| Uralo-Siberian | |
|---|---|
| (widely rejected) | |
| Geographic distribution | NorthernEurasia, theArctic |
| Linguistic classification | Proposed language family |
| Subdivisions |
|
| Language codes | |
| ISO 639-3 | – |
| Glottolog | None |
 | |
Uralo-Siberian is a hypotheticallanguage family consisting ofUralic,Yukaghir, andEskaleut. It was proposed in 1998 byMichael Fortescue,[1] an expert in Eskaleut and Chukotko-Kamchatkan, in his bookLanguage Relations across Bering Strait. Some have attempted to includeNivkh in Uralo-Siberian. Until 2011, it also includedChukotko-Kamchatkan. However, after 2011 Fortescue only included Uralic, Yukaghir and Eskaleut in the theory, although he argued that Uralo-Siberian languages have influenced Chukotko-Kamchatkan.[2]
Connections with the Uralic and other language families are generally seen as speculative,[3] including Fortescue's Uralo-Siberian hypothesis. Fortescue's observations have been evaluated by specialists as "inspiring" and "compelling" but are viewed as scattered evidence and still remain highly speculative and unproven and the soundness of the reconstructed common ancestors are challenging to evaluate.[4][5]
Structural similarities between Uralic and Eskaleut languages were observed early. In 1746, the Danish theologianMarcus Wøldike [da] comparedGreenlandic toHungarian. In 1818,Rasmus Rask considered Greenlandic to be related to the Uralic languages,Finnish in particular, and presented a list of lexical correspondences (Rask also considered Uralic andAltaic to be related to each other). In 1959,Knut Bergsland published the paperThe Eskimo–Uralic Hypothesis, in which he, like other authors before him, presented a number of grammatical similarities and a small number of lexical correspondences. In 1962,Morris Swadesh proposed a relationship between the Eskaleut and Chukotko-Kamchatkan language families. In 1998, Michael Fortescue presented more detailed arguments in his book,Language Relations across Bering Strait. His title evokes Morris Swadesh's 1962 article, "Linguistic relations across the Bering Strait".
Fortescue (1998, pp. 60–95) surveys 44 typological markers and argues that a typological profile uniquely identifying the language families proposed to comprise the Uralo-Siberian family can be established. The Uralo-Siberian hypothesis is rooted in the assumption that this distinct typological profile was, rather than an areal profile common to four unrelated language families, the profile of asingle language ancestral to all four: Proto-Uralo-Siberian.
None of the four families shows all of these 17 features; ranging from 12 reconstructible inProto-Chukotko-Kamchatkan to 16 inProto-Uralic. Frequently the modern-day descendant languages have diverged further from this profile — particularlyItelmen, for which Fortescue assumessubstrate influence from a language typologically more alike to the non-Uralo-Siberian languages of the region.
Several more widely spread typologically significant features may also instead represent contact influence, according to Fortescue (1998):
Apparently shared elements of Uralo-Siberian morphology include the following:
| *-t | plural |
| *-k | dual |
| *m- | 1st person |
| *t- | 2nd person |
| *ka | interrogative pronoun |
| *-n | genitive case |
Fortescue (1998) lists 94lexical correspondence sets with reflexes in at least three of the four language families, and even more shared by two of the language families. Examples are*ap(p)a 'grandfather',*kað'a 'mountain' and many others.
Below are some lexical items reconstructed to Proto-Uralo-Siberian, along with their reflexes inProto-Uralic,Proto-Chukotko-Kamchatkan (sometimesProto-Chukchi), andProto-Eskaleut (sometimesProto-Eskimo orAleut). (Source: Fortescue 1998:152–158.)
| Proto-Uralo-Siberian | Proto-Uralic | Proto-Chukotko-Kamchatkan | Proto-Eskaleut |
|---|---|---|---|
| *aj(aɣ)- 'push forward' | *aja- 'drive, chase' | *aj-tat- 'chase, herd' (PC) | *ajaɣ- 'push, thrust at with pole' |
| *ap(p)a 'grandfather' | *appe 'father in law' | *æpæ 'grandfather' | *ap(p)a 'grandfather' |
| *el(l)ä 'not' | *elä 'not' | *ællæ 'not' (PC) | *-la(ɣ)- 'not' (A) |
| *pit(uɣ)- 'tie up' | *pitV- 'tie' (FU) | *pət- 'tie up' | *pətuɣ- 'tie up' |
| *toɣə- 'take' | *toɣe- 'bring, take, give' (FU) | *teɣiŋrə- 'pull out' | *teɣu- 'take' (PE) |
Proposed cognates between the languages:[6][7]
| Proto-Yukagir | Proto-Eskaleut |
|---|---|
| *al 'below' | *atə 'below' |
| *amlə 'swallow' | *ama 'suckle' |
| *aŋa 'mouth' | *aŋ-va- 'open' |
| *cowinə 'spear' | *caviɣ 'knife' |
| *kin 'who' | *kina 'who' |
| *ləɣ- 'eat' | *iɣa- 'swallow' |
| *um 'close' | *uməɣ 'close' |
| *n’ə 'get' | *nəɣ 'get' |
| *ta 'that' | *ta 'that' |
| Uralic | Eskaleut[8] |
|---|---|
| *ila 'under' | *at(ǝ) 'down' |
| *elä 'live' | *ǝt(ǝ) 'be' |
| *tuli 'come' | *tut 'arrive, land' |
| *ke 'who' | *kina 'who' |
| *to 'that' | *ta 'that' |
| *kuda 'morning, dawn' | *qilaɣ 'sky' |
| *kuda 'weave' | *qilaɣ 'weave' |
According to Ante Aikio (who does not believe that Yukaghir is related to Uralic), the words glossed 'weave' and 'morning' in the last two rows, despite being homonyms in each language, are most likely unrelated. Such instances of coincidental homonymy between languages, which only very rarely happens by chance, suggest that some kind of contact most likely happened, but exact conclusions cannot be drawn with modern information.[8][9]
Fortescue suggested the following grammatical similarities to point to a relationship:
Proto-Uralic and Proto-Eskaleut number and case markers:[10]
| Proto-Uralic | Proto-Eskaleut | |
|---|---|---|
| nom./absolutive sing. | Ø | Ø |
| dual | *-kə | *k |
| plural | *-t | *-t |
| locative | *-(kə)na | *-ni |
| accusative sing | *-m | – |
| plural accusative | *-j/i | *-(ŋ)i |
| ablative | *-(kə)tə | *-kənc |
| dative/lative | *-kə/-ŋ | *-ŋun |
Yukaghir and Proto-Eskaleut verbal and nominal inflections:[6]
| Pronoun | Yukaghir | Eskaleut |
|---|---|---|
| trans. 1s | *ŋ | *ŋa |
| 3pl | *ŋi | *ŋi |
| 3 poss. | *ntə | *n |
| vialis | *-(n)kən | *-(n)kən |
| abl. | *-(n)kət | *(m/n)əɣ |
| all | *(ŋi)n’ | *-(m/n)un / *ŋus/*-ŋun |
| adv. loc./lative | *nə | *nə |
Some or all of the four Uralo-Siberian families have been included in more extensive groupings of languages (see links below). Fortescue's hypothesis does not oppose or exclude these various proposals. In particular, he considers that a remote relationship between Uralo-Siberian and Altaic (or some part of Altaic) is likely (seeUral–Altaic languages). However, Fortescue holds that Uralo-Siberian lies within the bounds of the provable, whereas Nostratic may be too remote a grouping to ever be convincingly demonstrated.
TheUniversity of Leiden linguistFrederik Kortlandt (2006:1) asserts thatIndo-Uralic (a proposed language family consisting of Uralic and Indo-European) is itself a branch of Uralo-Siberian and that, furthermore, theNivkh language also belongs to Uralo-Siberian. This would make Uralo-Siberian the proto-language of a much vaster language family. Kortlandt (2006:3) considers that Uralo-Siberian andAltaic (defined by him as consisting of Turkic, Mongolian, Tungusic, Korean, and Japanese) may be coordinate branches of theEurasiatic language family proposed byJoseph Greenberg but rejected by most linguists.
I would no longer wish to relate CK directly to [Uralo-Siberian], although I believe that some of the lexical evidence [...] will hold up in terms of borrowing/diffusion.
Fortescue's observations are encyclopaedic, and often innovative and inspiring, but the picture arrived at is one ofdisetcta membra [sic]
As a result, F's list of proposed cognates contains numerous gaps, with many stems not found in more than two or three languages among the languages being compared. Therefore, it is sometimes difficult to evaluate the soundness of the reconstruction, and more illustrative examples would have helped.