You are on page 1of 16
CHAPTER 26 Sakha and Dolgan, the North Siberian Turkic languages Brigitte PaAKENDORF AND EuGéNIE STAPERT 26.1 Introduction Sakha—which is the endonym of the people and their language—is more commonly known in the literature as ‘Yakut, It is the numerically and sociopolitically dominant indigenous language in the Sakha Republic (Yakutia), and its ‘geographical spread covers large parts of central and north- astern Siberia. Dolgan, on the other hand, is largely restricted to the southern edge of the Taimyr Peninsula and adjacent northwestern Yakutia, On the Taimyr Peninsula Dolgans ae settled primarily along the Kheta and Khatanga rivers, which flow northeast into the Laptev Sea. ‘The 2010Russian national census Nacional’ny sostav 2010: 12,21) counted 478,085 Sakha and 7,885 Dolgans. These are still predominantly rural, though the Sakha are more urban- ized: ~60% (284,834) of Sakha and ~77% (6,045) of Dolgans lived in rural settlements at last count. Among the Sakha, 406,126 (ie. approximately 85%) claimed to speak Sakha in response to the census, with 428,280 (90%) claiming @ knowledge of Russian (Vladenie jazykami naseleniem 2010: 158,152) Although there are thus stil substantial numbers of Sakha speakers, there isa trend toward Russian dominance: inthe 2002 census 93% of Sakha respondents claimed to know their heritage language and 87% claimed a knowledge of Russian, Nevertheless, in rural settlements Sakha stil quite healthy: her, itis the language of everyday use, and some ‘older speakers and most preschool children are monolingual (Pakendorf 2007: 2). Dolgan, in contrast, is clearly endan- gered: only 930 Dolgans (ie. less than 12%) claimed to speak Dolgan in 2010 (Vladenie jazykami korennyx 2010b: 2120). However, the sociolinguistic situation differs considerably from village to village, While Dolganisstill acquired by chil- dren in the remote villages of Syndassko and Popigay it is spoken fluently only by a few individuals over 60 years old in villages closer tothe city of Dudinka (Stapert 2013: 25). Both Sakha and Dolgan are written with a Cyrillc-based orthography with some additional graphemes for specific phones/phonemes. Sakha has been the object of extensive study, starting in the 19th century with the first reference ‘grammar (Bahtlingk 1851) and a multi-volume dictionary (Pekarskij 1907-30). Further landmarks are the two-volume reference grammar ofthe language (Ubrjatova, Korkina etal. 1982; Ubrjatova et al. 1995) and the ongoing publication ‘of the monumental Sakha explanatory dictionary (Slepcov 2004-15). Among the key sources concerning Sakha language contact are Kaluzyriski (1962) on the Mongolian elements in Sakha and Romanova etal. (1975) on Sakha-Evenki contacts Dolgan has been less widely studied; the most important ‘grammatical descriptions of this language are Ubrjatova (1985) and Artem'ev (2001, b), while Stachowski (1993, 1998) represents a rare lexicographical source. ‘This chapter's largely based on our oral corpora of Sakha and Dolgan’ (see Pakendorf 2007: 63-4 and Stapert 2013:7-9 for details). These contain approximately 29,400 and 17,700 words, respectively 26.2. Historical connections: Genealogy and contact Both authors are skeptical with respect to the genealogical relationship of the Transeurasian languages. Nevertheless, if one considers Transeurasian or Altaic languages to be an areal grouping rather than a language family, then Sakha and Dolgan belong to this grouping. » Eugénistapet compete information on Delganand wretethesectons| on phonology and on moreogy of ours and pronouns and Big Pardo complete formation on saa an wrote the resto the captr. "ie are grateful to the Wenner-Gren Foundation, the Max Planck Society (va the Department of Linguistics and the MPRG on Comparative Population Linguistics at the MP fo Evolitonary Antopoogy, both now ofan), and to the INEL project funded by the Union of the German deademes of Slences and Humanities, fr funding the fl trpe to com pile and check these corpora, and to al the Saha and Dlgan Individuals ‘tho conrbuted to our work. BP i furthermore gratefid to he CAREX [ASLAN (AN-1DLABX-O081) of Université de Lyon forts san support ‘within the program “Investstements Avene” (ANR-1IDEX-0007) ofthe French government operated by the National Research Agency (ANR) ‘Brg Planrt an anne Spe, Stn and Dan te Nar Soe kenge: he xr Gee Massa Langage et aon ed Nanve tess nt lesa Stey, fo ey Press (2020, © Bite Pkt aa Eoin Sten. or 10 Tones Ten oHeD628.0030027 RCS OOo REO ee THE NORTH SIBERIAN TURKIC LANGUAGES, Figure 26.1 Map of eastern Siberia showing the areas where Sakha and Dolgan are spoken ‘© Dynamique du Langage, UMR559%6, CNRS & Université de Lyon Sakha and Dolgan are classified among the Northeastern ‘Turkic languages, a heterogeneous group that is based on areal contiguity rather than on established genealogical ‘unity (ohanson 1998a: 82-3). Within this grouping Johanson (19984: 83) identifies the homogenous North Siberian group ‘comprising Sakha and Dolgan, and a heterogeneous South Siberian group. The North Siberian Turkic languages differ from their relatives in phonological and several structural features, as well as numerous lexical changes (Kaluayfski 1962; Pakendorf 2007; Pakendorf and Novgorodov 2009a;, Stapert 2013). Mutual intelligibility with other Turkic lan- _guages is thus severely restricted. 431 BRIGITTE PAKENDORF AND EUGENIE STAPERT ‘Table 26.1 Consonant phonemes Labial Alveodental Palatal Velar uvular Glottal Plosive pb td kg Nasal m a p > Trill r Fricative s W wo h) Affricate oe ‘Approximant j Lateral approximant L ‘The two languages are so closely related that Dolgan is sometimes considered a dialect of Sakha (eg. Voronkin 1999 154); however, on the basis of sociocultural and political fac- torsit canbe considered separate language (Arterev 2001 6, Stapert 2013: 62). Although Sakha and Dolgan are spoken over a vast geographical area, they are very homogenous, probably due to their recent spread, which is assumed to have taken place only in the late 17th and early 18th centuries (Dolgix 1960: 360; Warm 1996: 971-2; Pakendorf etal 2006: 348). Nevertheless, four dialect groups are distinguished for Sakha, based partly on geographic and sociocultural criteria (Voronkin 1999: 158): the central group, the Vilyuy group, the northwestem group (which in Voronkin’s classification includes Dolgan), and the northeastern group (Voronkin 1999: 154-5). InDolgan,an“upriver” dialect (ichegi is distinguished from a “downriver” dialect (alaragi)(Stapert 2013: 27). Sakha has undergone substantial contact influence history, both lexical and structural. Since Dolgan is derived from Sakha (Stapert 2013: 34-62), it shares some of these contact-induced changes, but has undergone additional changes under the influence of Evenki. During its history, Sakha has copied large amounts of lexical items from an unspecified Mongolic language. These include even basic lexical items such as kinship and body-part terms, a large number of verbs (Pakendorf and Novgorodov 2009a: 509), as ‘well as many derivational suffixes (Kahuzytski 1962; akendorf 2015). n addition, Sakha has undergone structural changes under the influence of Evenki (Pakendorf 2007: 95-270). Interestingly, as shown by molecular anthropological inves- tigations (Pakendorf et al. 2006; Pugach et al. 2016: $34), these contact-induced changes in Sakha took place largely in the absence of intimate physical contact (cf. Pakendorf 2007: 317-23). Dolgan has copied some lexical items and has also under- gone structural influence, including calques, rom Evenkisince its divergence from Saka, and has also come under Russian 432 influence (Stapert 2013: 380-57). In contrast tothe Sakina con- tact situation, the contact-induced changes in Dolgan went hand in hand with genetic admixture with both Evenks and Russians (Pugach 2016: S3A; cf. Stapert 2013: 69-76). 26.3 Phonology 26.3.1 Consonants Both Sakha and Dolgan have 16 consonantal phonemes’ and three allophones (in brackets), as shown in Table 26.1. In both languages there is variation between [de], (4), and [j], ea. [@aXtar] [d/axtar], and [jaxtar] ‘woman’, with the affricate pronunciation being more frequent in Sakha and the palatalized stop more frequent in Dolgan. similarly, [i], and[p] occur interchangeably in a restricted number of words, eg, [jax] ~ [jax] ~ fanax] ‘mouth’ ie] ~ [jel ~ [ine] ‘mother’. While in Saka the glide is standardized as ~fin Dolgan the palatal nasal fulfills this role, represented as in the orthography. Moreover, restrictions apply to the distribution of some phonemes: word-initially /r/ does not occur at al, /g/ ‘occurs in only a very limited number of native words (only gin ‘do’ in Saka, and a few more in Dolgan, eg. gine’ and gitta‘with), and /n/ occurs word-initally only in loanwords. Word-inital /n/ is restricted to mostly onomatopoeic words ‘of Turkic origin, whereas it is fairly common in words of ‘Tungusic and Samoyedic origin, e poo ‘hunchback’. The voiced consonants /b/, /d/, and /g/ do not occur in word- final position. » tn Section 263, we use PA for transcription Corespendences with the transription adopted throughout the res ofthe chapter are (where ferent: poi, Ge dnd, ere, o0, ye. Prevocalicaly [x] occurs before the low back vowels /a/ and /o/ in both languages, whereas (K] is found before all ‘other vowels, eg. zazr'snow’ vs kl ‘come’, Post-vocalically [occurs afterall low vowels in Sakha, whereas in Dolganit is restricted to low back vowels. While ths allophonic vari ation is represented graphemically in Sakha by «x= for {k] ‘and mah-a \wood-oss.3s ‘his wood. In word-inital position, (s] and [h] are possible in both spoken Dolgan and Sakha, but in writing, Sakha has standardized [5] with the grapheme , whereas Dolgan uses [h] (. In spoken Dolgan, fy] occurs only after low back vowels, i. ayz, but otherwise [g] is pronounced, ‘eg, begehe. Both [g] and [y] are represented by the letter . Consonant assimilation (progressive, regressive, and bidirectional) takes place between adjacent consonants ‘on morpheme boundaries, eg. at ‘horse’-LAr* ‘ru’ attar ‘horses’ For a fll overview see Stachowski and Menz (1998: 419-20). /p/, /k/,and /x/ voice before vowels, whereas /t/ voices before vowels only when itis adjacent to /t/ or /I. 26.3.2 Vowels Sakha and Dolgan have 20 vowel phonemes: eight short vowels, eight long vowels, and four diphthongs. They can bbe categorized according to height, roundedness, and frontness (Table 26.2). Diphthongs behave phonologically ‘according to their leading element and are therefore classi- fied as high vowels. For all features minimal pairs are attested, eg. at ‘horse’ ait ‘name’, it ‘dog’ ist ‘send! ‘grass’ uot ‘fire. ‘Table 26.2 Vowel phonemes Low High | Unrounded Rounded Unrounded Rounded Back | a 00: ik uur Front | ee: 00: iy yys ie ye + Throughout the chapter, morphemes in ioation ae presented intheir underving form, with capa leters indicating phonemes that undergo changes according to vowel hanmony and consonant asinlation rues. THE NORTH SIBERIAN TURKIC LANGUAGES, Vowels are combined according to the principles of palatal and labial vowel harmony. Palatal vowel harmony is wide- spread among Turkic languages (Johanson 1998b: 32-4), and applies within lexical stems (eg. kirsa ‘polar fox’, kiche ‘evening’) as well as across morpheme boundaries (eg kirsa-lar polar fox-r,kiche-er evening-1.). While in Sakha this principle applies very strictly, Dolgan allows for certain exceptions, eg harserda ‘morning’ vs. Sakha sarstarda, Labial vowel harmony is less common in Turkic languages Gohanson 1998b: 32-4) and also applies within stems and across morpheme boundaries (oyo-lor child: vs, kirsa-lar polar.fox-n). The only exception occurs after the high rounded vowels /u/ and /y/, after which low vowelsbecome unrounded, eg. tyrgen ‘speed’, omuk-tar foreigner-n. Similarly, muas ‘antler’ becomes muos-tar antler-r- and not *mucs-tor, since in diphthongs the initial high element is leading. 26.3.3 Phonotactics In native words the following syllable structures occur: V, VC, VEC, Vic, CV, CV,CVEC, CVC, CVCC. Consonant clusters are prohibited word initially, but word-finally they can occur in the combinations {rt} rt, [Ik [t, [mp], ink). Some stems containing a high vowel in the second syllable lose the high vowel in inflectional forms, e.g tayis- “exit” becomes tays-ar [exit-es(raeo.3sl] ‘she exits. This also occurs in the passive, where the suffix -(In changes into (Il, eg. tut-ullu-butsa [bulld-eass-er-rsr2-0s833c] “itwas built, 26.4 Morphology 26.4.1 Nouns Nouns have singular and plural number, whereby singular is unmarked, and plural is marked with a suffix (but see Section 265.2 for lack of plural agreement after numerals). Most common is -LAr, asin hirga sleigh’ > hirga-larsleighs’ Some words ae pluralized using the alternative suffix tar This phenomenon tsb Found in other Turkic languages Uohanson oo: 31-2) but in these cases the elson ofthe high vowel occurs ony before vowel ital sis Not that In Saha and Dolgan most fine tense markers ar ranspar- ently derived from partcples and are Mentcal with these inform. However our loses eflecthelr ychrani function, so that we loss the ‘sine form either as atense marker or asa participle. "This suffix was derived from the Mongolic phiral marker-t, which occured after & word final -n. The fix was coped from Mongo i Sakha, and was enforced bythe Turkic pial sufix LAr (Pakendort 208: 1-8. 433 BRIGITTE PAKENDORF AND EUGENIE STAPERT as in emeszs-ter ‘old women’ eme:ysin ‘old woman’. In ‘Sakha, this suffix is restricted to stems ending in -n, which typically refer to humans, but not always, eg. at-tar< atin ‘other, said in reference to ‘other districts. In Dolgan, on. the other hand, the association with humanness is stronger than the morphological restriction of ending in-n, eg, azta- ‘tar ‘women’ < d'agtar ‘woman’ (Sakha djaytal-lar), and duoytu-ttar doctors’ kirgittar ‘girls’. Finally, the associa- tive plural is derived using the proprietive suffix -LA:, which attaches only to names, eg, Jekim-ne:y kira uollara {Uekim-rsoex small son-toss.3e] ‘the youngest son of Jekim and family olgan and Sakha have eight cases (Ubrjatova et al. 1982: 130; Ubrjatova 1985: 122): nominative, accusative, dative, partitive, ablative, instrumental, comitative, and comparative, which occur in a possessive and a non- possessive declension (for a complete overview of suffixes see Table 26.3) Sakha and Dolgan deviate from the Turkic 26.4.2. Pronouns and possessive suffixes Possessive suffixes, which are identical for Sakha and Dolgan (Table 26.3), are used to mark possession on nouns, and they play an important role in definiteness, reference tracking in relative clauses, as well sin the verbal paradigm. In oblique «cases, possessive formant for the singular differ from these in the nominative. Possessive formants and case suffixes combine into the possessive declension represented in Table 26.3. Predicative suffixes are used to mark nominal and verbal predicates (Table 26), They ae the same for both languages. Sakha and Dolgan distinguish three persons in singular and tree in plural forthe free personal pronouns. Every person canbe inflected forall cases, using the case suffix ofthe non- possessive declension, but for Iss and 2s different stem is used in inflected forms. Since forthe dative case yet another stem is used, al three stems are given in Table 26: ‘Table 26.4 Predicative suffixes pattern in that they have lost the genitive and locative “Person Predicative suffix cases and have developed a partitive and comparative as well as a comitative that differs from the instrumental. 1° oa However, in Dolgan the status of the comitative is ques- 2c Gin tionable (cf. Ubrjatova 1985: 122), since it occurs only 4. ‘once in the spoken corpus (vs. 50 times for Sakna). A much more common way to express a comitative mening in If -Bit Dolgan is the use of the postposition gitta ‘with’ (which ao. att governs the accusative case) or the proprietive suffix us F r ‘Table 26.3 Case and possessive suffixes Non-Possessive Possessive Ise 236 36 In 2. an NOM — 0m “Oy oA ie sit “Lara DAT GA Bar “Gar (dighr/-Ar Bitar = -GItlgar——-LArigar Acct Bin Gin on -Bitin itn sarin PART -TA Bina (0ina -Bitina Gina sArina ABL -BittAn (oittan Bieta -Gteean—-LArittAn Ins, -Binan man -Bitinan—— Gen -LArInan com -BitinA:n/ —-GitinA:n/ -Bitinién——-ittntin COMP -rA:yAr -BitinAzyAr —-GitinAzyAr—_-LArinA:yAr Possesive declension only in Sak. 434 * throughout the chapter, when Sakha and Dlgan formally fer, the Sahl forms are presente first and are separated fom the Dolgan forms byaslash Dolgan and Sakha have a special dual inclusive pronoun bihilki, denoting ‘you and I, There are no special politeness forms, and 2s: is used forall addressees regardless of age or social status. Possessive pronouns can be used to reinforce possessive marked nouns (Table 26.5). Typically, these possessive pronouns are postposed, e.g. munsiay-pit bikiene {meeting- 1n. our] ‘our meeting’. However, in Dolgan they can be used as a modifier and occur in prenominal position: (2) Dolgan (Story_syn_marriage_APC_138) ‘orton miniene tohoyom barr then my —— birthpole-ross.ise esr ‘...and then there is my birth pole’ ‘The reflexive pronoun bee ‘self” in combination with pos- sessive suffixes is used with a purely reflexive meaning: it can also function as an emphatic marker (2). (2) Dolgan (Story syn_syndassko_PPK_26) cskuola-ta hhuoy-taruraha die-ge school.R-ross3s ween. pole house-oer iiiret-ler bejelere ——tut-an-nar teach-rasenen.3%. self-oss.3etbulld-scqcve-raso.3e. “There was no school, they taught us in a tent that they had put up themselves.” Demonstrative pronouns represent three degrees of distance: proximal, medial, and distal (Table 26.6). In emphatic use a THE NORTH SIBERIAN TURKIC LANGUAGES, prefix h(V}- is added to the demonstrative e.g, hu-bu [ou this] just this’, ha-manna{evri-here] just here, exactly here hol [ewes-that] ‘just that’. This form is particularly frequent in western Yaka (3). (6) Sakha (WvaP_039) hronnuk —hucbu—nehiliek-ke is ere wer-sucha evpithis nasleg-oer three only jal bar family vast “in this nasleg [administrative unit] there are only three families like that. Qualitative demonstratives (in this way’ ‘in that way’) are derived from the demonstratives with the adverbializing suffix -Lik eg. man-nik‘in this way, like this, itn-nik‘in this way, like this, on-muk' like that” (4) Dolgan (Story_tundra_syn_PPK_19) con-nuk —isker-bip-pit— bihigi-maynaj this-rov develop-sr2-tn. tm atfirst ‘Like this we grew up at first.” suffix bar-a-bin [go-mns-reo.1s] ‘go’ and hanac- ‘think’ in (think.res-reeo.tsc]'T think’. The third-person is both of which show the same distribution as -A and -:, eg. bar-ar [go-rs[se0.3:]] “s/he goes’ and hanir (thinkers: ‘s/he thinks’. The negative forms take the negation marker -BAt, the origin of is found in the negative present participle, e.g, bar- bap-pin[go-sc.s- mens] dont go’, bar bat-fr (go-e.s- so. ‘they don't go’. ‘The affirmative future tense is formed with the sufix-IAy (historically the future participle) and takes possessive subject-agreement marking, The singular forms have two variants, full and contracted, e.g, kér-i6y-iiy [see-rur- 0-25] «kid [See-r.oss.2e] ‘you wil ee, hdd -iay-a {go-ur-ross.3sc] = hidsds ia [go-r.oss.3sc) ‘she will go, kel- iey-tere [come-x-#ess3r1] ‘they will come’. In spontaneous speech the contracted forms are vastly more frequent than the fll forms. ‘The simple past is marked by the suffix -T and takes pos- sessive subject-agreement marking (with the exception of 3n., which takes the predicative agreement suffix); it is negated with the suffix -BA. Examples are bard-im [go-rsr1- 058.1] ‘I went’, bar-dea [go-rs-#0ss.3s] ‘s/he went’, ays ba-tim [go.out-nec-esr1-ross.1s] ‘didnt go out ‘Two distant past tenses are formed with the markers Bit (affirmative) and -BAtAy (negative), which are historically derived from the affirmative and the negative past participle, respectively Note that since the precise tense-aspect-modality (TAM) reading of these distant past tenses is obtained only through the choice of possessive or predicative subject agreement, we here gloss these suffixes simply as es2 (to contrast with the simple pastsufix-T, glossed a rst, andthe imperfect past discussed below, glossed as rsr3). The wit- nessed past takes possessive person marking, e.g. kergen tays-an bar-bit-im [spouse exit-seacve go-rs¥2-ross.1sc] “I ‘married’, kel-betey-e come-nez.st2-s0s.3s] ‘he didn't come’. ‘The unwitnessed past takes predicative subject agreement, eg leis bar-bt [work:sn.cvs go-rsra{rnc.3sc] ‘she went to ‘work, bier betey [give nssesraro sc] ‘she didnt give’, zohan er-bik-kinsij {when see-rsr2-raro.2so-0) ‘when did you see?” 438 In Sakha, the imperfect past has two affirmative variants, synthetic and analytic, with no difference in meaning, although the analytic construction—which isthe sole means ‘of expressing this tense in Dolgan—predominates in the oral corpus. The synthetic construction consists ofthe suffix -Ar (historically the present participle) taking possessive per- son marking, and the analytic construction consists of the lexical verb in participial form plus the auxiliary e-in simple bar arim|go-rx3-ross.Aso] I went (frequently, often)’, lire e-t-im [killeeserce auxorst1-oss.1sc] ‘used to kill/ killed many’. The negative exists only as an analytic con- struction, e.g. hugaha:-bat e-tim [come.close-ntcacr avx- pstl-ross.1s ‘didn't come close”. Finally, there is a remote past with pluperfect (11a) and resultative (11b) readings. Formally, itis parallel to the imperfect past, with the lexical verb carrying the suffix -BIt (historically the past participle) plus the auxiliary e- in simple past. (11) a. Sakha (tvaP_o18) fon Muyta tiksi-ge bide bara rice Muxta jetty-par one-muit go-smscvs Ihidgds-azr’gim-mit-im mim ini-ber werveure — do-esr2-ross.1se Ise front-oxr.1so bar-bit —boluot-tan tit iis lin gorsronce raftR-as four girl die-stacvs Hit ete, boluot adjd5an-an resrrzrce auxcsti-rss3s0 raft damage-stacvs ‘once | was going to goto the Muxta jetty, and four girls who had gone on the raft ahead of me had died ‘when their raft had broken.” b. Sakha (Efmy_s18) biligin darn udaskaya ——barbittara now Pick quarterR-oat g0-¥st2-70ss.3 is ayo oro bar bit three child spend.night-smicvs go-rsrrr ete averstT-r0ss.3s6 “Even now three children went to the quarter (utside the village), they have gone to spend the night.” 26.4.5.2 Mood Sakha and Dolgan have a large number of moods. Neither the indicative nor the hortative or immediate imperative have a dedicated mood suffix, but both the indicative and the imperative distinguish tense—the indicative seven, as described in the preceding section, and the imperative two: an immediate imperative and a remote imperative. The ‘Table 26.9 Hortative, imperative, and jussive suffixes THE NORTH SIBERIAN TURKIC LANGUAGES, Affirmative Negative $6 MININCL _AUGMAINCL SG MINANCL __AUGMAINCL Hortative mth “layin Wmim Amik {m-tAyly Imperative immediate 0 0 m0 m9 Remote 0 On mAs lm-Aer Op Jussive tin Tin-tar “BATIn “BATIn-LAr jussive is marked by the suffix -Tin and takes predicative (13) Dolgan (Story_Xet_Bear_JP 074) subject-agreement marking, and the singular immediate de -onton dale bariay ete imperative is expressed by the bare verb stem (Table 26.9). rreuthen _further.R_go-rur.rrer_ave-rsrI-ross.36 ‘The hortative marks a distinction between minimal and —kiiler—die-erin——algat-igy ‘augmented inclusive using the plural imperative marker), hhuman-r. house-acc.3e._destroy-runrcr ‘The distinction between an immediate and remote imperative e-te islikely tobe the result ofEvenki influence (Pakendorf 2007: 208-41), ‘The necessitive, which has been copied into Lamunkhin Even (Pakendorf 2009, 20148), is formed with a bimorphemic suffix -IAy-LAxy plus predicative person marking, and the past necessitive is expressed by addition of the auxiliary e~ ‘Another Sakha mood that has been copied not only into Lamunkhin Even, but also into Uéur Evenki (Myreeva 1964: 51) isthe assertive-presumptive marked by the suffix -TA and possessive subject-agreement marking, Although this ‘can mark probability, in spoken discourse it predominantly ‘expresses a strong, emphatic assertion (Pakendorf 2009: 90-2). ‘The voluntative-potential, marked by -(A:)/A plus pre- dicative person marking, expresses possiblity and hope and ‘can have an apprehensive meaning, The 2so/n. affirmative forms have developed into a remote prohibitive (Pakendort ‘and Schalley 2007) in both Sakha and Dolgan (12). (42) Dolgan (story_syn_marriage_APC_148) ‘onno dis-bin tohoyo-bun then say.rs-raco.tse_birth.pole-accse tih-er-egje-git fallcavs-vowsor-201, “Then I say: "Don't knock over my birth pole”. ‘The subjunctive mood is expressed by an analytic construc tion with the lexical verb carrying the future participle and the auxiliary in simple past tense; in Sakha, subject agree- rment is marked variably on the lexical verb (cf. 4a) or on tie auxiliary, whereas in Dolgan itis marked only on the auniliary (13). nuersrl-7058.356 "He would have gone further, would have destroyed people's houses. ‘There are two different conditional forms that mark a dif- ference between unreal and ral conditions. Unreal conditions| are expressed analytically, with the lexical verb carrying the present or the past participle plus possessive person marking, and the auxiliary buol- carrying the conditional suffix -TAr (14a), Real conditions are expressed by the suffix -TAy plus conditional person marking (historically, possessive-marked partitive case endings which themselves developed out of the Old Turkic locative (cf. Pakendorf 2007: 148-9 and refer- ences therein). This is frequently used in Dolgan (14b), and. it also (and more commonly) functions as the predicate of temporal adverbial clauses in both languages. (14) a, Sakha (XatR_095) criyj cepieki-ik ep-pitin abit light-sov say-estercrnoss. 230 buol-lar Dayar itiam nvcrsrl-t0ss336 perhaps. take-rurlsc ete die-bit aux-coworten3sc] _say-rsr2fmnen3sc] “If you'd said abit less, I might have taken it, he said’ b. Dolgan (Story_cure_syn_IMA_20) ‘onton buollayina ara biar-ig then re black liver-ross.2s6 iri-day-ina tuo beill-cono-3s6. rea ‘And then if your liver hurts... 439 BRIGITTE PAKENDORF AND EUGENIE STAPERT ‘There are furthermore two moods with epistemic functions, the presumptive and the assertive. The non-past and past presumptive are expressed by the present and past parti- ciple, respectively, carrying possessive person marking plus the auxiliary buol- marked for future third singular, eg, titer buol-uo [wait-re.rce-oss.3sc Aux-rT.35 ‘he is probably waiting, ist-bk-kt buol-uo [hear-rstrtcr-2m. aux ras} ‘you probably heard’, In addition, presumption can be expressed with the particle ini ‘probably’. The affirmative assertive is marked by the suffix -Ih plus predicative person marking (15a); the negative is expressed by the contracted form of the third singular future 1A plus the negative element suoy/huoy carrying predicative subject ‘agreement (15t). These forms express a strong assumption or decision, (05) a. Sakha (YmyE_158) oly barithi-gin inthat.case go-assrae0.2s6 “im that case you can go of course, b, Sakha (XatR_082) huog, hia ec take-rur3so ‘Saying no, I won't take it,. dien say.seacvs Iruoy-pun oo-oRED.AsG 26.453. Aspect/Aktionsart ‘Asis common in Turklc languages (Anderson 2004; johanson 1996b: 42), aspect and aktionsart in Sakha and Dolgan are mainly expressed via auxiliary verb constructions (16), in ‘hich the lexical verb carries a converb marker, while TAM and subject agreement are expressed on the auxiliary. ‘The auxiliaries also function as independent lexical verbs (Table 26.10). (16) Dolgan (Story_Syn_Holiday_MSA_04) onton bara tura-bit then go-snacve oun-ens-ti “then we go [for along time]’ While in Dolgan auxiliary verb constructions with hiit- largely still express a meaning of motion, in Sakha they have purely aspectual meaning, as seen by their occurring with stative verbs such as ‘live, ‘be calle, “be seen’, or “have a population’ ‘The unaccomplished aspect is expressed by analytical constructions consisting of the lexical verb carrying the simultaneous converb suffix and the lexeme ik carrying subject-agreement marking; in the past tense, this com- bines with the past tense-marked auailiary e-, eg. ati ilik- pin food-vazsiu.cve not yet-reeo.1sc} Thave not yet cooked’, Iel-eiikterine[come-swa.cvenot.yet- ile-le:-'to work’ from adjectives, e.g, ras ‘clean’ » rss “to clean’; from numerals, eg. uon ‘ten’ » uon-na:- ‘do ten times’ ikkis ‘second! » ikkis-te:- ‘do for the second time’; or from pronouns, eg, twoy ‘what’ > tuoy-ta:-‘do what” (Saka hestative), kim ‘who/what’ > kim-ne:- ‘do what’ (Dolgan hesitative), zanna ‘where’ > yannacla:-'go where. In Dolgan, _most loanverbs from Russian are integrated withthe help of this sufix, eg, zwonit-ta:- < zvont'‘to phone’, zameég-da:-« zameéat” ‘notice’. 26.4.6.2 Nominal derivation ‘Among the most productive nominalizing suffixes in Sakha and Dolgan is-Sit, which derives nouns referring toa person, ‘who performs an action on the object designated by the base noun, eg. bul-eut ‘hunter’ < bult ‘prey’, ngy-sit‘milker -ingy ‘cow ile-hit ‘worker’ = ile work’. Among the suffixes deriving nouns from verbs both Sakha and Dolgan have -:, eg, xay-i: ‘prison’ < zay- ‘lock up’. Furthermore, in Sakha bbut not Dolgan the suffix -(A:)hin derives abstract nouns from verbs eg, batta:-‘to push, squeeze’ > battachin ‘pres- sure’, Lastly, the suffix -(A:&l, which expresses habitual aspect (see Section 26.4.3), also functions as an agent nom- inalizer in Sakha (but not in Dolgan), eg. kimloh 3: “helper

You might also like