Nonmanual Gestures in American Sign Language
Evelyn McClave
California State University, Northridge

Non-manual gestures frequently co-occur with lexicalized signs in American Sign Language. This paper will present evidence that some prominent head movements observed among native signers are gestures rather than grammatical non-manual markers.
Head movements function grammatically in a wide range of contexts in ASL. For example, patterns of face and head movements together signal grammatical functions such as yes-no questions (Baker and Padden, 1978), rhetorical questions (Baker-Shenk, 1985), relative clauses (Liddell, 198), conditionals (Baker and Padden, 1978), and topics (Liddell, 1980).
The utterance under study is taken from a video of a Deaf linguist whose native language is ASL. In this segment the linguist is discussing data she collected from two deaf, Native American informants, male and female, for a study of what she refers to as non-handed signs. She mentions that she counted occurrences of manual signs for “yes” and “no" and compared them to occurrences of head movements for “yes” and “no”, the familiar nods and shakes, respectively. She states that the head movements for “yes” and “no” outnumbered by far the manual signs for those concepts. Then referring to non-manual head movements she signs

WOW SHOW-UP MANY
<----------------------------->
Wow! Many showed up.
(Following conventions for noting ASL, the English gloss for the signs are all in caps. The broken line indicates the duration of a lateral headshake. It is drawn under the notation to distinguish this head movement from a non-manual sign.)

The lateral head movement looks identical to the head movement for the negative non-manual marker. Yet all who viewed the tape, including seven native and near-native signers, did not interpret the utterance as negative, and the linguist herself later confirmed that no negation was intended. Despite what appears to be a negative, non-manual sign, all interpreted this utterance as affirmative. How can we explain this?
One approach would be to claim that context overrode what looks like the
non-manual negation signal. The signer has just stated that there were more tokens of non-manual signs for “yes” and “no" than manual signs for these concepts. So when observers see her shake her head as she signs “Wow! Many showed up,” they simply don’t believe that she means they didn’t show up. This explanation is unsatisfactory, however, because it assumes that the signer mistakenly used a non-manual marker – one as common as negation, which seems highly unlikely.
Another possibility is that the viewers noted the absence of the facial expression that often co-occurs with the negative headshake; that is, downturned corners of the mouth together with raised lower and upper lips (Liddell, 1980, p. 39). Yet the subject’s mouth is not in a relaxed, neutral position. Its configuration most closely resembles the non-manual adverb ‘mm’: the lips are together and pushed out but not puckered. This facial expression carries the meaning of enjoyment and relaxation (Liddell, 1980, p. 42). The co-occurrence of ‘mm’ and lateral headshakes has not been previously noted and discussed, however. Is this then a new non-manual marker comprised of the ‘mm’ facial expression and side-to-side headshakes? This is also unlikely because I have noted (but not videotaped) such lateral headshakes among native signers accompanied by smiles rather than the ‘mm’ expression in non-negative contexts. This suggests that the headshakes and ‘mm’ facial expression do not form a unit.
I would like to suggest that the signer was using a head movement pattern, a gesture, that is extremely common in American culture among hearing non-signers when assessing something either positively or negatively. The signer unwittingly assimilated the gesture from hearing Americans, and other native signers recognized the pattern and interpreted the head movement as a gesture rather than a non-manual sign.

Head movements of non-signers:
For a moment let us look at head movements among hearing Americans. Research has shown that movements of the head among hearing people during speech are not random (McClave, 2000). Candy Goodwin (1980) was the first to observe that lateral shakes of the head co-occurred with verbal assessments such as "It was so good.” The following examples are from hearing non-signers:

Aaron: You like that job?
Robert: Oh, that was a great job.
< ------------------------- >
Jane: I know. She drives her little yellow jeep
Valerie: I know
Jane: ya know
Valerie: She’s totally she’s so cool.
< ---------------------------->

Even though assessment head movements are below the level of conscious recall when they co-occur with speech, they are conventionalized movements in American culture. Of course, in isolation and out of context, lateral shakes mean "no" in American culture.
In ASL, neither WOW, nor SHOW-UP, nor MANY has a required non-manual component. Rather, the signer is using the conventionalized intensification head gesture simultaneously with her signs.
Uncertainty shakes
Another example of gesture co-occurring with ASL is mentioned by Reilly, McIntire and Bellugi (1994, p. 135). In the process of researching facial expressions, the authors recorded examples of "where" questions accompanied by lateral shakes. Reilly
et al. consider the shakes to be communicative gestures and "linguistically irrelevant" for ASL (1994, p. 135). ASL linguists agree that lateral head shakes are not required with WH-questions in ASL (Baker-Shenk, 1985; Valli and Lucas, 1992).
How then can we account for the lateral movements that Reilly et al. observed? It seems that head shakes that commonly accompany WH-questions among non-signing Americans were borrowed into ASL. Wh-questions and other expressions of uncertainty are often, though of course not obligatorily, accompanied by lateral shakes in hearing American culture. The following examples are from conversations of non-signers.
Valerie: ...I walked in to to work and Hasmeek says
"What," she goes, "what happened yesterday?"
<----------------------------------------------------> Valerie: The gra- the Graduate School of Education?
Jane: What does he do? What?
<----------------------------->

We have seen examples of signers using head movements for assessment and uncertainty that do not have a recognized linguistic function in ASL. These head movements are gestures, and evidence has been presented that these same movements are conventionalized among hearing, non-signing Americans in comparable linguistic environments. I hypothesize that the native signers immersed in American culture and attuned to the kinesic cues of speakers have assimilated these head gestures. What is more, it is likely that this process of assimilating gestures has been continuing for some time, and it is the likely source of some head movements that are now grammaticized in ASL.
It is not claimed that this process is the origin of all signs in ASL. It clearly is not. However, the close correspondence between some conventionalized, spontaneous gestures and some grammaticized features of ASL in the same linguistic contexts suggests that in some cases one evolved into the other. Direct quotes can be used to illustrate.Direct Quotes
When American speakers shift from indirect to direct discourse, they simultaneously change their head position (McClave, 2000). These head movements, often subtle and abbreviated, occur frequently among hearing, non-signers. Two examples follow:
Aaron: When they tell me
“Oh, don’t walk through this neighborhood. It’s so dangerous…”
head leftValerie: I said, “I had your son in my summer school class couple of years ago”
She goes, “That’s right!”
head right tiltThese shifts in head position occur in the exact environment in which we see body shifting in ASL. That is, as a signer starts a direct quote, the shift is marked with a change of head and body position. Among hearing people, the movement can be quite constrained and restricted to the head. Body shifting is grammaticized in this same linguistic environment in ASL. It is plausible that at some time in the past signers assimilated these very subtle but conventionalized movements from the hearing culture and the movements gradually underwent grammaticization in ASL.
Consistent with this hypothesis is the fact that head movement does not universally signal direct quotes. Vermeerbergen (personal communication) notes that there is no head movement with role shifts in Flemish Sign Language. Both American speakers and signers move their heads at the beginning of direct quotes. It is a movement pattern used in North America, and it is found in other cultures as well (McClave, et al., in preparation). It is likely that this pattern was adopted from the hearing culture and grammaticized in ASL. That expression of thought and language that is gestural is easily borrowed from spontaneous gestures of the hearing into signed languages since it is the most accessible to the Deaf.
Once adopted by signers the gestures may remain gestural in nature, that is, not conforming to the constraints of the linguistic system or they may eventually become grammaticized. Evidence of ongoing borrowing of lateral movements for assessment and WH-questions leads to the hypothesis that other spontaneous non-manual movements were borrowed in the past. The direction of borrowing from hearing culture into Deaf culture rather than the reverse is probable given the comparatively small number of Deaf Americans and the fact that the spontaneous gestures discussed here are used by most, if not all, hearing Americans.