What is a Lexical-Functional approach to Barayin Morphosyntax?

Last month, I had a book published on a fairly niche topic: Barayin Morphosyntax: A Lexical-Functional Approach. In the words of Facebook friend Jon Laurion, “I had to look up the definition of most of the words in the definition of Barayin Morphosyntax.” I’m not sure the blurb on the book cover or website is much help, so here is a bit more background to what this book is about.

Barayin is the name of a language — one of over 100 languages spoken in Chad. The Barayin are a group of about 6000 people, but they have four distinct subgroups: Jalkiya, Giliya, Komiya and Jalking. This book is actually only on the language used by the Jalkiya subgroup, but it’s not really about “language” in the everyday sense of the word. It’s not about how to learn the language, nor does it cover expressions and idioms that reflect the local culture. It’s about the grammar, or (more specifically) the morphosyntax.

Grammar, in the descriptive sense, is about the internal organization of how the sounds of the language (or signs in a sign language) are organized to convey meaning. The two aspects of Barayin grammar that this book is focused on are morphology and syntax (or morphosyntax). Morphology is the “shape” of words which basically looks at why words take particular prefixes or suffixes in different contexts. Syntax is about the order the words are found in, and how they pattern in hierarchically organized units that make up sentences and longer utterances. Other aspects, like phonology (analysis of the sounds) and pragmatics (how language gets interpreted in real world contexts), are not really discussed in this particular publication.

Okay, so that’s Barayin morphosyntax: how words (including their affixes) come together to express different meanings in Barayin. What about the “Lexical-Functional Approach”? The hyphenated adjective is a reference to Lexical-Functional Grammar (LFG). LFG is a way to model morphosyntax. At the most basic level, the linear order of words in a sentence are modeled by a tree diagram. In a tree diagram, the organization of words into units is shown by nodes (basically just dots) that connect to a higher node which in turn can be combined with more nodes to create more complex representations of how the parts relate.

Some very very basic tree diagrams (page 12)

Other information relevant to the morphosyntax is represented in two-part lists where one part labels a relevant grammatical category (like number, gender or tense) and the other parts give the value that a word or phrase has for that category (for example, NUMBER SINGULAR or NUMBER PLURAL or GENDER MASCULINE or GENDER FEMININE).

Attribute-value matrix used to represent features of the word “him” (page 36)

This simple way of representing grammatical features (called “attribute-value matrices”) sets up a quasi-mathematical method for relating parts of a sentence in more abstract ways. An equation could be written to say that an adjective has to have a suffix that has the same gender value as the noun it modifies, or that a verb requires any verb in its sentential complement to have a particular tense.

Equations to constrain the use of the “plural addressee marker” na

Without going into more detail, one of the satisfying aspects of this approach is that there is a computer program called XLE that allows you to program your model and test sentences from the language to see if the model parses the grammar of each sentence they way you predict. So the book also shows aspects of how the model was implemented in XLE to develop a type of computational grammar for Barayin.

Screenshots of the XLE parsing output (page 96)

Modelling the grammars of languages is part of a decades-long emphasis in the study of syntax that was mostly instigated by the idea that there must be some fundamental property that all languages have in common in order for them to be learnable. Results on that research project are mixed, and a crucial issue has been expanding the diversity of languages studied in order to test any hypotheses about supposed universals of language. This is why the series my book is published in focuses on endangered languages.

However, to set the context, it’s worth pointing out that modeling morphosyntax was not the primary or exclusive goal of working with the Barayin. The work began with an invitation from the community to help launch a literacy program, which involved some basic descriptive analysis of the language, which in turn lead to me, as a researcher, having more questions about how to compare Barayin with other languages and how to develop a more systematic understanding of how the grammar functions.

A well in Melfi, Chad