Glenda B. Claborne
Comm 515
Paper # 1

Are Gestures Nonverbal?

McNeill's Argument and Supporting Evidence

A common distinction between nonverbal and verbal behavior is based on a body/language perspective: nonverbal behaviors are communicated through bodily movements while verbal behaviors show through speech. Underscoring this distinction is the notion that body movements such as gestures operate through a separate channel and convey different meanings from those behaviors involving speech (see Feyereisen & de Lanoy, 1985 and Beattie, 1981, for reviews). McNeill (1985) challenged this perspective by arguing that gestures and speech develop synchronously from the same internal processes and structures. His argument suggests that if such is the case then gestures should be considered 'verbal'.

To support his assertion, McNeill (1985) gave the following evidence:

1. Gestures occur only during speech. McNeill cited a study which showed that speakers performed gestures thousands times more than listeners during a given period of time. He cited this study as proof that speech and gestures are so intertwined they must be restricted to the same thought processes occuring in the speaker's mind.

2. Gestures and speech have parallel semantic and pragmatic functions. McNeill discussed in detail the kinds of gestures he considers are equivalent to linguistic units in meaning and function providing further evidence that gestures and speech share the same computational stage. Referential gestures (iconix and various kinds of metaphorix) convey not only concrete images but also abstract concepts that are semantically parallel to sentences. Another class of gestures called beats do not have propositional content of their own but function more like metalinguistic indexes between breaks in speech or as prompts to get the speech going again.

3. Gestures synchronize with parallel linguistic units. Not only do gestures perform parallel semantic and pragmatic functions as speech, they also occur at the same time that the parallel linguistic units occur. To McNeill, this could only mean that "the linguistic unit and gesture belong to the same psychological moment and could arise from a common psychological structure activated on-line during speech" (p. 361).

4. Gestures are affected like speech in aphasia. McNeill noted that dissolution in speech had its parallel dissolution in gestures in cases of Broca's and Wernicke's types of aphasia. A disruption in and retention of a speech-related ability have their parallel disruption and retention in gestures in these types of aphasia. Broca's aphasics retain the ability to use referring terms and the corresponding ability to use referential gestures. However, Broca's aphasics also lose the ability to combine linguistic units into larger grammatical wholes, accompanied by the loss of ability to use beats which function to mark the interrelations between speech units. The reverse is true for Wernicke's aphasics. This parallel and simultaneous dissolution of function in speech and gesture when brain injury occurs suggests to McNeill a common neurological base for speech and gesture and therefore also a common computational stage.

5. Gestures develop parallel to speech in children. McNeill cited several studies which showed that both language and gesture development go through the same stages as a single system. Children coordinate their gestures and speech from an early stage of denoting concrete objects and situations, to expressing relations between objects, people and situations, and on to expressing abstract images and symbols. That children's gestures and speech follow the same developmental path is further evidence to McNeill that these two types of behavior arise from the same internal computations.

Critique

I find McNeill's (1985) arguments compelling but it is not clear to me whether establishing a common internal origin for two different sets of behavior would mean having to put these behaviors together in the same category as McNeill suggested by the title of his article. This to me is a definitional issue begging for more definitional questions. It is not clear to me whether a criterion is agreed on that if two sets of different behaviors (speech and gestures in this case) are proven to share a common computational stage then they should be classified in the same category (verbal behavior in this case).

There is really little debate about speech and gestures arising from closely related internal mechanisms as Feyereisen (1987) himself pointed out. I agree with Feyereisen (1987) and Butterworth & Hadar (1989) that McNeill relies too heavily on a single, complex, unitary notion of an inner speech to explain the mechanism for a common computational stage for speech and gestures. I agree with these critics that this doesn't explain the different patterns of association and dissociation between speech and gestures. I agree with the more plausible explanation that a specification of different levels or stages of interaction would explain the interrelationship of speech and gestures better.

McNeill (1987; 1989) charged Feyereisen, Butterworth and Hadar of taking an information-processing approach to the relation of speech and gestures, which naturally led them to view the relation in terms of thought as input and speech and gestures as outputs. To McNeill, this linear view naturally rejects any consideration of dialectical processes connecting speech to gestures. McNeill (1987; 1989) also pointed out that his critics were confused about the types of gestures on which McNeill(1985) based his arguments. Citing Kendon's (1983) continuum of gestures, McNeill (1987) emphasized that he was referring to those gestures on the leftmost end of Kendon's continuum which always accompanied speech while his critics referred to those gestures towards the righthand side of the continuum (pantomimes, emblems and sign languages).

I think that McNeill (1987; 1989) jumped too eagerly on the linear thinking charge because Butterworth & Hadar (1989) clearly used the examples of iconic gestures to illustrate the usefulness of their proposed stages and Feyereisen (1987) was clear in referring to gestures that usually accompany speech. Contrary to a rigid, linear way of looking at the close relation between speech and certain kinds of gestures, I think that McNeill's critics actually argued for flexibility in conceptualizing the common computational stage. Feyereisen (1987) and Butterworth & Hadar (1989) suggested conceptualizing the computational stage in terms of different levels of processing so as to allow for factors such as emotional and social inhibitions which would decrease the synchronicity and parallelism between speech and gestures.

So again, it's really less a question of whether speech and gestures do arise from a common computational stage and are therefore governed by a common psychological structure, but a question of just what constitutes this stage and structure. Among neuroscientists and psycholinguists, there is no question that that there are certain types of gestures that are controlled by the same area of the brain that also controls speech. Steven Pinker (1994), in his book The Language Instinct, called these gestures "linguistic gestures" (p. 302). However, Pinker (1994) noted that even with very specific, highly technological mapping of the brain, scientists still aren't sure whether it is exactly the damaged part of the brain, the surrounding area, or the interconnections of these areas with other parts of the brain that control a specific loss of speech or gesture. In the same way, communication scholars have not precisely mapped the cognitive and affective structures that they hypothesize underlie most of the overt behavior that they observe. The classification of behaviors into nonverbal and verbal based on a body/language distinction maybe arbitrary but I think that it is a futile exercise to argue about definitional issues based on incomplete pictures of internal states.

References