1. University of Virginia
  2. Arts & Sciences

Linguistic Anthropology Seminar -- 2008 - 2009

Fall 2008

Friday, August 29, 3-5pm (Brooks Hall Library)

Ellen Contini-Morava, Department of Anthropology, University of Virginia

"The Message in the Navel: (Ir)realisness in Swahili"

What does it mean to say that a language L "encodes" some semantic substance? One answer (Position A) is that (at least some) messages communicated by means of expressions in L can be interpreted according to that semantic substance. Another (Position B) is that specific grammatical forms in L conventionally, and invariably, signal information related to that semantic substance. According to Position A, the semantic substance exists independently of individual languages—presumably as a cognitive universal—and may find expression in a variety of ways without being regularly conveyed by any particular linguistic form. As such, it is in principle "encodable" in any language. For example, Chafe (1995:363) argues that "the realis-irrealis distinction…can be thought of as a covert semantic pressure that emerges in different languages in different ways", even in languages like English which lack overt morphosyntactic marking for this substance. Similarly, Givón (1994:268) lists "irrealis assertion" as a subcategory of epistemic modality that is universally applicable, even to such phenomena as propositions within the scope of the adverb maybe in English (as in Maybe she left, p. 271). In Position B by contrast, each language articulates its own grammatical content, and a given semantic substance may or may not be "encoded in" its grammatical forms. For example, Elliott’s (2000:57) survey includes "only languages where there is evidence of an obligatorily marked grammatical category for reality status". The question of the status of realis/irrealis as a "grammatical category" will necessarily be answered differently depending on how one conceives the relationship of meaning to morphosyntactic form (and of course depending on how one defines realis/irrealis).

I will argue that Position B is more methodologically sound, as it is tied to specific linguistic forms which serve as a control over the scope of the analysis. By contrast, Position A is at best an analysis of the messages that linguistic forms may be used to communicate. My data will be taken from Swahili, a Bantu language of East Africa, which encodes a number of distinctions related to modality from which messages of (ir)realisness can be inferred.

Suggested Reading

Elliott, Jennifer R. 2000. Realis and irrealis: Forms and concepts of the grammaticalization of reality. Linguistic Typology 4: 55-90.


Wednesday, October 15, 4-5:30pm (Brooks Hall Library)

Prashant Pardeshi, Kobe University

Peter Hook, University of Virginia

Two Perspectives on the Compound Verb in Indo-Aryan

1. "Synchronic exploration in search of a diachronic path: An areal-typological study on the grammaticalization of PUT/KEEP in Northeast, Central and South Asian languages" (Pardeshi abstract)


2. "Using Google to Find Parallel Functions in Divergent Forms: Perfectivity in Russian and Hindi-Urdu" (Hook abstract)


Friday, October 17, 1-3pm (Brooks Hall Library)

Tara Sanchez, Department of Anthropology, University of Virginia

"What the Multilingual Community Contributes to a Theory of 'Speech Community'"

The concept of 'speech community' is central to sociolinguistic research, yet researchers neither agree on its definition nor thoroughly understand its inner workings. In an effort to contribute to both endeavors, I summarize empirical findings from my own work in a multilingual community as well as those reported in the literature on various multivariety communities (e.g. bilingual, bidialectal, etc.), and relate them to previous proposals regarding the 'speech community'. The theory proposed here is primarily based on ethnographic observations and linguistic data (129 sociolinguistic interviews, as well as a diachronic corpus of written texts) collected by this author in 2003 on the Dutch Caribbean islands of Aruba, Bonaire, and Curaao. Interviews were conducted in Papiamentu (Iberian creole); Papiamentu speakers are also fluent in, and use daily, Dutch, Spanish, and English.

In this paper, I propose that 'shared linguistic norm' central to defining a speech community originate in more or less tangible 'subgroups' of same. In contact situations, such norms may be (are often?) the direct result of language transfer, and as such, 1) can be transparently traced back to their origin and 2) serve as identity markers of these linguistically defined subgroups. In monolingual communities, the connection to such subgroups (e.g. women, lower middle class, etc.) may not be as transparent, but it is no less real. In fact, I argue that this finding from the multilingual community may help explain the community-wide distribution of previously reported variants in monolingual situations.

Suggested Reading

Patrick, Peter L. 2002. The Speech Community. In J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes (eds.), The Handbook of Language Variation and Change, pp. 573-597. Malden, MA: Blackwell.


Friday, October 31, 1-3pm (Brooks Hall Library)

Suzanne Menair, Department of Anthropology, University of Virginia

"From Mischief to Misunderstanding: Verbal Play and Race Relations on Chicago's Trading Floor"

Suggested Readings

Bauman, Richard. 2004. Introduction: Genre, Performance, and the Production of Intertextuality. Pp. 1-11 of A World of Others' Words: Cross-cultural Perspectives on Intertextuality. Malden, MA: Blackwell.


Thursday, November 13, 4-5:15pm (Brooks Hall Library)

Eve Danziger, Department of Anthropology, University of Virginia

"Once More with Feeling: A Forbidden Performance of the 'Great Speech' of the Mopan Maya"

An utterance whose performativity cannot be suppressed even by quotation or staging achieves this effect through sustained indifference to the mental state of the proximate utterer. We can say that no relevant distinction of Goffmanian ‘animator’ from ‘principal’ exists in these cases. One genre of Mopan discourse that has this property is respect-greeting talk among kinsmen. Such talk cannot be simulated by those who are not in the appropriate relationships to one another, for fear of supernatural consequences. Above all, sexual partners may not engage in it. On one occasion early in my fieldwork however, a Mopan husband and wife did agree to demonstrate this genre to me. They were well aware that what they were doing was highly taboo, as their use of containment strategies makes clear. They insisted on anonymity, and they limited the extent of their violation, refraining from direct address to one another. Finally, at certain crucial moments in their speech, they giggled nervously. Laughter is actually forbidden in the context of respect-greeting speech, so that this constitutes a clear, if momentary (and apparently involuntary) intrusion of the performers’ animators into the personae of their principals. I propose from this example to approach a more general consideration of the conditions under which utterances that are immune to hedging by explicit quotation may sometimes nevertheless be mitigable by implicit quotation via double-voicing.

Suggested Readings

Du Bois, John W. 1986. Self-Evidence and Ritual Speech. In W. Chafe and J. Nichols (eds), Evidentiality: The Linguistic Coding of Epistemology, pp. 313-336. Ablex Publishing: Norwood NJ.


Friday, November 14, 3-4:30pm (Kaleidoscope)

Mark Turin, Cornell University & University of Cambridge

"The Linguistic Survey of Sikkim: Preliminary Results, Methodological Challenges, and Social Context"

From September 2005 to November 2006, under the auspices of the Namgyal Institute of Tibetology and in close partnership with the Government of Sikkim, India, I directed the first phase of a modern linguistic survey of Sikkim. During the survey process, the research team visited 105 government and private secondary schools across Sikkim to administer an extensive questionnaire on language use to students in classes 8-12. The preliminary results of these 17,000 completed survey forms offer insights into the process of language shift from indigenous mother tongues to regional vernaculars, the growing importance of linguistic heritage and feelings of group belonging over actual competence in specific languages, and the symbolic and practical steps taken by the state government to support linguistic diversity in Sikkim.

Dr. Turin's lecture is generously cosponsored by the University of Virginia Linguistics Program and the Center for South Asian Studies. A reception will follow the lecture.


Spring 2009

Wednesday, February 11, 3-5pm (Folklore Archive, basement of Brooks Hall)

Erik Larsen and Neal Fox (University of Virginia)
"A Comparative Optimality Theoretic Outlook on Loanword Phonology in Huave Dialects in Mexico"

This paper compares the success that the linguistic constraints proposed by Davidson and Noyer (see background reading) have in explaining the phonological changes inherent in loanword nativization in two dialects of Huave: San Mateo and San Dionisio del Mar. This comparison is made between the 469 loanwords collected from San Mateo by Davidson and Noyer and 246 loanwords compiled from sixteen stories transliterated from San Dionisio by Radin in 1929. Analysis reveals that the constraint set of Davidson and Noyer performs well in explaining the changes in phonological structure of Spanish loanwords in San Dionisio; it fully explains 77% of all changes. Loanwords left unexplained by their constraints can be resolved by newly theorized constraints, including COD, which restricts certain codas involving consonant clusters, and MAXCON:I-O, a faithfulness constraint governing the maintenance of individual phonemes from the Spanish source words. The addition of these constraints improves the total percentage of explainable phonological changes to over 86%. When applied to the original sample set from San Mateo, these two constraints, when combined with another new constraint called A-DROP, which results in the loanword dropping a fronted “a” from the Spanish source word, improves the percentage of words fully described by the constraint system from nearly 40% to over 50%. Patterns in vocalic change are evident, but more research is required to determine constraints governing their changes. A new category for organization of the constraint system is proposed which distinguishes between phonemic and segmental changes in phonology. Dialectical comparison reveals that members of the same language family share the same general constraint ordering, but some differences are to be expected.

Suggested Reading

Davidson, Lisa, and Rolf Noyer. 1997. Loan Phonology in Huave: Nativization and the Ranking of Faithfulness Constraints. In B. Agbayani and S. Tang (eds.), Proceedings of WCCFL15. Palo Alto: CSLI Publications.


Tuesday, March 24, 3-5pm (Brooks Hall Library)

Coulter George, Department of Classics, University of Virginia.

"Semitic Influence on the Greek of the Septuagint and New Testament: Evidence from the Use of Prepositions"

Ancient Greek had several competing ways of expressing when an event happened: all three oblique cases (genitive, dative, accusative) as well as a host of prepositional constructions can be used to indicate when, for how long, or within what time frame a given event took place. But the standard accounts of these constructions fail to capture many of the conditions that trigger the use of one rather than another. My ongoing monograph project is aimed at determining more precisely what these conditions were, especially in classical Attic prose authors (the corpus currently includes Thucydides, Xenophon, and Plato, a range of approximately 430–350 bc). To a large extent, the choice of construction is connected with the actionality of the verb modified by the construction, the pronominal and adjectival modifiers of the noun in the expression, as well as lexical quirks associated with individual nouns of time (e.g. the words for night and day take different constructions). The current paper, however, represents the beginnings of the second major part of this project, namely the extension of the study to the Koine Greek of the Hellenistic and Roman periods. Two major texts for our knowledge of the Greek of this time are the Septuagint (translated from Hebrew probably in the third and second centuries bc) and the New Testament (written largely by speakers with at least some fluency in Aramaic in the late first century ad). Debate continues over the extent to which the Greek of these Judeo-Christian texts differs from earlier Attic because of diachronic development in the Greek language on the one hand, or interference from Semitic on the other. On the basis of the clearer understanding of the Attic constructions attained in the earlier stages of this project, the paper at hand will examine how closely the temporal constructions from the Septuagint and New Testament match those of Hebrew and Aramaic as opposed to Greek texts of the period that are less likely to have been subject to Semitic influence.

Suggested Readings

Christidis, A.-F. 2007. A History of Ancient Greek: From the Beginnings to Late Antiquity, CUP, pp. 638-653.

Horrocks, Geoffrey. 1997. Greek: A History of the Language and Its Speakers, Longman, pp. 56-59 and 92-95.


Friday, April 3, 1-3pm (Kaleidoscope, Newcomb Hall)

Patrick McConvell, Australian Institute of Aboriginal and Torres Strait Islander Studies

"From Pama to Nyungan: Change in Australian Kinship Terminologies from the North-east to the South-west"

Geoff O'Grady, who recently passed away, bequeathed to us a great idea: the Pama-Nyungan family of languages covering most of Australia. One of his major concerns was the linking together of cognates whose meanings were apparently quite disparate, and providing innovative explanations often based on specific metaphors and associations in Australian indigenous thought. Sometimes these hypotheses were, on his own admission, adventurous to the point of lacking caution, but the debate engendered is the spur to progress.

When it comes to kinship, semantic change is commonplace both worldwide and in Australia. The specific focus here is the Pama-Nyungan languages. Some of the semantic changes are universal as aspects of kinship are universal but others reflect specific types of social system in indigenous Australia. So a single root may have different meanings in different languages and the problem is what meaning or meanings to reconstruct to various proto-languages, including proto-Pama-Nyungan itself.

Various approaches have been proposed to answer such questions. One was that of Dyen and Aberle for the Athapaskan languages, but that has been criticised as too mechanistic and not sufficiently attuned to regional cultural factors by Blust and others. A number of proposals both old and recent have suggested that certain types of changes to systems are unidirectional. Other approaches see changes in meaning as proceeding via changes in usage and pragmatics, and an intermediate stage of polysemy (Evans and Wilkins). If we put this together with what we know of the constraints on types of kinship systems worldwide and in particular regions like Australia, the hypotheses about how systems might have developed and particular terms changed meanings are considerably narrowed down together with our choices of proto-meanings.

This paper looks at the types of meaning change found in Australian languages and how this fits together with general theories of kinship semantics and semantic change. Then a couple of examples are examined in detail to illustrate how this approach can enable us to arrive a stronger hypotheses about the meanings of proto-terms and the systems they form part of. These examples are illustrated using the AUSTKIN on-line database and mapping system.

(1) The terms *tyam(p)V and *ngatyi which have reflexes in many Pama-Nyungan and in the the case of the first at least, Non-Pama-Nyungan languages. Reflexes of these roots have a number of meanings in different languages and sometimes in the same language including preponderantly Mother's Father, cross-cousin (related to MF by the alternate generation equivalence principle) and spouse/sibling-in-law (based on cross-cousin marriage). It is implausible however that there should be two terms with the same meaning (MF) in a single proto-language so ways out of this impasse are discussed.

(2) The term *kaala- means Mother's Brother (or younger Mother's Brother) in north-eastern languages and cross-cousin/spouse elsewhere. Here there are also languages in which there is polysemy with both meanings, following what is known as Omaha skewing. This provides the bridging context for the change from uncle to cousin.

Co-sponsored by the Department of Anthropology and the Linguistics Program.


Monday, April 13, 7:30-9:30pm (Brooks Hall Library)

Joel Robbins, Department of Anthropology, UC San Diego

"Some Things You Say, Some Things You Dissimulate, and Some Things You Keep to Yourself: Linguistic and Material Exchange in the Construction of Melanesian Societies"

This workshop will be a joint discussion with the Department of Anthropology Sociocultural Workshop of a precirculated draft paper, which can be accessed here.