On the Nostratic hypothesis

Jul 6, 2011 by

The Nostratic hypothesis has received much attention not only in professional academic circles but in the popular press as well and have been featured in Atlantic Monthly, Nature, Science, Scientific American, US News and World Report, The New York Times and in BBC and PBS television documentaries. Yet, despite all the ink spilt and all the buzz generated by this hypothesis, it remains outside consensus, in the sense that linguists who do not work on this hypothesis rarely accept it.

Since I’ve been asked by one of the readers to write about the Nostratic hypothesis, I will weigh in on the debate, but since long-range family relationships are not my speciality, I will focus here on describing the Nostratic hypothesis and reviewing some critiques of it.

Originally proposed by Holger Pedersen in an article on Turkish phonology in 1903, Nostratic hypothesis relates Indo-European languages to a number of other known families. The term “Nostratic” derived from the Latin nostrates ‘fellow countrymen’. According to Pedersen himself, Indo-European (which he refers to as “Indo-Germanic”, a term frequently used by German philologists in earlier days and by now obsolete) was most clearly related to Finno-Ugric and Samoyed (a small language family spoken in Siberia), and less clearly (or perhaps more distantly) to Turkish, Mongolian, and Manchu; to Yukaghir; and to Eskimo. Pedersen also believed that

“the Nostratic languages occupy not only a very large area in Europe and Asia but also extend to within Africa; for the Semitic-Hamitic [i.e., Afroasiatic] languages are in my view without doubt Nostratic”

In Pedersen’s days, Basque was considered to be related to Semito-Hamitic languages, so presumably it too was a member of the proposed Nostratic family. To translate this into modern day classification, we would say that Pedersen was positing a genetic relationship between Indo-European and the Uralic, Altaic, Yukaghir, Eskimo and Afro-Asiatic language families. Note also that Pedersen did not see this list of Nostratic languages as exhaustive, leaving room for adding more languages and language families:

“The boundaries for the Nostratian [sic] world of languages cannot yet be determined, but the area is enormous, and includes such widely divergent races that one becomes almost dizzy at the thought. (…) The question remains simply whether sufficient material can be collected to give this inclusion flesh and blood and a good clear outline.”

In Pedersen’s lifetime, the Nostratic hypothesis did not receive much attention, but after years of neglect it was resuscitated in the 1960s by two Soviet scholars Vladislav M. Illyč-Svityč (1934-66), a specialist in Indo-European, Altaic and Kartvelian languages, and Aaron B. Dolgopolsky (b. 1930), an Indo-Europeanist and Hamito-Semitist. They expanded the proposal to include two additional language families: Kartvelian (a language family spoken in the southern Caucasus region, including languages like Georgian and Mingrelian) and Dravidian (a language family spoken in southern India, including Tamil, Telugu and Malayalam). Moreover, Illyč-Svityč prepared the first comparative dictionary of the hypothetical Proto-Nostratic language, the presumed common ancestral language of the Nostratic languages, which was published posthumously. Their work has been continued by Vladimir Dybo, Vitaly Shevoroshkin and to some extent Sergei Starostin (who also worked on another controversial macro-family proposal, the Dené-Caucasian hypothesis).

The kind of evidence in support of the Nostratic hypothesis is the familiar methodology of comparative reconstruction: that is, compiling potential cognates and identifying sound correspondences. The list of proposed cognates includes Proto-Indo-European *b[h]or-, Proto-Afro-Asiatic *bar-, Proto-Uralic *pura, Proto-Dravidian*pur-, and Proto-Altaic *bur-  for ‘to bore, to pierce’; Proto-Indo-European *mer-, Proto-Afro-Asiatic *mar-, Proto-Uralic *mur- and Proto-Altaic *muru- for ‘to twist, to turn’; and other items (data from Bomhard 1996: 11-13). Note that all of these items in descendant proto-languages are themselves reconstructred.

Based on these sorts of data, certain sound correspondences of consonants in the initial position can be established. For example, the Proto-Indo-European *b[h]- corresponds to *b- in Proto-Afroasiatic and Proto-Altaic and to *p- in Proto-Uralic and Proto-Dravidian; the corresponding sound in (Proto-)Nostratic is reconstructed as *b-.

While Soviet scholars worked on the Nostratic hypothesis, in the West a similar but not exactly the same proposal has been put forward by Joseph Greenberg and his colleagues. They called their proposed macro family Eurasiatic and included not only Indo-European, Altaic, Uralic and Eskimo-Aleut but also Chukotko-Kamchatkan, and (with reservations) Nivkh. Note that Afroasiatic, Dravidian and Kartvelian languages are not included under the Eurasiatic proposal, even though Greenberg did not reject outright a relationship of Afroasiatic, Kartvelian or Dravidian to the other Eurasiatic languages, but he considered it to be a much more distant relationship.Thus, there is a certain overlap between the proposed Nostratic and Eurasiatic macro families. Given the significant overlap between the Eurasiatic and Nostratic proposals and the unclear evidence regarding Chukotko-Kamchatkan languages and Nivkh, it has also been proposed that Eurasiatic is in fact a branch of Nostratic.

In recent years, the work on Nostratic/Eurasiatic proposals continues: in 2008 an updated version of Dolgopolsky’s Nostratic Dictionary has been published online and a comprehensive two-volume set Reconstructing Proto-Nostratic has been published by another dedicated Nostraticist, Allan Bomhard. The Nostratic hypothesis even received some attention in the popular press when the New York Times published an article on the topic in June 1995.

But while the overall reaction to the Nostratic idea can be described as “skeptical fascination” (Salmons and Joseph 1998: 1), scientific opinions are divided. Supporters (cf. Vovin 1998, Ramer et al. 1998, Bomhard 1998, among others) contend that “Nostratic theory cannot be dismissed out of hand by a responsible historical linguist as something not being worthy of further discussion” (Vovin 1998: 257). Moreover, they argue that the assumption of Nostratic not only answers the age-old question of “what came before” but can also yield cogent solutions to language-particular problems in subgroups of Nostratic. Other scholars (cf. Vine 1998, Campbell 1998, Ringe 1998, inter alia) voice criticisms; yet others such as Merritt Ruhlen, propose similar but not identical classifications.

But perhaps the majority of linguists remain agnostic on the subject, taking it to be an issue that can be neither proved nor disproved with the means available to historical/comparative linguistics. And here is why. For smaller language groupings that have been proven beyond reasonable doubt, the proto-language is either known directly (as in the case of smaller families or branches of families, like Proto-North-Germanic, known as Old Norse, or Proto-Romance, known as Vulgar Latin) or reconstructed on the basis of descendant languages (e.g., Proto-Indo-European). For macro-families, the reconstruction of the proto-language is by necessity based on prior reconstructions of proto-languages of smaller language families; for example, the Proto-Nostratic must be reconstructed based on other reconstructed languages: Proto-Afroasiatic, Proto-Uralic, Proto-Altaic, etc. Of course, a reconstruction of a proto-language based on a reconstruction (rather than on actual descendant languages) introduces additional room for speculation and error, making the project more controversial. But this does not stop the more adventurous (or less conservative) linguists from trying. And the natural next step is hypothesizing about an even larger grouping of several known families and macro families – this is exactly what the Nostratic hypothesis is all about!

Thus, it is not the very idea of looking for distant relationships between languages that raises most eyebrows but the methodological issues: many linguists think that time depth constrains the utility and validity of the comparative method. In the words of Salmons and Joseph (1998: 4):

“As elsewhere in science, methodological flaws can undermine reliability and replicability of results. Just as acoustic phonetic investigations can be fatally compromised by background noise or sociolinguistic studies can be felled by problems in representativeness of sampling, so too can comparative linguistic houses be built on methodological sand.”

The chief methodological problem, aptly formulated by one of the strongest defenders of the Nostratic proposal, Allan Bomhard (1996: 5), is this:

“the problem is in knowing which languages to compare and in knowing what to compare since not all aspects of language are equally relevant to comparison. To be meaningful, comparison must strive to eliminate chance resemblances and to separate borrowings from native elements. This is often easier said than done.”

This difficulty of separating borrowings from true cognates is exacerbated further when we know little to nothing about the interactions between different linguistic groups. While attempts have been made to figure out which lexical items are less likely to be subject to borrowing (including work done by Dolgopolsky, one of the authors of the Nostratic hypothesis), such lists have drawn heavy criticisms from other linguists. And this is not the only problem with the comparative research into Nostratic hypothesis.

Other critics of the Nostratic/Eurasiatic work point out that the data from individual well-established language families that is used in Nostratic comparisons contains many errors: for example, Campbell (1998) demonstrates this for the Uralic data. Another potential problem concerns the kind of comparison done to evaluate possible cognates: much of the Nostratic work consists of binary comparisons which tend to lead to distortions (while in multi-language comparisons each language serves as a control on others). A further problem is raised by the focus on grammatical morphemes: some of these forms being compared are just two small to be reliably compared (the same is true to some extent of lexical forms being compared by Nostraticists as well).

In their defense, the proponents of the Nostratic hypothesis now employ complex statistical methods for lexical comparisons of multiple languages (cf. Oswalt 1998) and extend their work to conduct morphological and even syntactic comparisons. And indeed it appears that the future of Nostratic linguistics lies in developing mathematical methods for sifting through vast amounts of data and figuring out what portions of those data are due to relatedness and which are due to borrowing or even to pure chance.


Nostratic Hypothesis – References


Bomhard, Allan R. (1996) Indo-European and the Nostratic Hypothesis. Charleston, South Carolina: Signum Publishing.

Bomhard, Allan R. (1998) Nostratic, Eurasiatic and Indo-European. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 17-49.

Campbell, Lyle (1998) Nostratic: a personal assessment. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic: Sifting the Evidence. Current Issues in Linguistic Theory 142. Amsterdam: John Benjamins. Pp. 107-152.

Oswalt, Robert L. (1998) A Probabilistic Evaluation of North Eurasiatic Nostratic.  In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 199-216.

Ramer, Alexis Manaster; Peter A. Michalove, Karen S. Baertsch, and Karen L. Adams (1998) Exploring the Nostratic Hypothesis. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 61-84.

Ringe, Don (1998) A Probabilistic Evaluation of Indo-Uralic. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 153-198.

Salmons, Joseph C. and Brian D. Joseph (1998) Introduction. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 1-12.

Vine, Brent (1998) Indo-European and Nostratic: Some further comments. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 85-106.

Vovin, Alexander (1998) Nostratic and Altaic. In: Joseph C. Salmons and Brian D. Joseph (eds.) Nostratic. Sifting the evidence. Amsterdam: John Benjamins. Pp. 257-270.


Related Posts

Subscribe For Updates

We would love to have you back on Languages Of The World in the future. If you would like to receive updates of our newest posts, feel free to do so using any of your favorite methods below: