Multiple sources and multiple causes multiply explored

Multiple sources abound in language, at all levels of linguistic analysis (phonology, syntax, semantics, etc.), and in a range of historical pursuits, including etymology and variationist investigations. From a methodological standpoint, moreover, recognizing multiple sources is often good historical linguistic practice (contrary to inclinations towards neat and elegant solutions that satisfy Occam’s Razor). That is, if we can identify multiple pressures on some part of a language system, it cannot always readily be excluded that some or even all might have played a role in shaping a particular development; if all of the factors represent reasonable pressures that speakers could have been aware of and influenced by, excluding any could simply be arbitrary. In this paper, accordingly, I survey the breadth of multiple sources in a variety of areas of language change, and advance one particular consequence that multiple sources can lead to, namely the hypothesis that recognizing multiple sources can be a basis for positing proto-language variation that is realized in variation within single languages and across related languages.