Talks - Andrew Harvey

Abstract

Tanzania’s Rift Valley has been shown to feature linguistic diversity punctuated by remarkable convergences. This talk is a first effort at extending that analysis to the words people use to refer to their relatives. By way of context, this talk is a product of the University of Bayreuth course “Language and Meaning”, part of which was dedicated to kinship and kinship terminology across the African continent. Participants surveyed both lexical and ethnographic material from language communities across Africa, bringing together a wealth of material both lexical (kinship terms) and structural (kinship systems) in nature. This talk moblises this cross-continent data by employing it as a background against which languages of the Tanzanian Rift can be compared and contrasted, with the goal of asking questions about whether kinship terms and systems have been genetically inherited, affected by contact, or innovated in response to wider historical and social factors. Following an introductory survey of kinship terms and kinship systems of the Tanzanian Rift, the talk identifies patterns and attempts some initial investigations, using three kin relations as concrete examples: 1) siblings and cousins; 2) in-laws; and 3) the mother’s brother.

Abstract

This talk is given on the occasion of having a paper rejected from the linguistics journal Glossa. While rejection is nothing unusual – indeed, it is a common experience of academic publishing – some of the comments of one of the reviewers were particularly provocative, stating that my analysis of Gorwaa amounted to exoticisation of the language. Because the paper was rejected, I have had no formal channel by which to respond to these comments, which I take seriously, and consider objectionable. Because all of these comments have to do with how I represent the Gorwaa language through glosses, responding to them also presents a good opportunity to reflect on the choices I have made in describing the language, as well as to share some of my understanding of the Gorwaa language.

Abstract

Documentary linguistics is a major subfield of linguistics interested in the creation of lasting, multi-purpose records of languages. This talk provides an in-depth look at one such documentary linguistic project – a project documenting the Hadza language, funded by the Endangered Languages Documentation Programme (ELDP), which took place from 2019 to 2021. Following some brief context, both on the Hadza language and on the researchers, the talk 1) presents the documentary project in a stepwise fashion from preparation to project setup to daily activities to archiving; 2) reviews some outputs of the project, and 3) reflects on feedback from the Hadza people themselves who were involved in the documentary project.

Harvey, Andrew. 2022. Learning from, Talking about, and Reflecting on Data Loss: A Failure Report. Talk given at the conference Where Do We Go From Here? Language Documentation and Archiving in the International Decade of Indigenous Languages. The Berlin-Brandenburg Academy of sciences and Humanities. Berlin, Germany. 07/10/2022. DOI: 10.5281/zenodo.7155475

Note: This talk has not gone through a process of peer review, and findings should therefore be treated as preliminary and subject to change.

Abstract

Griscom and Harvey (2020) is a large, diverse archival record of language and cultural material from the Hadza people of Tanzania. Collected between 2019 and 2021 and employing a community-based model, while this collection represents a considerable contribution to Hadza language work (c.f. Leonard 2021), it is not perfect: during the course of collecting and processing the audiovisual materials, a considerable amount of video material associated with the collection was lost. This loss of data will be the focus of the talk.
Language documentation projects are complex undertakings (c.f. Bowern 2008: 48), and failures – including data loss (c.f. Tsutsui-Billins 2019) – are frequent. At the same time, it is not always easy for language documentarians to talk about their failures – this ultimately makes it easier for specific failures to occur again, at different times and in different projects. Modelling this talk on the failure report (c.f. Engineers Without Borders Canada 2017), the specific failure is examined in detail, the causes of the failure made explicit, and the lessons learned enumerated.
The project which resulted in Griscom and Harvey (2020) is, in many ways, a great success, both in terms of data collected, as well as the relationships developed throughout. With that said, focusing on the data loss which occurred is useful in several ways: it helps make what happened explicit such that learning can occur, it creates a space to talk about the dynamics of community-based work with marginalised communities, and it allows us to be reflexive in our praxis.

Abstract

I have spoken about names and naming in Gorwaa before, but much of the previous data consisted of Gorwaa people explaining how things worked in a general manner. New data, including interviews with Gorwaa parents about how their own children got their names, adds further detail to the existing account, as well as a perspective based on what Gorwaa people actually do, rather than what they say they do. Focusing on three questions – How are Gorwaa names given? Who are Gorwaa names given by?, and To what extent and in what ways are Gorwaa names and naming an endangered verbal art? – this talk provides additional detail, expands our understanding, and brings us closer to an informed account of this interesting and important aspect of the Gorwaa language.

Abstract

There is a legend among linguists that anyone who undertakes work on the Hadza language is doomed to fail. Provoked by this bizarre story, this talk uses the “Hadza Curse” as an entry point for a critical intervention in the relationship between researchers and the Hadza people, between traditions of Hadza scholarship and epistemic violence, as well as between language work and the transformative work of allowing the Hadzabe people to valorise, speak for, and represent themselves. Very early on, it will be established that the “Hadza Curse” is due to factors external to the Hadza people, and is a myth. Attention will then duly turn from imagined threats faced by linguists, to the real violence faced by the Hadzabe people. The talk will assert that the end to violence against the Hadzabe people is rooted in changing how they are portrayed to outsiders, and especially through challenging existing research practices. Recent cases of language work with the Hadzabe people are given as examples of what some of these transformations can look like.

Abstract

The present talk provides the first systematic analysis of the formal (phonetic and morphological) facet of interjections in Hadza. By using a prototype-driven approach to an interjective category and drawing on original evidence, the authors demonstrate that Hadza interjections closely comply with an interjective prototype. Hadza interjections meet most prototypical features and the exceptions attested fully match the exceptions that have previously been observed in other languages. Overall, the study corroborates the validity of the prototype postulated in scholarly literature, although certain refinements are also proposed.

Abstract

The linker in Gorwaa is a nominal morpheme with six different (sub)gender sensitive allomorphs. The bolded forms in slufitá wák “one lip”, desir‘eé’ “my girl”, hhawató baabá “father’s man”; kurkí “this year”, ayeemá uren “big lands”, and /aylá tleer “a long wedding song” are all linkers. Nouns do not show linker morphology in the following environments: unmodified subjects, unmodified objects in ‘second position’ (i.e. before the selector), and incorporated nouns. Nouns do show linker morphology everywhere else.Earlier accounts of linkers in Southern Cushitic languages characterise them as construct state morphology – a form taken by nouns in many Afroasiatic languages when possessed. Following a discussion of why the idea of possession is not adequate to characterise all occurrences of linker morphology, an alternative is advanced: that linkers are obligatorily present on all nouns with reference, but go unpronounced when at the right edge of the phonological phrase. Finally, new analysis made in Kerr (2020) is considered, as well as prospects for better understanding this pervasive phenomenon in Gorwaa.

Abstract

This talk provides a progress update on a collaborative project on riddling traditions in the Rift Valley that a group of RVN members have been working on for the last year or so. We outline our goals and rationale for the project, give some background about how our online collaboration has been organised, and we then present initial findings on four dimensions of our Rift Valley riddle research: (i) communicative ecologies; (ii) discourse structure; (i) grammatical and linguistic properties properties; and (iv) conceptual and stylistic features.

Harvey, Andrew. 2021. Remote but not distant: Lessons from fieldwork with Gorwaa, Hadza, and Ihanzu speaker communities. Talk given as part of the panel “Collaborative Linguistic Fieldwork During the Current Crisis and Beyond” at the fourth School of Languages Conference (SOLCON IV), University of Ghana, Ghana (Online). 05.11.2021. DOI: 10.5281/zenodo.5647954

Note: This talk has not gone through a process of peer review, and findings should therefore be treated as preliminary and subject to change.

Abstract

A seminal chapter by Kießling et al. (2008) identified a concentration of preverbal clitic clusters in the Tanzanian Rift Valley Area. Crucially they noted that preverbal clitic placement is neither a prominent feature of Nilotic nor the Bantu languages of the area and as such, they considered preverbal clitic clusters to represent an areal feature reflecting the sustained history of language contact in the area. The present paper builds on the foundation laid out by Kießling et al. (2008) and expands to include a broader range of languages, allowing for more detailed discussion and the inclusion of additional data.

Abstract

Language documentation has had a major impact on the trajectory of linguistics since the turn of the century, both in terms of what is considered linguistic data, as well as how that data is collected and used. Taking my work – 10 years of documentation in the Tanzanian Rift Valley Area – as a case study, this talk reflects on how my practice has been shaped by language documentation discourse, as well as how the local context has shaped how I do language documentation. Three central topics to be explored include the effect of a documentary approach on grammatical description, the emergence of literary genres from documentary materials, as well as the potential language documentation has for supporting the narrative sovereignty of marginalised speaker communities. Finally, the talk will provide some assessment on the future(s) of language documentation on the continent, and what language documentation in Africa must become to ultimately succeed. At the same time, the talk explores what collaboration can look like, actively pushing back against the idea of this work being the top-down pursuit of a “lone wolf” linguist. Examples will be provided as rich audiovisual material throughout, drawing on the verbal arts of the Gorwaa, Ihanzu, and Hadzabe people.

Abstract

The F31 Bantu languages of north-central Tanzania Nyilamba and Ihanzu do not feature labiodentals. This is remarkable in that all of the other languages of the intermediate genetic grouping to which Nyilamba and Ihanzu belong — the Takama branch — feature labiodentals robustly. With a specific focus on Ihanzu, this talk describes the lack of labiodentals in this language as a direct result of contact with the language isolate Hadza. Following a brief introduction to the Ihanzu language and its larger context, the talk reviews lexical comparative work in Masele (2001) to establish that, though every language of the Takama branch developed labiodentals independently either through sound changes or via the borrowing of lexical items, Ihanzu did not. This preservation of older structures is attributed to the existence of a parallel structure, that is, a lack of labiodentals in the phonemic inventory in the contact language Hadza. In order to attribute this preservation of a system lacking labiodentals to contact with the language isolate Hadza (which currently features the labiodental [f]), the talk establishes the adoption of [f] as a relatively recent effect of Hadza contact with the South Cushitic language Iraqw, and argues that at the time and place of the relevant contact with Ihanzu, Hadza did not feature this phoneme. Additional non-linguistic evidence of the nature of Ihanzu-Hadza contact is provided from oral histories, origin stories, and descent patterns of both the Ihanzu speaker community and the Bantu-speaking peoples of the wider region.In the context of this workshop, the talk provides a clear-cut example of the effects of a non-Bantu language (Hadza) on the sound inventory of a Bantu language (Ihanzu), within the larger ecology of the complex interchanges occurring in the Tanzanian Rift Valley Area. In addition, the proposed contact is recent enough to serve as a concrete case study, but also old enough (i.e. pre-colonial) to require recourse to linguistic and other non-direct methods of observation, such as oral traditions.

Abstract

Hadza has several clicks in its phonological inventory. Sands, Maddieson & Ladefoged (1996) and Sands (2013) describe Hadza with 9 clicks while Miller (2008) suggests that there are 12. The 4 types of Hadza clicks [ʘ, |, !, ‖] bilabial, dental, alveolar and lateral can be accompanied, in a contrastive way, by aspirated, glottal and nasal features [ʘ̰ , | ʔ, | h, ŋ|, ŋ| ʔ, ! ʔ, ! h, ŋ!, ŋ! ʔ, ‖, ‖ h, ŋ‖ h, ŋ‖ ʔ]. From an acoustic point of view, clicks can be described with 2 features [grave vs. acute] and [abrupt vs. noisy] following a proposal made by Traill (1994). In order to describe Hadza clicks in terms of gestures (with a measure of time) and to establish the relevant distinctive features, 4 women and 5 men native speakers were recorded using various instrumental techniques (Aerodynamic, acoustic, EGG, palatographic and Video). Some of the results show that there is a difference between the apical clicks [|, !] and the lateral click [‖] where the tongue blade extends backwards during production before the release. This back of the tongue movement is slower for aspirated clicks compared to non-aspirated ones. For noisy clicks, after releasing the anterior closure, there is an initial increase in loudness that increases to a peak and then decreases. It is a measure of how quickly the anterior part of the tongue moves away from its place of articulation. The back of the tongue moves away more slowly from the velum with the aspirated clicks compared to the non-aspirated ones. All the segments discussed here require accounting very precisely for the relative coordination of articulatory movements, the adjustments of the larynx and the movements of the velum. These coordination specifications are not directly related (one to one) with the phonological features.

Abstract

In this paper, we present a new analysis of Hadza tonology. In previous work, tone has been considered a property of the syllable (Sands 2013, Tucker et al. 1977), whereas we will argue that the mora is the tone-bearing unit. Based on data collected in 2019 and 2020 from four different Hadza-speaking areas, we argue that morae may be specified for H tone or unspecified (which are realized with low tone). In addition to presenting a typology of tone patterns in Hadza roots, we will also present a typology of the prominence patterns seen. Interactions of prosodic structure and tone will be discussed. For instance, a final 3sg.f PNG marker /-ko/ generally appears with high tone on nouns in a frame sentence (‘I saw X there’), or in list-intonation (when indicating continuation); in edge-final position, however,/-ko/ may appear with devoicing/elision and low pitch. Interestingly, despite having a very rich consonant inventory with contrasts in airflow mechanisms, aspiration, voicing and glottalization, Hadza does not appear to have any consonant/tone co-occurrence restrictions.

Abstract

When involving members of a speech community in the documentation of their own language, it is common to work with speakers who have some experience with formal education or technology. But what about communities whose speakers have very little access to both? This talk describes documentary projects spanning approximately ten years and involving four different speaker communities of the Tanzanian Rift Valley Area, all of which could be described as low-resource, marginalised, or existing outside of the larger power-structure. We aim here to assert that not only are Insider Researcher led projects in these contexts possible, but that they also result in objectively better documentations. This talk presents a mixed methods analysis showing how Insider Researcher led projects can produce documentations more comprehensive than those produced solely by outsider researchers, and thus of greater value to both linguists and fields beyond. We also show that, especially in the context of marginalised speaker communities, Insider Researcher led projects can produce documentations more emically meaningful than those produced solely by outsider researchers, and thus of greater value to the speaker community itself. Methods will include a quantitative review of our projects, showing how Local Researchers work in different patterns to outsider linguists (i.e. the authors), as well as a qualitative examination of the nature of some of these different patterns.

Abstract

The literature on language documentation and description commonly underscores the importance of fostering close relationships with speech communities and promotes the notion that active community involvement “lead[s] to better overall documentation” (Arka 2018; c.f. de Lima Silva and Riestenberg 2020; Vallejos 2014). To date, however, no detailed analysis has been conducted of the differences between documentary materials collected by community members and those collected by community outsiders. This talk considers two criteria commonly applied to good overall language documentations, explores how they may be measured, and assesses the documentary materials collected with varying levels of involvement from four speaker communities of northern Tanzania over the last decade: the Asimjeeg Datooga (Griscom 2018), Gorwaa (Harvey 2017), Hadza (Griscom and Harvey 2020), and Ihanzu (Harvey 2019).

Abstract

The South Cushitic languages Gorwaa (ISO639-3: gow) and Iraqw (ISO639-3: irk) – both spoken in Tanzania by approximately 130,000 and 500,000 people, respectively – are among the small number of Cushitic languages spoken in the country, setting themselves apart by their phonology, grammar, as well as the wider cultures of their speaker communities. Linguists working from both outsider (Harvey 2019) and insider (Alphonce 2020) perspectives have argued that Gorwaa and Iraqw naming traditions are important conduits of local history, and are under considerable pressure from Christian, Muslim, and Swahili naming conventions. While both of the above works attempt an internal typology of Gorwaa and Iraqw names and naming, no attempt has been made to place them within the larger context of Tanzanian names and naming systems. This talk begins with a detailed discussion of names and naming in Gorwaa and Iraqw, distinguishing several independent mechanisms within the larger whole, including inherited, circumstantial, and talismanic names. Then, drawing on a wealth of newly-available analyses of names and naming in other Tanzanian languages, the Gorwaa and Iraqw systems are compared and contrasted with those present in Bantu-, Nilotic-, and Hadza-speaking communities.

Abstract

This talk provides an initial account of our ongoing language documentation project, specifically in terms of where Hadza is spoken, who Hadza is spoken by, how Hadza is used. Due to the type of questions we are seeking to answer, which focuses on who speaks Hadza, rather than who has traditionally been characterised as Hadza, the picture of the Hadza people which emerges from our research methodology is strikingly different from many of those previously given. Sketches situating the Hadza language in its social context, and contrasts with past accounts will be drawn throughout. One goal of this talk is linguistic: discussing new findings on a language and speaker community for which there has been little documentation. Another goal is to address past assertions which have systemically erased large portions of the Hadza people.

Abstract

The concept of language documentation (especially as articulated in Himmelmann 1998), is experiencing radical changes in both theory and practice across the African continent (c.f. Seyfeddinipur (ed.) 2016). With a focus on the Tanzanian Rift Valley Area (largely as conceived by Kießling, Mous, and Nurse 2008), this talk examines how language documentation in Africa 1) is strongest when it takes into account the performativeness at the core of communicative practices; 2) has met with general failure when serving agendas of text-based maintenance and revitalization; and 3) shows great promise when it seeks to work within and strengthen traditions of orality (c.f. Ameka 2015).

Abstract

Gorwaa (South Cushitic, Afro-Asiatic) is an endangered language, spoken by approximately 133,000 individuals in the Babati district of central Tanzania. Analysis of the use of traditional Gorwaa names shows a continual decline and a reorientation toward Christian, Muslim, and Swahili naming conventions. As such, the naming tradition under examination may itself be viewed as also endangered. This is not trivial, as personal names and how they are bestowed provides a window into the cultural life of a people, as well as into historical language dynamics. Through interviews with Gorwaa people, listening to relevant life stories in the Gorwaa language and cultural material archive, and using a list of 750 Gorwaa personal names, collected during fieldwork in Babati district between 2012 and 2016, an ethnographic sketch of Gorwaa naming conventions is provided. Many of the semantically transparent names in the list refer to actions or events – a common theme throughout naming conventions on the African continent. More specifically, many personal names play a role in warding off evil, ensuring family continuity, as well as honouring figures of power. Extensive borrowings from Datooga will be considered as a testament to long-term, sustained contact in the area. Recent borrowings from Swahili and English provide insight into the nature of these relatively new contacts.

Abstract

A nomadic hunter-gatherer people who speak a click language, the Hadza people are a highly salient group within the Tanzanian Rift Valley Area, and have been a focus of academic inquiry in Tanzania for over 100 years. With a focus on the (often interrelated) fields of linguistics and anthropology, this talk traces the academic discourse on the Hadza people, and, specifically, how the former has shaped our conceptions (or misconceptions) of the latter. A dominant narrative to emerge is the conception of the Hadza people as members of an insular culture who have undergone minimal change. We would like to show how this concept has made its way into the popular imagination of outsiders, and the effects that this has had on the Hadza people today. We will then show recent challenges brought by linguists to this dominant narrative.

Abstract

New research suggests that Gorwaa (South Cushitic, Tanzania) is a non-configurational language, displaying radical pro-drop, freedom of word order, and the availability of (sub)constituents to occur discontinuously. This talk presents each of these phenomena, and evaluates competing theories for the phrase structure which underlies this non-configurationality (specifically, the Pronominal Argument model (Jelinek (1984), Baker (1996)), versus the Mirror Theory model (Adger, Harbour, and Watkins 2009). Key empirical evidence examined comprises a diverse range of morphosyntactic patterns, some of which have not been discussed in the South Cushitic literature to this point.

Abstract

The Tanzanian Rift Valley Area, as conceived by Kießling, Mous, and Nurse (2008), is unique on the continent in that it is the only place in which all four of the major African language phyla (Afro-asiatic, Khoisan, Niger-Congo, and Nilo-Saharan) have been in contact for a long time. In this same work, the authors identify 19 features (phonological, morphosyntactic, and pragmatic) which cut across individual languages and language phyla, and are therefore candidates for examples of areal convergence. Of these features, one of the most salient (and most extensively discussed) is that of the preverbal clitic complex: a series of functional particles which occur before the verb, and which carry out functions commonly conceived as verbal in nature. This talk examines the functional particles which occur before the verb in Ihanzu. Based on newly-conducted fieldwork, this talk seeks to add to our empirical knowledge of Ihanzu – an underdocumented language existing on the margins of the Tanzanian Rift Valley Area – as well as revisit the story of contact as told through these preverbal particles.

Abstract

The Rift Valley Bibliography is an open access digital bibliography of publications related to the languages of the Tanzanian Rift Valley Area (Kießling et al. 2008) and associated speech communities. Version 1.0 of the bibliography includes over 400 entries and is accompanied by a collection of PDF files that is available exclusively for members of the Rift Valley Network. During this introduction to the bibliography, we describe its contents and structure, suggest various ways that it can be used, and explain how members of the Rift Valley Network can contribute to future releases.

Abstract

The Tanzanian Rift Valley is unique on the continent in that speakers of all major African language phyla have been in contact here for quite some time (c.f. Kießling, Mous, and Nurse 2008, hereafter referred to as KMN 2008). KMN 2008, a foundational work on the study of the Rift Valley as a linguistic area, delimit the Rift as the area in which the following languages are traditionally spoken: Iraqw, Gorwaa, Alagwa, Burunge, (South Cushitic); Nyaturu, Rangi, Mbugwe, Nyilamba, Ihanzu, Kimbu, (Bantu); Sandawe (Khoisan); Datooga (Southern Nilotic); and Hadza (isolate). Caveats in this list include the fact that Datooga is comprised of a group of distinct varieties, as well as that the smaller “Dorobo” languages (genetic affiliation unknown), as well as the likely-extinct languages Aasax and Kw’adza (both probably South-Cushitic) should also be included here. In this talk, we seek to: 1) contextualize the Area in terms of i) geography and environment, ii) its speaker communities, and iii) previous research conducted pertaining to their histories, cultures, and languages; as well as 2) articulate the need for a Rift Valley Network at the current time, as well as describe the current makeup of the Network and the work conducted by its members.

Abstract

Ihanzu is a Bantu language of the Tanzanian Rift Valley Area about which very little is known. With no grammar, dictionary, or written texts, the language remains entirely undocumented. Based on recent research conducted in 2016 and 2018, this talk provides an initial report of the language, both in terms of its broader context (i.e. where it is spoken, by whom, endangerment status, etc.) as well as genetic and areal situation, and grammatical structure (including details on its phonology, morphology, and syntax). The talk closes with comment on immediate next steps in the inquiry, as well as longer-term prospects in the documentation, description, and further study of this language.

Abstract

Ten years after the publication of Gippert, Himmelmann, and Mosel’s (2008) ‘Essentials of Language Documentation’, many of their criticisms of traditional products of descriptive endeavours remain. This talk will focus on the two they bring up which are most closely related to what has been referred to as the Boasian trilogy – specifically, that i) “provided with a grammar and a dictionary it is still impossible to know how the language is spoken [e.g.] how everyday conversational routines look like […] or how one linguistically interacts when building a house or negotiating a marriage”, and ii) “grammars and […] dictionaries provide little that is of direct use to non-linguists [especially for this talk] researchers in other disciplines (history, anthropology, etc.) (p.19). Beginning with an examination of the traditional trilogy and the problems inherent in trying to incorporate all aspects of language therein, we will look at a broader body of descriptive products, including grammars of language use, paralinguistic repertoires, ethnographies of communication, and ethno-thesauruses. The talk will conclude by returning to the concept of ‘texts’ as the unifying body for these descriptive products, and what this expanded body of products means for language documentation.