Historical-comparative Linguistics (by Edward J. Vajda)
Linguists today hotly debate the issue of monogenesis vs. polygenesis. Did language arise once in the distant past, so that today's 5,000 languages are all descended from this original Mother Tongue? Or did language arise in several or even dozens of locations in prehistory, so that today's languages are descended variously from those multiple Mother Tongues? How can linguists help resolve this debate? One scientific way to study the origin of language is to try to prove historical relationships between languages. To find language families, that is, groups of languages descended from a common ancestor, linguists compare languages to find systematic differences or similarities.
This method of analysing languages is known as the comparative method; linguists using it are referred to as comparative linguists. Some languages are obviously related to one another, as shown by the presence of systematic differences--like the regular sound correspondence between English [T] and German [d]. Many such correspondences show up between the vocabulary of French and Spanish, on one hand, and Hebrew and Arabic, on the other, as well as between such geographically disparate languages as Hawaiian, Maori and Malagasy. No one would dispute that the languages in each of these groupings stem from a common ancestor. Many other languages seem totally unrelated: Navaho, English, Swahili. Or the evidence for their relationship is inconclusive: Japanese, Korean, Mongolian.
When comparative linguists discover a group of historically related languages, they try to reconstruct the original form of the ancestor language of each family, which they call a proto language (give example of Indo-European mother and daughter languages). Obviously, there is no way to prove the results, and proto-language reconstruction is risky business intellectually.
As we have noted, about 5,000 languages are spoken in the world today. Let's take a look at the map. There were considerably more languages spoken in the recent past before the expansion of Europeans to other continents. What seems also to be true is that these languages derive from a much smaller group of original languages.
Comparative linguists today hotly debate whether or not certain languages should be grouped together into families. In this regard, linguists are either lumpers or splitters.
Lumpers have have narrowed the number of proto-languages to about two dozen (see map): Indo-European, Uralic, Altaic, 4 families in Africa, a few in East Asia; perhaps only 3 in all of the Americas. Also, there are a few languages left over that seem not related to any others. They are called language isolates: Basque, Ket, Burushaski. These languages are probably remnants of larger families spoken in the distant past.
Splitters are far more cautious in drawing conclusions of genetic relationship. The map you received is one favored by the lumpers. If I had given you a world map devised by splitters, it would contain many times the number of basic groupings, and you would be very unhappy with me. For instance, instead of one family in Australia there would be at least five; and New Guinea would have over 70 families; and Amerindian is actually composed of a few dozen major groupings, each of which the splittes consider to be a separate family.
So the debate and the research goes on. Recently there have appeared linguists who might even be called "mega-lumpers", notably Stanford University's Joseph Greenberg. Greenberg and his colleagues are convinced they will eventually reconstruct the Mother Tongue of all languages, which they call proto-World.
This most recent theory of monogenesis, the proto-World theory, has evolutionary rather than religious overtones: Greenberg's hypothesis holds that the original language developed in Africa among early Homo sapiens. As Homo sapiens spread across the world, they took their language with them. That single language, which he calls the Mother Tongue or proto-world, diverged naturally over time into the several thousands of diverse forms spoken today.
So far no one has found conclusive proof that all existing languages are descended from a common source. But the more we learn, the more it seems that the lumpers are correct. And eventually it may very well be proven that there is a single mother tongue.