CC BY-NC-ND
59
Kirill Babaev
(Moscow, Russian State University for the Humanities)
Once Again on the Comparison of Personal Pronouns in Proto-Languages
The article discusses the importance of comparison of personal pronouns for studies dealing with long-range linguistic relationship. It is recognized that individual similarities between pronouns reconstructed for various low-level protolanguages may be ascribed to chance resemblance or borrowing, but the same is much less probable for comparisons between entire paradigmatic systems of pronouns. Based on a brief survey of the principal pronominal forms and markers in most of the protolanguages reconstructed for the world's language families, it is asserted that many of the current hypotheses on macrofamilies, such as Nostratic, Sino-Caucasian, Austric, and others, are indeed supported by such paradigmatic similarities.
For those scholars who engage in long-range comparison of the world's languages, personal pronouns have been a viable source of genetic data. Indeed, as was already noticed by the first Indo-Europeanists nearly two hundred years ago, personal markers, both syntactically independent and bound with verbs or nouns, are among the most stable elements of the language's morphology. The Ukrainian pronominal form мене 'me' and the Irish form tu 'thou' both sound very much the same as they must have sounded some 6000 years ago in the speech of Indo-Europeans, and bear the same meanings. Three personal pronominal forms are deservedly placed in the Swadesh list of the one hundred most stable lexical items in human vocabulary.
Moreover, personal pronouns are relatively rarely borrowed between languages, especially from a genetically different language stocks. It would be an extreme thing to say that they cannot be borrowed at all — since they can and are, as shown by a number of cases — but this happens much more rarely than with nominal or verbal lexical items. Instances of the first person singular pronoun 'I' being a loanword in a language are extremely scarce, probably no more than ten reliable examples known so far, and almost all of them can be explained by some peculiar sociolinguistic (or grammatical) conditions.
These two unique characteristics of personal pronouns — internal stability and resistance to external borrowing — make the analysis of personal pronouns, both lexical and morphological, extremely important for comparative linguistic studies.
Can personal pronouns serve as solid proof or, at least, a considerable heuristic argument in proposing distant genetic relationship between languages? This question has long been a subject of discussion in various fields of linguistics. Personal markers were a strong argument in favor of the Altaic theory advocated by numerous authors starting from J.-F. Strahlenberg ([1730]). They were one of the keystones in postulating the Nostratic macrofamily by H. Pedersen, and the Amerind family in the Americas by J. Greenberg ([2000]). In Papuan linguistics, personal markers arguably remain as the most important factor in positing linguistic relationship and elaborating classifications, most of which are not yet supported by conventional lexical comparative analysis ([Foley 1986: 3]). Pronouns are widely drawn upon elsewhere to prove genetic relationship, and, therefore, methodologically a positive answer to the above question would help us to validate a number of macrofamily hypotheses still under discussion.
Unsurprisingly, opponents of long-distance relationship claims have their own reaction. A number of respectable linguists have challenged the argument that personal pronouns may be essential for long-range genetic comparison. The most widespread and easily understandable objections are quoted as follows:
1. Personal pronouns can be borrowed. Well known examples of this kind are drawn mainly from closely related languages, like English, where they is a Scandinavian loanword, or Dravidian, where Kolami is claimed to have borrowed the 2nd person singular niv from Old Telugu ([Krishna-murti 2001: 91]). More distant languages also demonstrate sporadic examples: thus, Micronesian Chamorro borrowed Spanish yo 'I' ([Topping 1973: 107]), while Ambonese Malay adopted ose 'thou' from Portuguese Voce 'you (Sg.)'. Languages of New Guinea tend to "exchange" certain personal pronouns quite often ([Laycock - Z'Graggen 1975: 732, 737]), though this phenomenon has not yet been sufficiently studied.
Journal of Language Relationship • Вопросы языкового родства • 1 (2009) • pp. 37 - 48
2. Personal pronouns are, in fact, mostly very short lexical items usually consisting of CV roots which are hard to compare, because the chance of coincidence is very high ([Campbell 2003]). Moreover, it has been tentatively shown that pronominal patterns tend to crystallize an opposition of a nasal to a stop consonant as the basic phoneme of the pronominal root ([Nichols 2003]). Even if this hypothesis is not entirely correct, this pattern works in a great number of the world's languages.
In our opinion, the most responsible answer to both of these points is not to try and prove the opposite; rather it is to use personal pronouns for language comparison not just as isolated lexical items, but as a paradigmatic subsystem of morphology.
In the majority of the world's languages, pronouns constitute a balanced paradigm. The types of such paradigms are very well presented by M. Cysouw ([Cysouw 2003]), the most basic being the opposition of the two pronouns of the 1st and 2nd persons singular: the pronominal system is limited by this binary opposition in a number of languages, e. g. Golin and Salt-Yui in New Guinea ([Foley 1986: 70]). Other structural samples include the number opposition, with four basic members like I— we, thou — you. Still other systems are augmented even further, including 3rd person pronouns, usually adopted from the demonstrative domain; the introduction of the category of clusivity (inclusive / exclusive pronouns of the 1st person plural); more number categories (dual / trial / paucal &c.). Finally, personal pronouns can be declined for case, as in most Indo-European languages, adding even more forms to the pronominal paradigm. Nominative and oblique forms are frequently derived from suppletive roots, e. g., the Indo-European 1st person singular nominative *egHo(m) vs. oblique *me-.
It seems clear that in order to prove that the languages are genetically related, it is the entire pronominal paradigm that should be compared rather than isolated personal pronouns as disconnected lexical items. Adopting the latter approach, we may certainly encounter loanwords or mere coincidences, like Indo-European *me 'me' = Sumerian (Eme-sal) me 'id.'. However, comparison of entire paradigms drastically decreases the probability of either chance similarity or borrowing.
In those languages where personal pronouns consitute paradigms, there are no known cases of borrowing the whole paradigm from another language, provided exceptional situations with pidgin and mixed languages are excluded1 ([Thomason - Everett 2005]).
One of the most recent claims ([Campbell - Poser 2008: 214]) is that the paradigmatic approach to comparison is equally invalid for long-range comparison purposes. The main reason here is that, as the authors claim, personal pronouns are essentially iconic "phonosymbols", an assertion that is as insufficiently grounded as it is simply untrue; if it were so, pronominal comparison would be useless altogether, since pronouns would tend to share the same phonological features in all language families — which is clearly not the case, as can be easily seen from the chart below. Campbell & Poser also claim that pronouns are subject to analogical reformations, which is obviously correct (we state the same in [Бабаев 2008: 209]). However, analogical reformations usually have internal reasons and should not be expected to make one language's paradigm closer to that of another.
Paradigm comparison is also important for historical linguistics, because in long-range studies any comparison of materially short cognates with a CV or VCV structure, where the consonant is the only reliable sound to be compared, is dubious by definition. Therefore, comparison between proto-languages, aimed at reconstructing a 10 000-year-old or older linguistic situation, should be made between reconstructed cognate paradigms of personal pronouns rather than between personal pronouns as lexical roots.
Such attempts have, in fact, been made, some of them quite successful. It was noticed long ago that the pattern exhibited by personal pronoun paradigms in a number of language families in Northern Eurasia is remarkably similar; this pattern can be described as *mV for the 1s person vs. *tV for the 2nd person. The Nostratic/Eurasiatic theory proposes a genetic link between languages that present this pattern, namely, Indo-European, Altaic, Uralic, Yukaghir and, most probably, Chukchi-Kamchatkan. Indeed, (a) the linguistic probability of borrowing this pattern is practically zero; (b) the statistical probability of a chance similarity in such a geographical cluster is also close to zero; and c) a pattern like this is not attested for any other proto-language of any other family on Earth. This makes the "pronominal factor" one of the strongest bits of grammatical evidence for the Nostratic hypothesis.
Another example is the American pronominal pattern *na- (1st pers.) vs. *mi (2n pers.), first noted by A. Trombetti ([1905]) in a great number of the languages of both Americas and later de-
1 One well known case is that of the Thai language, which uses Chinese and English personal pronouns in specific sociolinguistic contexts ([Cooke 1968: 11 - 12]). However, they do not replace the common Thai personal pronouns.
veloped by J. Greenberg into one of the key arguments of his Amerind theory. Again, the spreading of the pattern is extremely wide, and the chance of borrowing or coincidence is lower than low.
The present article is inspired by a tentative list of reconstructed pronominal forms for the first two persons in the world's languages presented by M. Ruhlen ([1994])2. However, we find it necessary to continue Ruhlen's effort by substantially adding to the sketch pencilled by his pioneering work. Our task will be to present an updated and accurate comparative chart of personal pronoun paradigms, reconstructed for proto-languages of most of the world's language families, to be subject to further analysis.
To make our results as precise as possible, we will rely on the most recent data collected by scholars, especially concerning languages of Africa and the Pacific region. Although many families still lack thorough proto-language reconstructions, most of their pronominal systems have been at least tentatively sketched out since Ruhlen's work was published fourteen years ago. Apart from independent pronominal forms, we will try to include bound verbal and nominal affixes denoting person as well, provided they differ from the former.
Finally, it is necessary to give the appropriate reference links to etymological dictionaries and/or comparative grammars for each reconstructed form that we quote for a proto-language. Unfortunately, no such references are given in Ruhlen's list, which makes it harder to resolve all the problems that may emerge during one's analysis of it. We have attempted to find as many links as possible for the reconstructions, bearing in mind that for a number of language families (African for example) no comparative work exist at all, and for these cases we make our own tentative reconstruction based on the available dictionaries and word lists (these reconstructions are marked by a double asterisk).
The chart below will provide sufficient material for such an analysis. Along with the usual pronominal forms reconstructed for the family's proto-language, we also mark a number of peculiar forms from separate languages (or language groups) of the family in case they differ from the main family pattern — e. g., Svan forms are placed alongside the reconstructed Proto-Kartvelian ones, because Svan is considered the most morphologically archaic dialect of the family, and a number of judgments on Proto-Kartvelian pronouns has been made based exclusively on very peculiar Svan forms.
We have attempted to make our chart below as taxonomically correct as possible. The majority of the families listed go back to proto-languages which should have existed at roughly the same chronological period: 6 to 8 thousand years ago, considered by many of today's mainstream linguistics as the limit for comparative reconstructions. And, although a number of macro-families have also been recognized in the last decades, for some of them it is not always easy to reconstruct a single system of personal pronouns. Thus, the Afrasian pronominal reconstructions are still under serious discussion. Niger-Congo and Austro-Asiatic may be taken as even better examples: these two macrofamilies seem to be well established and widely accepted, but the personal pronoun systems for their proto-languages have not yet been verified. Therefore, we will only demonstrate pronominal forms for their subfamily proto-languages: Mande, Gur, Dogon, &c. and Munda and Mon-Khmer respectively. The Australian macrofa-mily is so far an unverified hypothesis (although widely accepted), and the New Guinean 'phyla' are even less so ([Foley 1986]). The most complicated situation lies in the Americas: the classification of solidly reconstructed families includes more than 60 taxonomic units ([Campbell - Mithun 1979: 39 -46]), and, although it has been reduced to around a dozen macrofamilies by Greenberg and Ruhlen ([2007: 6]), their theory is not universally accepted and still needs further verification.
The above chart, with its schematic information, cannot even pretend to be perfect, since each form in this matrix is subject to long and deep analysis per se. But even this sketchy picture still allows to make a number of interesting conclusions, based on the patterns of personal pronoun paradigms.
1. The languages of northern Eurasia show a pattern "M—T" which is found nowhere else in the world on the family proto-language level. Namely, these are Indo-European, Uralic, Yukaghir, Altaic and Chukchi-Kamchatkan languages. This phenomenon has been identified long ago and constitutes a part of the morphological evidence for the Nostratic hypothesis.
2. A similar pattern "M—S" is present in Indo-European, Kartvelian, Altaic and, arguably, Chukchi-Kamchatkan, where the form *yas 'thou' is of an unclear origin. In Indo-European, the distribution of
2 Editors' note: A similar list of 1st and 2nd person pronouns has been earlier produced by S. L. Nikolaev during his talk on the Nostratic Seminar in Moscow in April of 1985. The list was, however, much shorter, including only those major families across Eurasia, Africa, and North America for which at least some sort of protolanguage reconstruction had already been established or sketched.
"M—T" vs. "M—S" is morphological (the former is found in the pronominal system, the latter in the verb system), while in Altaic the variability is dialectal (Mongolic vs. Turkic and Tungus-Manchu). In both Indo-European and Altaic, however, the two systems are projected onto the proto-language level.
It is necessary to add that Nostratic languages also share the case paradigm for the 1st person pronoun: the form *mVn(V)- may be reconstructed for the Nostratic proto-language on the basis of Indo-European, Altaic, Uralic and Kartvelian languages ([Иллич-Свитыч 1971: 6]; [Бабаев 2008]).
Chart 1. Proto-language Personal Pronouns
Language Family 'I' 'Thou' 'We ' 'You (Pl.)'
Indo-European3 n *eg'ho(m) obl *me(ne) v st *-H- *tü, t(u)e v *-s- n *me(s), *we(i)-obl *ne/o- n *iü-obl *we-v *-te
Uralic4 n *mE obl *mEnV v st *-k n *tE obl *tEnV *mE *tE
Yukaghir5 mit tit met tet
Altaic6 n *bi obl *mi-n-, ? *ya n *si, *t'i obl *si-n-, ? *na *ba / *bu obl. *ma-n- ~ *mu-n- *t'a, *su obl. *su-n-
Turkic n *bi obl *män n *si obl *sän v prt *-y *bi-r v prt *-k *si-r
Mongolian obl *nama- ci < *t'i obl man- obl tan-
Korean na < *ya, ce ne <*na ne-huy
Japanese *a < *ya, wa(re) na(re)
Kartvelian7 *me(n) v subj *(x)w- *si v subj *x-v obj *g- *cwe(n) *tkve(n)
Svan8 excl naj / näj v incl subj lv incl obj gw-
Dravidian9 n *nyän obl*yan-v abs *-kV n *nin obl *nin-v abs *-ti n excl *nyäm obl excl *yam-n incl *näm obl incl *nam- n *nim obl *nim-
Chukchi-Kamchatkan10 *y3-m *yd-s obl *J9-n muri turi
Eskimo (W)11 rel -ma abs -qa rel -pdk abs -n rel -m-ta abs -(x)put/-wut v du -(x)puk v rel -pssi v abs -(x)si v du -(x)tsk
Elamite12 v (h)u obl un abs -k nu, ni v -(a)t(i) incl ela, elu excl nika, niku num
Chadic13 I *?an-i II *?i(ya) masc I *ka, II *ku fem *ki(m) incl *muni excl *na, *?yina *kuni, *kuna
3 [BEEKES 1995: 207 - 211].
4 [Redei 1988]; [Хайду 1985: 225, 234]; [Honti 1993: 124]; [Хелимский 1982].
5 [Николаева - ХЕлимский 1997: 166].
6 [EDAL 225, 1237]; [ITABASHI 1998: 130 - 135].
7 [klimov 1998], [Шмидт 1984].
8 [TUITE 1997: 18 - 26].
9 [DED 5160]; [Старостин 2006].
10 [МУДРАК 2000: 39 - 40, 97, 145 - 146].
11 [Bergsland 1986]; [МЕновщиков 1997].
12 [Дьяконов 1979].
13 [BLAZEK 1991: 37].
Language Family 'I' 'Thou' 'We' 'You (Pl.)'
Berber14 subj *9nakkw obj *Ty, *u(w) subj masc *kayy, fem *kamm obj masc *sk, ik, fem *am *na- *-tum-
Cushitic13 subj *?ani, obj *yi / *yu subj *(?an)-ti obj *ku / *ki subj *muni obj *na *(?an)-tinV
Omotic13 N *ta, *na-, S *?i N *ni, S *ya N incl *ni, excl *nu S *wV N *?antu, S *yV
13 Semitic *(?an)-äku, *(?an)-ä masc *(?an)-ta fem *(?an)-ti *hina, *nahnu/a, *na/i- masc *(?an)-tumu fem *(?an)-tina
Egyptian13 ('in)-k, -j masc -k, kw fem tm(t), -t ('in)-n tn
Sino-Tibetan15 *yä *nay, *kwV- — —
North Caucasian16 I (n) *zo (E *zö-, W *sa) II (obl) *nV/ *mV I *wo (E *wö- / *gwd, WC *wa) WC fem *ba II *yu-, *ywV- EC incl *Lä: EC excl *(nV)zi WC *ha EC *zwe WC *swa
Hurro-Urartian17 *(j)es- / *so- v -f- *we-v -u/o- *(j)es- / *so- *we-
Yeniseian18 I *?aj, II *b- I *?aw, II *kV- / ?Vk- *?ajay *?awoy
Na-Dene19 *swT / *xwT *ni, *wT (Tlingit) *na-, *nu-*ta-, *tu- *?uX-
Burushaski20 ja poss a- un obl gu- / gö mi ma
Basque21 ni v trans *-da- (h)i v trans masc *-ga- gu zu
Sumerian22 yae poss -yu *mV (?) zae poss -zu midi poss -me mizi, -ze
Nahali23 jo obl ey- ne jo, tye-ko ne(ko)
Nivkh24 ni ci excl nin du me-incl mer, mir cin
Ainu25 ku- obj 'en- / -'en- > e- excl abs ci-, -as-excl obj 'un-incl subj a-incl obj 'i- 'eci-, es-
Thai-Kadai26 *k(a)u *m(a)u rau, *tu miy, suu
14 [Дьяконов 1967: 218 - 227]; [Blazek 1991].
15 [Benedict 1972: 93]; [Starostin 1984].
16 [Starostin 1984]; [Nichols 2003: 293].
17 [Diakonoff - Starostin 1986: 81 - 82].
18 [Starostin 1984].
19 [Bengtson 1992].
20 [Berger 1974: 24].
21 [Trask 1997: 218].
22 [Дьяконов 1967: 62]; [Канева 2006: 59 - 61].
23 [Shafer 1941: 363]; [pinnow 1966: 189].
24 [Gruzdeva 1998].
25 [Vovin 1993: 79, 84, 106, 144]; [shibatani 1990: 25 - 31, 45 - 50].
26 [Benedict 1942: 586]; [benedict 1975: 203 - 208]; [li fang-kuei 1977: 361 - 363].
Language Family 'I' 'Thou' 'We' 'You (Pl.)'
Austronesian27 *(w)'aku / *ku *kaw, *(ka)su excl * '(k)ami incl *(k)ita *(ka)mu
Mon-Khmer28 *(V)ko / *ki, *?Vn / *?Vy *mi? / *mVj *hi? incl *bsn du *?a? *(b)na?, *pi?
Munda29 *iy *me excl *le incl *bi(n) *pe
Miao-Yao30 *ko(u), *way Yao * ?ja *koy, *kam Yao * mway *(m)pV Miao du *mu Yao *nV-
Andaman31 do(l)- / tu-Onge mi poss yu(l)- / yo- mu(l)-et- yu(l)- / yo-
Kusunda32 tsi tok nu nok
Papuan (general types)33 I nV II kV / gV / yV III tV / dV IkV/gV/yV II ma III na I ni II mV / pV III ki / ti I ki II mi III nik
West Papuan34 *tV *nV
Torricelli35 *k- *yV *t-,*k- *m- / *p- *y-
Trans- N. Guinean36 *na *ka *ni *ki
Australian37 *ya- *yin incl *yali excl *yay *nhurra
Algic38 39 *nV / *-Vn, v *-Vk v obj *-Vy *-Vm(?), *-Vt v obj *-Vs *-nän *-wäw
Keresiouan *(V)nä-
Hokan *nya / *nyi *mi / *ma
Penutian *n- *mV
Aztec-Tano40 *ne? *?eme *ta- *?eme
Chibcha *na-sV *mue-ya(nV)
Quechua *nu-qa *qa-m
Panoan *mi *no *mato
Nilo-Saharan41 *ägä *äh *ini incl *äna excl *am *wem
Kordofanian42 *ßi *ya excl *-r(i)-incl *-n(a)- *-ß(a)-
Atlantic43 *mi obj *ne *mo obj * 'i, *mV *tV (ceB.) *sV ? *nV
Mande (S)44 *$, *mä *i / *e, *6V excl *yi/5 incl *koa *kä
27 [ZoRC 1995: 1169, 1191]; [BENEDICT 1975: 203 - 209].
28 [CABATON 1905: 336 - 337]; [SHORTO 2006: 69]; [PINNOW 1966: 184]; [HAYES 1992: 161, 168]; [PEIROS 1998].
29 [PINNOW 1966: 166 - 167].
30 [Benedict 1975: 203, 205]; [Peiros 1998].
31 [Portman 1887: 5]; [Blevins 2007: 167]; [Burenhult 1996].
32 [Watters 2006: 44].
33 [WURM 1975: 191].
34 [CAPELL 1975: 678].
35 [LAYCOCK 1975: 768 - 770].
36 [Foley 2000].
37 [BLAKE 1979: 31 - 69]; [BLAKE 1988: 7]; [EVANS 1988: 103]; [EVANS 2003: 19]; [HARVEY 2003].
38 [PROULX 1985: 84 - 87].
39 American Indian language forms below are based on [Greenberg - Ruhlen 2007: 121 - 123, 229 - 230], unless footnoted.
40 [STEELE 1979: 447].
41 [EHRET 2001: 225 - 231].
42 [Schadeberg 1981: 182 - 184].
43 [Pozdniakov - Segerer 2004].
44 [BM£PHH 2006].
Language Family 'I' Thou' We' 'You (Pl.)'
Ijo45 *a, *i/s *i *W3 *o
Dogon46 mi poss n u- / uwV be, eme e poss ebV
Gur-Adamaua47 *mi / *ma *mo *-a(n) *-u(n) / *-ui(n)
**Gur48 *ma *u / *bV *tV *i, *nV
Ubangi49 *mi *ms *?ele *?sns
**Kwa50 subj *N obj *me *o, *wo *ne ? *mo / *mu
**Benue-Congo subj *N obj *mi *o, *u excl *tV
**Bantoid *N obj *me *a, (obj?) *wo *se *ne, *we
Bantu51 subj *Ni-obj *me subj *u / *we obj *ku *tu *sV *nV *mu
Khoisan52 N *mV, S *yV C *ti- / *ta- *a excl *e incl (du) *mV *-u
3. Another common paradigm, shared by the languages grouped within the Nostratic/Eurasiatic hypothesis, is the "K—M" pattern for the 1st person stative (intransitive) vs. relative (transitive). The Uralic intransitive opposition of the 1st person markers *-k (intransitive) and *-m (transitive) ([Xehhmckhh 1982]) is usually compared with the Turkic preterite (1st pl.) *-k and present *-m, and the Eskimo absolu-tive *-qa together with relative *-ma. The Hittite "series" of verb conjugation in *-hi and *-mi, usually regarded as the ancient Indo-European opposition of stative and transitive, might also be derived from the Nostratic type, since there are clues for the velar (or postvelar) origin of the Indo-European laryngeal pho-neme(s). The Nostratic postvelar form was proposed by Illich-Svitych ([Hhhhh-Cbhtbih 1971: 149]).
The syntactic function-related "K—M" opposition of the 1st person forms finds a similar correlation in the 2nd person, where the stative counterpart of "K" is usually "T". The pattern "K—T" in its stative aspect is found in Dravidian appellatives ([Zvelebil 1990: 35 -36]), in Indo-European (where *-t- is the 2nd person perfective affix in the same paradigm as the 1st person *H), as well as in Elamite and Afrasian families, which are usually considered as the closest relatives of Nostratic or Eurasiatic by long-range comparativists.
Again, it should be noted that this particular system is never reconstructed for any other proto-language around the world. This uniqueness, hardly representing the result of cross-borrowing, cannot be interpreted as anything other than a trace of genetic kinship between the languages of Eurasia.
4. Further evidence for another reconstructed macrofamily, Sino-Caucasian, or even broader, Dene-Caucasian, said to include North Caucasian, Sino-Tibetan, Yeniseian, Na-Dene languages and possibly some ancient tongues of Europe and the Near East (Hurrian, Urartian, Hatti, Basque) as well, is also found in the chart above. The pattern "S/Z— W" is easily traced in North Caucasian, Yeniseian, Na-Dene, and Hurro-Urartian languages. The primary consonant phoneme in the 1st person is not that clear, but it is definitely a coronal sibilant/affricate. Nahali data might be considered here as well.
It is interesting that the Sino-Tibetan proto-language paradigm does not fit this pattern. This discord might, however, be explained by the hypothesis that Proto-Sino-Caucasian, in fact, had two suppletive pronominal paradigms. Starostin ([1984]) explicitly showed that the second set of pronouns in North Caucasian should be reconstructed as *nV / *mV (1st person sg.) - *yu / *ywV (2nd person sg.), which correlates well with such Sino-Tibetan forms as, respectively, *ya and *kwV, as
45 [Williamson 2004].
46 [Bendor-Samuel & al. 1989: 176]; [Segerer 2002 -20 07].
47 [Boyd 1989: 207].
48 [Beacham 1968: 93]; [Casimir 1988: 57]; [Prost 1974: 646]; [Rennison 2000].
49 [Monino 1995].
50 [Segerer 2002 -2007].
51 [MEEUSSEN 1965]; [GUTHRIE 1967 - 1971, IV: 226 - 240]; [АКСЁНОВА - ТОПОРОВА 1990: 146-150; [NURSE -Philippson 2003].
52 [Greenberg 1963: 70 -74]; [Starostin 2003]; [Guldemann 2005].
well as Yeniseian prefixed pronouns *b- and *kV-. The Burushaski forms might also fit here, as proposed by Bengtson ([1997]) in his cognate list for a theory of Burushaski as a Sino-Caucasian language. It should be noted that, typologically, both Yeniseian and Burushaski share the two-paradigm system of personal pronouns with North Caucasian.
5. Languages of South-Eastern Asia and the Pacific seem to share the pattern "K—M" (for the opposition between the 1st and 2nd singular persons), as shown by Tai-Kadai, Miao-Yao, Mon-Khmer, and Austronesian data. These are the families grouped by P. Benedict into the hypothetical Austric macro-family. Additional similarities can be seen between single forms of pronouns in Austric languages (like *Vy 'I'); some of them are shared by Ainu, which is also considered to be related to Austric ([Bengtson - Blazek 2000]). The protoforms of pronouns for the Austro-Asiatic protolanguage are still foggy.
It is curious that the same "K—M" pattern also characterizes a number of languages in New Guinea: this is the type identified by Wurm ([1975]) as "Set II" of the Papuan personal pronouns. It is emphasized that set II is supposed to mark the most archaic strata of human language in New Guinea; nowadays, it is used in languages of the Torricelli phylum and those languages of the Trans-New-Guinea phylum which show a strong substratum influence. Although genetic relations between Austric languages of Asia and the Pacific have not yet been thoroughly studied, this coincidence might become a hint for any further studies in this regard.
6. Languages viewed under the Niger-Congo macrofamily umbrella show distinct similarities in the pronominal system. A pattern "M— W" for the 1st and 2nd person singular pronouns is displayed by most of the language families of Niger-Congo. Moreover, at least one 1st person plural pronoun can probably be reconstructed as *tV based on Bantu, Atlantic, and Gur languages. It seems to us that the Niger-Congo system actually consisted of two series of pronouns, one subject, one object (oblique), just as it looks like in numerous languages of practically all the families. The other 1st person singular pronoun *N (representing a variable syllabic nasal) most probably denoted subject forms opposed to the object (oblique) *mi.
Niger-Congo forms tend to show some basic similarity in the pattern (N [any nasal] — W), in the line of the hypothesis of their common genetic roots ([Blench 2007]). Kordofanian and Ijoid languages in this regard lie closer to Nilo-Saharan than to common Niger-Congo patterns. But there is still obvious lack in solid basis for the proto-forms, so these clues should be used carefully before reliable reconstructions are made for all sub-families of Niger-Congo.
7. Another striking similarity ("T—N", basically 1st pers. *tV vs. 2nd pers. *nV, with plural forms in *-k) lies between Kusunda, a language isolate of a forest gatherers' tribe of central Nepal, and the West Papuan languages whose speakers are thought to represent the latest wave of Papuan migration to the island of New Guinea ([Wurm & al. 1975: 947]). More anthropological and cultural similarities can be found between Papuans and Kusunda, so the two languages do deserve deeper linguistic comparison ([Whitehouse & al. 2004]).
It goes without saying that the more we delve into language prehistory, the more observations like these will be formulated. Non-evident cognates may well be hidden behind phonetic changes, impossible to uncover without a thorough analysis.
However, some more general comments might be even more valuable from the point of view of linguistic typology.
It would be incorrect to call the choice of phonemes for personal pronouns a linguistic universal. Neither in proto-languages nor in living ones does there seem to be any inclination towards a "resonant — stop" phonematic distribution like the one proposed by J. Nichols. From a statistical point of view, it is obvious that there are plenty of proto-languages with both the 1st and the 2nd person pronouns based on nasals (even such a phenomenon as sharing the same nasal sounds for both persons is quite common in living languages, e. g. Korean), or both formed by stop consonants (Miao, Afrasian, Aus-tronesian, Kusunda, &c.). There are also clear cases when both basic pronouns are built on the basis of sibilants. Although this statistical survey is relatively small, from a certain point of view it is more precise than one that would include all the living languages of the world, since the disproportionality in absolute language numbers among the various families would spoil the picture in the latter case. Thus, several hundred Austronesian languages will show phonetic values that will decisively overwhelm those of the single Basque language, but only because they are closely interrelated, while Basque is an isolate.
Subsequently, it is also incorrect to claim that "chance resemblance" can play an important part in pronominal comparison between languages of different families. There are absolutely no coinci-
dences in paradigm patterns between the languages which are not thought to be genetically related by modern long-range comparativists. One can certainly find similarities that may be hard to explain (like the above-mentioned Basque and Sumerian, or Papuan and Nivkh which share the "N—K" pattern). However, we are dealing here not just with single isolates instead of family proto-languages, but also with vastly diverse chronological periods: Basque and Sumerian data are separated by a time gap of at least 3000 years. It is only the comparison between chronologically close proto-languages of large families that may show us some clues in how to reconstruct deep linguistic prehistory.
References
Beacham 1968 — C. G. Beacham. The Phonology and Morphology of Yom. A Thesis Submitted to
the Faculty of the Hartford Seminary Foundation. Hartford. Beekes 1995 — R. Beekes. Comparative Indo-European Linguistics: an Introduction. Amsterdam -
Philadelphia: John Benjamins. Bendor-Samuel & al. 1989 — J. Bendür-Samuel, E. Olsen, A. White. Dogon / J. Bendor-Samuel
(ed.) II Niger-Congo Languages. NY - London: Lanham; pp. 169 - 177. Benedict 1942 — P. Benedict. Thai, Kadai and Indonesian: a New Alignment in Southeastern Asia II
American Anthropologist, vol. 44, N°. 4; pp. 576 - 601. Benedict 1972 — P. Benedict. Sino-Tibetan: a Conspectus. Cambridge. Benedict 1975 — P. Benedict. Austro-Tai Language and Culture. The Hague. Bengtson 1992 — J. Bengtsün. Notes on Sino-Caucasian / V. Shevoroshkin (ed.) II Dene-Sino-Cau-
casian Languages. Bochum: Brockmeyer; pp. 67 - 129. Bengtson 1997 — J. Bengtsün. Ein Vergleich von Buruschaski und Nordkaukasisch II Georgica, 20; pp. 88 - 94.
Bengtson - Blazek 2000 — J. Bengtsün, V. Blazek. Lexical parallels between Ainu and Austric
and their implications II Archiv Orientální, 68.2; pp. 237 - 258. Berger 1974 — H. Berger. Das Yasin-Burushaski (Werchikwar). Wiesbaden. Bergsland 1986 — K. Bergsland. Comparative Eskimo-Aleut phonology and lexicon II Journal de
la Société Finno-Ougrienne, vol. 80; pp. 63 - 137. Blake 1979 — B. Blake. Australian Case Systems: Some Typological and Historical Observations /
S. A. Wurm (ed.) II Australian Linguistic Studies. Canberra; pp. 323 - 394. Blake 1988 — B. Blake. Redefining Pama-Nyungan II Aboriginal Linguistics, vol. 1; pp. 1 - 90. Blazek 1991 — V. Blazek. The Microsystems of Personal Pronouns in Chadic, Compared with Afro-asiatic II Studia Chadica et Hamitosemitica. Aktes den internazionalen Symposions zur Tschadsprachenforschung, Frankfurt am Main, 6 - 8 Mai 1991. Köln; S. 36 - 57. Blench 2007 — R. Blench. Further Evidence for a Niger-Saharan Macrophylum II Advances in Nilo-Saharan Linguistics. Proceedings of the 8th Nilo-Saharan Linguistics Colloquium. Köln: Rüdiger Köppe Verlag; pp. 12 - 24. Blevins 2007 — J. Blevins. A Long Lost Sister of Proto-Austronesian? Proto-Ongan, Mother of
Jarawa and Onge of the Andaman Islands II Oceanic Linguistics, vol. 46, N°. 1; pp. 155 - 198. Boyd 1989 — R. Boyd. Adamawa-Ubangi / J. Bendor-Samuel (ed.) II Niger-Congo Languages. NY -
London: Lanham; pp. 178 - 215. Burenhult 1996 — N. Burenhult. Deep Linguistic Prehistory with Particular Reference to An-
damanese II Lund University Working Papers, 45; pp. 5 - 24. Cabaton 1905 — A. Cabaton. Dix dialectes indochinois recueillis par Prosper Odend'hal II Journal Asiatique, mars-avril 1905; pp. 265 - 344. Campbell 2003 — L. Campbell. How to Show Languages Are Related: Methods for Distant Genetic Relationship / B. Joseph, R. Janda (eds.) II The Handbook of Historical Linguistics. Oxford: Blackwell; pp. 262 - 282. Campbell - Mithun 1979 — Introduction: North American Indian Historical Linguistics in Current Perspective / L. Campbell, M. Mithun (eds.) II The Languages of Native America: Historical and Comparative Assessment. Austin & London: University of Texas Press; pp. 3 - 69. Campbell - Poser 2008 — L. Campbell, W. Poser. Language Classification. History and Method. Cambridge: Cambridge University Press.
Capell 1975 — A. Capell. The West Papuan Phylum / S. A. Wurm (ed.) II Papuan Languages and the New Guinea Linguistic Scene. (= Pacific Linguistics, Series C, N°. 38). Canberra; pp. 667 - 716.
Casimir 1988 — K. K. Casimir. Lexicon der Tagbana-Sprache. Dissertazion zur Erlangung des Grades eines Doktors der Philosophie an der Fakultät für Linguistik und Literaturwissenschaft an der Universität Bielefeld. Bielefeld.
Cooke 1968 — J. R. Cooke. Pronominal Reference in Thai, Burmese and Vietnamese II University of California Publications in Linguistics, 52.
Cysouw 2003 — M. Cysouw. The Paradygmatic Structure of Person Marking. Oxford: Oxford University Press,.
DED — T. Burrow, M. B. Emeneau. A Dravidian Etymological Dictionary. Oxford, 1984.
Dixon 1980 — R. Dixon. The Languages of Australia. Cambridge: Cambridge University Press.
EDAL — S. Starostin, A. Dybo, O. Mudrak. An Etymological Dictionary of Altaic Languages. HdO Vol. 8-1, 8-2, 8-3. Leiden: Brill, 2003.
Ehret 2001 — C. Ehret. A Historical Comparative Reconstruction of Nilo-Saharan. Köln: Koppe.
Evans 1988 — N. Evans. Arguments for Pama-Nyungan as a Genetic Subgroup, with Particular Reference to Initial Laminalization II Aboriginal Linguistics, 1; pp. 91 - 110.
Evans 2003 — The Non-Pama-Nyungan languages of northern Australia: Comparative Studies of the Continent's Most Linguistically Complex Region / N. Evans (ed.). Canberra: "Pacific Linguistics".
Foley 1986 — W. Foley. The Papuan Languages of New Guinea. Cambridge: Cambridge University Press.
Foley 2000 — W. Foley. The Languages of New Guinea II Annual Review of Anthropology, October 2000, Vol. 29; pp. 357 - 404.
Greenberg 1966 — J. Greenberg. The Languages of Africa. Indiana University, Bloomington. Mouton & C°, The Hague, Netherlands.
Greenberg 2000 — J. Greenberg. Indo-European and Its Closest Relatives: The Eurasiatic Language Family. Vol. 1: Grammar. Stanford: Stanford University Press.
Gruzdeva 1998 — E. Gruzdeva. Nivkh. München.
Guthrie 1967 -1970 — M. Guthrie. Comparative Bantu. An Introduction to the Comparative Linguistics and Pehisotry of the Bantu Languages. Vol. I (1967), II (1971), III (1970), IV (1970). London.
Güldemann, Elderkin 2005 — T. Güldemann, E. Elderkin. On External Genealogical Relationships of the Khoe Family / M. Brenzinger, C. König (eds.) II Khoisan languages and linguistics: the Riezlern symposium 2003. Quellen zur Khoisan-Forschung 17. Köln: Rüdiger Köppe.
Harvey 2003 — M. Harvey. Reconstruction of Pronominals among the Non-Pama-Nyungan Languages / N. Evans (ed.) II The Non-Pama-Nyungan languages of northern Australia: Comparative Studies of the Continent's Most Linguistically Complex Region. Canberra: "Pacific Linguistics"; pp. 475 - 513.
Hayes 1992 — L.V. Hayes. On the track of Austric. Part I: Introduction II MKS XXI; pp. 143 - 177.
Honti 1993 — Laslo Honti. Zur Morphologie ugrischer Personalpronomina II MSFOU 215, 121 - 127, Helsinki.
Itabashi 1998 — Itabashi Yoshizo ^ ^ H]. The Old Japanese Personal Pronouns as an Etymological Problem II UAJ, Vol. 70.
Klimov 1998 — G. A. Klimov. Etymological Dictionary of the Kartvelian Languages. Berlin: Mouton de Gruyter.
Krishnamurti 2001 — Bh. Krishnamurti. Comparative Dravidian Linguistics. Oxford University Press.
Laycock 1975 — D. C. Laycock. The Torricelli Phylum / S. A. Wurm (ed.) II Papuan Languages and the New Guinea Linguistic Scene (= Pacific Linguistics, Series C, N°. 38). Canberra; pp. 767 - 780.
Laycock - Z'Graggen 1975 — D. C. Laycock, J. Z'Graggen. The Sepik-Ramu Phylum / S. A. Wurm (ed.) II Papuan Languages and the New Guinea Linguistic Scene (= Pacific Linguistics, Series C, N°. 38). Canberra; pp. 729 - 763.
Li Fang-Kuei 1977 — Li Fang-Kuei [^ ^ Li Fänggui]. A handbook of Comparative Tai. Honolulu: University Press of Hawaii.
Meeussen 1965 — A. E. Meeussen. Reconstructions grammaticales du bantou. Tervuren: Musée Royal de l'Afrique Centrale (MRAC).
Monino 1995 — Y. MONINO. Le Proto-Gbaya : Essai de linguistique comparative historique sur vingt-et-une langues d'Afrique centrale. Louvain-Paris : Peeters.
Nichols 2003 — J. Nichols. Diversity and Stability in Language / B. Joseph, R. Janda (eds.) II The Handbook of Historical Linguistics. Oxford: Blackwell; pp. 283 - 310.
Nurse - Philippson 2003 — D. Nurse, G. Philippson (eds.) / The Bantu Languages. London - NY: Routledge Language Family Series.
Peiros 1998 — I. Peiros. Comparative Linguistics in Southeast Asia. Canberra: Pacific Linguistics.
Pinnow 1966 — H.-J. Pinnow. A Comparative Study of the Verb in the Munda Languages / N. Zide (ed.) II Studies in Comparative Austroasiatic Linguistics. The Hague: Mouton & C°.; pp. 96 - 193.
Portman 1887 — M. W. Portman. Manual of the Andamanese Languages. Delhi.
Pozdniakov- Segerer 2004 — K. POZDNIAKOV, G. Segerer. Reconstruction des pronoms atlantiques et typologie des systèmes pronominaux II Systèmes de marques personnelles en Afrique. (= Collection «Afrique et Langage», 8); pp. 151 - 162.
Prost 1974 — A. Prost. Le Notre, langue des Notaba dits Bulba IILes langues de l'Atakora. Bulletin de l'IFAN, T. XXXVI, sér. B, n°. 3,; pp. 628 - 659.
Proulx 1985 — P. Proulx. Proto-Algic II: Verbs II International Journal of American Linguistics, Vol. 51, N°. 1. (Jan., 1985); pp. 59 - 93.
RÉdei 1988 — K. Rédei. Uralisches Etymologisches Wörterbuch. Akadémiai Kiado, Budapest.
Rennison 2000 — R. Rennison. Koromfe-English/French/German Dictionary [Forthcoming].
Ruhlen 1994 — M. Ruhlen. First- and Second-Person Pronouns in the World's Languages II On the Origin of Languages: Studies in Linguistic Taxonomy. Stanford: Stanford University Press.
Schadeberg 1981 — T. Schadeberg. The Heiban Group. Vol. 1: The Survey of Kordofanian. Hamburg: Helmut Buske Verlag.
Segerer 2002 - 2007 — G. Segerer. Les marques personnelles dans les langues africaines [base de données en ligne du LLACAN: http://sumale.vjf.cnrs.fr/pronoms].
Shafer 1941 — R. Shafer. Nahali: A Linguistic Study in Paleoethnography II Harvard Journal of Asiatic Studies, v. 34.
Shibatani 1990 — Shibatani Masayoshi # ^ The Languages of Japan. Cambridge: Cambridge University Press.
Shorto 2006 — H. L. Shorto. A Mon-Khmer Comparative Dictionary. Canberra.
Starostin 2003 — G. S. Starostin. A Lexicostatistical Approach towards Reconstructing Proto-Khoi-san IIMother Tongue. VIII; pp. 83 - 128 ["81 - 126" in the on-cover TOC].
Steele 1979 — S. Steele. Uto-Aztecan: An Assessment for Historical and Comparative Linguistics / L. Campbell, M. Mithun (eds.) II The Languages of Native America: Historical and Comparative Assessment. Austin & London: University of Texas Press: pp. 444 - 544.
Strahlenberg 1730 — J.-F. Strahlenberg. Das Nord- und Östliche Theil von Europa und Asia, in so weit solches das gantze Russische Reich mit Siberien und der grossen Tatarey in sich begreifet in einer historisch-geographischen Beschreibung der alten und neuern Zeiten. Lübeck.
Thomason - Everett 2005 — S. Thomason, G. Everett. Pronoun Borrowing II Berkeley Linguistics Society, 27; 301 - 315.
Topping 1973 — D. M. Topping. Chamorro Reference Grammar. Honolulu: Hawaii University Press.
Trask 1997 — L. Trask. The History of Basque. London -NY: Routledge.
Trombetti 1905 — A. Trombetti. L'unità d'origine del linguaggio. Bologna.
Tuite 1997 — K. Tuite. Svan II Languages of the World / Materials, 139. München - Newcastle: LINCOM Europa.
Vovin 1993 — A. Vovin. A Reconstruction of Proto-Ainu. Leiden: Brill.
Watters 2006 — D. Watters. Notes on Kusunda Grammar: a Language Isolate of Nepal II Himalayan Linguistics Archive, 3 (2006); pp. 1 - 182.
Whitehouse & al. — P. Whitehouse, T. Usher, M. Ruhlen, W. S.-Y. Wang. Kusunda: an Indo-Pacific Language in Nepal II PNAS, Vol. 101, N°. 15, April 2004. http ://www.pnas.org/content/ 101/15/5692.full.
Williamson 2004 — K. Williamson. Proto-IjoidReconstructions [Forthcoming].
Wurm 1975 — Personal Pronouns / S. A. Wurm (ed.) II Papuan Languages and the New Guinea
Linguistic Scene. (= Pacific Linguistics, Series C, N°. 38). Canberra; pp. 191 - 217. Wurm & al. — S. A. Wurm, D. C. Laycock, C. L. Voorhoeve, T. E. Dutton. Papuan Linguistic Prehistory, and Past Language Migrations in the New Guinea Area / S. A. Wurm (ed.) II Papuan Languages and the New Guinea Linguistic Scene. (= Pacific Linguistics, Series C, N°. 38). Canberra, 1975; pp. 935 - 960.
Zorc 1995 — D. Zorc. A Glossary of Austronesian Reconstructions / D. T. Tryon (ed.) II Comparative Austronesian Dictionary, an Introduction to Austronesian Studies. Part I, fasc. 2. Berlin; pp. 1105 - 1197. Zvelebil 1990 — K. Zvelebil. Dravidian Linguistics : An Introduction. Pondicherry.
Аксёнова - Топорова 1990 — И. С. Аксёнова, И. Н. Топорова. Введение в бантуистику (Имя.
Глагол.) [An Introduction to Bantu Linguistics: Noun, Verb). М,: «Наука». Бабаев 2008 — К. В. Бабаев. Происхождение индоевропейских показателей лица [The Origins of the Indo-European Personal Markers). М. - Калуга: «Эйдос». Выдрин 2006 — В. Ф. Выдрин. Личные местоимения в южных языках манде [Personal Pronouns in Southern Mande) II Труды Института лингвистических исследований. Т. 2, Ч. 2. СПб: «Наука»; стр. 327 - 413. Дьяконов 1979 — И. М. Дьяконов. Эламский язык [The Elamite Language) II Языки Азии и
Африки. Т. III. М.; стр. 37 - 49. Дьяконов 1967 — И. М. Дьяконов. Языки древней Передней Азии [Languages of the Ancient Southwest Asia). М.
Иллич-Свитыч 1971 — В. М. Иллич-Свитыч. Опыт сравнения ностратических языков (семи-тохамитский, картвельский, индоевропейский, уральский, дравидийский, алтайский): Введение. Сравнительный словарь [A Tentative Comparative Dictionary of the Nostratic Languages (Semito-Hamitic, Kartvelian, Indo-European, Uralic, Dravidian, Altaic)). Т. I (b - К). М.: «Наука», Главная редакция восточной литературы. Канева 2006 — И. Т. Канева. Шумерский язык [The Sumerian Languages). СПб.: «Петербургское Востоковедение».
Меновщиков 1997 — Г. А. Меновщиков. Азиатских эскимосов язык [The Language of Asian
Eskimos) II Языки мира. Палеоазиатские языки. М.: «Индрик»; стр. 75 - 80. Мудрак 2000 — О. А. Мудрак. Этимологический словарь чукотско-камчатских языков [An Etymological Dictionary of Chukchi-Kamchatkan Languages]. М.: «Языки русской культуры». Николаева - Хелимский 1997 — И. А. Николаева, Е. А. ХЕлимский. Юкагирский язык [The
Yukaghir Language) II Языки мира. Палеоазиатские языки. М.: «Индрик»; стр. 75 - 80. Старостин 2006 — Г. С. Старостин. Еще раз к вопросу о личных местоимениях в дравидийских языках [Once again on personal pronouns in Dravidian languages) IIАспекты компаративистики. Вып. 2 [Aspects of Comparative Linguistics. Issue 2) (= Orientalia et Classica. Труды Института восточных культур и античности. Вып. XI [Papers of the Institute of Oriental and Classical Studies. Issue XI)). М. Хайду 1985 — П. Хайду. Уральские языки и народы [The Uralic languages and peoples). М.: «Прогресс». Хелимский 1982 — Е. А. ХЕлимский. Древнейшие венгерско-самодийские языковые параллели: Лингвистическая и этногенетическая интерпретация [The Earliest Hungarian-Samoyed Language Parallels: Linguistic and Ethnogenetic Imterpretation). М.: «Наука». Шмидт 1984 — К. Х. Шмидт. Типологическое сопоставление систем картвельского и индоевропейского глагола [Typological Comparison of Kartvelian and IndoEuropean Verbal Systems) II Вопросы языкознания, 33; стр. 48 - 57.
Резюме
Статья посвящена вопросу о том, насколько значимым является сопоставление данных личных местоимений и местоименных показателей для исследований в области дальнего родства языков. Утверждается, что, хотя отдельные, бессистемные сходства между местоимениями в праязыках неглубокого уровня могут быть объяснены случайным сходством или как заимствования, такая ситуация гораздо менее вероятна при сопоставлении целых местоименных парадигм. Основываясь на кратком обзоре основных форм личных местоимений и местоименных показателей, реконструируемых специалистами для подавляющего большинства праязыков, автор делает вывод, что многие из ныне существующих гипотез о дальнем родстве (таких, как ностратическая, сино-кав-казская, австрическая и т. п.) подтверждаются наличием как раз такого рода парадигматических сходств.
