Building Meaning in Context: A Dynamic Approach to Bantu Clause Structure

by Hannah Gibson (University of Essex)

When I first started on my doctoral research, I had access to two sentences from the Tanzanian Bantu language Rangi which seemed to show that the language allowed for post-verbal auxiliary placement. This word order was unexpected given that Rangi has a dominant SVO order. It is also unusual in the context of East African Bantu languages where auxiliary-verb order dominates.

I subsequently spent a year spent in Tanzania working with Rangi-speakers and Rangi-speaking linguists, and gathering data. Luckily for me, the two sentences I had started off with were indeed representative of the language. And the picture was in fact more complicated than that. While verb-auxiliary order is obligatory in the two future tense constructions, this order is ‘reversed’ in a range of syntactically-conditioned contexts: wh-questions, sentential negation, relative clauses, after subordinators and in cleft constructions.

Fast forward several years, and at the end of 2018, my monograph Building meaning in context: a dynamic approach to Bantu clause structure was published as part of the Philological Society’s Monograph series. The book adopts a parsing/production-based approach to modelling Bantu clause structure, employing the tools and assumptions laid out in the Dynamic Syntax framework. It includes a chapter which looks at this auxiliary placement alternation in Rangi in more detail, as well as examining a number of other features of Bantu morphosyntax.

Dynamic Syntax (DS; Cann et al. 2005, Kempson et al. 2001, 2011) is a grammar formalism that aims to capture this real-time parsing process. Under the DS perspective, linguistic knowledge is considered to be the ability to parse language in context, whilst syntax is considered to be the constraint-based way in which representations of content can be built up from words encountered in a string.

The book presents an overview of the key tools and mechanisms adopted by the Dynamic Syntax framework (Chapter 2). It then goes on to show the application of these assumptions to modelling the Bantu clause, drawing primarily on data from the East African Bantu language Swahili (Chapter 3). A number of key phenomena relating to Bantu morphosyntax – inversion constructions and passives (Chapter 4) and negation (Chapter 5) – are also examined.

Chapter 6 presents an account of the word order alternation found in Rangi that I set out to examine after seeing those two sentences all those years ago. There is also a chapter which goes beyond data from the Bantu languages and explores cross-linguistic similarities that emerge as a result of the formal account presented (Chapter 7).

One of the contributions made by the study of auxiliaries in Rangi, is that it further supports the predictive power of the DS framework. The formal tools used in this account of Rangi auxiliaries show parallels with accounts of distinct phenomena in unrelated languages – clitic placement phenomena in Cypriot Greek and Medieval Spanish, as well as cleft constructions in Japanese and ‘scrambling’ in Korean.

The book will be of interest to those working on African languages, and particularly the languages of the Bantu family. But also those interested in word order phenomena, lexicon-based formal approaches to modelling language, as well as those wanting to find out more about Dynamic Syntax.

TPS 116(2) – Abstract 1

Non‐Negative Word Order In Breton: Maintaining Verb‐Second

by Holly J. Kennard (University of Oxford)

This paper examines variation in Breton word order patterns in non‐negative utterances across speakers of different ages. Not only has there been some disagreement on how best to characterise unmarked word order in Breton, it has also been claimed that younger speakers of so‐called Neo‐Breton overuse subject‐initial word order under influence from French. Data from fieldwork provide a complex picture of word order variability. This seems to be driven by a number of factors, including the nature of the subject (lexical or pronominal), regional variation among older speakers, and a corresponding lack of regional features among younger speakers. Rather than overusing subject‐initial word order, the Neo‐Breton speakers tend to avoid this word order pattern when other word orders are available, such that the verb‐second pattern is being maintained.

DOI: 10.1111/1467-968X.12119