This book "supplies a vocabulary of English words and idiomatic phrases 'arranged … according to the ideas which they express'. The thesaurus, continually expanded and updated, has always remained in print, but this reissued first edition shows the impressive breadth of Roget's own knowledge and interests."
Danziger, Eve (2001) Relatively speaking: Language, thought, and kinship among the Mopan Maya, Oxford University Press, 125 pp, Oxford Studies in Anthropological Linguistics #26 (Announced on LINGUIST 12.1139; 12:1006; 12:393.3)
Rusty Barrett, University of North Carolina, Chapel Hill
The linguistic relativity hypothesis, the claim that the particular language one speaks influences the way in which one thinks about reality, has long been a central question in the broader study of language as it relates to thought (for reviews of the hypothesis, see Lucy 1992, 1997). Based on fieldwork conducted in Belize, Danziger considers the relativity hypothesis in the specific context of the use of kinship terms in Mopan Maya. By focusing on kinship terms, Danziger approaches the question of linguistic relativity using what Lucy (1997) has called the "domain-centered approach," examining the ways in which languages encode a particular domain of experienced reality. In reviewing research in the domain-centered approach, Lucy (1997: 299) notes that many studies in this area are problematic both because they fail to demonstrate "structural coherence" of the domain "on language-internal grounds" and because they tend to emphasize what one can say (reflective thought) rather than what is structurally salient (habitual thought). Danziger overcomes these problems through a combination of ethnograhpic study of Mopan kinship and empircal research on the acquisition of kinship terms by Mopan children. Through the ethnographic research, the language-internal structure of the kinship domain and the range of reflective thought regarding kinship are clearly outlined. The empirical research on acquisition, however, focuses on habitual thought, demonstrating a language-specific system for conceptualizing kinship. The combination of approaches also demonstrates that reflective thought (as evidenced through prototype categories) and habitual thought (in the traditional sense of the Sapir-Whorf hypothesis) are not mutually exclusive and can both function simultaneously in the minds of speakers.
Danziger proposes three possible semantic analyses for considering the relationship between language and an external domain of experience: monosemy, polysemy and homonymy. Following Whorf (1956), the monosemy approach assumes the linguistic relativity hypothesis and holds that specific languages will have specific means of encoding experienced reality. Under monosemy, the various classes of referents associated with a single lexical item will form a conceptual system specific to a given language. The polysemy approach uses prototype categories (e.g. Lakoff 1987, Palmer 1996, etc) in which a central referent is the prototypical member of the category. The polysemy approach need not oppose linguistic relativity unless one holds that the prototype members of a category will be universal. For the purposes of her analysis, Danziger defines the polysemy approach as claiming psychological primacy for the prototype of a category and assigning less psychological weight to more peripheral referents. The homonymy approach holds that there is no pscyhological connection between a lexical item and various classes of referents (such as with the English words "bank" meaning financial institution and "bank" meaning the edge of a river).
In the ethnographic sections of the book, Danziger determines how each of these three semantic analyses would model adult cognition with regard to kinship terms in Mopan. In order to ensure that the domain is defined in language- internal terms, Danziger examines the Mopan concept of "tzik" or respect which applies to particular social relationships. For the Mopan, one is required to formally greet senior relatives with a "tzik" greeting each time they meet or pass on the street. Absence of the "tzik" greeting when it would be expected is very significant, indicating either that incest has occurred between the two or that extreme anger exists between them. These "tzik" relationships include ties made through "compadrazgo" (co- parenting), such as godparents who accompany children through their sacrements in the Catholic church. Thus, the range of kinship terms Danziger considers goes beyond geneological ties and reflects a system particular to Mopan society. The ethnographic section also considers cases in which the actual kinship term for a particular individual may be under dispute in order to demonstrate how the particular terms might be viewed as fuzzy prototype categories.
Having laid out the structure of the Mopan kinship system, Danziger tests each of the three semantic approaches empirically by means of testing the stages of acquisition of kinship terms following the system of Piaget (1928). In the Piagetian analysis, children acquire kinship terms at particular stages, depending on the relational complexity of a given term. Children first pass through the absolute or categorical stage, in which they offer an absolute characteristic as the definition of a term (e.g. a "brother" is a "boy"). This is followed by the relational stage, in which children recognize that a term denotes a particular relationship, but where the relationship is seen as a property of a particular referent (e.g. recognizing that male children of the same couple are brothers, but not recognizing that having a brother also makes one a brother or sister). The final stage (the reciprocal stage) is that in which a child defines a term that recognizes the reciprocal nature of a term (e.g. in order to be a brother one must also have a brother or sister). Danziger demonstrates that according to the Mopan system of kinship each of the three semantic analyses (monosemy, polysemy and homonymy) predicts a different pattern for the acquisition of Mopan kinship terms by children.
Danziger tests these predictions, focusing on two particular terms, suku'un and tataa'. The Mopan term suku'un refers to older male blood relatives in the same generation as the speaker. As such, suku'un refers primarily to an older brother, but may also refer to older cousins, as well as uncles and nephews who happen to be in the same generation as the speaker. The term tataa' refers to older male relatives in a different generation from the speaker (who are not the speaker's father). Tataa' thus, refers primarily to one's grandfather, but may also refer to uncles and great-uncles. Thus, what English would designate as "uncle" may be either suku'un or tataa' depending on the difference in age between speaker and referent. Under the monosemy analysis these two terms should be acquired at the same stage because within the Mopan system they are equal in terms of relational complexity (older relative that is not a parent). The polysemy analysis, with it's focus on central referents, would predict that suku'un would pass through Piagetian stages earlier than tataa' because the central referent of tataa' (grandfather) is more relationally comples (X is parent of A; A is parent of Y) than the central referent of suku'un (brother) (X is a sibling of Y). Under the homonymy analysis the term suku'un must be analyzed as including those referents referred to in English as "cousins" (X child of A; A sibling of B; B parent of Y), which is more relationally complex than any of the referents for tataa'. Thus, the homonymy analysis predicts that tataa' will be acquired before suku'un. The empirical tests of Piagetian acquisition show that the two terms pass through acquisition simultaneously, indicating that the monosemy analysis best suits the relationship between Mopan kinship terms and their external referents. Thus, although in reflective conscious thought it is possible to determine a prototype referent for the particular categories, in habitual unconscious thought the categories function with sharp boundaries. Because it is precisely this habitual unconscious monosemic conceptualization that serves as the basis for the Sapir-Whorf hypothesis, Danziger offers her study as evidence in favor of the linguistic relativity hypothesis.
Perhaps the greatest achievement of this work is its unique combination of ethnographic and quantitative methodologies. It is this combination that allows the study to rise above the many problems facing anyone who attempts to examine the relativity hypothesis. It also begins to address the relationship between reflective and habitual thought, an area that will make it possible to begin to clarify the relationship between prototype semantics and linguistic relativity. This aspect of the book will be of greatest interest to linguists focusing on lexical and cultural semantics. The ethnographic research is itself a major contribution to Mayan ethnography. Danziger provides an excellent description of the Mopan system of kinship (and its role in establishing and maintaining social relationships) and her analysis of the importance of the 'tzik' concept of respect will be especially useful to anthropologists and Mayanists. In addition, the ethnographic data give special attention to the lives and viewpoints of Mopan women and thus represent a major contribution to the study of gender in Maya culture. The only disappointment in this work is its relatively short nature (104 pages of actual text). Given the richness of the analysis one might be left wishing for further information and more discussion of the theoretical implications of the research.
Lakoff, George. 1987. Women, Fire and Dangerous Things: What Categories Reveal about the Mind. Chicago: Chicago University Press.
Lucy, John A. 1992. Language Diversity and Thought. Studies in the Social and Cultural Foundations of Language No 12. Cambridge: Cambridge University Press
Lucy, John A. 1997. Linguistic Relativity. Annual Review of Anthropology 26: 291-312.
Palmer, Gary 1996. Toward a Theory of Cultural Linguistics. Austin: University of Texas Press.
Piaget, Jean. 1928. Judgement and Reasoning in the Child. Trans. M Warden. New York: Harcourt, Brace, Jovanovich.
Rusty Barrett is a visiting assistant professor at the University of North Carolina, Chapel Hill. His research interests are Mayan languages, linguistic anthropology/sociolinguistics, language and gender and phonology. In the fall he will be a visiting assistant professor in the linguistic department at the University of Michigan, Ann Arbor.