Festschrifts

With an analysis of the inflectional properties of Luiseno words, this paper builds on the examples offered in Aronoff 1994 of 'morphology by itself', of morphological generalizations not plausibly analyzed as anything other than morphology. All Luiseno words share four attributes---three of which are notional; the fourth serves to make the word accessible to the syntax. The value for the latter can be independent of the former, but it need not be. This interdependence is a purely morphological phenomenon.

This paper investigates the morphosyntax of the Icelandic DP, following work by Sigurdsson (1993, 2006); Julien (2005). In addition to providing an analysis of the basic structure of the Icelandic DP, this paper investigates two (apparent) movements within the DP: one involving the definite suffix -inn, and one involving pronominal possessors (e.g. minn 'my'). I show that the fronting of pronominal possessors cannot be analyzed as movement and instead must indicate a use of pronominal possessors as demonstratives. In contrast, the suffixed article does involve movement, and I argue that it is phrasal movement, contra the head movement accounts proposed by Sigurdsson (1993, 2006). I show that an analysis where the prenominal article and suffixed article occupy the same position cannot be maintained, which is surprising given that they never surface at the same time. Coupled with this is a requirement, that PP complements to N extrapose to the right edge of DP, which is thus another case of the strange behavior of complements to N in Scandinavian languages (see Hankamer and Mikkelsen (2005, 2008) on Danish). Finally, I briefly investigate the syntax of some cases in English where apparent semantic complements do not occupy a syntactic complement position.

Possessive (nominal) and indirect object (clausal) clitics are homophonous within the Balkan Slavic languages and Romanian. Pancheva (2004) shows that this syncretism is not just morphophonological but that the two types of clitics constitute identical feature bundles bearing dative case. Yet, these dative clitics seem to exhibit distinct behavior in the nominal and clausal domains: in Bulgarian the nominal clitics appear in second position within the nominal phrase while the clausal clitics are verb-adjacent and non-initial within the clause. It is puzzling that the same syntactic objects exhibit such different distributional patterns. I argue that in Bulgarian this seemingly distinct behavior follows from the interaction of a distributional constraint on dative clitics, NonInitiality within Spell-Out domains, and the different structural properties of the syntactic domains they are associated with. In particular, a number of constituents can be pre-clitic in clauses because various structural positions are available above the clitic, while in nominal phrases no comparable positions are available. Besides the direct consequences of this approach for the treatment of cliticization, it also provides an insight into the nature of Spell-Out domains, nominal and clausal structure, and the nature of syntax/PF interactions.

This paper examines inflection on fronted verbs in Danish. In both VP topicalization and VP left dislocation with resumption, the inflection on the fronted verb is governed by an in situ auxiliary, suggesting that the fronted VP originates as a sister to that auxiliary. This analysis is straightforward for VP topicalization, but fails for VP left dislocation. Yet, the two show identical patterns of inflection, down to systematic covariance in case of interspeaker variation. I take this as evidence that the two fronting constructions have the same syntax and vary only in whether the proform that mediates between the auxiliary and the fronted VP is overt or null. That in turn implies that verbal inflection can be governed at a distance, and that some mechanism other than the standard generative ones (affix-hopping, selection, and feature valuing) is involved.

This article examines suppletion and feature-conditioned allomorphy in Distributed Morphology. It discusses some empirical commonalities among these kinds of allomorphy, and then examines how they could be accounted for in DM. It then moves on to a particular case of irregularity, the Japanese verbal honorific. Some of these honorifics are shown to fit the criteria for suppletion and feature-conditioned allomorphy. However, it is then shown that the common notion of the cycle within DM cannot treat these honorifics as cases of allomorphy, suggesting that the phonological cycle must be larger than the version argued for in Embick 2010.

This paper proposes a unified model of the morphosyntax and morphophonology of the Modern Standard Arabic verbal system which attempts to preserve the empirical and analytical observations from recent Optimality-Theoretic approaches to templates in Semitic (Ussishkin 1999, 2000, 2005) as well as the observations from Distributed Morphology concerning argument structure and morphemic composition (Arad 2003, 2005). In doing so, a clausal syntax for Arabic is proposed which does not crucially rely on an Agr(eement) Projection as a landing site for subject movement. This is done using arguments from VP-adverb placement, negative clitic placement, and word order in perfective periphrastic verbal constructions in order to motivate the syntactic structure. This structure is then shown to pose a problem for modern theories of morphological linearization (Pak 2008; Embick 2010). Finally, the linearization problem is resolved by appealing to prosody as the mechanism for linearization, following recent proposals in morphophonology (Kramer 2007, Tucker 2011b). This move is motivated by data from Arabic Hollow Verbs which confirm the predictions the model makes with respect to allomorphic sensitivity of morphemes to each other over nonconcatenative (and therefore nonadjacent) distances. Finally, the implications of these findings for morphological and syntactic theory are discussed.

It has long been debated whether the object marker in Amharic is a reflex of object agreement (Amberber 1996; Yimam 2006) or a doubled object clitic (Mullen 1986; Yabe 2001). I re-open this debate, evaluating previous argument and developing several new arguments for the object marker being an object clitic. If agreement is analyzed as the Minimalist relation Agree, several diagnostics relating to case, the morphophonology of object markers, and what happens when no object marker is possible, all demonstrate that the object markers are doubled clitics. The paper thus serves as a case study in how to distinguish agreement and clitics using multiple diagnostics (cf. Preminger 2009).