General Information

Tentative Schedule

Schedule updated: 20/01/2022

※ Note that dates and times are in Japan Standard Time (JST), final schedule is subject to change. Also be aware that certain dates will be held at 5:00PM - 7:00PM (JST) or 6:00PM - 8:00PM, rather than the standard 10:00AM - 12:00PM (JST), in order to make presentation times more comfortable for those in certain time zones.

Program

The typology of click sounds: lessons from marginal click inventories
(February 19, 2022, 5:00PM - 6:00PM, JST)

• Presenter: Dr. William Bennett (Rhodes University)

• Abstract: A longstanding mystery in phonology has been the relationship between clicks and other types of consonants. A few languages employ very large numbers of distinct click consonants (e.g. !Xóõ, G|ui, N|u, !Xũ). Many more languages have a smaller number of click sounds (e.g. Nama, Xhosa, Hadza, Zulu). Perhaps no less common, but certainly much less well understood, are languages with small inventories of clicks, or more marginal use of these sounds. Afrikaans, for instance, is traditionally described as a language that lacks clicks; however, Namaqualand varieties of Afrikaans typically have a small set of non-standard lexical items that include clicks. These clicks tend to vary greatly in their phonetic realization; this obscures the phonological categorization of clicks in these inventories. This talk explores the typology of click consonant systems, with particular attention to what marginal click inventories can reveal about how these uncommon sounds are represented phonologically.

Applicatives and the argument-adjunct distinction in Bantu
(February 19, 2022, 6:00PM - 7:00PM, JST)

• Presenter: Dr. Kristina Riedel (University of Free State)

• Abstract: In this presentation I discuss work in progress on the argument-adjunct distinction in Bantu languages in the context of applicative constructions and prepositional alternatives in Southern Bantu. Two issues create problems for accounting for these constructions derivationally: the ambiguity between focus interpretations and interpretations involving the addition of an argument for certain types of applicatives (cf. Creissels 2004, Marten and Mous 2017, Gibson et al. forthcoming), and the fact that while valency clearly plays a role in some applicatives, in many cases it does not (irrespective of focus) (cf. Trithart 1983). These kinds of patterns seem to indicate a disconnect between applicative semantics (as far as these can be considered to be coherent) and applicative syntax on one hand, as well as more specific structural encoding of focus in Bantu, beyond conjoint and disjoint patterns and immediate-after-verb effects. Forker (2014) and other recent typological papers have argued that argument and adjunct status in syntax and semantics are separate, while Haspelmath (2014) questions argument status as a universal category. Andrason (2018) applies this to isiXhosa locative applicatives arguing for a scalar distinction between arguments and adjuncts in Bantu. Descriptive work on Bantu languages has linked object-behaviour of applied phrases to animacy and definiteness hierarchies (cf. Morolong and Hyman 1977, Duranti 1979, Hyman and Duranti 1982). While Generative work has typically derived the divergent syntactic properties from different structural configurations (cf. Zeller and Ngoboka 2006) or the result of Phases (McGinnis 2001), without trying to address the other types of semantic effects or the flexibility in structure and/or interpretation for some types of applicatives (du Plessis and Visser 1992, Creissels 2004, Marten and Kula 2014). In this presentation, I will explore how well a linker-type structure for the applicative can account for the data. This talk is partially based on joint work on Sesotho for-constructions with Hannah Gibson, and work on applicatives, prepositions and focus with Hannah Gibson, Lutz Marten and Maarten Mous.

A microparametric approach to typological correlations on focus marking strategies in Bantu morphosyntax
(February 12, 2022, 6:00PM - 8:00PM, JST)

• Presenter: Dr. Daisuke Shinagawa (ILCAA, Tokyo University of Foreign Studies) & Dr. Lutz Marten (SOAS)

• Abstract: This talk aims to discuss typological tendencies of focus marking strategies from a cross-Bantu perspective, based on the large-scale database on morphosyntactic variation of Bantu languages called Bantu Morphosyntactic Variation database, which is built around 142 morphosyntactic features or parameters. By examining inter-parametric correlation observed in BMV, we will provide a broader picture of micro-typological correlation between focus marking strategies and other logically independent grammatical features, namely main clause negation, double object constructions, and inversion constructions, from a cross-Bantu perspective.

The mistaken identity of the Jarawa who traveled north: toward a re-classification of “Jarawan Bantu”
(February 5, 2022, 10:00AM - 11:00AM, JST)

• Presenter: Dr. Chris Green (Syracuse University)

• Abstract: Jarawan languages have long fascinated Africanists, and particularly historical linguists and linguistic typologists. They stand near the cusp of the Bantoid-Narrow Bantu divide, with descriptions and classifications dating back to the early 20th century (e.g., Johnston 1921; Abraham 1940; Williamson 1973; Gerhardt 1982; Maddieson & Williamson; Shimizu 1983, among others) making claims that they belong in one group vs. the other. Until recently, however, remarkably little is known about them beyond wordlists, some of which contain only a few dozen words, and a single short descriptive paper (Gerhardt 1988). The most recent classifications based on extensive comparison (Piron 1995, 1998; Blench 2006) and advances in lexicostatistic modeling (Grollemund 2012, Grollemund et al. 2018, Grollemund et al. 2022) claim that they belong among the Group A Narrow Bantu, though details differ from scholar to scholar. In this paper, it is illustrated that data collected over the last three years on three Jarawan languages, and subsequent reanalysis of existing wordlists and the sound changes observed therein, do not support the same conclusion. From a lexical standpoint, it is shown that while Jarawan languages indeed share ~20% cognacy with Proto Bantu, these same cognates are also shared with Proto Grassfields (Hyman 1979/2007), and often with Proto Benue-Congo (de Wolf 1971); they cannot be shared Bantu innovations. From a morphological standpoint, it will be shown that Jarawan languages appear more akin to certain Grassfields languages. Lastly, in terms of diachronic sound changes, whereas North-Western Bantu languages (i.e., Group A) display two significant phonological innovations where *k > Ø and *g > k (Philippson and Grollemund, in prep.), these innovations are not observed in Jarawan languages. Thus, based on lexical cognacy, sound changes, and morphology, this paper revisits and bolsters arguments in favor of a re-classification of Jarawan languages among other Bantoid-non-Bantu languages, rather than within Narrow Bantu.

On Resumptive Pronouns in Yoruba
(February 5, 2022, 11:00AM - 12:00PM, JST)

• Presenter: Dr. Oluseye Adesola (Yale University)

• Abstract: Generative grammarians have been examining ways to provide a unified account for the patterns of occurrence of resumptive pronouns (RP) for more than five decades (Ross 1967 ... Hammerly 2021). Much of such work have been devoted to the occurrence of RPs in islands and how they ameliorate island violations. They were once thought to be rare in the subject position (Keenan 1985). At a point (McCloskey 2006), they were also thought not to be different from regular pronouns in any significant ways. In this paper, we show that Yoruba (and perhaps some other African languages: Akan, Igbo, and Edo) have two types of RPs: agreeing in resumptive pronouns derived via movement and a non-agreeing ‘resumptive’ pronoun which is base generated for EPP purposes. Agreeing resumptive pronouns are required to agree with their antecedents like regular pronouns in the language while the non-agreeing resumptive pronoun doesn’t agree in phi-features with its external antecendent. Thus, it patterns more like an expletive pronoun in the language. The paper also shows that Yoruba pronouns in general have less phi-features than are available in pronouns in other languages. Thus, the fact that Yoruba resumptive pronouns are less specified may contribute to why they occur more freely than resumptive pronouns in other languages. Also, the use of resumptive pronouns in Yoruba contributes to the absence of island effect and weak crossover effect in the language.

The parser consults the lexicon in spite of transparent gender marking: EEG evidence from noun class agreement processing in Zulu
(January 29, 2022, 5:00PM - 6:00PM, JST)

• Presenter: Dr. Jochen Zeller (University of KwaZulu-Natal)

  • joint work with Emanuel Bylund (Stellenbosch University, Stockholm University) and Ashley Glen Lewis (Radboud University, Haskins Laboratories)

• Abstract: In sentence comprehension, the parser in many languages has the option to use both the morphological form of a noun and its lexical representation when evaluating agreement. The additional step of consulting the lexicon incurs processing costs, and an important question is whether the parser takes that step even when the formal cues alone are sufficiently reliable to evaluate agreement. Our study addressed this question using electrophysiology to examine the parser's response to DP-internal noun class agreement violations in the Bantu language Zulu (Nguni; S.42), where both grammatical gender and number features are reliably expressed formally by noun class prefixes, but only gender features are lexically specified. We observed reduced, more topographically focal LAN, and more frontally distributed alpha/beta power effects for gender compared to number agreement violations. These differences provide evidence that for gender mismatches, even though the formal cues are reliable, the parser nevertheless takes the additional step of consulting the noun’s lexical representation, a step which is not available for number.

Tone, stress and vowel sonority interactions in Shingazidja (Bantu G44a, Comoros)
(January 29, 2022, 6:00PM - 7:00PM, JST)

• Presenter: Dr. Cédric Patin (University of Lille)

• Abstract: If the vast majority of Niger-Congo languages are tonal, several languages of the phylum no longer have a tone system. However, we know very little today about how a (Niger-Congo) language can lose its tones, and what can be the cause of such a loss. In this talk, I will argue that Shingazidja, a Bantu G44a language from the Comoros, is in the process of evolving its tonal system into an accentual system, and that a major actor of this ongoing evolution is a complex case of interaction between tones, stress and sonority.

Issues in Bantu Melodic Tone Assignment & Realization
(January 15, 2022, 10:00AM - 11:00AM, JST)

• Presenter: Dr. Lee Bickmore (University at Albany (SUNY))

• Abstract: Bantu tone languages generally exhibit both lexical and grammatical tones. Lexical tones distinguish nominal, verbal and adjectival roots, as well as a number of agreement markers and verbal prefixes. Grammatical tones are more complicated. They are often one of several morphological exponents expressing some constellation of inflectional features. For instance, in Cilungu, a Zambian Bantu language, some tense/aspect/mood/polarity (TAMP) combinations trigger a grammatical High tone (often referred to as a “Melodic Tone”) on the second mora of the stem, often in combination with various segmental affixes. Other TAMPs trigger a High tone on the stem-final vowel, while yet others trigger a High tone on the second and all subsequent TBUs of the stem. And some don’t trigger a grammatical High at all. While the existence of these melodic tones has been well known for a long time, what is much less clear is exactly when and where in the grammar they are generated and how they ultimately dock onto the appropriate verbal TBUs. In this paper, I consider the advantages and disadvantages of 3 different approaches to account for melodic tone realization: The first is the “All in the Phonology” approach where these tones are introduced and docked in the phonology. The second is the “Annotated Tones” approach where multiple types of floating High tones are introduced in the morphology and then linked in the phonology. The third is the “Single Tone” approach where one uniform floating High is generated in the phonology and then linked in the phonology. The two central questions which emerge in this regard are familiar ones: how much phonology if any does the morphology have access to, and how much morphological information must the phonology have access to. Preliminary conclusions to these questions and suggested lines of further study will be presented.

Complementizer Agreement with Superordinate Subjects in Setswana and Ikalanga
(December 18, 2021, 10:00AM - 11:00AM, JST)

• Presenter: Dr. Ken Safir (Rutgers University)

  • joint work with Rose Letsholo (University of Botswana)

• Abstract: [LINK]

Revisiting nominal licensing in Zulu
(December 18, 2021, 11:00AM - 12:00PM, JST)

• Presenter: Dr. Claire Halpert (University of Minnesota)

• Abstract: The questions of whether and how nominals are syntactically licensed in Bantu languages have been a matter of recent active debate (e.g. Diercks 2012; Halpert 2015, 2019; van der Wal 2015; Sheehan and van der Wal 2018; Carstens and Mletshe 2016; Pietraszko 2020). While a number of languages and phenomena throughout the Bantu family seem to lack evidence of the typical 'signature' of case-licensing (Diercks 2012), others show more straightforward case patterns (van der Wal 2015). Sheehan and van der Wal (2018) suggest the term Vergnaud Licensing to refer to nominals' requirements for particular syntactic configurations and show that Bantu anguages show differing behavior on a variety of Vergnaud Licensing metrics. In my own work, I've argued that the Bantu language Zulu has structural case effects that are largely obscured (1) by the prevalence extremely local case licensers (along the lines of lexical case assigners in Baker 2015) and (2) the location of structural case-assigning heads low in the clause. In particular, I argued that all nominals marked with the so-called augment morpheme were locally licensed and did not require low structural case. In this talk, I will complicate this view by investigating some environments that show hallmarks of structural case alternations/Vergnaud licensing, even for augmented nominals: passives Halpert and Zeller 2016, Halpert to appear), uncontrolled infinitives Halpert to appear) and possession. In passives and infinitives, external arguments become optional and are morphologically marked when they appear. I demonstrate that these marked overt external arguments are in fact structurally licensed in Spec,vP. In possessor raising, a morphologically marked postnominal possessor alternates with an unmarked prenominal possessor that can be targeted by A- and A-bar processes in the main clause. I propose that in all of these environments, a morphologically overt, acategorial Linker head is involved in licensing (cf. Baker and Collins 2006, Schneider-Zioga 2015, Pietraszko 2019). The requirement of this type of special licensing exactly in environments analogous to case-deprived environments in more familiar case-licensing languages in turn suggests that external arguments are structurally licensed the active, finite environments--even with the augment. As we look closely at more environments in a Bantu language like Zulu, the picture of nominal distribution and licensing becomes richer and more complex.

Grammatical tone and person marking in Rere
(December 4, 2021, 10:00AM - 11:00AM, JST)

• Presenter: Dr. Sharon Rose (University of California, San Diego)

  • joint work with Klaus Kim (UCLA) and Nina Hagen Kaldhol (UC San Diego)

• Abstract: Grammatical tone is manifest in the verbal systems of numerous African languages, often to express tense-aspect-mood (TAM) distinctions. In this talk, we present the argument marking patterns of Rere (also known as Koalib), a Kordofanian language of Sudan. Rere shows a particular pattern of argument marking whereby the verb has two templatic positions. Segmental markers in these positions express person, number and noun class. However, the grammatical role of the markers as subject/object switches if the object is 3rd person. This switch is expressed by tone, leading to minimal tone pairs. Tone is therefore independent and combines with the segmental markers to create grammatical role distinctions. Rere also uses tone to express TAM distinctions, and we show how the TAM tone patterns are partially subordinate to the argument-marking tone system. In addition, the extension of tone onto neighboring morphemes in an iterative or non-iterative fashion differs depending on the source of the tone.

Testing typologies of consonant inventories: the view from Africa
(December 4, 2021, 11:00AM - 12:00PM, JST)

• Presenter: Dr. Laura Downing (University of Gothenburg)

• Link to the abstract [pdf]

Two Unexpected(?) Aspects of the Shupamem A-bar System
(November 13th, 2021, 10:00AM - 11:00AM, JST)

• Presenter: Dr. Jason Kandybowicz (City University of New York)

• Abstract: In this talk, I discuss two aspects of the A-bar system of Shupamem, an understudied Grassfields Bantu language of Cameroon, that may or may not strike one as unexpected depending on their perspective. The first deals with the clausal integration of non-restrictive relative clauses. I show that non-restrictive relatives have the same properties as restrictive relatives in the language and argue that a range of considerations including illocutionary independence, scope relations with matrix negation and intentional verbs, VP ellipsis, pronominalization, binding, weak crossover effects, parasitic gaps, and split antecedents, among others, support the conclusion that Shupamem non-restrictive relative clauses are clause-internal nominally-integrated syntactic objects, as in Italian and Mandarin Chinese. This finding supports Cinque’s (2008) discovery that non-restrictive relative clauses come in both integrated and non-integrated varieties and typologically places Shupamem among the languages of the world that exclusively manifest the integrated non-restrictive relative clause structure. Is this unexpected? Given the limited number of languages presently known to manifest non-restrictive relative clauses of the integrated variety, we might conclude yes. But from another perspective (Cinque 2020), integrated non-restrictive relative clauses may be quite common cross-linguistically. The second aspect of the Shupamem A-bar system I discuss will likely strike most as surprising. I show that domains constituting cross-linguistically stable islands for movement are completely A-bar transparent in the language. Using considerations such as crossover effects, parasitic gap licensing, and reconstruction effects, I argue that A-bar movement from these domains (in the form of focus clefts and topicalization) demonstrably takes place. How unexpected is this finding? From a standard Generative perspective, it is highly surprising. However, recent work documenting the transparency of certain classic island configurations in African languages, may suggest that the pattern observed in Shupamem is more common than we think.

On Logophoric Pronouns in African Languages: Universals and Variation
(November 13th, 2021, 11:00AM - 12:00PM, JST)

• Presenter: Dr. Mark Baker (Rutgers University)

• Link to the abstract [pdf]

Bukusu object marking: at the interface of pragmatics and syntax
(October 16th, 2021, 11:00AM - 12:00PM, JST)

• Presenter: Dr. Michael Diercks (Pomona College)

• Co-author: Justine Sikuku (Moi University)

• Abstract: This talk describes a complex system of object marking (i.e. object clitics) in Bukusu that interacts with pragmatics/discourse semantics. We show that object markers can co-occur with (i.e. double) in situ lexical objects, but that this possibility is linked with 1) a discourse-given interpretation of the object, 2) the presence of a focused element inside the verb phrase, and 3) an emphatic interpretation of that focused element. That is, it appears that there is both a focus/givenness component of OM-doubling, but also an additional layer of emphasis (we show that both mirative readings and verum readings are possible). We propose that Bukusu (doubling) object markers arise via Agree relations generated by phi-features on a functional projections at the edge of vP, which include a focus operator and an operator contributing a conventional implicature generating the emphatic reading (Cruschina 2021). Therefore, rather than object marker (OM) doubling being driven by properties such as case, specificity, or linked with object shift as has been claimed for many clitic doubling languages, instead it appears that the closest empirical correlate of Bukusu OM-doubling is the semantics of focus and givenness in (for example) English intonation. Our analysis incorporates features of givenness/focus into a probe-goal system to account for intricate interactions of locality of Agree with interpretive effects of doubling.