theoretical Linguistics, Cambridge, 1968, p. 265—266), or by developing a single derivational schema, as is now increasingly believed; see note 27.

²⁷ Cf. P. S. Rosenbaum, The Grammar of English Predicate Complement Constructions, Cambridge, Massachusetts, 1967, p. 1—5, 71 ff.; R. Lakoff, Abstract Syntax and Latin Complementation, Cambridge, Massachusetts — London, 1968, p. 5, 17, 20—36.

²⁸ For the internal structure of Lithuanian complement clauses, see the author, Lietuvių kalbos prijungiamųjų sakinių struktūriniai tipai, „Baltistica“, VIII (1), 1972, p. 39. For the relationship between matrix and constituent sentences, cf. R. B. Lees, op. cit., p. 54—55; W. Hartung, Die zusammengesetzten Sätze des Deutschen (Studia Grammatica IV), Berlin, 1966, p. 62 ff.

²⁹ For Latin and English, see R. Lakoff, op. cit., 75—76; for English, P. S. Rosenbaum, op. cit. 71 ff.; sentences in the Romance languages with acc. c. infinitivo are likewise sometimes considered to be derived from subordinate clauses by means of embedding transformations; cf. M. Manoliu Manea, „Accusativus cum infinitivo“ dans les langues romanes (completive ou relative?), „Cahiers de linguistique théorique et appliquée“, VI, 1969, p. 125 ff. with references.

261

directly reflects it³⁰, the shared derivational scheme of GD and the explanatory clause shows that both constructions realize equivalent semantic content that can be segmented in the same way.

3.3.2. Object-type constructions are formed from sentences in which the object of the first (matrix) sentence coincides with the subject of the second (constituent) sentence, cf.: I see my father (my father returned) ⇒ I see my father having returned. Thus, the relation between these sentences differs from that involved in forming non-object constructions³¹. The rewriting rule for the noun phrase here is NP→NS; it shows that the noun phrase forming the basis of the construction includes both a nominal object and a sentence. From the same basic sentences, following the same derivational rule but applying a different transformation, subordinate explanatory clauses with a so-called proleptic subject can be formed: I see my father, that he returned. Such sentences are not characteristic of the modern literary language, but are fairly common in dialects and old writings, e.g.: nesa anus girdeia, jog g hie lieszuweis kalbeia VEE 78₂₉³². The external form of such sentences is close to the schema of the internal structure of object-type GDs. They may serve as synonyms of GDs, e.g.: Nesa kiekwienas girdeia | lieszuwiu sawa yūs kalbanczus VEE 84, cf. also: isz wido Egypczoniey ang žmona, jog buvo labey grėžy ChB (Reinh. 243) and regeia Egypsciakai aną Moterischkę, gražę sanczę BrB I Moz 12, 15 (sahen… das Weib | das sie fast schön war).

3.3.3. In old writings, the GDs themselves sometimes also have a proleptic agent subject expressed in the sentence by an accusative dependent on the verb.

Tu gerai pasznisti sawa Tie wę ir io Szmones, annus³³ stiprus sanczius, ir rustos schirdies BrB II Sam 17,8 (that they are strong). Paszima anis Nikanora amt Scharwo, ghi drauge³⁴ uschmuschtą BrB II Mak 15,28. (they knew Nicanor… that he too had been slain)³⁵. (Ptolemy) kaltinoia Alexandra, ghi bei giwato sawo ir kardalistes tikousi! BrB I Mak 11,11 (gave A. the blame | he had brought him close to death…). Jei tu nasrais sawa ischpaszisti IEsu | ghi santi Wieschpatimi BrP II 111³⁶.

GD with a proleptic subject is close to dative-with-participle constructions such as when we arrived, his servants ran to him BrB Jon 4,51, see LKK V, 10–11. Repetition of the object case, which indicates the semantic relation between the nominal constituent of the GD and the verb, also emphasizes the construction’s autonomy and its sentence-like character.

Constructions of this type are also important because they bring out the old relations between the object and non-object types. Just as with verbs of group 1,

³⁰ For semantic interpretation of internal, or deep, structure, cf. Ch. J. Fillmore, The Case for Case, Universals in linguistic theory, 1968; J. D. McCawley, Concerning the Base Component of a Transformational grammar, in: Foundations of Language, 4, 1968, p. 243–269.

³¹ Z. Harris similarly explains the difference between corresponding English constructions with the verbal -ing form, treating utterances such as He saw the car stalling on the tracks (i.e. object-type constructions) as formed according to the schema N₁ V₁ N₂ + N₂ V₂, and utterances such as I remember the dog barking there (i.e. non-object constructions) as formed according to the schema N₁V₁ (N₂ V₂), cf. Z. S. Harris, Co-occurrence and Transformation in Linguistic Structure, The Structure of Language, 1964, p. 178–179.

³² See E. Fraenkel, Kas., p. 156 with literature, Pietk. p. 87.

³³ Underlined above: iós.

³⁴ Underlined, written: ghi drauge prapulusi.

³⁵ Similarly BrB Jer 48, 30.

³⁶ Similarly BrP II 24; cf. also BrB II Mak 4,1 with to speak, see 2.1.2.3.

262

The accusative expressing the proleptic subject of the participial action sometimes also depends on a group 2 verb sakyti ‘say’, e.g.: Sakome išmintį Dievo paslaptų / anu vžusleptu SP II 205³⁷. However, sentences in which the correlative pronoun tatai ‘that’ emphasizes the dependence of the entire NP constituent, rather than its nominal member, on the verb are also used alongside these, e.g.:

IEsu Kristu / wisso swieto ischgelbethogi uszgimusi tatai (79) first of all, the Holy Angel coming from heaven proclaimed all this to the savior of the whole world BrP I 79–80. He told this to his own people, who were with him, saying that an army was coming against him, Karị atentį BrB II Mak 8, 12 (hielt er es seinen lüden fur… wie ein Heer komen würde).

By their internal structure, these constructions are non-object constructions. The correlative pronoun, which indicates the syntactic position of the GD in the sentence, is retained here just as it is sometimes retained in explicative clauses; cf. apsakė tatai, kad jis užgimė³⁸.

The autonomy of the GD is also clearly evident in those examples where the subject of the participial action is expressed in the sentence by a prepositional construction governed by the verb or by another case (not the accusative). Such constructions are likewise non-object in type, regardless of the verb’s valency and meaning.

Furthermore, we ought to know about these Wise Men, that they were not of the tribe of the Jews BrP I 153. And he commanded him to proclaim that he was the third lord, the king Karalis-taie BrB Dan 5,29 (das er der dritte Herr sey)³⁹. From this we can understand: you who believe in Christ, that he is the savior… BrP II 387. Therefore let us learn… to remain faithful to Christ, that he can save [us] from this distress BrP II 457⁴⁰.

3.3.4. The difference between the internal structures of the two GD types can also be formally defined in terms of dependency theory, a separate branch of generative grammar that has been developing in recent years⁴¹. The hierarchy of components in the internal structure of non-object GDs and explicative clauses is represented by the same dependency rules⁴²: V (* N); N (* S); these express that the nominal category (N) dependent on the verb (V) is realized as a clause (S). The same dependency relations can also be identified in the internal structures of corresponding constructions and explicative clauses in other languages⁴³. By contrast, the internal structure of object-type GDs should be represented by the rules V (*N), N (* NS), indicating that the N category is realized both as a noun and as a clause.

³⁷ idem, SP II 18²⁵.

³⁸ On the structure of explicative clauses with correlative words from a transformational perspective, see the author, “Baltistica” VIII (1), 1972, pp. 38 ff.

³⁹ See also BrP II 238; BrB Iz 16,6; I Mak 8,1.

⁴⁰ Similarly, BrP I 92; see also 2.1.3.2. with žvalgytis BrB Job 35,5 and daboties MT 117^s; see also 2.1.3.4. with skustis BrB II Mak 14,26.

⁴¹ Cf. D. G. Hays, Dependency Theory: a Formalism and some Observations, “Language” 40, № 4, (1964), pp. 511–525; J. J. Robinson, A Dependency-based Transformational Grammar, in: Actes du X Congress International des Linguistes, Bucarest, 1970, pp. 807–813

⁴² The rules are formulated according to J. J. Robinson’s notation (J. J. Robinson, Dependency Structures and Transformational Rules, “Language”, 46, № 2, 1970, pp. 265–266), except that category T is replaced by S to facilitate comparison with the rules for phrase-structure constructions. The lines marked with asterisks indicate required elements.

⁴³ On the identity, from the perspective of dependency grammar, of the internal structures of the French acc. c. inf. and the explicative clause, see Б. Ю. Розенцвейг, Об одном типе синтаксической интерфе-ренции, сб.: Фонетика, фонология, грамматика к 70-летию А. А. Реформатского, М. 1971, pp. 270–276; on the relation between corresponding German constructions and clauses according to valency theory (whose general principles are close to those of dependency grammar), see R. Große, Zum Verhältnis von Form und Inhalt bei der Valenz der deutschen Verben, in: Beiträge zur Valenztheorie, Halle (Saale), 1971, pp. 126–127.

263

3.3.5. The distinction between object and non-object GD does not always show up in their external structure. Many constructions with comment-taking transitive verbs in the 16th–17th centuries are, when considered separately, ambiguous, i.e. they may be formed from sentences related in different ways according to different derivational rules. For example, the GD regėjot ghy enti VEE 81 may be formed from the sentences regėjot [tai] (jis eina) according to the rule NP→S, as a synonym of regėjot, kad jis eina; on the other hand, it may also be formed from regėjot jį (jis eina) according to the rule NP→NS⁴⁴. The conditions for the syntactic homonymy of such GD arose long before the beginnings of writing, when the primitive object constructions converged with explanatory clauses of NP→S internal structure (initially asyndetic, later conjunctional). GD was linked to explanatory clauses by their dependence on the same transitive verbs. The relation between the nominal and participial members of GD was likewise of the same syntactic nature as that between the subject and predicate of the subordinate clause in an explanatory sentence. The most important difference in the external form of the two constructions was that in the GD construction the noun or pronoun expressing the secondary subject of the action was formally determined by the verb expressing the main action, whereas in the explanatory sentence its form was independent of that verb. In the latter case, the entire subordinate clause was subordinated to the verb. This difference, however, did not prevent the two constructions from being used in the same environment, especially since the accusative’s dependence on the verb was initially more semantic in nature.

As the nature of government changed and direct and more remote objects became differentiated, in constructions with verbs such as think, know, say, utter and the like (e.g. anis tars, mus nog jų beganczius BrB I Joz 8,6), the pronoun’s accusative mus could no longer be regarded as the direct object of the verb. Such GD either had to disappear, or the syntactic relations between their members had to be modified. Syntactic synonymy with explanatory clauses helped these constructions survive and even extended their use to other comment-taking verbs that did not govern the accusative but could take subordinate clauses of explanatory sentences. Having become equivalent to explanatory clauses, GD also occupied the syntactic positions in which they had previously been unable to occur. The more frequent and familiar it became, the more pronounced this process of analogy grew. As we have seen, in the 16th–17th centuries it had progressed furthest in the language of those authors who used GD most extensively.

With the spread of non-object-type GD, constructions with transitive comment-taking verbs of group 1 (see, hear, notice, feel, find and the like) also began to be used as equivalents of explanatory clauses with an NP→S type of internal structure. In such cases, the accusative of the noun or pronoun, which had the form of a dependent clause constituent, also represented the subject of the constituent sentence. Since the object use of these constructions also persisted alongside this use, they became syntactically homonymous. The type of many of them in old writings can be determined only from the lexical meaning of their members and from context. The contradiction between their differing internal structures and identical external structure made the grammatical form of the construction unstable and encouraged its further development.

⁴⁴ For the dual interpretation of the corresponding English constructions with the -ing form, cf. N. Chomsky, Current Issues in Linguistic Theory, in: The Structure of Language, p. 73; A Transformational Approach to Syntax, ibid., pp. 238–239.

264

4. Features of the Further Development of the Accusative with Participle

A review of the structure and use of GD in Old Lithuanian shows that the two main types of this construction—object and non-object—have differed from the very beginnings of writing in their internal structure and their synonymic relations with other constructions. This distinction is often revealed by the lexical and syntactic properties of the members of GD. As already noted, object constructions are formed only by transitive verbs capable of governing the accusative of concrete and abstract nouns. Non-object constructions, by contrast, can be formed by all comment-taking verbs regardless of case government, and often also by negative forms of transitive verbs that otherwise govern the genitive. The expression of the participial and nominal members also differs. For example, only non-object constructions feature participial links or reflexive pronouns occupying the position of the nominal member. However, in terms of the morphological forms of their main members, most GD of both types coincide at the beginning of the written tradition. The further differentiation of these types can be observed already during the period of the written tradition.

4.1. Non-object constructions

4.1.1. The further development of non-object constructions was greatly influenced by their synonymy with the spreading explanatory clauses. The participle, having become equivalent to the verb in the subordinate clause of an explanatory sentence, lost the syntactic motivation for its accusative form. The intensive use of non-object-type GD in Old Lithuanian further accentuated the discrepancy between the old nominal case form of the participle and its new verbal predicate. All this created favorable conditions for the establishment of indeclinable participial forms—the gerund.

The gerund entered GD as an already established means of expressing a secondary action performed by a different agent (or occurring spontaneously). The gradual spread of gerunds in the 16th–17th centuries has been discussed separately¹, so we will not examine it here; it should only be noted that in some 16th-century sources, especially in J. Bretkūnas’s writings, the first stages of this spread can still be discerned: gerunds usually occur only with neuter adjectives and pronouns (e.g. regesit tātai prassidėdant BrP I 15₁₀; isch̄tirsíu… ger essant BrB I Sam 20, 12), displacing older neuter participial forms (cf. tikra tariu esą BrB II Petr 1, 13), or are used in constructions without the first accusative (gird kẽikent BrB Pat 29, 24); occasionally they perform the function of a verbal copula in other cases as well (e.g. rassi tikrai tiesą sant BrB Moz 17, 4). The subsequent spread of gerunds was very rapid: in 16th-century writings they occur in only 3.4% of all GD, whereas in the 17th century they are found in 38%. In the 18th–19th centuries, the gerund form becomes the main and predominant form in non-object constructions. Non-object constructions with inflected participles gradually disappear, giving way to GD with gerunds, subordinate clauses, and other syntactic means.

4.1.2. The decline of non-object constructions with inflected participles and the establishment of gerunds are clearly evident when comparing texts from different periods translated from the same or similar originals. For example, in many passages where J. Bretkūnas uses GD with inflected participles, the 1701 Bytner New Testament transla-

¹ See the author, “The Accusative with Gerund in Old Lithuanian Sources,” Baltistica, IV(2), 1968, pp. 195–209 with references.

265

In [the Bible] or in L. J. Rėza’s Bible of 1824, we already find corresponding converbal constructions (see Table 3). Cf.:

And whom do you say that I am? BrB Luk 9, 20 – whom you say me to be Bt – whom you say me to be RB (Vos autem quem me esse dicitis? Wer saget jr aber das ich sey?)².

We know that he hears us BrB I Jon 5, 15 – we know him to be listening to us RB – we know | that he hears us Bt (Wir wissen | das er vns höret; scimus quia audit nos).

He had certain questions... about one... whom Paul said was alive. BrB Apd 25, 19 – whom he said to be alive Bt – said him to be alive RB (Paulus sagete | er lebete; quem affirmabat Paulus vivere).

But he thought that his brothers understood BrB Apd 7, 25 – he thought his brothers would understand RB – he thought | that his brothers... would understand Bt (Er meinet aber seine Brüder soltens vernemen. Existimabat autem intelligere fratres³).

You heard that he had been ill BrB Filip 2, 26 – heard him to be ill Bt – having heard that he had been ill RB (gehöret hattet | das er kranck war gewesen; audieratis illum infirmatum).

Of the 102 non-object constructions with participles used by J. Bretkūnas, 10 were replaced in S. Bytner’s translation, and 20 in L. J. Rėza’s Bible. More than converbal constructions, explanatory clauses push non-object GD with inflected participles out of use. The growing role of these clauses is especially evident in S. Bytner’s New Testament, where explanatory clauses introduced by jog (occasionally kad) replace more than half of all the GD used by J. Bretkūnas. Some GD, for example those with norėti, are not replaced by converbal constructions, but disappear altogether, and their place is taken by infinitival phrases, e.g.:

How he would like him to be called BrB Luk 1, 62 – how he would like him to be called Bt – would like to name him RB (quem vellet vocari eum; wie er jn wolt heissen lassen).

Similarly, GD with a personal or reflexive pronoun denoting the same participant as the verbal action gives way not to converbal constructions, but to a nominative with a participle (nominativus cum participio).

Those who say that they themselves are Jews BrB Apr 3, 9 – who say themselves to be Jews Bt – those who say they are Jews RB (die da sagen | sie sind Juden; dicunt se Judaeos esse).

The adversary... presenting himself as God⁴ BrB II Tes 2, 4 – presenting himself as God Bt – says that he is God RB (gibt sich für | er sey Gott; tamquam sit Deus).

4.1.3. As subordinate explanatory clauses and other synonymous constructions spread, the sphere of use of non-objective GD constructions generally narrows in the eighteenth and nineteenth centuries. For example, in S. Bytner’s 1701 translation of the New Testament, non-objective GD constructions with inflected participles are retained in only 5 of the 102 places where J. Bretkūnas used them, while non-objective GD constructions (with participles and gerunds) account for only about 15% of the GD equivalents in J. Bretkūnas. Similar data were obtained by comparison with J. A. Giedraitis’s 1816 translation, “Naują Įstatymą,” of J. Bretkūnas’s text.

² Similarly, Mark 8:27; Luke 9:18.

³ Cf. also Luke 5:24; 20:6; Rom. 3:8; I Cor. 14:23; II Cor. 8:22; 12:19; Rom. 4:18.

⁴ himself and being underlined; beside it – essans.

266

J. Bretkūno pavartotų GD atitikmenys S. Bytnerio (1701) ir L. J. Rėzos (1824) Naujojo testamento vertimuose

Table 3

Bt 1701

RB 1824

Non-objective

Objective

Non-objective

Objective

Accusative with participle

5

149

33

160

Accusative with gerund

10

6

20

11

Subordinate clause

68

14

20

5

Other constructions

19

12

29

5

Non-objective GD constructions were much better preserved in J. L. Rėza’s 1824 Bible, where participial and gerundial constructions account for about 50% of the GD equivalents used by J. Bretkūnas. Moreover, GD constructions in Rėza’s Bible often occur in places where J. Bretkūnas used other syntactic means. For example, in Acts alone, we find 36 non-objective GD constructions with participles and 13 with gerunds here—that is, even more than in J. Bretkūnas’s translation. However, it is highly doubtful that such a great abundance of non-objective constructions with inflected participles reflects the tendencies of the spoken language of the time. In the works of K. Donelaitis, written in a dialect in which GD constructions must have been very widespread, we see a completely different picture: only gerunds occur in non-objective constructions here. Apparently, the use of inflected participles in nineteenth-century religious writings was sustained by literary tradition, which particularly constrained changes in the language of sacred texts. GD constructions with inflected participles, customary in older literature, probably already had an archaic tinge at that time, and were used for stylistic purposes. Efforts to explain the linguistic devices characteristic of older writings partly account for the more frequent use of GD constructions with inflected participles in S. Daukantas’s works, which are otherwise rich in traditional syntactic devices (cf. inflected participles in the absolute dative, see LKK V 61); nevertheless, gerunds are found here too in most non-objective constructions. That non-objective GD constructions with inflected participles had already become rare in spoken language in the nineteenth century, and had disappeared altogether in many places, is evident from the writings of authors closer to popular speech, such as M. Valančius, A. Baranauskas, A. Strazdas, S. Stanevičius, and others, as well as from narrative folklore recorded at the time (folk songs preserved GD constructions with inflected participles much better owing to their metrical and rhythmic features). From the late nineteenth to the early twentieth century, when the literary language became firmly grounded in popular speech, non-objective constructions with inflected participles fell out of current use. They are not characteristic of the modern literary language and are used only occasionally in fiction as stylistically marked archaisms⁵.

4.1.4. In contemporary language, the main constituents of non-object constructions with adverbial participles are verbs from the same valency groups (and sometimes other wor-

⁵ Cf. Perhaps that is why, since the old days, when Gigi­s was still younger, everyone thought him to be a sorcerer KrėvRg 85; see also LTSR MA, Lietuvių kalbos gramatika, II, Vilnius, 1971, p. 363.

267

ds), which in the 16th–17th centuries formed corresponding constructions with inflected participles. The following are the most commonly used among them:

Group 1 verbs see, notice, hear, hear (of), listen to, feel, perceive, find, show, suspect, which in other cases also form object constructions, e.g.: The adviser saw his work being entirely in vain BsRP 84. I noticed that the clock had struck at quite the wrong time CvR III 234. I hear that your sweetheart has been drinking with someone else Nm. Father, having heard that he was hiring a maid, will probably hurry over CvR III 41. Andrius listens to the German woman repeating words he does not understand CvR IX 179. How did he feel his father would act? LKŽ IV 318. Kastis felt his eyes filling with tears ŠRagR III 21. (The young man) found everything coming to nothing BsMt II 207. This shows these motives to be very old (trš). How can you suspect me of being an egoist here! ŠRagR III 236.

Group 2 verbs know, understand, think, believe; say, speak, assert, report, which in GD with inflected participles are no longer used at all in contemporary language.

So that he might know the dawn was breaking JD I 201. He understood that his father had obtained a loan PaukšK 55. Kastis thought she was asleep ŠRagR III 124.⁴ I would not have believed Cipras to have remained like that over those few years. ŽemR II 118. Let the men then say that women do not know how to earn money VaižgRR II 1689. People said that the neighbor had sold the cow Krs. The daughter confirmed that her father had returned at almost the same time (trš.). The butler comes and reports that someone has arrived with a matter to attend to ŠRagR II 102.

With reflexive verbs of group 3 agree, expect adverbial-participle constructions are formed very rarely, and with boast, rejoice, complain, learn and the like they are no longer used at all. However, they are quite often linked to a predicate infinitive, e.g.: The sound of scythes being sharpened could be heard CvR III 190. From this it was evident that she was an excellent housekeeper Sln. And the young Vidmantas began to think their father guilty VaižgR VII 40.

Here, adverbial-participle constructions occupy the nominative position and in this respect correspond to GD in old texts following group 4 words. Because of their close connection with the nominative subject, their nominal constituent sometimes itself takes this case form, e.g.: a larger ship can be seen passing through SimonVK I 206⁶ (cf. a ship can be seen). In the nominative subject position they also occur with predicative neuter adjectives, passive participles, and certain verbs (seem, turn out, become clear and the like), e.g.: It was clear that a storm with hail was approaching MPs 43. It was not yet noticeable… that she had turned into a toad VaižgRR I 134–135. It is rightly said that the fox is cunning VaižgRR II 173. The guests think that no one in the world could guess the cook… VaižgRR 118. From this it follows that there were two priests LT I 345.

Thus, the accusative with an adverbial participle is found in contemporary Lithuanian under the same circumstances as non-object constructions with participles in 16th–17th-century texts. However, the scope of use of the non-object type of accusative with an adverbial participle is now considerably narrower. In corresponding cases, explanatory clauses synonymous with the construction are used much more often than in the older language.

4.1.5. The establishment of the adverbial participle in GD is part of the broader process of differentiation between attributive and semi-predicative participles, which also manifested itself in other constructions with oblique cases. For similar reasons, but much earlier, the dative of the pre-base stem of inflected participles became fixed in the absolute dati-

⁶ See ibid., p. 395.

268

in the construction, thereby laying the foundation for the participial category itself⁷. Just as in GD, the gerund also found its way into genitive-with-participle constructions; cf.: never experienced her being Jrk 112⁸.

A similar entrenchment of indeclinable participles (so-called gerunds, absolutives, etc.) in place of semi-predicative participles can also be seen in many other related languages. Latvian has gone still further along this path: there, present active indeclinable participle forms in -m have also become fixed and are widely used in GD constructions (cf. neredz saules uzlecam⁹). Indeclinable participle forms in the GD of the Slavic languages are found as early as the oldest records (cf. s. sl. vidiši učenika predaješte Supr 408)¹⁰, but their eventual establishment was hindered by the very decline of non-object constructions and their replacement by explanatory clauses. The use of the Latin accusativus cum infinitivo, whose meaning is close to that of the Lithuanian accusative with gerund, and which set the course for the formation of corresponding constructions in the Romance and, in part, Germanic languages, is probably not original either and may be explained by the entrenchment of the infinitive in place of older accusative forms of the participle¹¹.

4.1.6. In the development of Lithuanian, we also observe tendencies toward further changes in the syntactic relations of GD. Once the gerund had become established in place of the participial accusative, the dependence of P on N, which had been marked by agreement, faded completely in non-object constructions. Performing the same function as a verb in a subordinate clause of an explanatory sentence, the gerund became the main member of the NP phrase, representing its syntactic relations with V. Its uninflected form made it possible to use it without a nominal or pronominal accusative (or genitive). In Old Lithuanian, impersonal and other closely related constructions without N were among the earliest centers from which the gerund spread in GD; in sixteenth- and seventeenth-century written records they make up 14.1% of all gerund constructions, while in J. Bretkūnas’s writings, where this spread was only just beginning, they account for as much as about 65%. In present-day language, gerund constructions without an accusative are used in impersonal, indefinite-personal, elliptical, and other utterances, e.g.:

I realized it was necessary to wait ValPJ 29. I felt it dripping Grž. Then nearby I heard someone speaking ŽemR II 52. I saw them fighting Sk. I see it rising and greening on the noble little courtyard KlvD 32¹².

When used without a nominal or pronominal accusative, the gerund itself occupies the position of an object case in such instances; cf.: Having walked a little way, he heard a nose-snort, then a cough, and soon after that a frightened voice CvR III 61.

The nominal or pronominal accusative, having become syntactically unnecessary, comes to be regarded as dependent on the gerund. This dependence of N on P is most clearly seen in constructions with negative gerunds, which, like verbs, govern the genitive form of N, e.g.:

⁷ For more on this, see LKK V 15 ff.

⁸ See J. Jablonskis, Rinktiniai raštai, I, pp. 473—474; E. Fraenkel, Kas., pp. 84 ff.; LTSR MA, Lietuvių kalbos gramatika, II, Vilnius, pp. 363, 395.

⁹ J. Endzelins, Latviešu valodas gramatika, Rīgā, 1951, p. 1001 — 1002; LPSR ZA, Mūsdienu latviešu literārās valodas gramatika, I, Rīgā, 1959, p. 664.

¹⁰ W. Vondrák, Vergleichende Slavische Grammatik, ² II, 407.

¹¹ Cf. A. Hahn, Genesis of the Infinitive with Subject-Accusative, TAPA, 81 (1950), pp. 117 ff.

¹² Žr. dar J. Balkevičius, Dabartinės lietuvių kalbos sintaksė, V., 1963, p. 195; LTSR MA, Lietuvių kalbos gramatika, V., 1971, p. 395—396.

269

Those who observe this custom in their homes say that there are no fleas throughout the entire year MLLG II 179. Seeing that there would be no honest conduct, the fisherman did as that German lad wanted MLLG I 360 (Š). (cf.: you saw him not being a tiller of the soil VaižgRR I 443).

Constructions of this kind occur as early as in the old writings, e.g.: the rogues... among the Sacraments... there being no separation of anything / but one way being established MT 120₁₆. Unlike the corresponding constructions with participles, e.g.: you know, none of us being present BrB I Kar 5, 6 (see 2.2. 3. 4.), here there is no longer an overt syntactic relation between N and P; only the form of N depends on the converb, and not vice versa¹³. Similar relations are also found in constructions where a converb occurs with a partitive genitive, e.g.:

It was apparent that there was some kind of unrest in his soul VaižgRR I 128. At first we thought that we, too, would have such innocent jokes VaižgRR II 151. From house to house people could be seen walking in groups CvR IX 110.

The constructions of the type discussed here point to the emergence of the following model of syntactic relations in non-object-type GD:

(5) V → P → N

4.2. Object constructions

4.2.1. In the development of object constructions, the relation to complement clauses played a much smaller role, since the complement clauses synonymous with them and containing a proleptic subject (cf.: I see my father returning and I see my father, that he is returning, see 3.3.2.) are relatively rare and are entirely uncharacteristic of the literary language today. In these constructions, the case-government relation between V and N remained intact. As sentence cohesion grew stronger and its individual parts became more dependent on the verb, the participial member of the GD was also subordinated directly to the latter; see 2.1.5.(4). This model of syntactic relations in object GD, already found in the oldest writings, has remained essentially unchanged in present-day Lithuanian. However, with the formal separation of non-object GD, object-type GD with inflected participles came even closer to the corresponding constructions dependent on verbs that lack phrasal government, i.e. non-commentable verbs, while the participial accusative itself acquired a clearer qualitative meaning.

In present-day Lithuanian, object constructions with inflected participles are most often formed by the following commentable physical-perception verbs belonging to group 1: see, behold, catch sight of, hear, hear (something), notice, feel, sense, find, e.g.:

I would see him among the older ones milling about Maš AMB 78. We see the sheep’s thing lost VaižgRR II 158. And when he saw a little fire glowing in the distance, he set off in that direction LTR. Just let the forest be heard rustling JD II 551. He heard the little bee flying up. Vrn. Mother noticed me trembling SimonOBT 158–159.

¹³ In the old writings, transitional cases also occur in which the same verb governs both a construction with a participial genitive and one with a converb, cf.: S. Powilas... says that there is no creation and that it cannot exist (powiada że niemasz stworzenia / y że byc nie może) SP II 114₆. For he will see that their strength is gone, that they are no more, both shut away and abandoned BrB V Moz 32, 36 (wird ansehen / das jre Macht dahin ist / vnd beide das verschlossen vnd verlassen weg ist).

270

As if I feel some misfortune about to befall me… ŽemR IV 474. People found Gugė frozen KrevRg 254¹⁴.

In object constructions with non-commentable verbs, the principal verb is most often leave, meet, catch, keep, have, lead, carry, transport, throw, protect, help, (take), (lift), and the like, e.g.:

Jasiulis… left Marelė sighing heavily JD III 1210. The master… would catch her before she had even run out of the yard CvR II 59. Pi menú kept the horses hobbled all night. LKŽ VII 37. Purvino’s hired man brought her back utterly exhausted SimonoOBT 70. Tesaaugo he kept us asleep Šd (MLLG IV 45). Unable to feed the child, the girl was left wrapped by the roadside CvR III 14.

In such constructions, the accusative participle primarily marks the presence of a particular property of the object at the time of the verbal action¹⁵. Thus, the qualitative, adjectival meaning of the participle predominates here; cf. I found him tired | angry, or she led her out undressed | naked. On the other hand, in those object constructions where the verbal meaning of the participle predominated, converbs had already become widespread during the period of the written tradition.

4.2.2. Data from old writings suggest that the converb began to penetrate object constructions only after it had become established in non-object (impersonal) constructions or with neuter pronouns. J. Bretkūnas forms object constructions only with inflected participles; with rare exceptions, the same can be said of M. Daukša, B. Vilentas, and S. Vaišnoras. However, in seventeenth-century writings (e.g., K. Sirvydas’s “Punktai”), we already often find converbs in object constructions; this usage is particularly prominent in B. Chylinski’s Bible. In K. Donelaitis’s writings, as in the modern language¹⁶, converbs predominate and are freely used with an accusative dependent on the verbs see, hear, hear (of), feel, and the like, e.g.: we see them all returning sturdy in summer DnR 42; hearing the shining carriages rumble 28; he would hear the peasants changing 46; let us not fear that we will hear storms roaring 31; with summer and its warm days, we feel it returning 30.

Since object constructions after commentable and non-commentable verbs follow the same model of syntactic relations, the converb also began to penetrate the latter, especially with verbs of similar meaning. For example, once converbs had become established in constructions with find, conditions arose for using a converb in corresponding expressions with the non-commentable synonym of that verb, catch unawares (e.g.: many times a whirlwind caught me working DnR 79), or with its partial antonym, leave (e.g.: I left them making a stink DnR 132). In the modern language, converbs are most often found in constructions with non-commentable verbs such as leave, meet, come across, catch up with, seize, catch, nab, grab, and the like.

Those (girls)… left their rescuer asleep in a leap BsMt I 186. They met a man riding a horse BsV 299. And she found that child having feathers BsMt I 68. And that wolf, having run out, caught up with the fool while he was driving LB 242. Perhaps we’ll nab…

¹⁴ See also J. Jablonskis, op. cit., pp. 472 ff.; 614–615; J. Balkevičius, op. cit., pp. 193 ff.; LTSR MA, Lietuvių kalbos gramatika, II, pp. 362–363; on corresponding constructions with the genitive, see ibid., pp. 363–364.

¹⁵ Cf. A. Valeckienė, Predikatyvinis pažyminys kaip atskira sakinio dalis, LKK IX, p. 99 and footnote 13.

¹⁶ See J. Jablonskis, op. cit., pp. 321–322, 472; J. Balkevičius, op. cit., pp. 193 ff.; LTSR MA, Lietuvių kalbos gramatika, I, pp. 392 ff.

271

that thief while stealing J (LKŽ II 115) Gyrūnas gets upset when he catches him lying KrėvP III 11. Once his father caught him pulling up turnips in the garden LzPR II 138. He caught him while he was lying Upt.

As we can see, gerundial constructions are not formed with all verbs that can serve as the main members of the corresponding participial constructions, but only with those that are semantically associated with commentable predicates. Meanwhile, both participial and gerundial constructions may depend on group 1 commentable verbs denoting physical perception. Their meanings, however, are not quite identical. When the object of perception is the action itself, gerundial constructions are usually used; when it is a quality, participial (or adjectival) constructions are used. The meanings of the two constructions are already clearly distinguished in K. Donelaitis; cf.: but others among them… tumbled head over heels and, while crawling about in the yard, found Durakas DnR 99 and they found the bedding thoroughly damp everywhere 22. In contemporary Lithuanian constructions with the gerund, the dependent member likewise predominantly has a verbal meaning, whereas with the participle it has an adjectival meaning; cf.: I saw my brother returning KlvD 138 and the woman… saw the man coming CvR IX 114. Thus, object constructions with gerunds are semantically close to the corresponding non-object constructions (cf. 4.1.4.). This process of semantic differentiation in object constructions is by no means complete, and in many cases the gerund and the accusative of the participle may alternate here with only a very slight difference in meaning.

4.2.3. A similar differentiation of the corresponding constructions is also found in other Indo-European languages. In Latin, the accusativus cum participio and accusativus cum infinitivo after verbs of physical perception such as video, audio, etc., e.g. eum audivi cane-u tem / canere or eum vidi ingredientem / ingredi¹⁷, differ by almost the same shade of meaning as Lithuanian I hear him singing / singing or I see him enter / entering. A similar distinction between object-case constructions with the -ing form (close to our participles) and the infinitive also developed in English¹⁸.

4.3. The survey of the accusativus cum participio shows that the syntactic relations among its constituents changed in the same direction as in other participial constructions with oblique cases. These changes were primarily conditioned by the following broader regularities in the development of the syntactic system:

a) the gradual subordination of semi-predicative constructions to the verb, associated with increasing sentence cohesion and unity;

b) the strengthening of verbal government and the differentiation of governed constructions from adjoined ones;

c) the differentiation of forms expressing semi-predicative and attributive relations.

Received

1973.III.12

¹⁷ See O. Riemann, Syntaxe latine d’après les principes de la grammaire historique, Paris, 1927, p. 536. Cf. also the difference in meaning between the infinitive and the participle in Plautus’s writings, as noted by A. Hahn: the infinitive means “perceive that,” while the participle means “perceive someone doing something.” See A. Hahn, Genesis of the Infinitive with Subject-Accusative, TAPA, 81 (1950), p. 125.

¹⁸ Cf. K. Yamakawa, The Two Constructions: “Accusative and Participle” and “Genetive and Gerund,” “Annals of the Hitotsubashi Academy,” Tokyo, 1957, V. 8, № 1, p. 44—45, 47, 62.

272

В. АМБРАЗАС

STRUCTURE AND USE OF THE ACCUSATIVE WITH PARTICIPLE (ACCUSATIVUS CUM PARTICIPIO) IN OLD LITHUANIAN

Abstract

The structural and genetic basis of the accusative with participle (hereafter ACP) is formed by double-accusative constructions, in which the first accusative, of a noun or pronoun, depends directly on the verb, while the second accusative, of a noun, pronoun, or participle, is linked to the first by a semi-predicative relation, e.g.: Kat sakei ie tawa Seser? BrB I Moz 12, 19. Of all the varieties of the double accusative, the ACP became the most widespread in Old Lithuanian and underwent a distinctive course of development.

Relics of the prepositional-less accusative used for an indirect object with the verbs kalbėti, sakyti, byloti, skelbti, rašyti, girdėti, klausyti, žinoti, mąstyti in sixteenth- and seventeenth-century sources (e.g.: kalbame išzminti ChB I Kor 2, 6; girdeiom… darbus BrP I 231) support A. A. Potebnya’s and E. Tangl’s hypothesis that the nominal constituent (N) of the construction originally depended on the verb (V), and allow a single primary model of the syntactic relations in the ACP to be reconstructed: V→N←→P. This model is motivated by certain ancient features of the Indo-European sentence: the semantic nature of the relation between case forms and the verb, the freer use of appositive constructions akin to clauses, and the function of a secondary predicate characteristic of participial forms that are nominal in origin.

The analysis of the syntactic form and use of the accusative with participle in this study is based on almost all currently available written sources of Lithuanian from the sixteenth and seventeenth centuries. For the classification and statistical characterization of the verbs serving as the main component of the ACP, see Tables 1 and 2.

Almost half of the ACP constructions in Old Lithuanian texts are formed with verbs that govern an object in the accusative, both with concrete and abstract meanings. These include verbs of physical perception matyti, regėti, veizdėti, išvysti, girdėti, išgirsti, klausyti, jausti, pajusti, rasti and the like; verbs of indication rodyti, ženklinti; and some verbs of mental activity: pažinti, atminti, laikyti. Most of these still retain a direct object relation with the nominal constituent of the ACP. However, in almost 41% of these constructions, the accusative noun or pronoun can no longer be regarded as a direct object of the verb. In such cases, the aforementioned verbs of physical perception take on the historically secondary meaning of mental perception.

ACP constructions with verbs of mental activity, which usually govern only abstract objects, are also widely attested: žinoti, suprasti, manyti, mąstyti, dingoti, sapnuoti, tikėti, nulemti and others. A direct object relation between the verb and an accusative noun or pronoun is found here only in comparatively rare cases of an archaic character.

18. 520

273

The direct dependence of the first accusative (N) on the verb is already often replaced in the earliest sources by clausal government of the construction as a whole. As a result, the ACP occurs with verbs that cannot, and apparently never could, govern the accusative case, for example, with reflexive verbs of supposition tikėtis, tartis (palaima padetą tikis DP 596), with impersonal verbs of state rodytis, regėtis, rastis (regissi [iums] mane walgantį BrB Tob 12, 19), and others. In Old Lithuanian writing, the ACP also depends on a neuter participle functioning as a predicate, e.g., žinoma, giedama, rašoma, kalbama, randama… sakyta, parašyta, regėtina (yra, buvo), and in some cases even on nouns garsas, stebuklas, nusitikėjimas, sententencija, dimos, gromata. The widespread use of the participial link (e.g.: kosznas tare, diena sancze BrP 337) also attests to changes in syntactic relations, which can be represented by the schema V→(N←—→P).

As a result of sentence centralization, changes in syntactic relations are also observed in constructions that have preserved the object dependence of the accusative noun (N) on verbs regėti, matyti, jausti, rast and the like. The participle, which is related only to the noun, also enters into a relation with the verb, and the construction assumes the typical form of a three-member phrase:

The participle in the construction under study has all the structural properties of a verb. Active participles are used here as equivalents of simple finite verb forms, and passive participles as equivalents of periphrastic passive forms. Unlike in many other Indo-European languages, active and passive participles also occur in future-tense forms. ACP constructions with participles of the verb būti are especially widespread; they form combinations corresponding in function to compound predicate forms and account for about 20% of all ACP constructions in Old Lithuanian writing.

An examination of the syntactic structure of the ACP leads to the conclusion that, in the complexity of its grammatical form, the semantic diversity of its verbs, and the breadth of its sphere of use, the Old Lithuanian construction surpasses corresponding constructions in other related languages. In Old Lithuanian, the possibilities for forming and using this construction, inherited from the Indo-European linguistic community, developed especially richly, as a result of which the ACP came to perform functions for which many other related languages generalized infinitival constructions and subordinate clauses.

In its deep structure, revealed by transformational analysis, the Old Lithuanian ACP corresponds to a subordinate complement clause. Most ACP constructions in sixteenth- and seventeenth-century sources are formed according to the derivational rule NP→S. ACP constructions with object dependence have preserved the deep structure of the type NP→NʹS, which is also characteristic

274

for complement clauses with a proleptic subject. The internal structural affinity between the ACIP and the complement clause provided the basis for their widespread synonymous use and played an important role in the further development of the construction under study.

The transformation of syntactic relations within the construction created favorable conditions for the spread of indeclinable participial forms in -nt and -us-, the so-called gerunds. Gerunds enter the ACIP as an already established, formally marked means of expressing semi-predicative adverbial relations. They first appear with neuter adjectives and pronouns and in constructions without a first accusative; however, in the seventeenth century they spread very rapidly to constructions with a first accusative of masculine and feminine nouns as well. Whereas gerunds are found in only 3.4% of constructions (a ratio of 76: 2.177) in all the sixteenth-century texts available for study, in the seventeenth century the figure is already 38% (349: 571).

In eighteenth- and nineteenth-century written sources, constructions with gerunds continue to spread and gradually become dominant. ACIP constructions after many verbs fall out of use and are replaced by the corresponding gerund constructions or subordinate clauses.

The ACIP in the modern literary language differs from the corresponding Old Lithuanian construction both in its structure and in its place within the system of syntactic means. At present, constructions with declinable participial forms are used only when the first accusative is the object of verbs of sensory perception (matyti, regėti, išvysti, pastebėti and the like), discovery (rasti, aptikti, suikti), and certain factitive verbs that lack phrasal government (vesti, siųsti, palikti, suikti and the like). The participle’s accusative here functions as a predicative attribute and has a double syntactic relation (with the first accusative and with the verb); it alternates with the gerund, cf. matau tėvą sėdintį (sėdint). When the verb governs the construction as a whole and the first accusative cannot be regarded as its direct object, only gerunds are used, e.g. žinau tėvą sėdint (in such cases, the participle’s accusative is possible only as a stylistically marked archaism).

Changes in the structure of the ACIP, which proceeded in the same direction as in participial constructions with the dative case, were conditioned by certain more general patterns in the development of the syntactic system:

a) the gradual subordination to the verb of semi-predicative constructions, associated with the growth of the sentence’s internal unity (compactness);

b) the differentiation of forms expressing semi-predicative and attributive relations;

c) the narrowing of the scope of the prepositionless accusative as a result of a clearer distinction between direct- and indirect-object forms.

Received 12.III.1973

18*