Colloquiums are held in 145 Straub, 3:00-4:30

Feburary 10

Professor Spike Gildea
Hierarchical alignment as a typological category

 There is debate in the literature over the validity of the category  Hierarchical Alignment (cf. Zúñiga 2006). In a recent statement from one side of this debate, Creissels (2009) points out: “In the definition of  ‘ergative alignment’ and ‘accusative alignment’, ‘alignment’ refers to possible similarities between the behavior of S and that of A or P, and  from a strictly logical point of view, this definition of alignment  leaves just two possibilities: either S = A /= P, or S = P /= A.   Consequently, the proliferation of terms including ‘alignment’ as one of  their components cannot be justified on the basis of this definition...”

 By this definition, Creissels is forced to exclude Hierarchical  Alignment as a possible type, because “the coding characteristics of A  and P are determined by their relative ranking with respect to some  hierarchy” and therefore S cannot consistently align with either A or P.  Similarly, when S does not form a consistent category (as in split  intransitive systems), that system must also fall outside the domain of  alignment typology. This presentation argues that alignment typology is better conceived of  in terms of CORE ARGUMENTS rather than A, S, and P.  Given that  foundational decision, semantic alignment (split intransitive systems)  again belong to alignment typology, and hierarchical alignment may be  seen in an interesting new light.  I am confident that this could be  shown for any language that has hierarchical properties, but for this  talk, I will focus only on indigenous languages of Latin America.  The simplest examples are the hierarchical prefix sets of the Cariban  and Tupí-Guaranían families, in which the transitive verb agrees with a  Speech Act Participant (SAP) in preference to third person, regardless  of grammatical role.  In Cariban, the form of the SAP prefix varies  depending on grammatical role of SAP: one set of prefixes marks SAP A  and SA, another marks SAP P and SP (Gildea 1998). Thus, for prefixes,  SAP SA aligns with SAP A and SAP SP aligns with SAP P, creating a  hierarchical split S system. Verbal person agreement prefixes in most  languages of the Tupí-Guaraní family (Jensen 1998) are quite parallel to  the Cariban pattern, aligning SAP SA with SAP A (acting on 3P), and  aligning SAP SP with SAP P (being acted on by 3A). Mayan language  Huasteco shows a similar system for transitive verbs, but for  intransitive verbs, the S category is a unitary entity.  The Huastec S  aligns with SAP P (as opposed to SAP A), creating a hierarchical  absolutive category. ).  In all three cases, the existence of aligning  morphological forms is clear, but not the centrality (nor the unity) of  S, A, and P. The last two examples remove A and P from the equation altogether.  In  Reyesano (Tacanan, Guillaume 2009) the verbal prefixes index SAP  regardless of role.  Third person A is indicated by a separate verbal  suffix, allowing speakers to deduce the role of an SAP despite the lack  of explicit marking.  The alignment between intransitive and transitive  clauses is thus purely hierarchical: S aligns with the higher-ranking  transitive core argument, regardless of role).  The opposite alignment  is attested in linguistic isolate Movima (Haude 2009): The two core  arguments of a transitive clause are the PROXIMATE, which is internal to  the VP, distinguished from the external OBVIATIVE by multiple  phonological and syntactic properties.  Semantic roles are coded via  direction morphology, with the DIRECT verb indicating that the proximate  argument is agent and the INVERSE verb indicating that the proximate  argument is patient.  The S of an intransitive clause aligns perfectly  with the obviative of the transitive clause (leaving out any role  whatsoever for A and P).  In sum, core arguments continue to show the grammar of alignment even  in those languages where there is no clearly identifiable (and/or  unified) S, A and P.  Therefore alignment typology needs to be able to  include categories of alignment patterns that reference factors beyond  S, A and P, including verbal subcategorization (for semantic alignment)  and relative position of an argument on the referential hierarchy (for  hierarchical alignment), and sometimes both together.

 Creissels, Denis. 2009.  Ergativity/Accusativity Revisited.  Presented at the eighth meeting of the Association for Linguistic Typology,  Berkeley, July 25.

 Gildea, Spike. 1998. On reconstructing grammar: comparative Cariban  morphosyntax. Oxford Studies in Anthropological Linguistics, vol. 18.  Oxford: Oxford University Press.

 Guillaume, Antoine. 2009. Hierarchical agreement and split  intransitivity in Reyesano. International Journal of American  Linguistics 75.29-48.

 Haude, Katharina. 2009. Hierarchical alignment in Movima. International  Journal of American Linguistics, 513-532.

 Jensen, Cheryl.  1998.  Comparative study: Tupí-Guaraní. Handbook of  Amazonian languages, v. 4  ed. by Desmond C. Derbyshire and Geoffrey K.  Pullum.  New York : Mouton de Gruyter.

 Zavala, Roberto. 1994. Inverse Alignment in Huastec. Función  15/16.27-81

 Zúñiga, Fernando.  2006.  Deixis and Alignment: Inverse Systems in  Indigenous Languages of the Americas. Amsterdam: John Benjamins.