Abstract
This article is a continuation of the analysis of the Karaim -p edi- past tense presented, for the first time in scholarly literature, in Németh (2015). In the latter paper, this verbal category was described on the basis of a few South-Western Karaim examples, only, and was termed plusquamperfectum ii. In this paper the description of its semantic scope has been refined based on an analysis of recently discovered North- and South-Western Karaim examples as well as on a further evaluation of Turkic (including Eastern Karaim) data. Importantly, it is argued that the practice of (also) expressing habitual events by means of this verbal category (which is quite an unusual feature in the Turkic linguistic world) is a consequence of contact-linguistic factors, namely the influence of the Polish language, in which Western Karaims were (and still are) proficient. Finally, to obtain a complete picture of its evolution the -p edi- pluperfect is placed in the broader context of the Karaim past tense system. Following Németh (2015) and Németh (2019), this paper is the third in a series of articles introducing previously undocumented grammatical categories of Western Karaim.
1 Introductory Remarks
Western Karaim – a moribund Kipchak Turkic language that used to be spoken in two dialectal variants – still lacks a proper historical linguistic description. This is mainly because only a handful of Western Karaim sources date from before the 19th century and, in fact, most of them were only discovered in the last few years. Hence, for decades there was no reliable linguistic material on which to base any in-depth historical research.1 As a result, the analysis of newly discovered written records older than the 19th century provides us with an opportunity to supplement the current historical description of Karaim with new observations.2
In the first of a series of articles introducing hitherto undocumented Western Karaim grammatical categories, we presented seven examples of the SWKar. -p edi- past tense (Németh 2015). The category in question is constructed by attaching the -dy simple past forms of the auxiliary verb e- ‘to be’ to the -(y)p perfective converb form of the root verb. For example:
1. [ajt- ‘to say’ + -yp → ajtyp ‘having said’] + [edi ‘was’ ← e- ‘to be’ + di] →
2. ajtyp edi > ajtybedi (-p- > -b- voicing caused by the intervocalic position).
We concluded that this construction was mainly employed to express actions that had come to an end before another past action took place, the latter being usually expressed with the -dy simple past tense forms. Hence, we termed it plusquamperfectum ii – in contrast to the -ġan edi- plusquamperfectum i. However, we left open the question of its exact semantic scope and its relation to other Karaim past tenses owing to the scarcity of available examples and also because of the fact that two of the seven examples clearly referred to habitual actions performed in the past and therefore conveyed rather progressive than pluperfective meaning (Németh 2015: 222–233). Moreover, these two forms were documented out of context in sentences in which there was no mention of another past action that would have served as a chronological reference point (see examples 15 and 16 below).
The main goal of the present paper is to revise and refine the above description and answer the open questions. This is done by:
1. evaluating further examples of this grammatical category,
2. showing contact-linguistic factors that influenced its semantic scope, and
3. presenting it against the background of the Karaim past tense system (primarily by way of a comparison with the -ġan edi- pluperfect).3
The linguistic material adduced below was recently discovered in the North-Western, South-Western Karaim translations of the Pentateuch copied in the 18th–19th centuries. As a comparison, Eastern Karaim data of similar age will be presented, too. Given that we are dealing here with a grammatical category that was seldom used not only in Karaim but also in other Turkic languages, the Karaim material discussed in this paper is also modest in size.4 Nevertheless, one highly advantageous circumstance that facilitates our work is that the above translations of the Bible allow us to specify the context of each grammatical form very precisely. Just how beneficial this circumstance is has been already shown in Németh (2019).5
2 The Turkic Linguistic Background
A general overview of the use of this verbal category in other Turkic languages is provided in Németh (2015: 216–220). In the present analysis, in order to recapitulate the results of this comparative research in a concise way, we refer to the summarizing table presented in Németh (2015: 220), see table 1.
However, we ought to supplement the latter with some valuable additional information. Firstly, as we know from Grunin (1967: 375) and Pritsak (1959b: 84), the construction in question was also present in Armeno-Kipchak, which is an important fact given the high degree of language relatedness between Armeno-Kipchak and Karaim. Secondly, we can improve the description of the respective Uyghur verbal construction: as demonstrated by Rentzsch (2005: 115–118), in Modern Uyghur the -(i)vidi- (< -p edi-) forms express “a postterminal aspect (in other terms, a perfect) with the additional feature that the validity of the event is cancelled at the time of speaking” (Rentzsch 2019: 3).6
Johanson (1999: 178) describes the -p edi- forms as perfect-like (so-called postterminal) “pluconstatives”7 which, he writes, “may occur in so-called ‘progressive contexts’ without being ‘progressives’ themselves”. He provides the following examples to document his definition: Krch. kelip edi ‘had come’, Kzk. kelip edi ‘had come’, ǯuqlab edi ‘had fallen asleep, was asleep’, ölüb edi ‘had died, was dead’, and aryb edi ‘had got tired, was tired’. Based on the above, we supplemented Table 1 with the information that the Krch. -p edi- construction also played the role of a pluperfect (see also Johanson 1995: 95; in Johanson 2021: 721: remote past). The role of expressing resultative past (in Karachay) is mentioned by Seegmiller (1996: 23), too, a work not cited in our previous article.
Finally, it is worth mentioning that the -p edi- (> -bedi-) past form was likewise used in the Chagatay texts, see, above all, Eckmann (1966: 179), Blagova (1994: 344–345), and Schönig (1997: 221ff.). It was employed in narratives for events that take place prior to those expressed by the simple (-dy) past.
3 The Available Western Karaim Data
3.1 Introductory Remarks
In the interlinear morpheme-by-morpheme glossing below (edited according to the Leipzig Glossing Rules) the construction in question as well as the -ġan edi- pluperfect and the -r edi- imperfect are labelled analytically as a compound of the past tense forms of the auxiliary verb e- ‘to be’ and the respective non-finite forms of the main verb. Suffixes performing a derivative role as well as non-productive morphemes are not labelled.
3.2 North-Western Karaim Forms
This is the first time the -p edi- tense is documented for North-Western Karaim. The material is excerpted primarily from mss. ADub.iii.73 and TKow.01. ADub.iii.73 contains a translation of the Pentateuch into North-Western Karaim, and six of the examples presented below were found there in 2015–2018 after a careful editing of the entire Torah (folios 1 ro – 342 vo), see Németh (2021). As a next step, after reading the Torah in ms. JSul.iii.01, further South-Western Karaim examples were collected (see, 3.3 below) and the respective Biblical verses from ADub.iii.73 and JSul.iii.01 were compared with the content of further two North-Western Karaim sources, namely: TKow.01 and Mickiewicz and Rojecki (1889).
We can locate the linguistic material quite precisely in time and space. Manuscripts ADub.iii.73 and TKow.01 were copied by a person called Simcha ben Chananel (born in Trakai, died in Kukizów, in present-day Western Ukraine, on 27 March 1723; the hazzan (head priest) in Kukizów from ca. 1709). He worked on ms. ADub.iii.73 in spring 1720 (see 342 ro – 342 vo), whereas TKow.01 was finished on 7 December 1722.8 The translation of Genesis published by Mickiewicz & Rojecki (1889) originates from the 19th century.
(1) alybedim (Genesis 12:19)
b) Nek ajttyj tuvduġumdu ol da alybedim any özüḿa qatynlyqqa da haligine muna qatynyj alġyn da barġyn. (TKow.01: 19 ro)
(2) qalybedi (Genesis 18:11)
a) Da Avraham da Sara qartlar ediĺar jetkenĺar künlerǵa qartlyqta qalybedi bolma Saraġa jol qatynlarġa kibik. (ADub.iii.73: 24 ro)
b) Da Avraham da Sara qartlar ediler jetkenler künlerge qartlyqta qalybedi bolma Saraġa jol qatynlarġa kibik. (TKow.01: 25 vo)
(3) beklebedi (Genesis 20:18)
a) Ki bekĺaḿa beklebedi Adonaj bar qursaq ašyra üvünd́a Avimeleḫnin Sara üčün qatyny üčün Avrahamnyn. (ADub.iii.73: 28 vo)
b) Ki bekleme bekĺabedi Adonaj bar qusaq ašyra üvünde Avimeleḫnin iši üčün Saranyn qatynynyn Avrahamnyn. (TKow.01: 30 vo)
c) Ki bekĺaḿa bekĺabedi H bar qursaq ašyra üv́uńun Avimeleḫnin Sara qatyny ü ´čuń Avrahamnyn. (Mickiewicz & Rojecki 1889: 24)
(4) barybedi (Genesis 24:62)
Da Jicḥaq keldi kelmekt́an ki barybedi Beʾer Laḥaj Roʾiǵa da ol oltururedi jerind́a ol tüšlüknün. (ADub.iii.73: 36 vo)
(5) atabedi (Genesis 26:18)
a) Da qajtty Jicḥaq da qazdy ošol qujularyn ol suvlarnyn ki qazdylar vaḥtlarynda Avrahamnyn atasynyn da japtylar alarny Pelištim ölüp sortun Avraham da atady alarġa atlar kibik ki atabedi alarġa atasy anyn. (TKow.01: 41 ro)
(6) ajtybedi (Exodus 15:9)
a) Ajtybedi parʿo dušman quvajym jetejim ülešejim olǯa tolsun alardan klegim suvurajym qylyčymny tasetsin alarny qolum. (ADub.iii.73: 110 vo; TKow.01: 115 ro – 115 vo)
(7) belgilebedi (Exodus 21:8)
a) Eger jaman körünse közlerinde bijinin ki özüne belgilebedi any da juludu any jat ulusqa erklenmesin satma any tanġanynda anar. (TKow.01: 124 ro – 124 vo)
(8) ajtybedi (Deuteronomy 9:25)
a) Da tüšüp jalbardym alnynda Adonajnyn qyrq kün da qyrq ke ´ča ašyra ki tüšüp jalbardym ki ajtybedi Adonaj tasetḿa sizni. (ADub.iii.73: 297 vo)
b) Da tüštüm jalbarmaqbyla alnynda Adonajnyn ošol ol qyrq künnü da ošol ol qyrq kečeni ašyra ki jalbarmaqbyla tüštüm ki ajtybedi Adonaj tasetme sizni. (TKow.01: 309 ro)
3.3 South-Western Karaim Forms
The South-Western Karaim material presented in Németh (2015) can be supplemented in the present paper with further material found in manuscript JSul.iii.01, which also contains a translation of the Pentateuch. It was copied in 19th-century Halych by Jeshua Josef Mordkowicz (born in Halych in 1802, died in Halych on 23 July 1884), a copyist, translator, and hazzan. First, we will present the equivalents of examples 2, 3, 5–8 referred to above. In order to facilitate the semantic analysis of forms from different manuscripts used in the same Biblical context, the numbering of the examples below tallies with the numbering introduced in 3.2.
(2) qalybedi (Genesis 18:11)
c) Da Avraham da Sara qart ediler jetkenler künlerge qalybedi bolma Saraġa jolu qatynlarnyn. (JSul.iii.01: 17 ro)
(3) beklebedi (Genesis 20:18)
d) Ki bekleme beklebedi H bar qursaq asyra ivine Avimeleḫnin isi ücün Saranyn qatynynyn Avrahamnyn. (JSul.iii.01: 30 vo)
(5) atabedi (Genesis 26:18)
b) Da qajtty Jicḥaq da qazdy osol qujularyn ol suvlarnyn ki qazdylar vaḥtlarynda Avrahamnyn atasynyn da japtylar alarny Pelištim elip sortun Avraham da atady alarġa atlar kibik ki atabedi alarġa atasy anyn. (JSul.iii.01: 27 ro)
(6) ajtybedi (Exodus 15:9)
b) Ajtybedi parʿo dusman quvajym jetejim Jiseraʾelni ilesejim olǯa tolsun alardan klegim suvurajym qylycymny tasetsin alarny ḥali qolumnun. (JSul.iii.01: 74 ro – 74 vo)
(7) belgilebedi (Exodus 21:8)
b) Eger jaman körindi ese ol közlerinde bijinin ki ezine belgilebedi any da juludu any jat ulusqa erklenmesin satma any tanġanynda anar. (JSul.iii.01: 80 vo)
(8) ajtybedi (Deuteronomy 9:25)
c) Da tisip jalbardym alnynda H-nyn osol qyrq ol künde da osol qyrq ol kecede ki tisip jalbardym bosatlyq qoladoġac anyn ücün ki ajtybedi H tasetme sizni. (JSul.iii.01: 202 ro)
Besides the six examples presented above, in JSul.iii.01 we can find further instances of the use of -p edi- forms:
(9) qojubedi (Genesis 28:18)
Da künlej turdu Jaʿaqov erten bylada da aldy osol ol tasny ki qojubedi baslary tibine da qojdu any maceva da qujdu jav basy istine anyn. (JSul.iii.01: 30 ro)
(10) kelibedi (Exodus 15:19)
Ki kelibedi aty parʿonun markavy byla da jalan atlylary byla tengizge da qajtardy H alar istine osol suvlaryn ol tengiznin da uvullary Jisra ʾelnin bardylar quru byla ortasynda ol tengiznin. (JSul.iii.01: 74 vo)
(11) alybedi (Numbers 12:1)
Da cajnav sözledi Mirjam da Aharon Moše ücün isleri ücün ol kušlu qatynnyn ki aldy ki kušlu qatyn alybedi. (JSul.iii.01: 157 ro)
(12) qaḥirlenibedi (Deuteronomy 9:19)
Anlyq byla ki qorqtum alnyndan ol acuvnun da ol qaḥirnin ki qaḥirlenibedi H siznin istine tas etme sizni da qabul etti H tefilemni menim daġyn olda keretni. (JSul.iii.01: 199 vo)
(13) acuvlanybedi (Deuteronomy 9:20)
Da Aharonda acuvlanybedi H astry tasetme any da tefile ettim daġyn Aharon icin ol vaḥtta ol. (JSul.iii.01: 199 vo)
(14) alybedi (Deuteronomy 24:3)
Da ḥorlasa any ol er ol sondraġy da jazsa anar get bitigi da berse qoluna anyn da sirse any ivinden jemese ki else ol er ol sondraġy ki alybedi any ezine qatynlyqqa. (JSul.iii.01: 213 vo)
4 Evolution of the Middle Western Karaim -p edi- pluperfect – the Relation of Internal and Contact-Induced Phenomena
4.1 Semantic Analysis of the Western Karaim -p edi- Forms
At the outset, let us present a concise explanation of the Biblical context in which the respective forms appear. The English translation of the analysed word forms is highlighted with bold letters.
(1) Gen 12:10–20. Abraham went to Egypt and foresaw that the beauty of his wife Sarai (i.e., Sarah) would rouse the desire of the Egyptians, who might then kill him. To save his life, he begged his wife to pass herself off as his sister. Pharaoh took her into his harem, but he was smitten by God with sickness and (hence) learned the truth. Then Pharaoh asked Abraham why he told him that Sarai was his sister. Pharaoh said that it was because Abraham had not told the truth that he had taken her for wife as a result of which God smote him with great plagues. Pharaoh ordered Abraham to take his wife and go.
(2) Gen 18:1–12. Abraham’s guests (three men) conveyed the Lord’s promise that Sarah would have a son. The reader is informed by the narrator that Abraham and Sarah were old at the time the visitors came and that conception could not take place because Sarah had ceased to share her bed with Abraham.
(3) Gen 20:1–17. Abraham made the mental reservation that Sarah was not his wife but rather was his sister and hence she was taken to the harem of Abimelech, King of Gerar. Abimelech learned the truth in a dream through divine communication and gave Sarah back to Abraham. Then Abraham prayed to God to heal Abimelech and his household, because previously – in order to preserve Sarah’s honour – God had plagued them with illness preventing Sarah from giving birth.
(4) Gen 24:1–63. Abraham sent his servant to bring to Isaac, his son, a wife from his own homeland. The servant travelled to the place called Aram Naharaim, found Rebekah, and took her to Abraham’s abode. Isaac met her after he returned from Beer-Lahai-Roi, where he had gone before.
(5) Gen 21:25–30, Gen 26:1–22. Like his father Abraham, famine forced Isaac to migrate and he went to Gerar, whose king was Abimelech, King of the Philistines. Isaac’s flocks and herds multiplied, and the Philistines envied him. Abimelech’s men seized Abraham’s well (see Gen 21:25–30, Gen 26:15) and the king called on Isaac to leave. Isaac’s servants once more dug the wells which the servants of Abraham had dug and gave them the same names Abraham, his father, had given them.
(6) Exo 15:1–10. After having crossed the Red Sea and after having destroyed Pharaoh and his army, Moses and his people sang The Song of Triumph (Exo 15:1–18). In this song the events of Exo 14 are eulogized. First, in Exo 15:1 and Exo 15:4–5, Pharaoh’s defeat and death is described. Then, in Exo 15:9, the listener is informed of what Pharaoh had said before his defeat when he decided to pursue Moses and his people. Cf. example 9 below. In Exo 15:10 the listener is told that Pharaoh was then covered by the sea.
(7) Exo 21:6–11. The initial part of Exo 21 sets out the laws for female servants. In Exo 21:8, it is stated that if a female servant does not suit the man who had allotted her for himself as his concubine and redeemed her, her relatives might redeem her or she might be sold to another Israelite.
(8) Deu 5:1, Num 13:16–33, Num 14:1–5, Deu 9:22–28. Moses called his people (Deu 5:1) and repeated to them all the laws and judgements of God. He recalled all the previous occasions when the children of Israel had not obeyed God’s commandments. Moses reminded them that he prayed to God because previously God had said that he would destroy the children of Israel because of the events that had taken place at Kadesh-Barnea (described in Num 13:16–33, Num 14:1–5).
(9) Gen 28:1–18. Isaac sent Jacob off to Paddan-Aram to marry into his mother’s family (Gen 28:1–2). On his way there Jacob chanced on a place and lay there for the night and took a stone for his pillow. In his dream he saw a vision (Gen 28:12–17). When he woke up, he recognized that the place he had stayed overnight was God’s house and set the stone, the stone he had taken for a pillow, as a pillar and anointed it with oil.
(10) Exo 14:23–29, Exo 15:1–19. After having crossed the Red Sea and after having destroyed Pharaoh and his army, Moses and his people sang The Song of Triumph (Exo 15:1–18). In this song the events of Exo 14 are eulogized. Exo 15:19 is a link verse in which the narrator recapitulates the events described in Exo 14:23–29, namely that after Pharaoh and his hosts had entered the Red Sea, God brought back the waters and made them drown.
(11) Num 12:1–2: Num 12:1–2: In Num 12:1, the Biblical narrative informs the reader that Miriam (Moses’ sister) and Aaron (Moses’ brother) complained that Moses had taken a Cushite woman for a wife.
(12) Exo 34:1–28, Deu 5:1, Deu 9:12–19. Moses called his people (Deu 5:1) and repeated to them all the laws and judgements of God. He recalled the moment he went up to Mount Sinai (see Exo 34:1–28) for the second time to receive the second set of the Tablets of the Law. He beseeched Good to have mercy on the children of Israel, because God had become furious because of their sinning against him.
(13) Exo 32:1–4, Exo 32:9–13, Deu 5:1, Deu 9:12–21. Moses called his people (Deu 5:1) and repeated all the laws and judgements of God. He recalled the moment when, in the time he went up into Mount Sinai to receive from God the Ten Commandments, the Israelites cast a golden calf as an idol for their worship under the leadership of Aaron (Exo 32:1–4). Moses beseeched and pleaded with God that they be spared (Exo 32:11–13) when he learned that God had become angry with Aaron and had decided to destroy Aaron and the people of Israel (Exo 32:9–10).
(14) Deu 24:1–4. The initial part of Deu 24 describes the right of divorce. In particular, it explains that a divorced woman whom another man had taken for a wife and whose marriage to that man had later also ended through divorce or through his death, cannot then marry her first husband again because she is in disgrace.
The main function the adduced examples have in common is that they all refer to an action that took place earlier than another past action.11 In addition, all the -p edi- forms listed above refer to completed actions12 and they are primarily employed in narrative contexts. Used in conjunction with the simple past forms, the construction’s role is to shift the time-frame of the narrative backwards in time (to be then followed, again, by simple past forms operating in the anterior time-frame). In other words, the -dy simple past is guided by chronology, whereas the -p edi- tense refers to relative chronology.
The above philological data suggest that that the semantic scope of the WKar. -p edi- past tense should not be defined on the basis of so-called focality – a notion that Johanson (1999: 176, 2000: 110, 121) considers to be key to describing the -p edi- past in the Kipchak Turkic languages. In Johanson’s view, the semantic range of this construction is “restricted to what is still relevant of the event” from the respective point of view (i.e., it is a “focal” category). To exemplify that, he quotes the semantic opposition of Krch. “ketib edi ‘had left (and was still gone)’ or ‘was gone’ vs. ketgen edi ‘had gone’ […]” (Johanson 2000: 120–121), in which ketib edi is defined as a focal (“+foc”) form, whereas ketgen edi is supposed to be its nonfocal (“–foc”) counterpart.13 As we see, however, this mental category has little relevance to the Karaim -p edi- forms: some of them behave, so to speak, in compliance with the above juxtaposition (such as, for instance the use of qalybedi ‘had ceased’ in example 2); some of them, in turn, convey exactly the opposite meaning, see barybedi ‘had gone’ meaning, after adding the context to it, ‘had gone and returned some time ago’ in example 4. Csató (2000: 746) states that in Karaim “focality oppositions which are so central in other Turkic languages are lost both in non-anteriority and anteriority”, which, in her view, “may be the result of a process of simplification, which is partly due to the fact that Karaim has ceased to be used as a full-fledged vernacular.” Since the latter observation was based on modern data, it is worth stressing that the data analysed here stems from a time period when Karaim was still a fully-fledged vernacular.
A comparison of the verses analysed above with their equivalents in the other available Western (and Eastern) Karaim translations of the Torah allowed us to make the following observation: In the vast majority of cases in which no -p edi- forms were used in the respective verses it is the (purely temporal) -dy simple past tense that was applied by the translators (see tables 2, 4). This demonstrates the semantic similarity between these two tenses, while bearing in mind the reservations noted above.
However, none of the forms listed above express habitual events, which is especially noteworthy in light of the two South-Western Karaim verb forms presented in Németh (2015: 222; see examples 15, 16 below) that are used for repeated and habitual events, i.e. for events usually expressed by the WKar. -r edi- imperfect tense forms. This seems to be an anomaly in light of the fact that such use of the -p edi- past is alien to other Turkic languages. In Németh (2015) we refrained from giving a definite answer to the question of how these two forms relate semantically to the other five examples in which the role of the -p edi- forms was exactly the same as it is in sentences 1–14 above. However, having now much more reliable data at our disposal, let us take a closer look at these two forms once more and propose a contact-linguistic explanation of this phenomenon.
Firstly, let us remind the reader that examples 15 and 16 were found in a notebook of Józef Sulimowicz, in which he gathered Karaim linguistic material from other native speakers and supplemented it with his own observations and explanations.
(15) barybedim
Buvalo barybedim mamaba saharha. (Sulimowicz 1969)
(16) ystyrynybediler
Kacanes, davno Karajlar ystyrynybediler zerette. (Sulimowicz 1969)
In both sentences the informant, more precisely Rachela Eszwowicz (born 1869, died 1946), refers to actions that took place in the distant past. In sentence no. 16 this is explicitly indicated by the use of the adverb davno ‘long ago’, but in the case of example no. 15 we must place the content of this utterance in the context of the following genealogical facts: Rachela’s mother, Salomea Eszwowicz was born in 1837 and died in 1903 (Anna Sulimowicz, personal communication on 02 Jan 2019), whereas Rachela was the only one of Salomea’s three children who lived to be an adult. Rachela was the half-sister of Józef Sulimowicz’s mother and she was one of the oldest members of the Karaim community of Halych at the time when, in 1933, Józef Sulimowicz enrolled at the University of Warsaw to study Turkology (see Sulimowicz, A. 2013: 5). This means that it was in the early 1930s, i.e., ca. 30 years after the death of Rachela’s mother, when Józef Sulimowicz had the opportunity to record Rachela’s linguistic material (later, in 1969, this material was copied in Józef Sulimowicz’s notebook). Thus, in example 15 the respective event is recalled after the passing of at least 30 years.
A reference to the distant past is, however, certainly not an intrinsic semantic component of the Tkc. -p edi- tense. In the Turkic languages this feature is rather implied by the context and is a semantic concomitant of its pluperfective role. In Western Karaim the appearance of this role was, in our view, a result of the influence of the Polish verbal system: although Polish pluperfect was typically used to express a past action that preceded another past action, less dominantly it also denoted actions that took place in the distant past (Klemensiewicz & Lehr-Spławiński & Urbańczyk 1955: 373; Klemensiewicz 2002: 620), regardless of its perfective or imperfective meaning.17 This mirrors the semantic scope of the above South-Western Karaim forms. Our assumption is all the more valid as the Polish language unquestionably had an impact on Rachela Eszwowicz’s idiolect – just as it had a strong influence on South-Western Karaim in general, see, e.g. Németh (2010; 2011a: 62–76; 2022 (forthcoming)).
The latter data might also suggest that – at least, to be on the safe side, in Rachela Eszwowicz’s idiolect – the main semantic component of the Karaim -p edi- past was intended to convey the anterior character of an action in relation to another past action, regardless of the fact of whether the action was habitual or non-habitual in its nature. Note, that even though both sentences are quoted out of context, in each case we can safely say that the relevant actions that “had been performed in distant past” are logically contrasted with the state of their not being performed both in the recent past and at the moment of speaking.
4.2 The Relation of WKar. -p edi- to the -ġan edi- Pluperfect
It is already Johanson (1999: 178–179) who wrote that the “functional difference” between the -ġan edi- and -p edi- forms in the Kipchak languages is “subtle” (he refrained from discussing it in the cited paper). But what makes a proper comparison of the two Western Karaim pluperfects even more difficult is that we know of only a few 18th- and early 19th-century forms that could formally be interpreted as the -ġan edi- pluperfect.
The fact is that forms like aryġan edi (see sentence 17 below) can morphologically be interpreted in two ways: on the one hand, as the 3rd person -ġan edi- pluperfect form of ary- ‘to become tired’ or, on the other, as the word aryġan ‘tired’, formally a -ġan participle of the above-mentioned verb (lemmatized in KarRPS 77), used with the simple past form of the copular verb e- ‘to be’. We must be aware of this syncretism during our analysis and we believe that some of the 18th–19th century -ġan edi- forms we found actually functioned as predicative nominals rather than as verbal predicates.18
But, first, let us take a look at this linguistic material in its entirety:
(17) aryġan edi (Gen 25:29)
a) Da biširdi Jaʿaqov aš da keldi ʿEsav ol tüzd́an da ol edi aryġan. (ADub.iii.73: 38 ro; TKow.01: 40 ro)
b) Da bisirdi Jaʿaqov as da keldi ʿEsav ol tizden da ol aryġan edi. (JSul.iii.01: 26 vo)
(18) tutqunlanġan ediĺar (Gen 39:20)
a) Da aldy biji Josefnin any da berdi any ol zyndan üvǵa orunġa ki tutqunlary ol bijnin tutqunlanġan ediĺar da edi anda ol zyndan üvd́a. (ADub.iii.73: 65 ro; TKow.01: 67 ro)
b) Da aldy biji Josefnin any da berdi any ol zyndan ivge orunġa ki tutqunlanġanlary ol meleḫnin tutqunlanġan ediler da edi anda ol zyndan ivde. (JSul.iii.01: 44 ro)
The word form tutqunlanġan ediĺar ‘were imprisoned’ is repeated two more times in exactly the same textual and grammatical context in Gen 40:3 (in the singular form tutqunlanġan edi) and Gen 40:5 – in mss. ADub.iii.73 (65 vo), TKow.01 (67 vo), and JSul.iii.01 (44 ro, 44 vo). We refrain from repeating such data here.
(19) quǯurajmaġan edi (Exo 12:39)
Da biširdiĺar ošol ol ḥamurnu ki čyġardylar Micrid́an jajmalar macalar ki quǯurajmaġan edi ki sürüldüler Micriden da bolalmadylar kečikme da daġyn azyqta qylmadylar özlerińa. (ADub.iii.73: 106 ro–106 vo)
(20) titinlengen edi (Exo 19:18)
Da tavy Sinajnyn titinlengen edi barysy anyn ücün ki endi anyn istine šeḫinasy H-nyn ot byla da kötirildi titini anyn titini kibik ol kirec pecnin da qaltrady bar ol tav astry. (JSul.iii.01: 79 ro)
The question of which of the above -ġan edi- forms are finite forms and which are to be treated as nominal predications poses something of a challenge. Not only the context, but also the equivalents of the -ġan edi- forms in the standard Hebrew Bible, as well as the nominal predicative forms serving as their counterparts in some of the North-Western and Eastern Karaim translations of the Bible (see Table 3) point rather to nominal predication and the absence of a pluperfective meaning in the case of examples 17 and 18 (as well as in verses Gen 40:3 and 40:5, in which the form attested in example 18 is repeated). With respect to the form tutqunlanġan ediĺar in example 18 above, we should also mention that the alternative nominal predicative form would be tutqunlan-ġan-lar e-di-ĺar ‘were imprisoned’ (= be.imprisoned-ptcp-pl + be-pst-3pl), with the -lar plural also attached to the participial form, which would then evidently be a copula form. But the use of the additional plural suffix is not obligatory in this case and thus its absence in other forms is not decisive. Secondly, it is worth emphasizing – even though it should not be treated as a decisive argument, either – that in the Eastern Karaim ms. bsms 288 as well as in the North Western Karaim translation of Genesis published by Mickiewicz & Rojecki (1889) in verses Gen 25:29, Gen 39:20, Gen 40:3, and Gen 40:5 the present tense copula suffix forms are used to express the meaning that the -ġan edi- past tense form was meant to convey in ADub.iii.73. These forms, which are also presented in Table 3, are as follows: EKar. ary-ġan-Ø ‘is tired’ (= become.tired-ptcp-3sg.cop; bsms 288: 29 ro), tutqun-durlar ‘are prisoners’ (= prisoner-3pl.cop; bsms 288: 44 vo), tutqun-dur ‘is a prisoner’ (= prisoner-3sg.cop; bsms 288: 44 vo), tutqun-dyrlar ‘are prisoners’ (= prisoner-3pl.cop; bsms 288: 44 vo), see Jankowski et al (2019: 39, 66, 67, respectively), and Mod.NWKar. ary-ġan-Ø ‘is tired’ (= become.tired-ptcp-3sg.cop; Mickiewicz & Rojecki 1889: 31), tutḥunlan-ġan-Ø (= be.imprisoned-ptcp-3sg.cop; Mickiewicz & Rojecki 1889: 51), and tutḥunlan-ġan-lar (= be.imprisoned-ptcp-3pl.cop; Mickiewicz & Rojecki 1889: 51).
The situation is somewhat different in case of the forms used in sentences 19 and 20. After consulting the context, the structure of the original Hebrew text, and the Karaim equivalents used in other manuscripts they can still be interpreted in two ways: as nominal predicates and as pluperfective verbal forms. However, one interesting observation in this context is that the -ġan edi- past form quǯurajmaġan edi (28), which features in manuscript ADub.iii.73, was replaced by the same copyist and author (i.e. by Simcha ben Chananel), with the nominal predicative construction tüvül-Ø quǯur-Ø ‘is not leavened’ (= is.not.3sg.cop + leavened; TKow.01: 110 vo) in manuscript TKow.01, while in Eastern Karaim bsms 288 it is the -p edi- pluperfect (see example 23) that is used in the very same context. Hence, the use of the negative suffix -ma- in ADub.iii.73 (instead of the negative particle tüvül) suggests the need to treat example 29 as a -ġan edi- pluperfect.
Thus, if we compare the limited available material with the role which the -ġan edi- past plays in the Turkic linguistic world – see, above all else, Juldašev’s (1965: 167–184) analysis for an exhaustive description (without, however, any Karaim data being adduced) – it seems valid to speculate that the -ġan edi- pluperfect was employed to express a single, completed event, or a state which resulted from a completed action, which took place prior to another action in the past.
In other words, the -p edi- pluperfect apparently functioned as a narrative past, whereas the -ġan edi- pluperfect was primarily marked with a resultative and perfective shade of meaning. For this reason, depending on which semantic component of the -p edi- past was foregrounded by the speakers, it was either the -dy simple past or the -ġan edi- past or even the -r edi- imperfect (used to denote habitual, progressive actions in the past – in contrast to the -ġan edi- past, which was restricted to single and completed events) that gradually replaced it in the finite verbal system.
∵
Finally, it is ought to mention that the use of the pluperfect constructions in the Western Karaim translations of the Torah bears no relation to the Hebrew forms used in the original. Firstly, as a grammatical category the pluperfect is alien to Biblical Hebrew. Secondly, the Hebrew equivalents of the Karaim forms adduced above can appear both with and without a waw-consecutive, see, e.g. וָאֶקַּח wåʾeqqaḥ ‘so I might have taken’ (Gen 12:19) in example 1 andחָדַל לִהְיֹות and ḥåḏal lihyōṯ ‘had passed being’ (Gen 18:11) in sample 2.
4.3 Comparative Eastern Karaim Data
We should begin by pointing out that we found no examples of the EKar. -ġan edi- pluperfect (see its description by Prik 1976: 134–135) in the translation of the Torah edited by Jankowski et al. (2019). But one viewpoint that tallies with our above observations was that recently forwarded by Jankowski (2019: xx) with reference to EKar. -p edi- forms documented in this manuscript. In his view, this category “expresses action in the past related to another subsequent action or point of view, similarly to the past perfect.” As far as the verses analysed above are concerned, in ms. bsms 288 the discussed verbal construction is used in examples 1–3, see:19
(1) Nučun ajttyŋ qyz qardašymdyr ol da alyp edim any özümä ḫatqnlyqqa da endi muna ḫatynyŋ al da ket. ‘Why did you say, ‘She is my sister,’ so that I had taken her for my wife? Now then, here is your wife. Take her and go.’ (Gen 12:19)
(2) Da Avraham da Sara qartlar jetkänlär künlärgä tyjylyp edi bolma Saraġa töräsinčä ḫatynlarnyŋ. ‘And Abraham and Sarah were old, advanced in days. It had ceased to be with Sarah according to the custom of women.’ (Gen 18:11)
(3) Ki tyjma tyjyp edi H [ačmaġïn] barča qursaqnyŋ [evinä Avimeleḫniŋ] säbäbi učun Saranyŋ ḫatyny Avraha{m}nyŋ. ‘For the Lord had surely closed all the [apertures of the] wombs of [the house of Abimelech] because of Sarah, Abraham’s wife.’ (Gen 20:18)
The Pentateuch copied in that particular manuscript contains a further ten examples of the -p edi- construction:
(21) […] Da barča terägi ol jerniŋ bolmastan burun jerdä da barča otu ol tüzniŋ bitmästän burun ki jaġdyrmajyp edi H Täŋri ol yer üstünä da adam joq edi išlämä ol jerni, […] ‘[…] And when no bush of the field was yet in the land and no grass of the field had yet grown, because the Lord God had not caused it to rain on the land, and there was no man to work the ground, […]’ (Gen 2:5)
(22) Da Avimeleḫ juvyqlašmajyp edi aŋar da ajtty. ‘But Abimelech had not approached her, and he said:’ (Gen 20:4)
(23) Da edi nečik ki tavustu Jičḥaq alġyšlama šol Jaʿaqovny da edi tek čyqma čyġyp edi Jaʿaqov qatyndan jüzläriniŋ Jičḥaq atasïnïŋ da ʿEsav qardašy keldi avlamaġyndan. ‘As soon as Isaac had finished blessing Jacob, when Jacob had scarcely gone out from the presence of Isaac his father, Esau his brother came in from his hunting.’ (Gen 27:30)
(24) Birgäsinä kimniŋ ki tapsaŋ šol Täŋriŋni tirilmäsin qaršysyna qardašlarymyznyŋ tanyġyn özünä nedir birgämizgä da alġyn özüŋä da bilmädi Jaʿaqov ki Raḥel oġurlap edi alarny. ‘Anyone with whom you find your gods will not live. In the presence of our brothers point out what I have that is yours, and take it.’ But Jacob did not know that Rachel had stolen them.’ (Gen 31:32)
(25) Atam ant berip edi maŋa muna men ölärmin zerätimdä ki qazdym özümä jerindä Kenaʿannyŋ anda kömgin meni da endi čyġajym endi da kömäjim šol atamny da qajtajym. ‘My father had made me swear, “Behold, I die. In my grave which I have dug for myself in the land of Canaan: bury me there.” And now let I go up and bury my father, and let me return.’’ (Gen 50:5)
(26) Da symarladylar Josef učun demä ataŋ symarlap edi ölmäsindän burun demä. ‘And they ordered [a messenger] to Joseph, saying, “Your father had commanded before dying, saying:”’ (Gen 50:16)
(27) Da piširdilär šol ḥamurny ki čyġardylar Mysyrdan kömäčlär mačalar ki äččimäjip edi ki sürüldilär Mysyrdan da bolaj almadylar kečikmä da daġyn azyq qylmadylar özlärinä. ‘And they baked unleavened cakes of the dough which they brought out of Egypt, for it had not leavened; because they were driven away out of Egypt, and could not tarry, neither had they prepared for themselves any provisions.’ (Exo 12:39)
(28) Da aldy Moše šol süjäklärin Josefniŋ birgäsinä ki ant bermä ant berip edi oġlanlaryna Jisraʾelniŋ demä saġynma saġynsa Täŋri sizni da čyġarġajsyz šol süjäklärimni mundan birgäŋizgä. ‘And Moses took the bones of Joseph with him, for he had solemnly sworn the sons of Israel, saying, “God will surely remember you, and you shall carry up my bones with you from here.”’ (Exo 13:19)
(29) Da Moše jüz jigirmi yašar kiši edi ölgänindä sönükmäjip edi közü da solumajyp edi qyzyly janaġynyŋ. ‘Moses was a hundred and twenty years old when he died; his eye had not dimmed and the red colour of his cheeks had not faded.’ (Deu 34:7)20
We found no Eastern Karaim examples expressing habitual events with EKar. -p edi- which, to a certain extent, tallies with our view regarding the contact-induced nature of the Western Karaim anomaly, given that Polish exerted no influence on Eastern Karaim.
5 Final Thoughts on Middle and Modern Western Karaim Pluperfects
As the linguistic material shows, both categories were already only moderately productive in the 18th century and became very rarely used in the 19th century. This chronology appears to follow precisely the same time frame as that marking the disappearance of the Polish plusquamperfectum: it was rendered obsolete in the 19th century and the early 20th century, but the process of its disappearance was already underway in the 18th century (Andersen 1987: 27; Klemensiewicz 2002: 620). The strong structural influence of Polish on Western Karaim, which also influenced the finite verb forms of the latter, is a fact (see it discussed, e.g., in Csató 1996, 2000; Németh 2010, 2011a: 62–76). Hence, it seems likely that both the weakening of the original system of Turkic pluperfects in Karaim and the anomalous evolution of the semantic scope of SWKar. -bedi- was due to the influence by the Polish adstratum. Seen in this light, quite telling is the fact that Mykolas Firkovičius, who was a native speaker of Karaim and Polish, with an excellent command of his native tongues (born 1924, died 2000), also left the -ġan edi- pluperfect undiscussed in his grammatical description (Firkovičius 1996). This suggests that Firkovičius perceived this verbal category as non-productive (nor did he discuss the -p edi- past in his grammar; apparently for the same reason). The only -ġan edi- verbal form we managed to find in this work is, in fact, described as a -ġan participle, see:
(30) barhan edim
Ḿeń šaharha barhan edim, anyn ü ´čuń siźǵa ḱečiktim. (Firkovičius 1996: 109)
In his grammar, Musaev (1964: 277) claims that the -ġan edi- past is rarely used, but the examples he adduces must be taken with a grain of salt: we managed to localize the original attestation of one of them, namely Mod.NWKar. **śuvǵan edim t́uǵal ´čamni **‘I loved my beauty’ (Musaev 1964: 277), and it turned out that the actual fragment, documented by Kowalski (1929: 58), is śuv́ar edim t́uǵal ´čamni ‘I used to love my beauty’. It is the second time we found proof of Musaev misquoting data in his grammar (see Németh 2013: 135); in this case he changed an -r edi- (imperfect) form into a -ġan edi- (pluperfect) form.
Csató (2000: 737) provides two late-20th-century examples, which show that, to certain extent, the -ġan edi- past continued to function, at least in some idiolects, as a productive category until recent times. However, the latter might be also true with respect to the -p edi- pluperfect, given that in the 1960-ies Józef Sulimowicz’s interlocutors (and J. Sulimowicz himself) were aware of its existence (see Németh 2015: 225). Even more importantly, however, examples of the use of both pluperfect forms in the 20th century show that the appearance of these forms in 18th- and 19th-century Bible translations reflects their natural use in spoken Middle Western Karaim. The latter is also supported by the adduced Turkic comparative material.
Finally, one terminological remark is due. Since the -ġan past tends to be termed as preterite ii (see, e.g., Pritsak 1959a: 335), from purely terminological point of view it would be perhaps more appropriate to term the -ġan edi- past as plusquamperfectum ii, whereas it is the -p edi- past which should rather be termed plusquamperfectum i.
Acknowledgements
The research upon which this publication is based has been awarded funding from the European Research Council (erc) under the European Union’s Horizon 2020 research and innovation programme (grant agreement number 802645). The task of preparing the transcription and the English translation of manuscript ADub.iii.73 as well as the process of collecting the required linguistic data from it has been financed by the National Science Centre of Poland (Pol. Narodowe Centrum Nauki), research project agreement number umo-2015/17/b/hs2/01498.
Editorial Symbols
| […] | missing portion of text |
| [abc] | damaged and reconstructed fragment |
| {abc} | interlineal or marginal addition in a manuscript by another hand |
| abc > cba | borrowing, internal development |
| abc → cba | derivation |
| **abc | nonexistent, hypothetical, or erroneous form |
Abbreviations
abl = ablative | acc = accusative | adj = adjective | an = action nonimal | arch. = archaic. | Arm.-Kipch. = Armeno-Kipchak | art = article | BHeb. = Biblical Hebrew | cond = conditional | cop = copula | cvb = -p converb | dat = dative | Deu = Book of Deuteronomy | EKar. = Eastern Karaim | Eng. = English | ex. = example | Exo = Book of Exodus | fn. = footnote | fut = future | gen = genitive | Gen = Book of Genesis | Heb. = Hebrew | imp = imperative | indef = indefinite | instr = instrumental (case) | int = intensifying particle | Kirg. = Kirghiz | Kmk. = Kumyk | Kklp. = Karakalpak | Krch.(-Blk.) = Karachay(-Balkar) | Kzk. = Kazakh | lit. = literally | loc = locative | masc = masculine | Mod.NWKar. = Modern North-Western Karaim | ms. = manuscript | mss. = manuscripts | neg = negative suffix | Nog. = Nogai | nom = nominative | NWKar. = North-Western Karaim | opt = optative | pass = passive | perf = perfect | pers = personal ending | pl = plural | Pol. = Polish | poss = possessive | pperf = pluperfect | pred = predicative | pst = simple (Tkc. -dy) past | ptcl = particle | ptcp = -ġan participle | Russ. = Russian | sg = singular | SWKar. = South-Western Karaim | Tkc. = Turkic | Ukr. = Ukrainian | Uyg. = Uyghur | Uzb. = Uzbek | WKar. = Western Karaim
References
Primary sources
ADub.iii.73 = A translation of the Torah, the Book of Ruth, the Book of Jeremiah, Ecclesiastes, and the Book of Esther into North-Western Karaim. The Torah was copied between 25 Mar 1720 and 31 May 1720, the other books were copied after 31 May 1720 and before 27 Mar 1723, in Kukizów by Simcha ben Chananel (died 27 Mar 1723). Stored in Warsaw in the private archive of the late Aleksander Dubiński (1924–2002) under the accession number ADub.iii.73. 385 folios. Critically edited in Németh (2021).
bsms288 = A translation of the Tanakh (without 1–2 Chronicles) into Eastern Karaim. Copied in four volumes in the 18th century. Volumes 1 and 4 (i.e., Pentateuch, The Song of Songs, Ruth, Lamentations, Ecclesiastes, Esther, Psalms, Proverbs, Job, Daniel, Ezra and Nehemiah) were critically edited in Jankowski et al. (2019). Stored in the Cambridge University Library under the accession number bsms288. 620 folios.
Gaster Hebrew ms170 = An Eastern Karaim translation of a large portion of the Torah (Genesis 1:1 – Deuteronomy 32:51) and the Book of Lamentations (4:11–5:22). Copied in the beginning of the 19th century. Stored in the John Rylands Library of The University of Manchester Library. Described by Jankowski (1997), its full text compared with bsms 288 in Jankowski & Aqtay & Cegiołka & Çulha & Németh (2019). Full text available at: https://luna.manchester.ac.uk/luna/servlet/s/a0jk2m (accessed 4 Oct 2019). 265 folios.
JSul.iii.01 = A translation of the Torah and of some fragments of the books of Joshua, Judges, 1–2 Kings, Isaiah, Jeremiah, Ezekiel, Hosea, Joel, Obadiah, Micah, Habakkuk, and Zechariah (i.e., Torah and Haftarah) into South-Western Karaim. Copied by Jeshua Josef Mordkowicz (1802–1884) in Halych in the 2nd half of the 19th century. Stored in Warsaw in the private archive of the late Józef Sulimowicz (1913–1973). 298 folios.
Mickiewicz, Zacharjasz [= זכריה מיצכיוויץ] & Rojecki, Elijahu [= אליהו רואצכי] (transl.). 1889. ספר חמשה חומשי תורה מתורגם ללשון קדרי לבני מקרא קודש המורגלת בפי קהלות הקראים הדרים בגלילות רוסיא. Хамиша хумше тора, т. е., Пятикнижие Ветхаго Завѣта, переведенное на караимское нарѣчіе посредствомъ Захарія Михайлова Мицкевича и Илья Исаева Роецкаго. Vilnius.
Sulimowicz (1969) = A notebook of Józef Sulimowicz with notes in Polish on Karaim grammar and lexicography. Stored in the private archive of Józef Sulimowicz (1913–1973) in Warsaw.
TKow.01 = A translation of the Torah into North-Western Karaim. A partially vocalised sister manuscript of ADub.iii.73 mistakenly described by Kowalski (1929: 289) and Zajączkowski (1939: 94) as a manuscript created in 1723 in Derażne. Copied by Simcha ben Chananel, the copyist of ADub.iii.73; it was finished on 7 December 1722 A.D. Kept in Kraków in the private archive of the inheritors of the late Tadeusz Kowalski’s (1889–1948) private archive. 351 folios.
Secondary sources
Andersen, Henning. 1987. From auxiliary to desinence. In: Harris, Martin & Ramat, Paolo (eds.), Historical development of auxiliaries, 21–51. [= Trends in Linguistics. Studies and Monographs35]. Berlin, New York, Amsterdam: Mouton de Gruyter.
Aqtay, Gülayhan. 2009. Eliyahu ben Yosef Qilci’s anthology of Crimean Karaim and Turkish literature. Critical edition with introduction, indexes and facsimile. [= Yildiz Dil ve Edebiyat Dizisi 8]. Vol. 1–2. İstanbul: Eren Kitabevi.
Baskakov, Nikolaj Aleksandrovič & Zajączkowski, Ananiasz & Šapšal, Seraja Markovič (eds.). 1974. Karaimsko-russko-poĺskij slovaŕ. Słownik karaimsko-rosyjsko-polski. Moskva: Russkij jazyk.
Benson, Joseph. 1857. The Old Testament and New Testament of our Lord and Saviour Jesus Christ. (According to the present authorized version.) With critical, explanatory, and practical notes: The marginal readings of the most approved printed copies of the New Testament with such others as appear to be countenanced by the original Greek: A copious collection of parallel texts; summaries of each book and chapter; and the date of every transaction and event recorded in this part of the sacred oracles, agreeably to the calculations of the most correct chronologers. Vol. 1. New York: T. Carlton & J. Porter. [Available on-line, without page numbers indicated, at https://biblehub.com/commentaries].
Berta, Árpád. 1998. West Kipchak Languages. In: Johanson, Lars & Csató, Éva Ágnes (eds.), The Turkic Languages, 301–317. London, New York: Routledge.
Blagova, Galina Fedorovna. 1994. «Babur-Name». Jazyk, pragmatika teksta, stiĺ. K istorii čagatajskogo literaturnogo jazyka. Moskva: Vostočnaja literatura RAN.
Csató, Éva Ágnes. 1996. Some typological properties of North-Western Karaim in areal perspectives. In: Boretzky, Norbert & Enninger, Werner & Stolz, Thomas (eds.), Areale, Kontakte, Dialekte. Sprache und ihre Dynamik in mehrsprachigen Situationen, 68–83. [= Bochum-Essener Beiträge zur Sprachwandelforschung24]. Bochum: Brockmeyer Verlag.
Csató, Éva Ágnes. 1998. Über die finiten Verbformen des gesprochenen Nordwest-Karaimischen. In: Demir, Nurettin & Taube, Erika (eds.), Turkologie heute – Tradition und Perspektive, 47–53. [= Veröffentlichungen der Societas Uralo-Altaica48]. Wiesbaden: Harrassowitz Verlag.
Csató, Éva Ágnes. 2000. Some typological features of the viewpoint and tense system in spoken North-Western Karaim. In: Dahl, Östen (ed.), Tense and aspect in the languages of Europe, 723–751. [= Empirical Approaches to Language Typology20–6]. Berlin, New York: De Gruyter.
Çulha, Tülay. 2006. Karaycanin Kisa Sözvarliği. Karayca–Türkçe Kisa Sözlük. [= Dil ve Edebiyat Dizisi6]. İstanbul: Sanat Kitabevi.
Eckmann, János. 1966. Chagatay Manual. [= Indiana University Publications. Uralic and Altaic Series60]. Bloomington, IN: Mouton.
Firkovičius, Mykolas. 1996. Mień karajče ürianiam. Vilnius Danielius.
Friedman, Richard Elliott. 2003. The Bible with sources revealed. A new view into the Five Books of Moses. New York: HarperOne.
Grunin, Timofej Ivanovič. 1967. Dokumenty na poloveckom jazyke XVI v. (sudebnye akty kamenec-podoĺskoj armjanskoj obščiny). [= Pamjatniki piśmennosti vostoka3]. Moskva: Nauka.
Gülsevin, Selma. 2016. Karay Türklerinin Dili (Troki Diyalekti). [= Türk Dil Kurumu Yayinlari1165]. Ankara: Türk Dil Kurumu.
Hahn, Reinhard F. 1998. Uyghur. In: Johanson, Lars & Csató, Éva Ágnes (eds.), The Turkic Languages, 379–396. London, New York: Routledge.
Jankowski, Henryk. 2014. Two Karaim religious poems by Isaac ben Abraham Troki. Karaite Archives 2: 35–57.
Jankowski, Henryk. 2019. Introduction. In: Jankowski, Henryk & Aqtay, Gülayhan & Cegiołka, Dorota & Çulha, Tülay & Németh, Michał, The Crimean Karaim Bible. Volume 1: Critical edition of the Pentateuch, Five Scrolls, Psalms, Proverbs, Job, Daniel, Ezra and Nehemiah, ix–xxxix. [= Turcologica119]. Wiesbaden: Harrassowitz Verlag.
Jankowski, Henryk & Aqtay, Gülayhan & Cegiołka, Dorota & Çulha, Tülay & Németh, Michał. 2019. The Crimean Bible. Vol. 1. Critical edition of the Pentateuch, Five Scrolls, Psalms, Proverbs, Ezra and Nehemiah. Vol.2. Translation. [= Turcologica119]. Wiesbaden: Harrassowitz Verlag.
Johanson, Lars. 1995. Mehrdeutigkeit in der türkischen Verbalkomposition. In: Erdal, Marcel & Tezcan, Semih (eds.), Beläk Bitig: Sprachstudien für Gerhard Doerfer zum 75. Geburtstag, 81–101. [= Turcologica23]. Wiesbaden: Harrassowitz Verlag.
Johanson, Lars. 1999. Typological notes on aspect and actionality in Kipchak Turkic. In: Werner, Abraham & Kulikov, Leonid (eds.), Tense-aspect, transitivity and causativity. Essays in honour of Vladimir Nedjalkov, 171–184. [= Studies in Language Companion Series50]. Amsterdam, Philadelphia: Benjamins.
Johanson, Lars. 2000. Viewpoint operators in European languages. In: Dahl, Östen (ed.): Tense and aspect in the languages of Europe, 27–187. [= Empirical Approaches to Language Typology20–6]. Berlin, New York: De Gruyter.
Johanson, Lars. 2021. Turkic. Cambridge: Cambridge University Press.
Juldašev, Achnef Achmetovič. A. 1965. Analitičeskie formy glagola v tjurkskich jazykach. Moskva: Nauka.
KarRPS see Baskakov, Nikolaj Aleksandrovič & Zajączkowski, Ananiasz & Šapšal, Seraja Markovič (eds.) (1974).
Klemensiewicz, Zenon & Lehr-Spławiński, Tadeusz & Urbańczyk, Stanisław. 1955. Gramatyka historyczna języka polskiego. Warszawa: Państwowe Wydawnictwo Naukowe.
Klemensiewicz, Zenon. 2002. Historia języka polskiego. Warszawa[7th edition]: Wydawnictwo Naukowe PWN.
Kocaoğlu, Timur (in collaboration with Frikovičius, Mykolas). 2006. Karay. The Trakai dialect. [= Languages of the World. Materials458]. München: Lincom.
Kowalski, Tadeusz. 1929. Karaimische Texte im Dialekt von Troki. [= Prace Komisji Orjentalistycznej Polskiej Akademji Umiejętności11]. Kraków: Polska Akademia Umiejętności.
Matthews, William Kleesman. 1975. Russian historical grammar. London[reprint]: University of London.
Medvedjev, Fedir Pylypovyč. 1955. Istoryčna gramatyka ukrajinśkoi movy. Korotkyj narys. Častyna perša. Charkiv: Charkivśkyj deržavnyj universytet imeni O. M. Goŕkogo.
Musaev, Kenesbaj Musaevič. 1964. Grammatika karaimskogo jazyka. Fonetika i morfologija. Moskva: Nauka.
Musaev, Kenesbaj Musaevič. 1977. Kratkij grammatičeskij očerk karaimskogo jazyka. Moskva: Nauka.
Nadžip, Èmir Nadžipovič. 1960. Sovremennyj ujgurskij jazyk. Moskva: Izdateĺstvo vostočnoj literatury.
Németh, Michał. 2010. O wpływach polskich na język Karaimów łuckich. LingVaria 5.2: 199–212.
Németh, Michał. 2011a. Unknown Lutsk Karaim letters in Hebrew script (19th–20th centuries). A critical edition. [= Studia Turcologica Cracoviensia12]. Kraków: Jagiellonian University Press.
Németh, Michał. 2011b. Zwięzła gramatyka języka zachodniokaraimskiego z ćwiczeniami. [= Prace Karaimoznawcze1]. Poznań: Katedra Studiów Azjatyckich.
Németh, Michał. 2013. Ananiasz Zajączkowski’s doctoral thesis. The original manuscript of Sufiksy imienne i czasownikowe w język zachodniokaraimskim. Folia Orientalia 50: 115–157.
Németh, Michał. 2014. An early North-Western Karaim Bible translation from 1720. Part 1. The Torah. Karaite Archives 2: 109–141.
Németh, Michał. 2015. A historical morphology of Western Karaim. The -p edi past tense in the south-western dialect. Acta Orientalia Academiae Scientiarum Hungaricae 68/2: 215–228. DOI: 10.1556/062.2015.68.2.6.
Németh, Michał. 2018. An early North-Western Karaim text copied before 1700. A linguist’s contribution to the biography of Josef ha-Mashbir. Almanach Karaimski 7: 83–98.
Németh, Michał. 2019. A historical morphology of Western Karaim. The -a-d- continuative in the north-western dialect. International Journal of Eurasian Linguistics 1/2: 268–308DOI: 10.1163/25898833-12340017.
Németh, Michał. 2020a. Gustaf Peringer’s Karaim Biblical Material Revisited. A Linguistic Commentary on a Text Sample from 1691. In: Szczyrbak, Magdalena & Tereszkiewicz, Anna (eds.). Languages in contact and contrast. A Festschrift for Professor Elżbieta Mańczak-Wohlfeld on the occasion of her 70th birthday, 295–309. Kraków: Wydawnictwo Uniwersytetu Jagiellońskiego.
Németh, Michał. 2020b. Middle Western Karaim. A critical edition and linguistic analysis of pre-18th-century Karaim interpretations of Hebrew piyyutim. [= Languages of Asia22]. Leiden, Boston: Brill.
Németh, Michał. 2021. The Western Karaim Torah. A critical edition of a manuscript from 1720 Vol. 1– 2. [= Languages of Asia 24]. Leiden, Boston: Brill.
Németh, Michał. 2022 (forthcoming). Slavic impact on Karaim. Forthcoming in: Greenberg, Marc L. (editor-in-chief): Encyclopedia of Slavic Languages and Linguistics Online. Brill].
Peake, Arthur S. (with the assistance of Grieve Alexander James) (eds.). 1920. A commentary on the Bible. New York, London: Thomas Nelson & Sons.
Prik, Oĺga Jakovlevna. 1976. Očerk grammatiki karaimskogo jazyka (krymskij dialekt). Machačkala: Dagestanskij gosudarstvennyj universitet imeni V. I. Lenina.
Pritsak, Omeljan. 1959a. Das Karaimische. In: Deny, Jean & Grønbech, Kaare & Scheel, Helmuth & Togan, Zeki Velidi (eds), Philologiae Turcicae Fundamenta. Vol. 1, 318–340. Wiesbaden: Franz Steiner Verlag.
Pritsak, Omeljan. 1959b. Das Kiptschakische. In: Deny, Jean & Grønbech, Kaare & Scheel, Helmuth & Togan, Zeki Velidi (eds), Philologiae Turcicae Fundamenta. Vol. 1, 74–87. Wiesbaden: Franz Steiner Verlag.
Pritsak, Omeljan 1959c. Das Neuuigurische. In: Deny, Jean & Grønbech, Kaare & Scheel, Helmuth & Togan, Zeki Velidi (eds), Philologiae Turcicae Fundamenta. Vol. 1, 525–563. Wiesbaden: Franz Steiner Verlag.
Rentzsch, Julian. 2005. Aspekt im Neuuigurischen. [= Turcologica65]. Wiesbaden: Harrassowitz Verlag.
Rentzsch, Julian. 2019. Review of the cumulative habilitation treatise A history of the Karaim language during the period from the second half of the 17th century up to the second half of the 19th century (based on an analysis of newly discovered written sources) by Dr. Michał Németh. Mainz (8 September 2019).
Schönig, Claus. 1997. Finite Prädikationen und Textstruktur im Babur-name (Heiderabad-Kodex). [= Turcologica31]. Wiesbaden: Harrassowitz Verlag.
Seegmiller, Steve. 1996. Karachay. [= Languages of the World. Materials109]. München, Newcastle: Lincom Europa.
Sulimowicz, Anna. 2013. Mieczek i Siunek. Awazymyz 38.1: 4–13.
Tentzel, Wilhelm Ernst (ed.). 1691. Monatliche Unterredungen einiger guten Freunde von allerhand Büchern und andern annemlichen Geschichten; allen Liebhabern der Curiositäten zur Ergetzligkeit und Nachsinnen heraus gegeben. Leipzig: Fritsch.
Zajączkowski, Ananiasz. 1931. Krótki wykład gramatyki języka zachodnio-karaimskiego (narzecze łucko-halickie). Łuck: Aleksander Mardkowicz.
Zajączkowski, Ananiasz. 1932. Sufiksy imienne i czasownikowe w języku zachodniokaraimskim (przyczynek do morfologii języków tureckich). [= Polska Akademja Umiejętności. Prace Komisji Orjentalistycznej15]. Kraków: Polska Akademia Umiejętności.
Currently, the main grammatical descriptions of the dialects of Karaim (iso 639-3 code in the unesco Atlas of the World’s Languages in Danger), i.e. of the now extinct Eastern (Crimean) Karaim with its variants and Western Karaim with its two (i.e., north-western and the extinct south-western) subdialects are as follows: Kowalski (1929), Zajączkowski (1931), Pritsak (1959a), Musaev (1964, 1977), Prik (1976), Firkovičius (1996), Berta (1998), Kocaoğlu (2006), Aqtay (2009: I 33–47), Németh (2011a: 21–76, 2011b), and Gülsevin (2016: 49–131). Karaim word-formation has been described by Zajączkowski (1931: 20–23; 1932) and Çulha (2006: 18–32).
The oldest texts written in Western Karaim originate from the second half of the 17th century (see them described in Németh 2020b: 1–5). These are relatively short religious poems recorded in Hebrew script. Three of the six currently known handwritten texts composed before the end of the 17th century have been the subject of critical editions (Jankowski 2014, and Németh 2018). Additionally, the first three verses of the Torah collected as linguistic material by the Swedish Orientalist Gustaf Peringer Lillieblad (1651–1710) and published in 1691 by Tentzel (1691: 572–575) have been analysed in depth by Németh (2020a). The 18th century, in turn, saw the emergence of translations of entire Biblical books into Western Karaim (see below), and a heyday of the paytanic and paraliturgical literature written in the Karaim language.
The latter verbal category was only described in some of the above-mentioned works, see: Musaev (1964: 276–278), Prik (1976: 134–135), Berta (1998: 312), Csató (1998: 52; 2000: 737), Aqtay (2009: 40), Németh (2011a: 126–128; 2011b: 48), Gülsevin (2016: 94).
In his unpublished notes written in 1969, Józef Sulimowicz (1913–1973), a Polish Turcologist and a native speaker of South-Western Karaim described this tense as only being very rarely used, see Németh (2015: 216, 225). In fact, it took the present author several years of philological work to come across the examples presented below, which also indicates how common this category was.
The Biblical context of the respective verses was primarily checked against Benson (1857), Peake (1920), and Friedman (2003).
As far as we know, this characteristics of the -p edi- past is not mentioned in other descriptions of Uyghur, see, for instance, Pritsak (1959c), Nadžip (1960), and Hahn (1998).
Johanson does not provide a definition of this term; the only clarification we find is “cf. pluperfect vs. perfect” (see Johanson 1999: 178). We assume that this term refers to an utterance relaying an event that took place before another event.
For more information regarding Simcha ben Chananel himself and ms. ADub.iii.73, see Németh (2014, 2015b). For some more details on TKow.01, see References below.
ADub.iii.73: damaged text. Reconstructed on the basis of: TKow.01: ajttyj, JSul.iii.01: ajttyn, bsms 288: ajttyŋ.
ADub.iii.73: damaged text. Reconstructed on the basis of: TKow.01: qatynlyqqa, JSul.iii.01: qatynlyqqa, bsms 288: ḫatynlyqqa.
The only uncertainty might lie in the interpretation of the use of NWKar. alybedim in example 1 (Exo 12:19). The past event which succeeds chronologically the event expressed with the -p edi past form appears in Exo 12:17, i.e. two verses before the -p edi form itself is used: Da ḥas[tala]tty Adonaj ošol parʿonu {ullu ḥastalyqlar} [byla] da ošol elin üvün[ün] Saraj üčün qatyny üčün Avramnyn ‘And the Lord made Pharaoh and the people of his house [sick with] {great diseases} because of Sarai, Abram’s wife.’ (ADub.iii.73: 17 ro, see Németh 2021). However, one pivotal circumstance is that from the perspective of the speaker (i.e., Pharaoh), there is a causal link between the fact that he had taken Sarah for a wife and (hence) God’s smiting him with plagues and the actual succession of events leaves no doubts. In general, the past events that are related to each other do not have to appear in the same utterance, see, for instance, Juldašev’s work (1965: 194) for respective examples in the Turkic languages.
This accords with the basic meaning of the WKar. -p converb which denotes an action that is finished, see A. Zajączkowski’s description edited by Németh (2013: 217–218, 134–136).
It is, however, not entirely clear how focal the Kipchak -p edi forms are from Johanson’s perspective. In Johanson (1999: 176) he considers them to be “of lower focality” and quotes Krch. ölüb edi ‘had died’ as an example, but in Johanson (2000: 110) the same construction is described as “high-focal”, without any reference to his article of 1999 and a Karachay example, namely ketib edi ‘was gone’, is given here, too. The same is the case with the Tkc. -ġan edi construction: he considers it “nonfocal” in Johanson (2000: 120–121), but then it is classified as “of lower focality” in Johanson (1999: 178). What makes Johanson’s analysis less compliant with a discussion based on philological data is that in his highly theoretical works the respective verbal forms are adduced and defined without documenting the context, i.e. without providing entire utterances (see, e.g., Johanson 1995, 1999, 2000). This gives rise to situations such as, for instance, both Kzk. kelip edi and Kzk. kelgen edi being translated as ‘had come’ in an article devoted to the specific features of these grammatical categories (see Johanson, 1999: 178, 179).
< Ukr. бувало ‘it used to happen (used for expressing that an event used to happen repeatedly, though irregularly)’.
SWKar. kacanes ‘formerly, once’ is built on the basis of Slavonic pattern, i.e. kacan ‘when’ + -es (<Pol. -ś, Ukr. -сь id.) particle building indefinite pronouns, cf. Pol. kiedyś ‘formerly, once’.
< Ukr. давно ‘long ago’, Pol. dawno id., Russ. давно id.
Even though South-Western Karaim was shaped also by other Slavonic languages, we have mainly concentrated here on Polish because the Russian pluperfect disappeared well before the oldest Karaim manuscripts analysed in this paper were written (Matthews 1975: 208). On the other hand, although it continued to remain a productive category in spoken Ukrainian for a somewhat longer time, it did not perform the function of expressing actions performed in the distant past (Medvedjev 1955: 176–178).
We are also restrained by the fact that if we are to contrast their use accurately, the linguistic material must originate from the period when both pluperfects were still productive categories, preferably the Middle Western Karaim (i.e., before the early 19th century) period (Modern Western Karaim data is discussed in the final chapter of this paper). Seen in this light, we are fortunate to have at least some data at our disposal.
Given that the present article is devoted primarily to Western Karaim – and also so as to comply with the length restrictions set by the editors – the Eastern Karaim linguistic material is presented without a morpheme-by-morpheme analysis.
As we see, there is no trace of the -Vp edi- > -Vbedi- change in ms. bsms 288. The same is true for manuscript Gaster Hebrew ms 170, see e.g. eri ki alyp edi any özünä ḫatynlyqġa ‘his man, who had taken her for himself for a wife’ (Deu 24:3), see Jankowski et al. (2019: 337, fn. 833) and example 14 above.