|
Reply:
|
There are actually several parts to this question. One part has to do with *when* (that is, how long ago) human or human-like languages began to be spoken (and what *that* means). The other part has to do with whether this beginning of human language occurred only once in human history (the theory of monogenesis of language) or many times.
Let me first address the latter question. There is a generally-accepted methodology for reconstructing the undocumented history of now-living languages (including sometimes documented but no longer spoken ones, such as Tokharian for Indo-European). This usually (but not always) involves mostly the reconstruction of words of a predecessor state of the group of languages, and sometimes of (aspects of) its syntax. With the tools that currently exist, the furthest back that such reconstructions have been able to go (and remember: they are only partial reconstructions) is several thousand years, say, about 6000 years at most. Many linguists are trying to attack this problem from several points of view and with differing methodologies and assumptions, but the fact remains that results with any degree of acceptance in the field have only attained about 6000 to 10000 years time depth before the present.
One result, based on various intersecting lines of reasoning, that has achieved a certain degree of acceptance, is that the modern form of human language (which is what the book you cite is referring to and which is all but universally accepted as being a single [undoubtedly socio-culturally influenced] evolutionary entity) developed at least approximately 50-60,000 years before the present; other estimates go as far back as half a million years, and the consensus seems to be between 100,000 and 300,000 years for the beginnings of the evolutionary steps towards human languages, culminating in modern-looking man about 100,000 years ago.
One older and currently little-used tool to try to quantify language change is called glottochronology (it is based on the nearly universally-accepted comparative method developed in the last 200 years in linguistics). Results from cases where the predecessor language is documented (such as Latin and the Romance languages) to a greater or lesser degree indicate that, of the so-called 'basic vocabulary' (a supposedly universal inventory of 100 or 200 terms originally developed by Morris Swadesh in the 1940s and 1950s and basically consisting of the 100 [or 200] most common terms found in the languages of the world), approximately 17 or 18% is lost during each millennium. 'Lost' here means that the term is replaced by a completely new term, not one derived from the original by normal language change; that is, a new term is created or, more usually, perhaps, borrowed from another language. Rather simple mathematical calculations show us that in about 20,000 years, if these figures are roughly correct, two originally identical languages will have been reduced to only 4% (approximately) of their original common vocabulary which they still share. On the other hand, making some reasonable assumptions about the phonological structure of languages in general, it has been shown that we can expect *any* two human languages to have about 4% of their vocabulary which will share a similar phonological shape and similar meanings in the two languages. Bottom line: using such a methodology, assuming it is tested and accepted, we still could only reasonable prove in this way that two languages were 'genetically' related *if* they had separated more recently than 20,000 years ago. Unfortunately, nearly everyone is agreed that the real date for the original development of human language, in more or less its modern shape, is many tens of thousands of years earlier than 20,000 B. P. Now, it has been suggested (no doubt correctly) that if we use morphology and syntax more in our reconstructions (although this has proven to be correspondingly more difficult to do) that we could push back the earliest date of relationships provable (by such methods) to the satisfaction of at least a substantial number of linguists. No one, to my knowledge, has ever suggested that such methods could ever get back the order of magnitude more necessary to 'prove' (or even suggest) monogenesis.
Bottom line: the monogenesis of human language, even if it is correct, is not even close to provable, even in principle, by any methods currently known. And this is despite the fact that, using fossil evidence and DNA comparisons with closely-related primate species (some only a few million years removed from a common ancestor), most linguists feel comfortable assuming that the development of human language had, for all intents and purposes, culminated in modern human language at some point between approximately 50,000 and 200,000 years before the present. (Please note that I say this, and concur in this belief, despite the fact that I find so-called 'Whorfian' arguments from art and other cultural artefacts to be wholly unconvincing and unsatisfactory.)
On the separate issue of the complexity of human language, nearly all linguists agree that:
1) *Any* normal human being can and will learn *any* human language natively, if raised and spoken to in that language in a community that uses that language. (I don't believe that there are *any* linguists who would dispute this.)
2) All human languages share many aspects of their structure [whence the so-called 'universals of language']. (Same comment.)
3) All human languages are equally complex, in a basic sense which, nevertheless, in many aspects remains to be fully specified. (*Not* all linguists are in agreement with this statement, although I don't doubt that the majority is.)
4) Human language is genetically innate (in the ethological sense of the term), and not just culturally transmitted. (Many, perhaps most, linguists do *not* agree with this. I do.)
I hope this (overly long) answer to your question helps.
Jim
James L. Fidelholtz
Posgrado en Ciencias del Lenguaje
Instituto de Ciencias Sociales y Humanidades
Benemérita Universidad Autónoma de Puebla, México
|
|
