Every Wikipedia entry can link to the same (or, at least similar) entry in another language. This is known as an inter-language link. Analyzing inter-language linking behavior we help bring networks of languages and content to the surface, with results that range from expected to less obvious dependencies. We find strong preferential attachment to English (en), French (fr), German (de), Spanish (es) and other languages, suggesting intensive inter-language activity as a result of time-demographic affordances (launch of Wikipedia version; size of language, Internet and Wikipedia users). Furthermore, distinctive geographic and linguistic clusters exist, for example, for East Asian languages (Chinese (zh), Japanese (ja), Korean (ko), and Vietnamese (vi)), Latin languages (Spanish (es), Portuguese (pt), and Catalan (ca)) and Slavic languages (Russian (ru), Slovak (sl), and Polish (pl)). In addition, we find that if a language node is connected only to a few others, that language’s preferential attachment is to nodes that are geographically and/or linguistically closer (for example, Javanese (jv) and Sundanese (su) to Indonesian (id) rather than English (en)). Inter-language linking analysis provides a new mechanism to explore global dependencies in the co-evolution of languages and content. This is fascinating for both small and large languages as it gives insight into the evolutionary dynamics of languages under this particular institutional technology of interaction. The question which languages contribute and which languages receive suggests that here we have a new mechanism of knowledge globalization, one that supersedes language boundaries as the limits to externalism, and therefore socio-cultural and economic evolution.
View at MIT Press Journals.