We are pleased to announce the following confirmed invited speakers at ACAL50, listed in alphabetical order:
Muhammad Abdul-Mageed (The University of British Columbia, iSchool & Department of Linguistics)
title: Computational language variation
abstract: Human language varies across a very wide host of dimensions such as geography, age, gender, and social class. Although classical methods (e.g., interviewing native speakers) have enabled us to understand how certain languages vary across one or more such dimensions, there are now novel computational and machine learning approaches that can both expedite and advance the study of language variation at scale. In this talk, we introduce the theory underlying a number of these approaches and discuss their utility. We then illustrate these theoretical underpinnings using a range of massive-scale datasets, including from all North African and Western Asia Arabic varieties. In the context of the talk, we will probe some of the thorny social and political issues associated with machine intelligence and language variation.
Kofi Agawu (Princeton University)
title: Iconicity in African musical thought & expression
abstract: African thought and expression achieves sameness through repetition, variation and imitation; I use Peirce’s term, iconicity, to describe this widespread modality. Based on resemblance and analogy, icons leverage similarity between signifier and signified, and so belong to a category of firstness that renders them basic, unavoidable, but available for elaboration. Iconicity models the impulse to transpose properties across domains while maintaining their notional equivalence. This paper assembles instances of musical iconicity form West and Central Africa and considers the analytical issues they raise. Examples include communal lamentation (Borna), talking drums (Akan, Yoruba), a griot’s instrumental imitation (Labé), vocal imitation of instrumental music (Wagogo), recreation of traditional dance on the piano (‘Agbadza’), and war chanting (Banda Linda). We are lead to pose a broader question — What kind of gift, legacy, constraint, or opportunity is iconicity? — which takes us beyond music, prompting us to attend to how African creativity prioritizes semiotic patterns.
Hagit Borer (Queen Mary University of London)
title: Mapping complex words to phonological representations:
Lessons from (comparative) Semitic
abstract: Combinatorial approaches to word structure bring into sharp focus the mapping of syntax to phonological form (PF). Semitic languages are famously incompatible with such approaches, as all verbs (and many nouns) contain a root (with 2-4 consonants), a template (for morphological class), and a vocalic melody (for Tense, Aspect, Voice); e.g., Hebrew hiqlid ‘typed’ contains the root Q,L,D, the template h-, and the vocalic melody i-i (Past). Prosodic constraints model linearization of C and V, and co-occurrence restrictions indicate a syntax such as (1), where selection is local: Roots select Templates; Templates select Tense; Roots don’t select Tense. This predicts that Passive Voice (2) will interrupt Template and Tense; however, Arabic (but not Hebrew) seems to exhibit non-local selection for (2). The difficulties vanish if one embeds syntactic locality in the mapping between syntactic structure (including word structure) and PF.
(1) [Tense [Template [Root]]] (2) [Tense [Passive.Voice [Template [Root]]]]
Joash Gambarage & Lisa Matthewson (The University of British Columbia)
title: The Bantu-Salish connection in determiner semantics
abstract: This talk provides a semantic analysis of the D(eterminer)-system in Nata (Eastern Bantu, E45), and compares it with the strikingly similar D-system of St’át’imcets (Lillooet Salish). Our core proposal is that Ds in both languages encode whether the speaker believes the noun phrase’s referent exists in the real world. Following Gambarage (2019), we assume that the D category in Nata is instantiated by augments/pre-prefixes, with a semantic contrast between presence vs. absence of an overt D. We show that neither Nata nor St’át’imcets Ds encode well-known distinctions like definiteness or specificity. Instead, they encode existence. Despite these parallels, the two D-systems are not identical. In St’át’imcets, existence is asserted based on the speaker’s personal knowledge; in Nata, overt Ds are also used for entities which are surmised to exist, or are future possibilities. We address the implications of our findings for the question of universals and variation in determiner semantics.
Maxwell Kadenge (University of the Witwatersrand)
title: African languages as archives of African Indigenous Knowledge Systems:
Some reflections on the taboos and names of the Shona people of Zimbabwe
abstract: Indigenous Knowledge Systems (IKS), also known as traditional knowledge or ethnoscience, is a body of knowledge of ‘indigenous’ people that encodes their experience of a particular location and how they respond to challenges in it. This body of knowledge manifests in different facets of life, including onomastics, agriculture, health, security, relationships, environmental management, development, education and rites of passage. For example, Shona taboos and totems serve as codes of conduct to preserve the natural environment, resolve conflict, and maintain the integrity of relationships. Shona names such as ‘Torture’, ‘Anywhere’, ‘Reward’ and ‘Network’ are terse expressions that communicate a people’s experience, worldview and belief system. They are loaded with meaning and identify their bearers uniquely by evoking a people’s history and everyday reality. This presentation explores how African languages are repositories of speakers’ IKS as demonstrated by the sociocultural meanings embedded in Shona taboos, totems and names of the people.
title: A multi-level linguistic approach to oral narrative development of South African languages
abstract: This talk explores the interplay between linguistic theory, linguistic structure and acquisition in Bantu language development by examining how children speakers of several languages of South Africa develop pragmatic and discursive cues during production of oral narratives. The investigation of lesser-studied languages, such as Bantu languages, further provides much-needed information on the cognitive issues at work before a child is a competent speaker. I discuss the complexity of discursive development from a multilingual perspective which incorporates speech and co-speech gesture. Though late language acquisition shows that oral narratives gain in linguistic complexity with age, children nevertheless do not attain the full mastery of adult speakers. Oral narratives reveal the cognitive factors involved in language processing from morpho-syntax through to pragmatics, and the linguistic structure and culture of a language impacts how the developmental trend is distributed across the languages under investigation.
Bert Remijsen (University of Edinburgh)
title: Evolutionary phonology and pathways of quantity in West Nilotic languages:
evidence from phonetics, phonology & morphology
abstract: A central hypothesis of Evolutionary Phonology (Blevins 2004) is that phonetic patterns that are hard to produce or perceive may develop with the effect of avoiding syncretism in morphological paradigms. The study of quantity in four languages that belong to the West Nilotic subgroup of the Nilo-Saharan language family (Dinka, Shilluk, Nuer, and Thok Reel) provides support for this hypothesis; the relevant evidence comes from the phonetics, phonology and morphology of quantity. In these languages, vocalic suffixes have been lost diachronically, with several typologically remarkable phenomena developing as a result. The best-known of these is three-level vowel length (Andersen 1990), attested in all four languages. In addition, in Shilluk, there is evidence of vowel shortening and of floating quantity; while floating autosegments are well established in the study of tone, the Shilluk phenomena represent a comparable phenomenon in the realm of quantity.