Models of linguistic change

From the field of historical linguistics or diachronic linguistics, a series of models or theories have been proposed to explain the linguistic field, a complex phenomenon that tries to define, under general parameters, the process of modification and transformation that all languages undergo. In short, it is a matter of establishing universal principles that explain why languages change over time following a common pattern.

In contrast to what is commonly known as ‘linguistic variation’, where the focus is on synchronic aspects, models of linguistic change focus on diachronic modifications, that is to say, those that take place taking into account the time factor. 

Family tree theory

In the middle years of the 19th century, the German linguist August Schleicher proposed a system of language classification similar to botanical taxonomy, tracing groups of related languages and classifying them in a genealogical tree.

In this sense, the so-called Stammbaumtheorie model was based on the consideration of the history of language as a tree: there is a proto-language (which represents the trunk), and from this the different groups of languages branch off, which can be complete families (Slavic, Germanic, Indo-Iranian languages), or individual languages (Greek).

This is the classical model of language diffusion, which was further developed in the study of Indo-European languages. As a support for his theories, Schneider composed the fable of The Sheep and the Horse, an artificial text reconstructed in the Proto-Indo-European language, conceived as the proto-language.

Wave theory

In historical linguistics, the so-called wave theory (Wellentheorie), advocated in the late 19th century by Johannes Schmidt, is a model of linguistic change in which new features of a language spread from a central point in ever weaker concentric circles, similar to the ripples formed in water when a stone is thrown (hence the name).

Thus, the wave theory puts the focus on each concrete linguistic innovation: the phenomenon has a center of irradiation and an area of expansion in the form of concentric waves; we call the expansion of each phenomenon or change an isogloss, and these isoglosses overlap each other. Thus, one language or language family is linked to another because they share a certain number of isoglosses.

Innovation A  Innovation B  Innovation C  Innovation D

As this issue is somewhat complex, let us try to clarify it through an example: in Latin, there were two different ways to refer to the adverb ‘more’: plus and magis.

  • The plus isogloss includes French (plus) and Italian (più).
  • On the other hand, the magis isogloss includes Catalan (mes), Spanish (más) and Portuguese (mais).
  • In Romanian, on the other hand, both isoglosses converge (mai mult, plus).

The theory of lateral areas

This is an important derivation of the wave theory proposed by the Italian linguist Matteo Bartoli in the first half of the 20th century.

The main premise of this theory revolves around the idea that geographically isolated areas tend to retain archaisms. This means that an area that has been isolated for cultural or historical reasons is more likely to preserve older states of the language because, being isolated, it is not influenced by innovations that affect the native language in the rest of the territory.

In an area where one language or a family of languages is spoken (e.g. Latin in the Mediterranean basin), the central territories are highly exposed to innovations, so that changes spread more rapidly among them, as they are better communicated with each other. On the other hand, the innovation does not reach the lateral or peripheral areas, which are not connected to each other and preserve the previous stage. Let’s see it with an example:

  • Latin had a diphthong au, which in Romance languages tend to evolve to o: aurum (lat.) > oro (Sp., It.) or (Fr.), etc.
  • On the other hand, the diphthong is only preserved in the lateral areas, geographically more distant from the Italic peninsula: ouro (Galician-Portuguese), aur (Romanian).


In view of this, we can conclude that all these models share a common conception of linguistic change:

  • It acts under certain conditions in a specific linguistic context.
  • It has no exceptions: exceptions are interferences with other laws.
  • It operates without limitations in a given context and for a given period of time.
  • It presupposes the same result if the starting conditions are the same.


However, this idea of a blind, universal and absolute law that is fulfilled in a context has been quite nuanced. From the field of sociolinguistics, it has been shown that there are numerous exceptions to changes that were previously considered universal, so there are more determining factors than a precise phonetic context. This is because languages do not behave in a uniform manner, since every language is dialectalized and has varieties. In this sense, sociolinguistics shows that these varieties influence linguistic change:

  • Diastratas: the same language can vary according to the class or social stratum to which its speakers belong.
  • Diaphases: the same speaker can modify their way of speaking according to the speech act itself (a more or less elevated register of the language).
  • Diatopics: (the dialects themselves): the same language varies according to the region in which it is spoken, giving rise to language substrata and substrates.
  • Diachronics: are the evolutionary or historical varieties themselves.


Carlos Sánchez Luis