Two New Sign Languages - UCLA Center for Behavior, Evolution

advertisement
#Word count limit: 2,500 words, counting the abstract and text only, and 15-18 references#
Penultimate draft (3500 words):
Arbib, M.A., 2007, New Sign Languages and Language Evolution, Current Directions in Psychological Science (to
appear).
New Sign Languages and Language Evolution
Michael A. Arbib
Computer Science Department, Neuroscience Program, and USC Brain Project
University of Southern California
Los Angeles, CA 90089-2520
arbib@pollux.usc.edu
Abstract
Human language is far more than speech and its derivatives such as writing. Human signed languages like
American Sign Language are fully expressive human languages, and speakers normally accompany their speech
with facial and manual gestures. Thus any theory of language evolution must address these integral roles that
manual signs and gestures play today. What are the capabilities of the human brain that make it possible for humans
to learn language while other creatures can not? How much structure must the social environment offer a child to
acquire language? We probe these questions by studying two sign languages of recent vintage. Nicaraguan Sign
Language developed in just 25 years while Al-Sayyid Bedouin Sign Language developed over at most 70 years, and
are still developing. We examine the emergence and dynamics of these languages to advance discussion of what
supports society offered to allow these communities to exploit the human brain’s readiness for language in novel
ways.
DESIGN FEATURES AND LANGUAGE EVOLUTION
(Hockett, 1987) listed "design features" common to all languages. Every language consists of discrete units which
can be combined in open-ended ways to express novel meanings – yielding discreteness and combinatorial
patterning, respectively. Duality of patterning refers to the way in which meaningful units are composed from a
smaller set of meaningless units as when discrete sounds combine to form words. Moreover, many Deaf people
communicate by grammatically combining signs, based on hand shape and trajectory, possibly accompanied by
expressive facial movements. Their sign languages are fully expressive human languages yet share certain features
which distinguish them from spoken language, due to the modality of transmission (Sandler & Lillo-Martin, 2006).
Nonetheless, (Stokoe, 1960) demonstrated that American Sign Language (ASL) also has duality of patterning –
meaningless handshapes, locations, and movements combine to form a large set of signs.
However, something called a design feature of language may actually reflect more general capabilities. For
example, a frog’s behavior exhibits discreteness and combinatorial patterning, being composed from discrete motor
schemas like orient, jump, lunge, and snap. The Mirror System Hypothesis (Arbib, 2005; Rizzolatti & Arbib, 1998)
suggests that much of what enables the human brain to support language rests on mechanisms which evolved to
support praxis and protolanguage rather than language per se: Praxis comprises skills that include manipulation of
objects towards some practical end; whereas protolanguage is a form of communication posited to be used by
communities of our distant ancestors who – transcending the limited vocalizations and gestures of nonhuman
1
primates – shared a stock of communicative symbols (call them protowords) which could be added to repeatedly,
but has no syntax. Here we leave implicit the role of the mirror system and instead present hypothesized
evolutionary stages that start with the capacity for imitation.
Apes have a capacity for simple imitation. Gorillas learn elaborate feeding strategies but may take months to do
so. (Byrne, 2003) implicates imitation by behavior parsing. The young ape comes to recognize relevant subgoals
and by trial and error derives action strategies for achieving them. By contrast, humans are capable of complex
imitation, perceiving when and how a novel action may be approximated by a composite of known actions and
employing this perception to perform an approximation to the observed action, which may then be refined through
practice. This grounds the claim for two evolutionary stages:

A simple imitation system for grasping, shared with the common ancestor of human and apes.

A complex imitation system for grasping, unique to the hominim line leading from that common ancestor to
humans.
Each of these changes would have been of evolutionary advantage in supporting the transfer of novel skills
between the members of a community. But what of communication? Monkey vocalizations are innately specified
(though occasions for using a call may change with experience), whereas a group of apes may communicate with
novel gestures, perhaps acquired by ontogenetic ritualization whereby some fragment of an action comes to serve
for communication about that action (Tomasello, Call et al., 1997). This supports the argument for two further stages
of hominim evolution yielding:

Protosign: a manual-based communication system breaking through the fixed repertoire of primate
vocalizations to yield an open repertoire.

Protolanguage: an expanding spiral of conventionalized manual, facial and vocal communicative gestures.
The transition from complex imitation to protosign may have involved pantomime first of manual praxic actions
and then of non-manual actions (e.g., flapping the arms to mime a flying bird). Further evolution would yield
protosign as a system of conventional gestures that initially simplify, disambiguate (e.g., to distinguish "bird" from
"flying") or extend pantomime but in due course come to replace it (Arbib, 2005). This capability for protosign may
then have provided essential scaffolding for protospeech and the human language-ready brain. Finally we have:

The transition from protolanguage to language: the development of syntax and compositional semantics
One influential view is that (i) each human is innately endowed with a Universal Grammar for an autonomous
syntax that combines words into well-formed sentences irrespective of their meaning; and (ii) that the child need
simply hear a few sentences to “set the parameter” – such as whether or not the language is S-O-V (having sentences
based on the form Subject-Verb-Object) – for each key principle to acquire the grammar of her first language
(Baker, 2001).
There is no space to develop my counter-argument (Arbib, 2005) here in any detail, but the key hypothesis is
this: Once humans have brains that support both complex imitation and protolanguage (combining speech and
gesture), they have brains that can learn and use language. Complex imitation combines the ability to perceive that a
novel action may be approximated by a composite of known actions associated with appropriate subgoals with the
ability to employ this perception to perform an approximation to the observed action, which may then be refined
through practice. Both parts come into play when the child is learning a language; while the former predominates in
adult use of language. However, a modern language combines a vast lexicon with a large number of constructions
2
which determine how words and morphemes may be combined to create new meaning. The catch, then, is that it
may well have taken early Homo sapiens many tens of millennia to develop enough lexemes and constructions for a
tribe’s communication system to qualify as a language rather than a protolanguage (though no rigid threshold is
claimed).
But were tens of millennia really necessary? We now turn to two new sign languages which appeared in tens of
years rather than tens of millennia to see if there is something about the environment in which they emerged which
short circuited the historical path to language.
TWO NEW SIGN LANGUAGES
A deaf child may reach six years of age with less than 10 words of oral vocabulary, yet deaf babies raised among
signers follow the same linguistic milestones as a hearing child acquiring a spoken language. A deaf child raised
apart from deaf signers will often develop home sign, an idiosyncratic set of gestures (some pantomime-like, others
more conventionalized) used to communicate with family members. Typically, such a child will have a small
“vocabulary” of home signs together with a few strategies for combining signs into longer messages (GoldinMeadow, 2003).
Before the 1970s, deaf Nicaraguans had little contact with each other and no shared sign language emerged. An
elementary school for special education of the deaf was opened in Managua, Nicaragua in 1977, followed by a
vocational school in 1981. Students continued their contact outside school hours, and by the mid-1980s deaf
adolescents were meeting regularly on weekends. They began to develop a new gestural system, in part by
consolidating their idiosyncratic home signs, which soon expanded to form a rudimentary sign language (Kegl,
1994) from which emerged Nicaraguan Sign Language (NSL). As of 2005, there were about 800 deaf signers of
NSL, ranging from 4 to 45 years old.
Al-Sayyid Bedouin Sign Language (ABSL) is another recent sign language which, unlike NSL, developed within
the family structures of a pre-existing socially stable community. The Al-Sayyid Bedouin group is in the Negev
region of Israel, but ABSL is distinct from Israeli Sign Language. The group descends from a single founder, who
arrived from Egypt 200 years ago. About 150 people with congenital deafness have been born into the community
within the past three generations. Not only the deaf members of the community but also many of its hearing
members communicate by means of ABSL. Signers readily use ABSL to converse about topics as diverse as social
security benefits, construction techniques, and fertility (Sandler, Meir et al., 2005).
And there is no reason to believe that either of these sign languages has finished developing. Clearly, the signs
employed in each language are different from spoken words. However, not only are the signs different, so too are
the constructions which combine them. Both NSL and ABSL employ constructions different from those of the
circumambient Spanish and Arabic, respectively. Here are examples:
Motion Description in Nicaraguan Sign Language
(Senghas, Kita et al., 2004) grouped 30 deaf Nicaraguans into cohorts based on when they were first exposed to
NSL. Signed expressions were recorded while subjects described a video clip in which the cat Sylvester, having
swallowed a bowling ball, wobbles rapidly down a steep street. These were compared to co-speech gestures (the
hand movements which normally, and usually unconsciously accompany speech) produced by 10 hearing
Nicaraguans describing the event in Spanish. This event combines manner (rolling) and path (descending). Senghas
et al. found that all the Spanish speakers' gestures and 73% of the first-cohort’s expressions included manner and
3
path simultaneously. By contrast, less than a third of these expressions for the younger, second- and third-cohort,
signers were of this kind – evidence of an emergent feature of NSL which made the segmented, sequenced
construction their preferred (but not exclusive) means of expressing motion events. Moreover, NSL now has the XY-X construction, not seen in the gestures or speech of Spanish speakers, to express simultaneity of X and Y.
Word Order in Al-Sayyid Bedouin Sign Language
Sandler et al. (2005) studied eight signers of the second generation of ABSL signers when spontaneously
recounting a personal experience or describing single events portrayed in video clips In the vast majority of
sentences recorded, the grammatical relation between subject (S), object (O), and verb (V) was given by the order
S-O-V, though subject or object could be omitted. For example, in the description WOMAN APPLE GIVE; MAN
GIVE [‘‘The woman gave an apple; (she) gave (it) to the man’’], the first clause is S-O-V, and the second clause is
IO(indirect object)-V. The only hearing subject they studied, bilingual in Arabic and ABSL, shows no influence of
the local Arabic’s S-V-O order in his signing. Hence, the robust word-order pattern exhibited by the data
demonstrates an independent development within ABSL.
EMERGENCE OF THE NICARAGUAN DEAF COMMUNITY
Is the implication of the rapid rise of ABSL and NSL that once the brain of Homo sapiens achieved its present
form, a mere 2 or 3 generations sufficed for emergence of a full human language? To the contrary, both ABSL and
NSL owe their rapid development to the existence of social matrices that would not have been present “in the
beginning”. I believe that the researchers involved with ABSL and NSL would agree, though some of the popular
press discussing their work has made suggestions to the contrary. A detailed history by (Polich, 2005) may help us
see not only the social forces at work in the emergence of NSL, but also the crucial role of individuals. She
emphasizes that the increased opportunities for adolescents and young adults to get together, which distinguished the
1980s from the 1950s, played a vital role in forming the Nicaraguan Deaf community and NSL. Ruthy Doran, a
teacher who, starting around 1980, gave deaf adolescents regular opportunities to socialize with each other, told
Polich: “There wasn't a sign language at the time ... But we were able to understand one another. We would … use a
lot of the gestures … And when everything else failed, we would write words down, or else act it out.” Thus the
community being formed included hearing people who spoke Spanish, while some who could not speak had at least
some words of written Spanish.
Ad hoc use of signs differs greatly from using grammar to flexibly create new sentences, but the creation of NSL
did exploit prior innovations of other language communities. News of existing sign languages reached Nicaragua in
the late 1970s (Polich, 2005). It was a revelation to see that signs could constitute a full, expressive language. Alas,
teachers were unable to get approval to use Costa Rican sign language for instruction in Managua. However, the
idea of sign language was now available to some members of the nascent community. Polich also records the
formation in the late 1980s of the Nicaraguan Association to Help and Integrate the Deaf. The early officers were
deaf young adults who could speak Spanish but were also part of the community that was creating NSL. Polich
helps us see individuals as embedded within a developing community in which increasing social interaction
supported the growth of NSL while the increasing “bandwidth” of communication in NSL strengthened the
community and increased the participation of those who could not speak Spanish.
Early members of the Deaf Association told Polich that one individual, Javier Gómez López, "taught" all the
others the sign language (in its early form). He had been given a sign language dictionary during an athletic trip to
4
Costa Rica in the late 1970s. He became dedicated to making sign language a functional communication system for
himself and his friends, and to sharing this knowledge with other deaf persons in Nicaragua. He was active in
workshops which decided which variations of signs should be adopted as "standard". The Royal Swedish Deaf
Association provided funds for officers of the Deaf Association to travel to other Central American countries for
regional meetings of deaf people. The Swedes also financed the collection of entries for a professionally published
dictionary for NSL, professional sign language instruction for teachers, and basic literacy training for the deaf.
Thus, although families play no role in transmission of NSL, the schools and Deaf Association has been essential
in building the Nicaraguan Deaf community, fostering and transmitting NSL in a community setting. By contrast,
the development of ABSL benefited from family and community structures very different from those in Nicaragua.
BROADER IMPLICATIONS
ABSL shows the influence of socio-cultural traditions of the larger society on such things as the signs chosen to
represent religious holidays – for example, the sign for MONTH derives from MOON and one sign for Thursday is
MARKET because Thursday is market day. Yet, in constructions like S-O-V, the cultural development of the
structure of ABSL seems to be internal to the sign language, and not influenced by the dialect of Arabic that is the
language of the larger community. However, I shall suggest that it is crucial to distinguish the transfer of ideas about
what to communicate and general properties of language that aid such communication from the specific forms
adopted by the language.
(Wray, 1998) argued that an important mechanism in language evolution was fractionation, replacing the
protowords of early protolanguages by words denoting their “components”. As protowords fractionated, syntactic
constructions emerged to recapture the meanings of protowords – but then also supported new combinations and
provided space for the creation of new words. I argue that this mechanism is at work in NSL. Senghas et al. (2004)
observed such a two-fold scenario in the decomposition of path and manner, and the development of the X-Y-X
construction to retrieve the simultaneity of the original gesture. My hypothesis is that the key capacity here is
complex imitation, rather than a design feature specific to language, and certainly not a setting of parameters in
some innate Universal Grammar.
Once one has the idea of pantomime, inventing new pantomimes is easy; and once one has complex imitation,
the sharing of successful pantomimes and their ensuing conventionalization is also straightforward.
But what converts a collection of protosigns from a protolanguage to a language? (Goldin-Meadow, 2005) asserts
that ABSL arose with no influence from any established language, signed or spoken. But ABSL developed in a
community which included Arabic speakers who had to find low-energy ways to sign what they wanted to say. The
resulting system has many novel features not present in Arabic but this is very different from asserting there is no
influence. Such outside influences providing an immense compression of the accumulation of cultural traditions. As
we saw, too, the community that developed NSL knew about Spanish and other sign languages – which is very
different from saying that many individuals possessed this knowledge. All that is required is that this motivates one
individual to develop a new sign or construction that others not only comprehend but also use. There is then a
“critical point” in terms of enough people using this innovation for its use to become widespread rather than
remaining localized or dying out. This is an example of cumulative cultural evolution which (Tomasello, 1999)
attributes to a uniquely human ratchet effect whereby “individual and group inventions are mastered relatively
faithfully by conspecifics, including youngsters, which enables them to remain in their new and improved form
5
within the group until something better comes along.” This is possible because human beings are biologically
adapted for culture in ways that other primates are not. Tomasello argues that the key adaptation is one that enables
individuals to understand other individuals as intentional agents like the self, but I think that complex imitation
would suffice without invoking any “Theory of Mind”. In any case, the point is that human infants begin to engage
with other persons in various kinds of joint attentional activities involving gaze following, social referencing, and
gestural communication. This supports forms of cultural learning allow human beings to, in effect, pool their
cognitive resources both contemporaneously and over historical time in ways that are unique in the animal kingdom.
Note that "a few words of Spanish" (relating to Ruthy Doran's account) or a sign dictionary and a few signs
picked up in an athletic event trip (relating to Javier Lopez) just gives a drive to develop a large vocabulary in a way
that was closed to early humans developing protolanguage. The seeds for a hierarchical syntax in NSL come from
the influence of deaf Spanish speakers in the “early days”. Interestingly, as the expressiveness of NSL grows, so the
influence of Spanish speakers wanes – the notion of hierarchy can now propagate within the growing complexity of
NSL. Again, I distinguish the “idea” of a property of language from its form in a particular language. The use of
writing may offer a useful comparison point. Writing systems have been in use for only a few millennia and we have
every reason to believe that non brain changes were required to support literacy (though the syndromes of dyslexia
show that not all human brains are equally well prepared to match speech to writing. Moreover, many societies have
lasted till modern times with no written form for their spoken language. Yet, once the notion of developing a writing
system enters` a given community, then the development of that writing system (whether phonetic or as a syllabary)
may proceed in just a few years, even if much effort is required to decide how to choose letters or syllables to
approximate the sounds of the language.
In conclusion, then, I argue that the key properties available to protohumans as well as to the NSL and ABSL
communities were basic brain mechanisms for complex imitation and a critical mass of people with whom to
“ratchet up” useful innovations. However, such ratcheting may well have proceeded slowly in small communities.
Discovering that one can acquire an endless array of symbols and constructions, though, would not have been
apparent to early humans but can be made readily apparent in a supportive community such as that which has
nurtured the emergence, and continuing development of AABSL and NSL.
Acknowledgements: My thanks to Susan Goldin-Meadow, Sotaro Kita, Asli Özyürek and Wendy Sandler for
their most helpful comments on an earlier draft. Alas, space limitations mean that only a fraction of these comments
are adequately addressed in the present paper. Stay tuned for “Invention and Community in the Emergence of
Language: A Perspective from New Sign Languages,” to appear in Foundations in Evolutionary Cognitive
Neuroscience, S.M. Platek & T. Shackelford, Eds., Cambridge, Cambridge University Press (to appear).
.
REFERENCES
Arbib, M. A. (2005). From Monkey-like Action Recognition to Human Language: An Evolutionary Framework for
Neurolinguistics (with commentaries and author's response). Behavioral and Brain Sciences, 28, 105-167.
Baker, M. (2001). The Atoms of Language: The Mind's Hidden Rules of Grammar. New York: Basic Books.
Byrne, R. W. (2003). Imitation as behavior parsing. Philosophical Transactions of the Royal Society of London (B),
358, 529-536.
Goldin-Meadow, S. (2003). The Resilience of Language: What Gesture Creation in Deaf Children Can Tell Us
About How All Children Learn Language. New York: Psychology Press.
Goldin-Meadow, S. (2005). Watching language grow. PNAS, 102, 2271-2272.
6
Hockett, C. F. (1987). Refurbishing Our Foundations; Elementary Linguistics from an Advanced Point of View.
Amsterdam and Philadelphia: John Benjamins.
Kegl, J. (1994). The Nicaraguan Sign Language project: an overview. Signpost 7(1), 40-46.
Polich, L. (2005). The Emergence of the Deaf Community in Nicaragua: "With Sign Language You Can Learn So
Much". Washington, DC: Gallaudet University Press.
Rizzolatti, G., & Arbib, M. A. (1998). Language within our grasp. Trends in Neuroscience, 21(5), 188-194.
Sandler, W., & Lillo-Martin, D. (2006). Sign Language and Linguistic Universals. Cambridge: Cambridge
University Press.
Sandler, W., Meir, I., Padden, C., & Aronoff, M. (2005). The emergence of grammar: systematic structure in a new
language. Proc Natl Acad Sci U S A, 102(7), 2661-2665.
Senghas, A., Kita, S., & Özyürek, A. (2004). Children Creating Core Properties of Language: Evidence from an
Emerging Sign Language in Nicaragua. Science, 305, 1779-1782.
Senghas, A., Özyürek, A., & Kita, K. (2005). Response to Comment on "Children Creating Core Properties of
Language: Evidence from an Emerging Sign Language in Nicaragua". Science, 309, 56.
Stokoe, W. C. (1960). Sign Language Structure: An outline of the visual communication systems of the American
deaf. Buffalo, NY: University of Buffalo.
Tomasello, M. (1999). The Human Adaptation For Culture. Annu. Rev. Anthropol., 28, 509-529.
Tomasello, M., Call, J., Warren, J., Frost, T., Carpenter, M., & Nagell, K. (1997). The ontogeny of chimpanzee
gestural signals. In S. Wilcox, King, B. & Steels, L. (Ed.), Evolution of Communication (pp. 224-259). Amsterdam/
Philadelphia: John Benjamins Publishing Company.
Wray, A. (1998). Protolanguage as a holistic system for social interaction. Language & Communication, 18, 47-67.
Recommended Readings
Arbib (2005), cited above, not only presents the Mirror System Hypothesis but also contains many Commentaries
arguing pro and con aspects of the hypothesis. My Response shows how these insights enrich the theory.
Bellugi, U., Poizner, H. & Klima, E.S. Language, modality and the brain (Trends Neurosci. 1989, 12: 380-388)
show that signed languages exhibit the complexities of linguistic organization found in all spoken languages, and
that the left cerebral hemisphere in humans is similarly specialized for signed as well as spoken languages.
Goldin-Meadow, S. (2003) The Resilience of Language: What Gesture Creation in Deaf Children Can Tell Us
About How All Children Learn Language (Psychology Press, 2003) uses the study of home sign to tease out
essential design features of language.
7
Download