It seems easy to claim that one language is different form another, so that it should be labelled differently. Linguist tried to devise different criteria for labelling one language as unique, including mutual intelligibility.
Max Weinreich wrote that “a language is a dialect which has an army and a navy” and after this quotation Campbell (193) explained how the definition of language is not entirely a linguistic matter. For example, although Norwegians and Swedes understand each other, their languages have different names. Even more strikingly, in China there are many dialects and understanding between their speakers varies significantly. In accordance with the mutual intelligibility criterion, a linguist would categorize them as separate languages, but all those dialects are by recognized by the Chinese government as the single Chinese language. The mutual intelligibility is not reliable even to distinguish a dialect from a language. The criterion states that “[e]ntities which are totally incomprehensible to speakers of other entities clearly are mutually unintelligible, and for linguists they therefore belong to separate languages” (ibid). Campbell illustrated the problem by explaining that Portuguese speakers understand Spanish very well, which does not apply to the extent to which Spanish speakers understand Portuguese. In Europe during nineties, the naming of new languages, such as Croatian, Serbian, Bosnian and Montenegrin, came after the disintegration of the former Serbo-Croatian language.
However, dialectology plays an important role in historical linguistics. One of the attempts to explain change was given by Johannes Schmidt, who in 1872 published the improved version of the “Wave Theory”, a model:
[W]here changes were said to emanate from a centre as waves on a pond do when a stone is thrown into it, where waves from one centre … can cross or intersect … waves coming from other … centres. Changes due to language contact (borrowing) were seen as analogous to successive waves crossing one another …. (Campbell 189)
To discuss which definition of language can be used in the synchronic empirical study, we must mention Chomsky, who distinguished “externalized” and “internalized” language, or E-language and I-language, respectively – in the synchronic language states. To Chomsky, E-Language is “normative”, which is his term for the sociopolitical language, it is “the construct … understood independently of the properties of the mind/brain”; conversely, I-Language is “some element of the mind of the person who knows the language, acquired by the learner, and used by the speaker-hearer”.
In the context of historical linguistics it is significant that I-Language consists of elements that create a generative system of language. This idea is important, because linguistic competence in humans is not a mere transfer of sets of utterances (“normative” language), “but rather … the coming into being in the mind of the acquirer of a system for generating linguistic representations” (ibid). Put in another way, it is possible to contrast “aspects of the output conditioned by features of the grammar and those … conditioned by extragrammatical factors” (ibid). By including this into the study of historical linguistics, the discipline improved form the stage of deciphering dead languages to modern theories and significant relationships between language, mind, reality and society.