Urituyacu River, Loreto, Peru

Fieldwork Forum (FForum)

Department of Linguistics
University of California, Berkeley

When?Tuesdays 9:30am - 11am
Where?1303 Dwinelle Hall
What?We are a working group dedicated to the critical examination of methodologies in language documentation, description and revitalization, as well as to the linguistic and ethnohistorical analysis that falls out from that work. Our aim is to learn from and ultimately improve upon methods for carrying out more rigorous, insightful and ethical linguistic and cultural fieldwork, and to help researchers implement those methods.
How?Fieldwork Forum is made possible through a Working Group Grant provided by the Townsend Center for the Humanities at the University of California, Berkeley.
Who?FForum is organized by Erin Donnelly and Zachary O'Hagan. We welcome all those interested in linguistic fieldwork, with all levels of experience, including those in other departments.

Spring 2014

Apr 22Eric Campbell (University of Texas at Austin)
Exploring Morphosyntax and Culture in Texts
Fieldwork is the means to carrying out language documentation, and documentation is the basis for carrying out grammatical description, which in turn advances linguistic theory and typology. Naturalistic texts about culturally relevant material give us the best chance to capture how a language is really used by speakers (see discussion in e.g. Evans & Dench 2006). My goal in this talk is to present an analysis of a Zenzontepec Chatino (Otomanguean, Mexico) text in order to highlight some of the most characteristic aspects of the structure of the language. The text is kwentū nu n-tzaʔą=ūʔ ‘A story about those who had animal spirit companions’. Though nagualism (Foster 1944), a belief that certain people could transform themselves into animals or natural forces for good or evil purpose, is fading out in Mesoamerican culture and worldview, it is a still prevalent theme in Zenzontepec Chatino folklore.

Apr 29Discussion Session: Fieldwork Advice

May 6Daisy Rosenblum (Berkeley; UCSB)

Jan 28Luís Miguel Rojas Berscia (Radboud Universiteit)
The Syntax and Semantics of Causative Constructions in Balsapuerto Shawi
Among the valency changing phenomena in the languages of the world, one of the most studied has been causation. All the languages in the world manifest causative events which, cognitively, can be characterised as two events: a causer and a causee. They can be also manifested lexically, morphologically or periphrastically. Moreover, many types of causation can be found among languages: most of them have direct and indirect types, which tend to appear as marked elements; however, recently, a new category has been established, associative. Like other languages of the world which manifest direct causation in a more compact fashion and indirect causation in a less compact fashion, Balsapuerto Shawi behaves in the same way. Causation in Balsapuerto Shawi manifests itself, then, as a continuum, ranging from lexical causatives to periphrastic causatives, conveying a specific meaning for each of these types. Balsapuerto Shawi, as well as other Amazonian languages, presents a complex system of causation, showing not only morphosyntactic markers for direct and indirect causatives, but also for associatives. The latter constitute an interesting phenomenon in South American languages, since many of them present special makers for them, which is typologically uncommon. The associative marker in Shawi is not an associative applicative (Guillaume & Rose 2010). For this reason, Balsapuerto Shawi, following my hypothesis, would not be far from belonging to this big group of Amazonian languages with complex systems of causation.


Feb 4Andrew Garrett, Nico Baier, Ronald Sprouse (Berkeley)
Archiving Your Data with the CLA
This presentation will introduce a new web interface for archiving your data and other work with the California Language Archive. The interface is open to anyone at Berkeley. There will also be a hands-on demonstration of how to create a collection for your project with the CLA and upload your data.


Feb 11Pilar Valenzuela (Chapman University)
Noun Classification in Shiwilu
Shiwilu (a.k.a. Jebero) is one of the two extant members of the Kawapanan linguistic family from Peruvian Amazonia. Among the salient features of Shiwilu grammar are a set of manner/instrumental prefixes, various applicative constructions, a pragmatically motivated “ergative” marker, a suffix that may either increase or decrease the verb valency, and a noun classification system. The latter is the topic of the present talk. Shiwilu has a closed set of some seventeen classifiers. These are bound, mono-syllabic roots that are usually identical to/truncated versions of independent nouns referring to plant or body parts, or other natural elements. They bear resemblance to conventional classifiers in terms of the meaning properties they denote and their semantic transparency (Aikhenvald, 2000). As has been reported for other languages of the region (Payne 1987, Derbyshire & Payne 1990, Seifart 2007, among others), the Shiwilu system combines derivational and agreement-like functions; in this sense, it can be said to be a representative of a “Northwestern Amazonian type” of nominal classification (cf. Seifart & Payne 2007). A second prominent characteristic of the Shiwilu system (which is not shared by its sister language Shawi) is that the same set of classifiers appears in various morpho-syntactic environments such as demonstratives, quantifiers, adjectives, nouns, and verbs. When incorporated into the verb, classifiers may play the S, O, A, or locative applicative functions. Finally, three classifiers are formally identical to the suffixes used for patient, instrumental, and agent nominalization.


Feb 18Judith Tonhauser (Stanford University; Ohio State University/CASBS)
Cross-linguistic meaning variation: The view from Paraguayan Guaraní and English exclusives
A cross-linguistically viable theory of meaning not only captures the meanings of utterances across typologically unrelated languages, but also makes predictions about the nature of cross-linguistic variation. In this talk, I explore the structure and interpretation of the exclusives =nte ‘only’ and -año ‘alone’ in Paraguayan Guaraní (Tupí-Guaraní), and identify far-reaching similarities in their meanings to the English expressions only and alone, respectively. These empirical observations serve as a starting point for asking why one might expect to observe these strikingly similar meanings in typologically unrelated languages and what kind of theory of exclusives not only captures these meanings but also makes predictions about the cross-linguistic variation one might expect to find in this domain.


Feb 25Roberto Zariquiey Biondi (Berkeley; Pontificia Universidad Católica del Perú)
Speech genre, addressee's perspective and pragmatic markedness in Kakataibo (Pano, Peru)
The present paper offers a characterization of the resources that speakers of Kakataibo (Pano, Peru) find in their language for situating the information they present from the perspective of their addressees, in terms of a distinction between information that is proximal to them and information that is not. From the perspective of conversational analysis, the semantic contrasts to be discussed and illustrated in this paper can be understood by means of the notions of epistemic status and epistemic stance (Heritage 2012a,b), which have to do with the knowing status of speech act participants and its encoding in grammar. What these forms do, as we will see throughout this paper, is to encode the expectations of the speaker about the perspective of the addressee in relation to the information presented in an utterance, producing constructions that can be analyzed as multiple perspective constructions (Evans 2005), in which the perspective of the addressee and the perspective of the speaker are encoded in the same utterance. The most salient characteristics of the mechanisms used in Kakataibo to encode addressee's perspective have to do with the diversity of the relevant forms, which belong to various morphosyntactic paradigms, and with the semantic and pragmatic features that play a role in the configuration of the distinction between information that is proximal to the addressee and information that is not. Another interesting fact in relation to how this category is manifested in the language has to do with the presence of various portmanteau morphemes, which combine the encoding of addressee's perspective with the encoding of other meanings related to modality, mirativity and type of speech act. Although similar systems have been described for other South American languages (Kogi: Arwako-Chibchan, Colombia, Bergqvist 2012; Andoke: isolate, Colombia, Landaburu 2007; Awetí: Tupian, Brazil, Drude 2005; Yurakaré: isolate, Bolivia, Gipper 2011; among others), the diversity and the semantic and pragmatic properties of the morphological forms involved make the Kakataibo data highly interesting from a cross-linguistic perspective.


Mar 4Jonathan Manker (Berkeley)
Video Documentation of Critically Endangered Languages
This presentation will discuss the results of the first stage of the Hän Video Documentation Project which began in summer 2013. The goal of this project was to produce three short films of Hän Athabascan speakers narrating videos of themselves performing culturally relevant activities, which included making beaded moccasins, cooking frybread, and preparing salmon. The Hän Athabascan language of eastern Alaska is a critically endangered language, with only 6-8 speakers who no longer use the language on a regular basis, which presented unique challenges. This presentation will explore the strategies used to make this a useful means of language documentation given these difficulties. Additionally, I will consider some interesting phonological, syntactic, and semantic data that resulted from the format of this documentation. The presentation will also include viewing portions of the three video drafts.

Mar 11Mark Sicoli (Georgetown University)
Two Modes of Language Documentation and their Affordances: The Zapotec-Chatino Survey and The Multimodal Interaction in Lachixío Zapotec Corpus
The affordances of language documentation efforts are grounded on the materials and effects that they produce, enabling and constraining future research questions and revitalization efforts, which are important early considerations in research design. In this talk I present two models of language documentation conducted within the Zapotec-Chatino language family of southern Mexico. The first is the Zapotec-Chatino Survey, a documentation of 103 population centers that generated parallel transcribed recordings using a survey instrument based on 1301 lexical items in multiple inflectional frames to generate about 2500 utterances for each town (around 260,000 utterances in the corpus). The work was originally funded by INALI (The National Institute of Indigenous Languages of Mexico), conducted under the auspices of the PDLMA (The Project for the Documentation of the Languages of Meso-America) (2007-2010) with the results currently being archived through a grant from the NSF Documenting Endangered Languages Program. The second project is the Multimodal Interaction in Lachixío Zapotec Corpus, a 40-hour transcribed video corpus of conversations in everyday life from one Zapotec speaking town in the highlands of Oaxaca. The corpus was recorded during fieldwork sponsored by the Max Planck Society (2008-2009) and transcribed, translated and archived under a NEH Documenting Endangered Languages Fellowship (2011-2012). I will describe each of these documentation efforts focusing on relationships between their original goals and the materials produced through their different field methodologies and community involvements, exploring the question of how each is fitted to different types of research questions and can inform different aspects of language maintenance and revitalization. I will discuss several research projects that have spun off each documentation effort with implications for both language specific and comparative studies.

Mar 18Leanne Hinton (Berkeley) and Advocates for Indigenous California Language Survival
The Breath of Life Language Workshop for California Indians
Every other year in early June since 1996, the Linguistics Department has partnered with the Advocates for Indigenous California Language Survival to present the Breath of Life Language Workshop for California Indians. In this intensive one-week workshop, about 60 participants from tribes all over California come to access the campus archives to find and utilize materials on their language, culture and history for purposes of language revitalization. They explore the archives, work on their languages with linguistic mentors, and attend presentations on linguistics and language teaching, learning and revitalization. In this presentation at Fforum, the mission, history and structure of Breath of Life will be reviewed, and the role of the linguistic mentors will be discussed. Graduate students and faculty of linguistics at Berkeley and elsewhere are invited to participate in this year's workshop as mentors, to be held June 1-7.

More information about the workshop and application forms can be found on the website of the Advocates of Indigenous California Language Survival:



Apr 8Martine Bruil (Berkeley; Leiden Universiteit)
Understanding Ecuadorian Siona Verb Classes: A Synchronic and Diachronic Account

Ecuadorian Siona, a Western Tukanoan language spoken in the eastern lowlands of Ecuador, has an interesting verb class system consisting of three classes that behave differently with respect to tense marking. The first class is a small class that consists of a set of verbs that are frequently used, such as ba'ije ‘to be, to live’, saije ‘to go’, and daije ‘to come’. One of the distinctive features of this class is the suffix -i that appears in the present tense, infinitives, nominalizations, and serialized verbs. A second feature is that all verbs in this verb class are monomoraic. The second class is much larger and the verbs in this class never carry the suffix -i. All verbs in this class have either a bimoraic root or stem. The third class is a small set of two bound verbs: the copula a and the future suffix/bound verb si. These verbs only have present tense morphology, which is similar to the past tense morphology in the other verb classes.

Other Western Tukanoan languages show a very similar verb class system and one of the analyses that was proposed is a voice analysis. Johnson & Levinsohn (1990) and Wheeler (1987) analyze the suffix -i in Secoya and Colombian Siona, respectively, as a middle voice marker that disappears in past tense. One argument for this analysis is that the verb class with the -i suffix contains many verbs that seem to have decreased valence pattern in comparison with the verbs from the larger verb class.

However, I will argue in this talk that the suffix -i is not a voice marker. In my view, the suffix is the remains of an imperfective suffix that was used in an earlier stage of the language to use inherently perfective verbs in the present tense or other imperfective contexts. The language seems to have had a split between eventive verbs that were inherently perfective and stative verbs that were inherently imperfective. This system is still found in Kubeo, an Eastern Tukanoan language (Chacón 2009). The -i suffix was only maintained with monomoraic verbs in order to create bimoraic stems. The bimoraic verbs did not need this suffix for this purpose and therefore lost it. The decreased valence pattern of many of the i-verbs can be explained by the fact that many verbs in the non-i verb class are derived transitive or causative verbs. I analyze the bound verb class as the remains of the original stative verb class. The advantage of this analysis of the verb class system is that it explains more aspects of the system, such as the use of the suffix -i and the expression of tense.


Apr 15Terrence Kaufman (University of Pittsburgh)
The Investigation of Sound Symbolism in the Preparation of Dictionaries for Meso-American Languages
Sound symbolism is the more or less direct encoding of meaning through specific phonological material. In general, the relation between sound and meaning in the lexicon of a language is arbitrary. However, in apparently all languages, some (a significant but small fraction of) roots and submorphemic strings encode meaning, at least in part, through the assignment of meaning to certain features, segments, and segment combinations. The most obvious kind of sound symbolism is imitative -- where sound in language imitates sound in nature. This is commonly called onomatopoeia. But there are many kinds of sound symbolism: not only are sounds encoded, but also sights, feelings in the body, states of mind, and size. All languages have onomatopoetic (phonomimetic) roots, but not all have symbolic roots and submorphemic strings that do not relate to sound. The sound features used to encode meaning may be restricted to those found in the non-symbolic lexicon, but they may also include sounds used only in symbolic expressions. The symbolic roots in a language may be integrated into the morphosyntactic system, or they may be syntactically isolated, like vocatives or adverbs of manner.

Systemic anomalies that may be associated with sound symbolism:

  1. The sounds that occur in symbolic roots may use features and feature combinations not permitted in the ordinary lexicon; they may be subject to different phonological rules than the ordinary lexicon, or exempt from some otherwise general phonological rules.
  2. The presence of a semantic link between certain sounds and certain meanings may render the sounds in question more resistant to change than the sounds that arbitrarily encode meaning: in the latter case sound change is regular and exceptionless.

I will talk about how to do the research, and what you will find once you do it.