Thursday, January 21, 2010

My thoughts on the Nostratic theory



Though I'm not a qualified linguist, I'm well read on the subject of linguistics and taken courses in the field.
Based on all the knowledge that I have in the subject, I have decided to post my opinions on what I feel is an exaggerated hypothesis.

Those who have a basic insight into major language families and their histories will know what I'm writing about.
The Nostratic theory is a hypothetical Super family of various language families thought to be related to one another going back thousands of years. Most of these language families are major language families spoken in the world including Indo-European, Altaic, Uralic, Dravidian, South Caucasian, Afro-Asiatic and some Native American families (Native American languages are diverse and grouped into various families unrelated to each other).

All the mentioned families save for Dravidian, South Caucasian and some Native American families are major families in the world in terms of number of languages and speakers.

From all my knowledge on linguistics which goes down a great deal, I believe some supporters of the Nostratic theory (both linguists and non-linguists alike) are getting carried away from all the findings of the last few centuries.

The discovery of the Indo-European family for example is no small finding. Numerically and geographically speaking this is the largest family on Earth spoken in almost every continent as a first or second language. It's also an extremely diverse family.
The discovery of it's existence has connected many dots for the first time in human history and was a major step in discovering the existence of various language families using various comparison techniques.

From there, I feel linguists are getting their hopes too high that by connecting one set of dots may help them connect even more dots and answer many unanswered questions.
Explaining the existence of a large family like Indo-European does not explain why so many languages sometimes with shared geography are not related to one another.

This is from where the over-excitement comes in. Because linguists have managed to fit many distinct major and diverse languages into large families, they could be pushing harder and overestimating what could very well be coincidental similar sounds in root words found in the mentioned language families.
As an example, some linguists have gone ahead to try and include Burushaski, a language isolate spoken in the far northern areas of Pakistan, into the Indo-European family.

A close friend of mine and well known Internet blogger, Robert Lindsay has an education in linguistics and appears to be a supporter of the Nostratic hypothesis.
Though I respect his views, I do not find myself in agreement with him. According to him about 12,000 common cognates (core vocabulary) have been found between proposed Nostratic languages.

Though identifying a proposed language family usually begins with this step supposedly called phono-semantic sets, I find the overuse of this technique especially in the wrong area can lead to misconception.

My view is also shared by linguists who do not support the Nostratic theory. The grammatical evidence used for this family by supporters also seems a hypothesis of it's own that most likely stems on common geography.
All these hypothetical macro families share a common geography, which is the most likely cause of them having similar sounding supposed core vocabulary or 'root words' (an alternation of roots words in a language due to outside influence cannot be considered a root word or at least a pure root word).

For example there is an Indo-Uralic hypothetical family consisting of Uralic and Indo-European.
Not only do the proposed birthplaces of these two language families share a common geographic area, but even many of their modern members have a common location.

The map below shows the proposed birth places of Uralic (marked in blue) and Indo-European (marked in red):



Finnish, Hungarian and Estonian, the most widely spoken Uralic languages today belonging to the Finno-Ugric branch of Uralic are spoken in the heart of Europe, a predominantly Indo-European speaking continent.
They are surrounded by Indo-European languages and have been used to compare proposed similarities between Uralic and IE based on similar sounding words which most likely were influenced by the IE languages that have been surrounding them for centuries.

Another comparison has been made between IE and Semitic, a subbranch of the Afro-Asiatic family.
It should be noted that no other subbranch of Afro-Asiatic has been compared to IE except for Semitic, which shares a common geography with the IE speaking world unlike other subfamilies of Afro-Asiatic.
In this comparison between the two for example, the linguist comparing them admits the same similarities between Semitic and other proposed Nostratic languages have been hard to find.

This is another piece of possible genetic evidence that pro-Nostraticists would take into consideration.

The map below shows the close geography of Semitic (spoken mainly in North Africa and the Arabian peninsula) to Indo-European:


There has also been a proposed Ural-Altaic family, which most linguists today have discounted on the basis there's not enough genetic relationship between the two.
Again the proposed relationship between Uralic and Altaic coincides with a relatively common geography between the most likely birthplaces of these two language families which is in Northern Asia.
An indirect relationship between the two may exist since both show common grammar and vocabulary to Dravidian, Sumerian and some Native American language families in certain areas.

From here comes important questions for those who use theories like Indo-Uralic, Indo-European-Semitic and Alto-Uralic (better known as Ural-Altaic) as evidence for a common Nostratic family.
Why are certain branches within the proposed Nostratic family hypothesized to have a common sub grouping such as Indo-Uralic or Alto-Uralic or Semitic-Indo-European?
Is the common geography between these subfamilies coincidental or a likely cause for common sounding 'roots words' (remember root words that may be influenced cannot be considered root words or pure root words at least) and genetic relationships?

Why is there no Indo-Altaic hypothesis? Or Afro-Uralic hypothesis? Or an Afro-Altaic hypothesis? Even if they are all close enough to push into one hypothetical family, why has there been no deep comparison between Afro-Asiatic and Uralic or IE and Altaic or Afro-Asiatic?

A bizarre chart of the proposed Nostratic family. Note how it is lumped alongside the proposed Borean family. Also note how the Paleo-Siberian languages have been lumped together as "one family:"


For established language families, you may pick any two languages of that particular family and compare their core vocabulary along with genetic grammar.
This has not been done with proposed Nostratic languages despite the theory being enhanced as far back as the 1950s.

Many linguists have criticized the use of supposed common core vocabulary between Nostratic languages to propose common origins and I find myself in agreement with them.
The method of using similar sounding core vocab should be used only as a starting point to propose a common relationship, but genetic evidence must be added to confirm this otherwise it is accepted that the core vocab is either borrowed/influenced (hence not pure roots words anymore) or merely coincidental.

From all the supposed genetic evidence found which could also very well be coincidence or a result of common geography for thousands of years, I have challenged that there should be a comparative study on every subbranch of Nostratic.

Another chart of the proposed Nostratic family. The arrangement and inclusion of some language families in this chart is also disputed amongst linguists:


Some of the theories linking IE to Semitic also includes the similar sound of certain consonants. This could also be the result of borrowing due to the common geography between the two language families. In another post I had concluded that the use of Dravidian-sounding consonants in Indo-Aryan languages and some Iranic ones was from influence, again due to common geographic proximity.

Until full genetic comparisons between all members of Nostratic languages can be made, no credibility should be given to this theory.
I have not touched much on certain Native American language families and South Caucasian languages because I have not read on them as much, so I cannot really comment on those yet.

I also believe that in the future some evidence of migration of supposed Nostratic speakers is found to prove a separation from the supposed birthplace of Proto-Nostratic to the birthplaces of the modern families that Nostratic supposedly consists of such as Afro-Asiatic, Indo-European, Dravidian and others.

1 comment:

  1. Hi Robert. What I meant was by common geography they are located closer to each other within the same area.
    About the comparisons between IE and Afroasiatic you might be referring to Indo-European and it's closest relatives. That book uses Semitic as a reference to compare Afroasiatic IE. From what I remember the source claimed that only Semitic connects to IE. And even an Assyrian claimed this that Semitic connects to IE out of all Afro-Asiatic subbranches.
    And what I meant by Indo-Uralic and alike were hypothetical families and similarities. Though such a family is not proven to exist, they are still pointing towards (coincidental) similarities between the two. If there is a Nostratic family, ALL it's sub groupings must be compared to one another.
    I also don't agree with you that core vocabulary cannot be replaced with borrowing words. Like the Urdo word for ground/surface is derived from Turkic or the world for planet/world is also borrowed.
    Words for core vocabulary might not change most of the time, but they can easily be replaced provided the borrowed word does not bring in grammar conflict with it.

    I agree we cannot prove Nostratic nor disprove it anymore than a judge & jury can 'prove' a convict innocent/guilty. 99% of those kinds of cases rely on evidence. In the case of Nostratic there's evidence to support it, but also evidence to go against it including cultural, genetic and geography.

    But another good question would be if a language family like Altaic originated outside of the Altay mountain region, could a trail(s) of this languages roots be found leading away from the Altay mountain region going backwards?

    What do you think about grouping the Paleo-Siberian languages into one? Or do you disagree with just that particular chart?

    ReplyDelete