Our approach to present time and the theoretical and methodological contribution of a historical sociolinguistic theory is resolved on a first proposal, from the empirical sociolinguistic research on the theory of linguistic change in progress. An empirical sociolinguistic theory of change is the most adequate for a historical reconstruction, in that it leads to hypotheses more easily confirmed or refuted. In the recent past, the linguist has assumed, due to unquestionable methodological priority, the greatest degree of abstraction in his analysis of the grammatical descriptions of the successive states of a language (and of a family of languages), in order to obtain the evolution and the internal (and comparative) reconstruction of the sequence of grammars. From this point of view, he has achieved a regular and systematic analysis of the results of linguistic change. The present dilemma lies in the functional character of such an idealization, for it excludes the development of historical change, and requires different data and a different methodology for the investigation of the real process of change. The delimitation between the level of research and the level of the researched reality allows us to assume an abstract model that completes and reviews previous formulations, through the «heterogeneization» of the homogeneity. In particular, a historical sociolinguistics involves the following dimensions: a) The temporal (apparent), geographical, social and functional synchronic variability is the direct translation of linguistic change in progress, and the details of any evolution can only be studied fruitfully in the overall context of the speech community; b) The speaker is basically multilectal, for his sociolinguistic competence reaches far beyond his own lect (having a larger receptive than productive control), and he constantly revises the grammar throughout his life; c) A multilectal or polylectal grammar is based on an implicational scale of the rule systems of any speech community that is not unequivocally bilingual, i.e. where there is a speech-continuum with no clearly-defined break; d) (Synchronic) multilectal grammar, based on the sociolinguistic foundations of multilingualism and multilectism, is the most adequate theory for historical linguistics; e) The hypothesis of the regularity of linguistic change must be related to the principle of transition and embedding of linguistic change in progress. The question of the transition supposes a constant and regular revision of the multilectal grammar of successive generations of speakers, from the shifts of a passive competence to an active one. The phenomenon of embedding explains that it affects neither uniformly all the implicated forms nor simultaneously all the (sub)groups of speakers; f) Multilectism is the last general and historical principle of linguistic change, since speech is a technically open form of social behaviour. Social interaction is the tool of that principle; g) The systematic covariation of linguistic data and social factors through the probabilistic treatment of a quantitative paradigm, constitutes descriptively the main key to a correct understanding and explanation of the historical process of linguistic change, and h) Multilectal and historical studies involve a general sociolinguistic or interdisciplinary treatment, and assume special contributions by pragmatics and psycholinguistics.