Old Lithuanian “añskat”, “šiskat”, “tàskat” and Cognates

11straipsniai / articles danieL Petit École normale supérieure, Paris oLd Lithuanian AñSKAT, šÌSKAT, TàSKAT and cognates senosios lietuvių kalbos añskat, šskat, tàskat ir su jais susiję žodžiai 0. introduction first introduced by the greek grammarians (chrysippos, apollonios dyskolos) and later elaborated by the german psychologist karl bühler (1934: 80), the notion of ‘deixis’ has been much debated throughout the 20th century1. although intuitively appealing, it is an artifact, which encompasses various types of gestural reference and covers a wide range of phenomena with different semantic features. among the most striking expressions of deixis, the so-called ‘thetic judgements’ occupy a special position, in that they simply report the existence of a situation (e.g. engl. there is a cat sleeping in my room), while ‘categorical judgments’ simultaneously assert the existence of an object and describe its action or state (e.g. engl. a cat is sleeping in my room = there is a cat + it is sleeping in my room). this distinction is well known2. however, a subclass of thetic judgments seems to have been generally overlooked in the standard textbooks of linguistics and in most specialized monographs and articles dealing with deixis3. Many languages present special deictic particles, which draw the attention of a speech partner to some event or state of affairs (e.g. Lat. ecce, russ. vot, fr. voici, voilà ‘here is/are’). by their most salient syntactic and semantic features, especially by 11 see for example charles fillmore (1997) or stephen Levinson (2006) for a recent overview. 12 see for example hans-jürgen sasse (1987). 13 an exception is Levinson (2006: 111), who uses the term ‘presentative particles’, otherwise very rare in english-language linguistics. ALL_62_63_maketas.indd 11 2011.06.17 08:23:46 12 danieL Petit acta Linguistica Lithuanica LXII–LXIII their reference to a speech partner, such particles differ from the other deictic particles (e.g. Lat. ibi, russ. zdes’, fr. ici, là ‘here’). in scholarly literature, they are sometimes named ‘presentative particles’; this technical term goes back to the description of the biblical hebrew particle hinneh ‘here is/are’ and has been afterwards applied to the french particles voici, voilà. as i tried to show elsewhere4, the notion of ‘presentative particles’ is a powerful and attractive notion which could account for the use of some deictic particles of the baltic languages, especially Lith. šta, Latv. re, raũ, lũk and še ‘ecce’. the aim of this paper is to discuss the origin of a subclass of presentative particles that is widely attested in old Lithuanian and in some Lithuanian dialects, but remains unexplained from an etymological point of view. 1. descriPtion in old Lithuanian literature and in some modern Lithuanian dialects, we find a set of presentative particles, which obviously share some common morphological features, but still lack a convincing etymology. descriptively, these particles are a compound of a demonstrative stem (anàs, šs, resp. tàs) and an ending -(s)ka(t) or -(s)ga(t). in table 1 all relevant forms are listed and briefly described. judging from the main uses of these different particles in Lithuanian, it is clear that they belong to the category that i have briefly presented above under the name of ‘presentative particles’: their basic meaning corresponds to that of the Latin particle ecce. to begin with, some of these particles may introduce a nominal form. examples with šskat: [1] Wolfenbüttel Postilla (WP 118 2–3 [1573]): ſchiſkat tar/naite weſʒpaties. ‘here is the maid servant of the Lord!’ [2] jonas bretkūnas, Postilla (bP i 84 21 [1591]): Jei tawe gandin Smertis bei prapůlims / tadda ſchiſkat tawa gelbetoias Kriſtus. ‘if death and perdition frighten you, then behold your saviour, christ!’ With šsgatės: [3] simonas daukantas (LkŽ XiV 914): šisgatės piemuo! ‘here is the shepherd!’ 14 daniel Petit (2010). ALL_62_63_maketas.indd 12 2011.06.17 08:23:46 senosios lietuvių kalbos añskat, šskat, tàskat ir su jais susiję žodžiai 13straipsniai / articles tabLe 1 stem ansstem šisstem taswith voiceless consonant -k- -ka šiska daukantas -kat anskat bretkūnas šiskat Wolfenbüttel Postilla, bretkūnas taskat Wolfenbüttel Postilla, bretkūnas, klein, sapūnas, Mielcke, clavis germanico-Lithvana, Lexi con Lithuanicum, nesselmann, kurschat -kate šiskate Wolfenbüttel Postilla -kategi taskateghi Wolfenbüttel Postilla with voiced consonant -g- -g tàsg dialects of Papilė, tirkšliai -ga ansga juška šiga juška, dialects of alsėdžiai, skaudvilė šsga kossarzewski, dialect of raudėnai tàsga Pabrėža, juška, dialects of Plungė, raudėnai, Žemaičių kalvarija -gat tàsgat dialects of kvė darna, rietavas, salantai -gatės añgatės dialects of kuliai, Plateliai, salantai añsgatės alsėdžiai, gargždai, grūšlaukis, stakliškės šgatės sereiskis, dialects of salantai, skuodas šsgatės daukantas, juška, kossarzewski, dialects of alsėdžiai, gargždai, Plateliai, salantai, Veiviržėnai, Žarėnai tàsgatės juška, Lazdynų Pėlėda, baronas, dialects of grūšlaukis, ylakiai, kvėdarna, Luokė, Plungė, rietavas, salantai, seda,tirkšliai, truskava, tverai, Viekšniai, Žarėnai -gati tàsgati dialect of Žemaičių kalvarija -gaties šsgaties dialect of Žarėnai -gatis šsgatis niedermann, senn, brender and salys tàsgatis Pabrėža, Valančius, juška, dialects of kretinga, Mosėdis, Plateliai, raudėnai, salantai, Žemaičių kalvarija -gačiuos tàsgačiuos dialect of Šatės ALL_62_63_maketas.indd 13 2011.06.17 08:23:46 14 danieL Petit acta Linguistica Lithuanica LXII–LXIII With tàskat: [4] Martynas Mažvydas, Gesmes Chrikſcʒoniſkas (MŽ 447 4 [1570]): Motriſchk taſkat sunus tawa! ‘Mother! there is your son!’ [5] jonas bretkūnas, Biblia (bb song 5, 2 [1590]): taskat balsas mana Prieteliaus tuskinanczio. ‘there is the voice of my friend who is knocking.’ in this construction, the presentative particles are regularly followed by a nominative. this nominal construction is predominantly attested to in old Lithuanian religious literature; an influence of the biblical phraseology is possible. however, some instances may be found in later sources as well. example with tàsga: [6] antanas juška, Литовскій словарь ([1897–1922], cf. LkŽ XV 961): Tàsga triba jo dirvos! ‘here is the edge of his field!’ a construction with an indirect interrogative clause is also found, but seems to be very rare. to judge from the material collected from the LkŽ and old Lithuanian literature, very few instances can be produced, and most of them are doubtful. for example: [7] jurgis Pabrėža’s writings (LkŽ XV 961): Tasga, koks vėl reikalas! ‘Look over there! What a story!’ obviously, we are dealing here not with an indirect interrogative clause, but with an independent particle tasga followed by an exclamation koks vėl reikalas! ‘what a story!’. We are not yet able to determine whether this impossibility to use these particles with indirect interrogative clauses is due to the scantiness of our documentation or reflects a significant feature. the same particles may be used as sentence-opening particles, introducing independent verbal predicates. this construction is abundantly attested not only in old Lithuanian literature, where one might reckon with the influence of biblical phraseology, but also in the modern dialects, where such an influence is precluded. examples with añskat: [8] jonas bretkūnas, Biblia (bb 22 kings 4, 25 [1590]): Ansſkat ateia Sunamitiene. ‘behold! the woman from sunam is arrived.’ ALL_62_63_maketas.indd 14 2011.06.17 08:23:46 senosios lietuvių kalbos añskat, šskat, tàskat ir su jais susiję žodžiai 15straipsniai / articles With šskat: [9] Wolfenbüttel Postilla (WP 158 30 [1573]): Jr ſchiſkat Rit met ankſti iſch wiſsi ſʒali muras paſsikiele. ‘and behold! early morning arose from all countries a wall.’ [10] jonas bretkūnas, Biblia (bb judg 15, 16 [1590]): Schiſkat anis gul pulkais. ‘behold! they are lying gathered together.’ [11] jonas bretkūnas, Biblia (bb Lk 7, 27 [1590]): Schiſkat ſiuncʒiu Angel mana. ‘behold! i am sending my angel.’ With šskate: [12] Wolfenbüttel Postilla (WP 195 12–13 [1573]): A meſs ſchiſkate, ka/iamis miname. ‘and we – behold! – are trampling down with our feet.’ With tàskat: [13] Wolfenbüttel Postilla (WP 16v 28–29 [1573]): A taſkat artinaſi iumis karaliſ/te Dangaus! ‘and behold! the word of the realm of heaven is coming to you!’ [14] jonas bretkūnas, Biblia (bb gen 29, 6 [1590]): Ir ſchitai taſkat jo dukte Rahel ateit su avimis. ‘and behold! his daughter rahel is arriving with her sheep.’ [15] jonas bretkūnas, Postilla (bP i 303 7 [1591]): Todelei mielas Krikſchcʒonie tu taſkat eſſi krikſchtitas. ‘therefore, o dear christ! behold! you were baptized.’ With tàskategi: [16] Wolfenbüttel Postilla (WP 117r 10 [1573]): Taſkategi panas Die=‖was tawe paaukßtinaija. ‘behold! the Lord god has raised you.’ With tàsgatės: [17] dialect of tryškiai (LkŽ XV 961): Tàsgatės kare buvo, ne juokai! ‘behold! it was in war, is not a joke.’ ALL_62_63_maketas.indd 15 2011.06.17 08:23:46 16 danieL Petit acta Linguistica Lithuanica LXII–LXIII 2. MorPhoLogy from a morphological point of view, it is obvious that we are dealing here with two different formations. as far as i can see, forms with a voiceless consonant (-ka-) are exclusively old Lithuanian: they occur in the Wolfenbüttel Postilla (1573) and in the works of jonas bretkūnas (ca 1590), daniel klein (1653, 1654) or kristupas sapūnas (1673). they are still mentioned in some 19th century dictionaries, where they were probably taken from older sources: we have for example taskat, Adv. siehe da in the dictionaries of georg nesselmann (1851: 95) and friedrich kurschat (1883: 450). on the other hand, forms with a voiced consonant (-ga-) occurred only more recently. to my knowledge, their oldest attestations go back to the writings of jurgis Pabrėža (1771–1849) and Motiejus Valančius (1801–1875). there is thus a chronological difference between both types of endings: forms with -kaare older than forms with -ga- . this first impression is corroborated by the dialectological distribution. in old Lithuanian, forms with -kaare best attested in the Wolfenbüttel Postilla (WP 1573), but occurrences are found in various texts based on other dialects as well, e.g. Martynas Mažvydas, bretkūnas or Mikalojus daukša. it is thus likely that the ending -karepresents a common Lithuanian inheritance. the picture is radically different with the younger forms with -ga-, since they are restricted to a little number of Low Lithuanian (Žemaitian) dialects: they are predominantly attested in the north of a line drawn between klaipėda and kuršėnai (e.g. in kretinga, skuodas, telšiai, rietavas, Luokė, Papilė). from older literature one may mention the writings of Pabrėža and Valančius, both born precisely in this area (skuodas, resp. kretinga). it is clear that the voiced ending (-ga-) is a local innovation of the northern Low Lithuanian dialects. however, it can hardly be considered a regular evolution of an older -ka- . 3 . etyMoLogy since the 19th century, the etymology of these particles has been much discussed. in a brief note, the Latvian master jānis endzelīns (di ii 499 [1913]) explained the ending -skat in añskat, šskat, tàskat as cognate with the stem of the Latvian verb skatît ‘to see, to look’. this explanation was also adopted by eduard hermann (1926: 387). another view was proposed by adalbert bezzenberger (1877: 175–176), who saw in añskat, šìskat and tàskat compound forms of the demonstrative pronouns añs (anàs), šìs, resp. tàs in combination with different ALL_62_63_maketas.indd 16 2011.06.17 08:23:46 senosios lietuvių kalbos añskat, šskat, tàskat ir su jais susiję žodžiai 17straipsniai / articles deictic particles. he derived the first particle -kafrom the Pie coordination *ke ‘and’ (cf. gr. τε, oind. ca) and compared the second particle -t (from -te) with the ending of Lith. nèt ‘even’. in a slightly different way, ernst fraenkel (1921: 3, 33, cf. LeW i 11 [1962]) derived añskat, šskat and tàskat from the demonstrative pronouns añs (anàs), šs, resp. tàs in combination with a deictic particle -ka- . however, he did not trace back the ending -t to another deictic particle, but to a shortened form of the 2nd singular personal pronoun in the dative case (-ti ‘to you’ > -t). the same view was taken by Peter arumaa (1933: 35) and more recently by Wojciech smoczyński (2007: 17, 640, 661). none of these scholars seems to have supported any of these explanations by precise historical or typological evidence. it would perhaps not be unfair to apply to all of them the skeptical judgement pronounced by alvils augstkalns (1934–1935: 67) on fraenkel’s proposal: Diese Vermutung läßt sich aber, soweit ich sehe, auch durch nichts erklären . a thorough examination of all the implications involved by these etymologies is still to be conducted. Let us begin with endzelīns’ proposal. in support of his claim that the ending of añskat, šskat and tàskat goes back to a verbal form cognate to Latv. skatît ‘to see, to look’ one could mention the Modern french presentative particles voici, voilà. these particles were created in Middle french from the imperative of the verb voir ‘to see’ in combination with the local particles ci ‘here’, resp. là ‘there’. in contrast to fr. voici, voilà from vois ci, vois là (imperative + local particle), Lith. añskat, šskat, tàskat would display the opposite word order (local particles *anà, *š, *tà + imperative of *skatīti = Latv. skatît). from this point of view, the regular use of these particles with the nominative (as in ſchiſkat tarnaite weſʒpaties ‘here is the maid of the Lord’ WP 118 2–3 [1573]) would be secondary. a similar case would be the greek presentative particle ἰδού ‘ecce’, which goes back to the imperative form of the aorist ἰδεῖν ‘to see’: it was originally followed by noun phrases in the accusative, but in hellenistic greek it came to be used with noun phrases in the nominative (e.g. ἰδοὺ ὁ ἄνθρωπος ‘ecce homo’ jn 19, 5). however attractive it may appear at first glance, endzelīns’ etymology faces a serious difficulty: the segmentation proposed by endzelīns (*anà, *šì, *tà + *skat-) does not account for the forms with a voiced consonant (cf. tàsgat, etc.), nor for the forms without final dental (cf. šiska, tàsga, etc.). it is likely that the ending -t is not inherent to the structure, but is rather an adventitious element added to an already constituted stem, which suggests that, the connection with the verbal root of Latv. skatît must be seen as mistaken. i now turn to bezzenberger’s and fraenkel’s etymologies. Whichever solution one prefers, both have in common the idea that -ka-, resp. -gaare originally ALL_62_63_maketas.indd 17 2011.06.17 08:23:46 18 danieL Petit acta Linguistica Lithuanica LXII–LXIII deictic particles added to the demonstrative stems of añs (anàs), šs, resp. tàs. although rare, a deictic particle -kaoccurs in some Lithuanian dialects, instead of standard Lithuanian -ki-. a particle -gahas left some traces in old Lithuanian (e.g. nesanga ‘because, for’, betaiga ‘but’, cf. oPruss. anga ‘if, whether’) alongside the more frequent particle -gi-. the sibilant before these deictic particles must belong to the demonstrative stems. it should probably be interpreted as the masculine nominative singular ending (añs = anàs, šs, tàs), and indeed in many cases the context implies agreement with a masculine singular noun, see for instance the following examples: [18] Martynas Mažvydas, Gesmes Chrikſcʒoniſkas (MŽ 447 4 [1570]): Motriſchk taſkat sunus tawa! ‘Mother! there is your son!’ (agreement of taskat ‘this one there’ with sunus ‘son’). [19] simonas daukantas (LkŽ XiV 914): šisgatės piemuo! ‘here is the shepherd!’ (agreement of šisgatės ‘this one here’ with piemuo ‘shepherd’). however, counter-examples are relatively frequent, either with feminine gender nouns, as in the following examples: [20] Wolfenbüttel Postilla (WP 118 2–3 [1573]): ſchiſkat tar/naite weſʒpaties! ‘here is the maidservant of the Lord!’ (disagreement of šiskat ‘this one here’ with tarnaitė ‘maid servant’). [21] antanas juška, Литовскій словарь ([1897–1922], cf. LkŽ XV 961): Tàsga triba jo dirvos! ‘here is the edge of his field!’ (disagreement of tasga ‘this one there’ with triba ‘edge’). or with plural nouns: [22] jonas bretkūnas, Biblia (bb judg 15, 16 [1590]): Schiſkat anis gul pulkais. ‘behold, they are lying gathered together.’ or even without any substantive at all: [23] dialect of tryškiai (LkŽ XV 961): Tàsgatės kare buvo, ne juokai! ‘behold, it was in war, it was not a joke.’ ALL_62_63_maketas.indd 18 2011.06.17 08:23:46 senosios lietuvių kalbos añskat, šskat, tàskat ir su jais susiję žodžiai 19straipsniai / articles this might be the result of a secondary extension, once añskat, šskat, tàskat have been reanalyzed as uninflected particles. as a parallel, one could mention the hittite presentative particles kāša, kāšma ‘look here, behold’ and āšma ‘lo, behold’, which, according to a standard etymology, go back to the common gender nominative singular of demonstrative pronouns kāš ‘this one here’ (< Pie *ḱós, cf. Lith. šs < *ḱi-), resp. *āš ‘this one’ (< Pie *h1ós), in combination with a deictic particle *-(m)a (cf. hitt. imma ‘truly, really, indeed’, namma ‘then, once again, furthermore’), but may be used more widely as petrified adverbial formations. one may likewise assume that the first part of añskat, šskat and tàskat is a frozen masculine singular nominative ending. the forms were reanalyzed as uninflected adverbs or particles. it should be noted that forms without sibilant are attested in some Lithuanian dialects: šiga, šgatės (beside šsgatės), añgatės (beside añsgatės); they must be seen as renewed adverbial formations. the ending -t remains as the main problem. this is precisely the point where bezzenberger’s and fraenkel’s explanations diverge. i shall first discuss fraenkel’s etymology, since nowadays it turns out to predominate in the standard handbooks. his claim is that the ending -t reflects a second singular personal pronoun in the dative case -t (from -ti ‘to you, for you’). although not mentioned by fraenkel nor by any scholar having endorsed his opinion, this etymology could apparently be strengthened by a vast amount of typological evidence, showing a phraseological predilection of presentative particles for being linked with second singular or plural personal pronouns in the dative. in Latin, for example, the presentative particles ecce and em ‘here is / here are’ are frequently followed by the dative singular tibi ‘to you, for you’. examples with Lat. ecce tibi ‘here is / here are’: [24] cicero, Pro sestio 89: Ecce tibi consul, praetor, tribunus plebis noua noui generis edicta proponunt! ‘see now the consul, the pretor, the tribune of the people proposes new edicts of a new sort!’ [25] cicero, ad atticum 7, 19: Ecce tibi ante diem tertium Nonas Februarias mane accepi tuas litteras! ‘but here, this 3rd of february, i have received in the morning a letter from you!’ With Lat. em tibi ‘here is / here are’: [26] Plautus, asinaria 880: Em tibi hominem! ‘there is your man!’ [27] Plautus, amphitryon 778: Em tibi pateram, eccam! ‘there! your bowl, see!’ ALL_62_63_maketas.indd 19 2011.06.17 08:23:46