Transparency

The notion of transparency and its antonym, opacity, have been used in different domains of linguistics, including the study of linguistic change. The way the term ‘transparency’ is interpreted differs across authors. Whereas the typological, functionalist literature typically defines transparency in terms of semantic recoverability from surface forms, approaches within the generative framework focus on recoverability of abstract, underlying syntactic or phonological structures from surface representations. What is common to all definitions is that they suggest that transparency corresponds to language optimization. Therefore, transparency can plausibly be hypothesized as a factor driving change as speakers of a language might be expected to aim for more transparency over time. However, once we examine the status of transparency in the context of language change in some detail across frameworks, we are led to conclude that there is little evidence suggesting that transparency can truly be at the origin of substantial linguistic changes. If transparency had the potential of triggering change as an independent causal factor, we might expect for such change to be more systematic and for changes leading to increased opacity to be blocked. Furthermore, it would often remain unclear why a change happens when it does. The picture that emerges instead is that something in the data that language learners are exposed to must change beforehand for increases in transparency to be observed. Rather than transparency being an intrinsic causal factor driving language change, increases in transparency may primarily be the results of constraints on first and second language acquisition.