No Uralic language family?

Jaakko Häkkinen
(29. January 2006)

Angela Marcantonio has recently claimed that there is no evidence for such entity as the Uralic language family. One might even stop to think about it for a few seconds, were it not that Marcantonio instead of Uralic language family presents Ural-Altaic language family.

However, it happens to be the case that Altaic unity (including at least Turkic, Mongolic and Tungusic, sometimes added even Koreanic and Japanic) has not been proved to be genetic but only areal-typological in nature: language families compared are not related, but the similarities are due to long-durating contacts.

This means that by the comparative linguistic method used at the normal criticism-level, to which I here give an index value 100, Altaic language family is far from plausible. Instead, the existence of Uralic language family is indisputable.

Marcantonio's great error is that her criticism-level is not objective or equal: comparing the Uralic data she applies criticism-level 400, while comparing the Ural-Altaic data she applies criticism-level 25. Such a twisted method is a natural consequence of biased premises; with this I mean that Marcantonio has decided even before the study that there shall be no Uralic language family, but instead Ural-Altaic language family.

Naturally such an unbalanced method is totally worthless and unscientific. Moreover, Marcantonio's statistic method is similarly twisted. Her understanding about the reconstructive method is also rather weak: she has no understanding about the sound laws and the sound correspondences.

These points have already presented in reviews on Marcantonio's book:

Ante Aikio's critique
Johanna Laakso's review

Marcantonio also believes that historical, archaeological or genetic evidence could disprove the linguistic evidence; a supposition which is naturally incorrect (see the other texts in these pages).

So there is, after all, the Uralic language family?

Uralic language family stands clearly apart from other language families - this is evident to everyone who takes a look at the actual Uralic data. Modern Uralic languages share a great number of common features. There are also similarities between Uralic and Altaic language families, but if we stay at the criticism-level 100, we definitely cannot postulate Ural-Altaic proto-language on this basis. The common features are most plausibly explained by areal contacts, just like the similarities between distinct but typologically Altaic language families.

Back to the Main Page