Language Evolution

The distribution of language family sizes

When the sizes of language families of the world, measured by the number of languages contained in each family, are plotted against their size ranking, it is seen that the distribution approximates a line defined by the formula y = ax-b. It is suggested that this apparent power-law distribution of language family sizes is of relevance when evaluating over-all classifications of the world's languages, for the analysis of taxonomic structures, and for developing hypotheses concerning the prehistory of the world's languages. It seems that three different major generating models will all eventually lead to power-law distributions: preferential attachment (in the case of networks), the Galton-Watson stochastic branching process (or some version thereof), and (some version) of the sand-pile model. General ingredients in a simulation procedure will be discussed and the results of an initial attempt to simulate the distribution of language family sizes using the second of the models mentioned will be shown.