Edited by Lotte Sommerer and Evelien Keizer
[Studies in Language Companion Series 221] 2022
► pp. 395–428
It is generally accepted that coordinated compounds known as dvandvas are attested in some languages but not in others. For instance, they are very common in Japanese (e.g. dan-zyo ‘lit. male-female = male and female’) but unattested in English (e.g. * male-female), which instead requires phrasal coordination. The present paper uses Competition Theory to explore where this cross-linguistic variation comes from, focusing on the unattestedness of dvandvas in English. This theory assumes that cross-linguistic variations occur when morphology and syntax compete to realize an underlying structure; languages are classified into two types in a macroparametric manner, depending on whether they prefer morphological or syntactic realization. Taking a competition-theoretic approach, this paper claims that the unattestedness reflects the nature of English as a particular type of language, which is determined by a macroparameter. Furthermore, Competition Theory predicts that dvandvas have phrasal counterparts in English. The correctness of this prediction will be verified by observing that a particular type of coordinated phrase exhibits the defining property of a dvandva regardless of their different surface forms. It will be proposed that the relevant type is derived in a specific way, which results in its dvandva status.