In principle, there is nothing wrong about comparing languages; it seems a very natural way of approaching historical linguistics, and in fact it is difficult to imagine any type of linguistic research without some kind of inter-language comparison. The problem arises when the task of comparing grammatical or phonetic features derives into the establishment of supposed laws and the reconstruction of supposed proto-languages. It was in the field of Indo-European (IE) studies that the comparative method was first applied in a systematic way, and where it was carried to its apparently logical conclusions: a perfectly organized world of rules and genealogical trees based mainly on the analysis of the extant texts of ancient languages. Proto-Indo-European (PIE), the most famous of these proto-languages, and somehow the jewel of the crown of the comparative method, was described as a structurally complete language, with inflections, declensions and its own phonetic system, which was eventually 'improved' with of a series of mysterious phonemes, called laryngeals (see this post and this one for more on laryngeals), which are as absurd as they are essential for the 'perfection' of the reconstructed proto-language. That reminds me of a famous etching by the Spanish painter Francisco de Goya (1746-1828): El sueño de la razón produce monstruos (="The sleep of reason produces monsters"; see picture on the left).
Personally, I think it is impossible to reconstruct a theoretical proto-language from a series of supposed 'descendant' languages. First, because very probably this proto-language never existed as such, and those 'descendants' are rather the result of a complex net of interactions; second, because even if there was something resembling a proto-language common to a series of subsequent languages, the possibilities of actually reconstructing the 'structure' of this proto-language are quite remote, especially because the interpretation of written texts is not a good indicator of something as complex as the history of a language. It is true that the data obtained in the last couple of centuries via the comparative method can still be useful and illustrative; they are indeed the product of thorough study and erudition. However, the idea of a perfect tree-like diagram of languages and proto-languages must be abandoned. There are still many linguists today who believe in these immaculate ideas but, fortunately, some other scholars have criticised them in a variety of ways.
In this blog I have already pointed out some of the weak points of traditional (comparative) linguistics. Right now I'm in the process of reading, as part of my research, a series of articles which deal with this topic, for example by authors such as Gessman or Caflish. One of the things that Gessman has shown is that the famous Grimm's Law, designed to explain some important features of Germanic languages, is rather implausible. But Jakob Grimm and his followers, e.g. Karl Brugmann, lived in the 19th c., and therefore they did not have the tools and the perspectives that we have today in order to analyse linguistic diachrony.
Recently, I have read an interesting article by the American linguist Andrew Garrett: Convergence in the formation of Indo-European: Philogeny and Chronology (2006). After analysing some phonological and morphologiccal features of ancient Greek dialects, he comes to the conclusion that the idea of a Greek proto-language derived from a common IE proto-language is not tenable. The linguistic materials from ancient Greek dialects point in a completely different direction, and this could be also applied to other IE branches. (p. 139): "the familiar branches arose not by the differentiation of earlier higher-order subgroups - from 'Italo-Celtic' to Italic and Celtic, and so on - but by convergence among neighbouring dialects in a continuum"; (p. 141): "detailed analysis reduces the dossier of demonstrable and uniquely Proto-Greek innovations in phonology and inflectional morphology to nearly zero"; (p. 139): "I will suggest that conventional models of IE philogeny are wrong". I think Garrett's innovative ideas about the formation of Greek and IE are highly interesting, and they may open interesting new lines of research in historical linguistics. I agree with him completely when he says: (p. 139) "Convergence together with loss of intermediate dialects in the prehistoric continuum, has created the historical mirage of a branchy IE family with its many distinctive subgroups". - The mirage of order, structure, rules, laws.
In his article, Garrett also deals with other topics, mainly the philogeny and chronology of IE. Even though in the first part of the article he expounds the groundbreaking ideas referred to above, in the rest of the article, quite surprisingly, he sticks to the traditional paradigm, using a series of arguments such as linguistic palaeontology and his own version of catastrophism, which he calls 'systems collapse'. I found this quite contradictory, even paradoxical. In any case, these things deserve further scrutiny (and criticism), so I'll be talking about them in a future post (this one is already quite long).
- GARRETT, Andrew (2006). «Convergence in the Formation of Indo-European subgroups: Phylogeny and chronology», in P. Forster, and C. Renfrew, eds. Phylogenetic methods and the prehistory of languages, Cambridge: McDonald Institute for Archaeological Research, 139-151.
- GESSMAN, Albert (1990). «Grimm's Law - Fact or myth?», in Language Quarterly 28: 3-4. (first published in 1974).