Referring to persons in Maya conversation* This paper examines how speakers of Yucatec Maya refer to and talk about other people in the course of ordinary talk. As Schegloff (1996) noted, the vast majority of reference to persons takes the form of self-reference ‘I, we,' or reference to an addressee, ‘you.' But a great deal of conversation is also given over to talk about other, non-present people, typically refered to in the ‘third person.' Common third person expressions include pronouns, deictics, proper names and titles, and descriptions of various sorts. Roughly speaking, the more interactants know about the persons they refer to, the more possible ways they could formulate the reference. Different refering expressions activate different sorts of background knowledge. If we use a full proper name, for instance, the activated knowledge may include descent (eg my last name comes from my father's last name), affiliation (eg hyphenated last names among the married), ethnicity (Santoro vs. Zapata vs. O'Larkin) and so on. By contrast, if John Larkin is refered to as Jack or with the nickname JT, then there is an overlay of familiarity suggesting that the speaker has special knowledge or access to the individual. A kinship term like “my mother's brother” obviously encodes a great deal of relational information, which is organized around the speaker as the center of a kin schema (although adressee or another person would serve just as well). If on the other hand, we refer to this individual as “the guy in front of the house,” then we have selected a spatial frame of reference with highly specific entailments (Levinson 2003). In selecting among these (and other) ways of refering to a person in talk, the speaker in effect formulates the individual from a certain perspective. The combination of individuated reference plus formulation under a perspective is what I have called construal (Hanks 2005). Construal is an interactive, indexical phenomenon which must ultimately be studied at the token level. We will say that a speaker who refers to a “third' person unavoidably construes him. In this respect, it is similar to actual usage of honorific forms in a language like Japanese: a speaker has access to several alternate construals of the situation, the referent or the addressee, and the selection of honorific form unavoidably signals a social construal. As Ide (1989, 2005) has shown, in order to select honorifics, a speaker of Japanese must “discern” the current field of utterance in order to attune herself to normative expectations of propriety or elegance, also known as wakimae . Similarly, in construing an individual, the speaker tacitly discerns the utterance situation, thereby attuning to prefered or normative expectations. In English, I usually refer to my parent by a possibly shortened kin term (my mother, mom, my father dad) and not by full proper name (Anne M. Hanks, William Hanks) except under special circumstances. I refer to a judge by title plus last name (Judge Jones) and not by nickname or kin term. Some of the discernment is clearly guided by situational relevance, but there are other factors as well. In Hanks (2005) I argued that deictic construal in Yucatec Maya is guided in part by rules of thumb that simplify referential practice, making selection of deictics in ordinary situations more automatic, less calculative than it might otherwise be. In the present paper, I argue that reference to persons in Maya is guided by normative preferences as to which properties of persons make the best or prefered basis for refering to them. Roughly speaking, if the referent is mutually known to be a kinsman of either Spr or Adr, it is the kin term that is prefered. Yucatec usage appears different from English as described by Schegloff (1996), but similar to Tzotzil (Haviland, forthcoming), Tzeltal (Brown,forthcoming) and Lao (Enfield, forthcoming). In Maya, I will suggest that there is a preference hierarchy of expressions that guides speakers, as shown in (1), where the leftmost term is the more highly ranked: The paper is presented in two parts. Part 1 summarizes the linguistic resources in Yucatec Maya for making reference to persons, with notes on the kinds of background knowledge they imply when used. Part 2 then presents a series of transcribed interactions in Maya, in which the linguistic resources are used in locally initial recognitional reference. The evidence indicates that when Maya speakers construe persons, they tend to do so in terms of kinship, in contrast to American English speakers' preference for proper name. ----------------------- |