The Routledge Handbook of Theoretical and Experimental Sign Language Research

Josep Quer, Roland Pfau, Annika Herrmann, Josep Quer, Roland Pfau, Annika Herrmann

• Publisher: Routledge
• Year: 2021
• ISBN:
  9781317624271
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Sign Language
• Index: Languages & Linguistics

The Routledge Handbook of Theoretical and Experimental Sign Language Research bridges the divide between theoretical and experimental approaches to provide an up-to-date survey of key topics in sign language research. With 29 chapters written by leading and emerging scholars from around the world, this Handbook covers the following key areas:

• On the theoretical side, all crucial aspects of sign language grammar studied within formal frameworks such as Generative Grammar
• On the experimental side, theoretical accounts are supplemented by experimental evidence gained in psycho- and neurolinguistic studies
• On the descriptive side, the main phenomena addressed in the reviewed scholarship are summarized in a way that is accessible to readers without previous knowledge of sign languages

Each chapter features an introduction, an overview of existing research, and a critical assessment of hypotheses and findings. The Routledge Handbook of Theoretical and Experimental Sign Language Research is key reading for all advanced students and researchers working at the intersection of sign language research, linguistics, psycholinguistics, and neurolinguistics.

Chapters 5, 18 and 19 of this book are freely available as a downloadable Open Access PDF at http://www.taylorfrancis.com under a Creative Commons [Attribution-Non Commercial-No Derivatives (CC-BY-NC-ND)] 4.0 license.

1

Sign language phonology

Theoretical perspectives

Harry van der Hulst & Els van der Kooij

1.1 Introduction

In this chapter, we provide an overview of the ‘phonology’ of sign languages (admittedly biased toward our own work, but with ample discussion of, and references to work by many other researchers). In Section 1.2, we will first focus on presenting constraints on sign structure. In doing so we use our own model for the overall structure of signs as a frame of reference, although our objective is not to argue for it or imply that this is the only defensible proposal.¹ Then, in Section 1.3, turning to rules, we will argue that sign languages do not seem to display a type of phonological rules that is typical of spoken languages, namely rules that account for allomorphy. To explain this, we will distinguish between what we call utterance phonology and grammatical phonology. While sign languages certainly display utterance phonology, as we will show (effects of automatic assimilations, co-articulation, and reduction processes), we will argue that we do not find what we will call grammatical phonological rules which account for allomorphic alternations, although we will consider possible objections to this claim. While the point of establishing phonological structure in signs rests on the claim that signs consist of meaningless parts (like the consonants and vowels in spoken languages are meaningless parts), in Section 1.4, we discuss views that question the alleged ‘meaningless’ character of all phonological building blocks. Recognizing meaning-bearing units² will provide a possible explanation for why sign languages seem to disallow allomorphic rules. We will then propose that the lack of grammatical phonological rules that regulate allomorphy in sign phonology is compensated for by rules of a different kind (which might be called phonological). Such rules account for systematic form-meaning relationships internal to alleged ‘monomorphemic’ signs.³ Section 1.5 presents some of our conclusions.

1.2 Basic units and constraints

Up until 1960, sign languages were not regarded as fully-fledged natural languages that possess morpho-syntactic structure and an independent level of phonological structure. Recognition of phonological compositionality, which was due to the groundbreaking work of William Stokoe (Stokoe 1960), suggested that sign languages display duality of patterning which had long been identified as a pivotal property of spoken languages (Martinet 1955; Hockett 1960). Stokoe proposed a transcription system for signs that replaced holistic drawings and verbal descriptions by a finite number of graphic symbols for what he perceived as separate meaningless parts or ‘aspects’ of the sign: the handshape, the movement of the hand, and the location in front of or on the body.⁴ The ideas of Stokoe were further developed in the sense that other properties of signs were added to his list of major units. Also, the major units were subsequently decomposed into distinctive features. We will provide examples of such additional proposals in subsequent subsections.

While it is possible to recognize and transcribe phonetic properties of signs, it simply does not follow automatically that these distinctions have a reality in the mind of the signer in terms of storage in memory and language processing. Groundbreaking work on these issues was done by a research group at the Salk Institute during the 1970s, resulting in the Klima and Bellugi volume (1979), a must-read for sign language researchers. Performing recall experiments, they showed that percepts of signs in short-term memory are compositional. Importantly, they also showed that errors were in the direction of formational similarity and not meaning. Studying ‘slips of the hands’, they argued for the likelihood of compositionality in the articulatory phase (see Gutiérrez & Baus, Chapter 3). Finally, they showed that American Sign Language (ASL) users can make judgments about what they considered well-formed or ill-formed for ASL, which supports formal compositionality in the lexicon.

We will now turn to our review of phonotactic constraints in sign languages. It is well known that spoken languages can differ quite dramatically in the constraints that specify the inventory of segments and the ways in which these segments can be combined, despite the fact that almost all constraints follow universal ‘markedness’ principles which regulate symmetry in inventories, sonority sequencing in syllable structure, and assimilation in sequences, and the general fact that ‘more complex structures’ imply the presence of ‘less complex structures’. While the phonological form of sign languages is likewise subject to phonotactic constraints, there is much less evidence for cross-linguistic differences in this respect, although, arguably, there are currently not enough data from typological cross-linguistic studies to know to what extent different sign languages can differ.

Putting aside for the moment the question whether sign languages have units that formally or functionally can be compared to such notions as vowel, consonant, or syllable (see Section 1.4 for discussion), we would find evidence for language-specific (context-free⁵) constraints if sign languages differed in their inventories of handshape, movement, or locations. Context-sensitive constraints (assuming that the context does not have to be linear) would capture restrictions on the manner in which these major units can be combined to make up signs. While constraints of such kinds certainly exist, it seems to be the case that sign languages only display minor differences in terms of their inventories of the major units (as was already shown in Klima & Bellugi’s (1979) comparison of ASL and Chinese Sign Language), as well as in the ways in which these units combine into signs. But, again, only few systematic typological studies have been done.⁶ However, the sets of major units as found in sign languages do not seem to be random (just like vowel and consonant inventories are not random), and while this perhaps can be explained on the basis of phonetic principles of articulation and perception (as has also been argued for spoken languages), it would seem that an account in terms of smaller building blocks (features) provides insight into these inventories. Thus, even though perhaps most constraints appear to generalize over sign languages as a whole (rather than differentiating between them), they nonetheless, like in spoken languages, provide a window on the compositional structure of the major sign units in terms of features. Given a feature analysis, we can, for example, make distinctions between simpler and more complex units, which allows for implicational statements. As in the case of segmental inventories of spoken languages, the occurrence of more complex units implies the occurrence of simpler and more frequently occurring units, which are often called ‘unmarked’. Given the right set of features, simpler (and thus less marked) units have a simpler featural structure (see Battison 1978; Sandler 1996b).

(1)	Examples of unmarked handshapes, locations, and movements^7
	a.	Handshapes: B ( ), S ( ), 1 ( )
	b.	Locations: Neutral space, Chest
	c.	Movements: Downward path, End contact (hand touching a location)

Given that signers are able to judge whether or not a given handshape belongs to t...