Associate Professor of Anthropology
College of Staten Island, CUNY
RAGIR@ulster.net

Creole and signed languages inevitably appear when some critical number of language-ready children and/or adults engages in social activities over an extended period in schools, on plantations, on whaling ships, and in the African commercial centers for the slave-trade. New language-speakers utilize semantic content drawn from the local content and form of activities around them. Current descriptions of the growth of signed languages in deaf communities suggest that the emergence of new languages, both signed and spoken, is a function of group-cohesion (Kegl et al. 1998; McWhorter 1997). The emergence of a Creole from a plantation or trade Pidgin would appear to be quite different from the emergence of language in early human populations, where only a minority of individuals might have been "language-ready." However, the indigenous signed systems among the deaf in hearing-speaking populations are natural experiments in language genesis, and they provide important insights into the constraints on forming a body of linguistic practices before languages became ubiquitous in hominid populations.

If new languages can form in deaf-signing communities, then we might expect the same capacity to exist in hearing-speaking communities. The fact that grammars unfold and acquire formal marking as new members of a community learn a sign system suggests that self-organizing properties of aggregate activities may be at least as important for the evolution of language as the innate properties of mind (Batali 1998). Steele (1998) hypothesized that, as in other biological communities, cooperative interaction in hominid groups was self-organizing. Computer simulations confirmed that, goal-oriented communicative behavior increased in its shared content and complexity over time, and that hierarchically embedded levels of semantic structure emerged spontaneously. Both Steele (1998) and Batali (1998) suggested that an innate language-specific structure or acquisition devise was not necessary to account for the origins of syntax, rather syntax originated spontaneously – an embedded heirarchy of compositional levels – based on the pressure to convey more information with limited resources of time, memory and processing power.

Deacon (1997) argued that:

These self-organizing properties of language and mind suggest rapid episodes of language emergence in language-ready populations rather than the gradual selection for language.

Perhaps the best known historical example of an indigenous sign language community is that of Martha’s Vineyard, whose members contributed substantially to the formation of American Sign Language during the 19th century (Groce 1985). Because a congenital deafness which affected roughly 4% of the total population was concentrated in isolated upland farming settlements, every large upland family contained one or two deaf members. No good description of the sign language was made prior to its virtual disappearance at the end of the 19th century. However, it was used by both hearing and deaf inhabitants of Martha’s Vineyard for at least two hundred years, and the social context in which it developed and flourished has been carefully reconstructed. No economic, political, or social prejudices separated deaf from hearing citizens on Martha’s Vineyard; the majority of the deaf married, and the number of deaf-deaf marriages was substantially lower than in New England. Intermarriage among the islanders kept the proportion of deaf in the population high, and for over two hundred years, the three upland farming communities, each consisting of 100-300 hundred persons, included at least ten deaf and many hearing children all learning sign from a prior generation of native signers.

Unlike the 18th and 19th century community on Martha’s Vineyard, there were no indications of indigenous sign languages or congenital deafness in Kegl et al.’s (1998; Kegl & McWhorter 1997) survey of the northwest province of Nicaragua. Prior to 1980, when the Sandanista government established the first schools for the education of the deaf, most of the deaf children and adults were linguistically isolated within their own families and used home signs. The home signs were more discrete and conventional than mimetic gesture but tended not to combine multiple discrete signs in a relational way that resembled grammatical structures. Only a small enclave of four deaf adult siblings had developed what Kegl and Iwata (1989) described as a "signed pidgin." The shape and meaning of their signs were stable and combined in telegraphic phrases with little internal structure; the enclave was very isolated and the system static. Schaller and Sacks (Sacks 1999) documented a group of five deaf Mexican siblings who had developed a similar elaboration of home sign that they used to communicate with their Spanish-speaking father and each other. Even after two decades of use, the siblings had not developed a stable phrase structure, nor had they systematized the meanings of many of their gestures. These studies have demonstrated neither grammatical structure nor a stable lexicon will spontaneously emerge in the absence of the right kind of social setting, but that within an appropriate setting the input to individuals need not be a fully formed language to begin language genesis.

Based on her experience in Nicaragua, Kegl (1998) proposed that a critical number of about ten children was required to generate a language. I should like to add the hypothesis that a ‘language-ready’ community needs not only a critical mass but also historical continuity in order to generate a language. McWhorter (1997) documented the growth of semantic marking of grammatical categories in trade languages or "pidgins" in well-established historical trade systems, even though no children contributed to the formation of these "languages." His evidence directly contradicts Bickerton’s (1990, 1998) hypothesis that spontaneous grammatical structuring can only appear with the first generation of children to learn the pidgin as a native language. McWhorter’s careful historical documentation of the expansion of grammatical structures in Pidgins – in the centers for shipping slaves to the new world, on plantations or in mines using migrant male workers, and on whaling ships – gives us the best insight into the self-organizing properties of language from the perspective of a community rather than that of an individual.

Senghas (1995, 1997) described the development of grammatical structures in Nicaraguan Sign Language that emerged over a period of twenty years. After a few years, while the signed peer-group jargon of the school yard was still the major source of semantic and grammatical information, each new class learned to sign from the children who were already familiar with the school community’s systematic, but limited, ordering of semantic roles and conventionalized signs. Over a decade of school admissions, the language pool became lexically enriched; grammatical marking of verb arguments, spacial inflection, size and shape specifiers, and person agreement began to appear in young (under 6.6 years) and medium aged signers (6.7-10 years) who entered during the second decade of the school (Kegl et al. 1999:221). Year after year, the youngest signers streamlined gesture-shape and movement, and formalized emergent sets of grammatical markers, while the older children appeared to contribute much of the new lexicon (Senghas 1995). The most proficient signers were the school community’s youngest and newest members. The founding cohort and individuals older than ten when they first entered the school never attained the proficiency of these young signers.

A review of earlier studies of signed pidgins in Papua New Guinea (Kendon 1980abc), on Providence Island ( Washabaugh 1980ab, 1986), and in Guatamala (Shuman 1980ab) confirms the hypothesis that lexical and grammatical complexity increases only in an open language pool – with the entrance of new cohorts of signers. The single cohort signed pidgins never developed a grammar, although some of the practitioners attained stable and fluent gesture-shapes and movements. In all cases of limited pidgins used between hearing and deaf signers, the number of signers was relatively small, few if any new members entered the community, and the sign system disappeared after the death of the last deaf-signer. These group-constraints on the emergence of indigenous sign languages – the contextual variables that affect whether or not a community develops a formal system with grammar – have profound implications for various theories about language origins.

Rather than evolve, languages and grammars appear self-organizing within a community of language-ready individuals. I propose that there are at least two interacting levels: (1) the self-organization of the central and peripheral nervous systems during ontogeny in response to the way that the child participates in a community (Deacon 1997; Edelman 1987; Elman et al. 1996; Karmiloff-Smith 1992; Kauffman 1993; Oyama 1985), and (2) the self-organization of the exchange of information constrained by the medium and intentional content of the representational system (Bates & Goodman 1999; Bates & Elman 1996; Chomsky 1988; Hemelrijk 1999; Parrish & Edelstein-Keshet 1999; Steels 1998). The comparison between indigenous sign language communities in which the number of speakers and the lack of historical depth constrain semantic and grammatical complexity can give us insight into the formation of linguistic communities and the dynamics of language emergence in prehistory.

Of course the deaf are easily identified in a speaking community, which makes their association with other deaf individuals relatively straightforward. As far as we know, language-readiness in a population without language does not provide language-ready children or adults with identifiable characteristics, and a critical mass would have to be achieved more or less by chance. How are groups of 50 to 150 hominids – the most likely size of hominid hunter-gather bands (Dunbar 1996, 1998) – to secure the interaction of a critical mass of "language-ready" individuals necessary for generating standardized, peer-group jargons or pidgins that can become grammatically structured as new individuals entered into the pool of language-users. The limited linguistic systematicity found in closed-groups of signers that were smaller than the presumed "critical mass" (Kegle 1998; Shaller and Sacks 1999) suggests that open, dynamic systems of communicative practices are fundamental to the sign languages, and that a language-ready brain is likely to be essential for both pidgin and language genesis. If language can be generated only after a dynamic community of language-ready individuals exists, then the brain cannot evolve language capacity through the selection for individuals with grammatically complex speech or for the successive stages of grammatical complexity differentiating linguistic practices of groups (Batali 1998).

Three assumptions often underlie discussions of language origins: (1) that natural selection for fundamental cognitive functions and for language shaped the evolution of language; (2) that the syntax and semantics of the earliest languages differed from modern languages because of species-level differences in brain structure or organization; and (3) that language emerged once, like a species, and then diverged from this original primitive state. There are strong indications that these three assumptions are not true. The brain and languages appear to be self-organizing, and new languages grow in semantic and syntactic complexity so long as the interacting pool of speakers can attain a critical mass and remain open to new members either through births or adolescent dispersal. Although general encephalization and global increases in cortical connectivity occurred during human evolution, neurologists and psychologists have found no unique, language-specific cortical sub-structures to account for language syntax or semantics (Calvin 1996; Donald 1998; Jerison 1995; Edelman 1987). Furthermore, since deaf communities lacking linguistic models have repeatedly organized new sign languages, there is reason to believe that large-brained hominid communities could have generated vocal languages independent of each other.

Finally, studies suggest that language formation takes place over a historical rather an evolutionary time span, and that ontogenetic development, rather than phylogenetic selection, is the central formative process affecting the neurological specialization underlying language. A common criticism of language origin hypotheses has been that they are too speculative – that there is no way of disproving a hypothesis by observing the proposed processes of language formation in nature, in the laboratory, and in archaeological and fossils remains. However, excellent descriptions of communication systems among linguistically isolated deaf individuals in speaking communities exist. When we understand why signed languages emerge spontaneously over historical time in some communities with deaf individuals and not in others, then we can begin to understand the constraints on language genesis in hominid evolution.

 Conference site: http://www.infres.enst.fr/confs/evolang/