Distributed Morphology Today: Morphemes for Morris Halle

7. Cycles, Vocabulary Items, and Stem Forms in Hiaki

7 7.1 Introduction The analysis of arbitrary morphological classes has a number of architectural implications in Distributed Morphology (Halle and Marantz 1993). There is no central repository of Saussurean ‘words’ in the framework—no soundmeaning pairings that are the building blocks for both phonological and semantic sentence-level representations. Instead, there are separate lists. One list contains all syntactic and semantic information necessary for the derivation of a well-formed LF representation, and forms the input to the syntactic derivation . A second list, the Vocabulary, describes the phonological realizations that are inserted as exponents of particular syntactic terminal nodes, following all syntactic operations. This raises the question of where class features are located. What elements do rules which are sensitive to class membership refer to? Are they sensitive to properties of the abstract syntacticosemantic formatives of the first list? Or are they instead sensitive to properties of the phonological exponents, the Vocabulary Items in the second list? In Distributed Morphology, any features which are syntactically (and possibly semantically) active must be a property of the abstract morphemes which are input to syntactic derivation. Embick and Halle (2005) treat even Latin conjugation class features in this fashion, attaching them to the roots in the first list, input to the syntax. In contrast, we will argue that class features in Hiaki are not properties of roots in the syntax but rather are properties of Vocabulary Items, the phonological exponents inserted at the end of the syntactic derivation. Irregular morphophonological rules (readjustment rules) apply to a particular class of Vocabulary Items in the appropriate morphosyntactic environment. Classifications of this kind play no role in the syntactic/semantic computation, but are crucial in triggering the application of the appropriate morphophonological rule to yield the correct surface form in such cases. The existence of such Cycles, Vocabulary Items, and Stem Forms in Hiaki Heidi Harley and Mercedes Tubino Blanco 118 Chapter 7 morphophonological classifications, irrelevant to syntax, is thus an argument against the lexeme, as such, and in favor of the DM-style separation of the two lists: List 1, input to the syntax (the source of the Numeration in Minimalist syntactic theories), and List 2, Vocabulary Items which simply realize the output of the syntax. Further, the Hiaki case provides a clear argument for Vocabulary Insertion applying to Root elements (l-morphemes, in Harley and Noyer 2000’s terminology), as well as to f-morphemes. In addition, the notion of a phasal cycle within the word proves useful in permitting a simple statement of the relevant conditioning context for the application of morphophonological rules. The Hiaki case presents many of the same morals for the architecture and the notion of “stem” as the Latin perfect does, as elucidated in Embick and Halle 2005, but some of the issues arise in even starker relief due to the cross-categorial nature of the stem classes, and especially due to the interaction of the stem classes with suppletion, in particular with the suppletive roots of Hiaki. 7.2 Hiaki Stem Classes In Hiaki (Yaqui), lexical stems have bound and free alternants. The bound forms are used as the base for affixation of (broadly speaking) derivational morphology, while the free forms are the base for affixation of (broadly speaking ) inflectional morphology, and also of course can stand alone, without any affix. The bound and free stems for poona ‘play, beat’ and kiima ‘bring (pl. obj)’ are illustrated in (1), the latter affixed with a derivational suffix. (1) Free Bound a. poona pon-tua play play-caus ‘is playing’ ‘is making (someone) play’ b. kiima kima’a-tua bring.pl bring.pl-caus ‘is bringing (things)’ ‘is making (someone) bring (things)’ An almost-complete list of the free-stem-selecting and bound-stem-selecting suffixes are listed in (2a) and (2b) respectively: (2) Bound-stem and free-stem suffixes of Hiaki (Harley and Tubino Blanco 2010) a. Hiaki verbal suffixes that require the bound stem (in no particular order) -tua (caus) -‘ea (desid) -su (compl) -se/-vo (go) -pea (desid) Cycles, Vocabulary Items, and Stem Forms in Hiaki 119 -tevo (icaus) -ri (obj.ppl) -la (ppl) -taite (inch) -‘ii’aa (desid) -ria (appl) -tu (become) -ri (ppl) -naate (inch) -hapte (inch) -roka (quot) -vae (prosp) -le (consider) -wa (pass) -ne (irr) -yaate (cess) -siime (go along) -sae (dir) -na (pass.irr) b. Hiaki verbal suffixes that require the free stem (again in no particular order) -k...