What\u27s with \u27why? On the non-existence of why-clefts in Nukuoro

Crosslinguistically, there exists ample evidence that some ‘why’ questions are unique among wh-questions in their syntax. While the idiosyncrasies of ‘why’ are diverse, they may be unified under an account in which ‘why’ is base-generated in the left-periphery (Collins 1991, Rizzi 2001, Ko 2005, Stepanov and Tsai 2008, Thornton 2008). Rizzi observes that for Italian, only ‘why’ may co-occur with focused elements (2001). Other wh-phrases must compete with ‘why’ for a focus position. Meanwhile, in some wh-in-situ languages, ‘why’ is shown to pattern differently with respect to scrambling effects (Ko 2005, Stepanov and Tsai 2008). Korean and Japanese have a rigid constraint against LF movement of wh-elements across a scope bearing particle. Thus, wh-phrases may not scramble to surface after such a particle. Yet Ko shows that ‘why’ may follow scope bearing particles, suggesting that ‘why’ need not undergo LF movement across them because it is base generated high (2005). Given the growing body of literature on the crosslinguistic behavior of ‘why’, it is surprising that none have examined the behavior of ‘why’ questions in a language which employs a wh-clefting strategy. This paper seeks to fill this empirical void examining the status of ‘why’ in Nukuoro [nkr], an SVO Polynesian Outlier language of Micronesia. Drawing on novel field data, I argue for a high initial merge position for gu aha ‘why’ in the language. In Nukuoro, a wh-in-situ (ex: 1a) and wh-clefting strategy (1b) are in free variation as wh-question formation strategies. The core evidence for a cleft analysis of Nukuoro wh-questions comes from the genitive relative construction (GRC), which marks surface subjects of relative clauses as possessors (see Otsuka 2010, Herd et al 2011, Baker 2012). That this same genitive marking appears on subjects in wh-questions reveals an embedded relative clause. (1) a. koe ne hegau aha i de hale goloa anaahi you pst buy what loc def.s house supply yesterday ‘What did you buy at the store yesterday?’ b. ni aha aau ne hegau anaahi pl what your pst buy yesterday ‘What did you buy yesterday?’ (2) gu aha koe/ *aau ne hegau denga daagoli anaahi why you/your pst buy def.pl taro ‘Why did you buy taro yesterday?’ Puzzlingly, as (2) illustrates, ‘why’ questions using gu aha do not allow for a genitive subject, suggesting that gu aha questions are not in fact clefts. Using diagnostics such as the distribution of gu aha ‘why’ in focus contexts, intervention effects, and the ability of gu aha to appear in canonical adverbial positions, I argue for a why-in-situ account of Nukuoro gu aha. In particular, I propose that gu aha is externally merged in the left periphery and does not undergo any additional movement (cf. Rizzi 2001, Ko 2005). In addition, I contrast the patterning of gu aha with a synonym, hiidenga aha ‘what reason’ and show that why-in-situ is subject to language-internal variation. Like English why, hiidenga aha may be clefted and patterns with regular wh-adjuncts. I argue that gu aha behaves similarly to how come in English, which has also been argued to start off high (Collins 1991). This intralinguistic variation supports and extends the broader observation that why-in-situ is parameterized between languages, some languages (ex: Italian) externally merge ‘why’ in the left periphery, while others (ex: German) do not (Stepanov and Tsai 2008, Thornton 2008)