Language change as a scientific construct of a probabilistically organized information system

Abstract The present paper discusses language change from an information and systems theoretical point of view, taking on a diachronic perspective. It is argued that human language has to be regarded as a probabilistically organized information system in which synchronizations of linguistic systems of individuals create unstable (dynamic, ever-changing) collective levels (“language systems”). Therefore, probabilistic organization of language processing on an individual level leads – via bottom-up structure – to probabilistic organization of language systems as a whole. If we thus regard linguistic objects like e.g. a Saussurean sign as generally unstable and defined by probability distributions even from a synchronic point of view, we must understand language change (diachronic developments) as probabilistic as well. Therefore, language change in its “classical sense” (a change in linguistic objects) has to be reinterpreted as a change in probability distribution. Nevertheless, the term language change and its meaning then still lack exactness regarding some details; so we have to use this term carefully and be aware of its weaknesses. With a close look at language as an information system with both a synchronic as well as a diachronic dimension, we finally have to admit that language change is a scientific construct serving as a – sometimes quite useful – simplification within the linguistic field.