THE IDEA THAT iconic visible gesture had something to do with the origin of language, particularly speech, is a frequent element in speculation about this phenomenon and appears early in its history. For example, Socrates hypothesizes about the origins of Greek words in Plato's satirical dialogue, Cmtylus. Socrates' speculation includes a possible role for sound-based iconicity, as well as for the kinds of visual gestures that deaf people employ. Plato's use of satire to broach this topic also points to the fine line between the sublime and the ridiculous, which has continued to be a hallmark of this sort of speculation. In this article I present recent evidence that supports the idea that language first arose as visible gesture. This evidence is culled from several lines of research, including studies on the neurological underpinnings of gesture (e.g., mirror neurons; gestural communication among African apes; the cognitive basis of signed languages; and the emergence of new signed languages).
Two common themes running through gestural theories of the origin of language are that iconic visible gesture is more natural than speech as a communication device and that it solves the problem of accounting for the origin of completely arbitrary signs. How did English speakers (actually Latin speakers and perhaps even speakers of earlier Indo-European languages) come to agree on "rose" as the name of a flower that would smell just as sweet if it were known by another name? As mention of the word's Latin lineage suggests, we can to some extent trace the histories of individual words, but where did the first spoken words come from, and how did these apparently completely arbitrary symbols come to be associated with their referents? Perhaps at some stage in the evolution of language signs were simply iconic and pantomimic illustrations of the things they referred to. One could then imagine a stage during which incidental sounds, especially those that were also iconic or onomatopoeic, came to be associated in a gestural complex with the visible sign and its referent. Subsequent to this, the visible sign could either wither away or come to be used as a visual adjunct to the now predominant spoken word.
Other elements have also entered the basic argument from findings that have accumulated in paleoanthropology, primatology, neurology, and linguistics, including the following: the apparently late appearance of a fully modern vocal tract and the early appearance of fully modern upper extremities, the seemingly greater facility for languagelike behavior by apes in the gestural as opposed to the vocal medium, the discovery of gesturally responsive mirror neurons in the Broca's area homologue in monkeys, the working out of the linguistic structures of signed languages of deaf people, and, especially in this regard, Stokoe's notion of semantic phonology (Armstrong, Stokoe, and Wilcox 1995). This article introduces these elements of an evolving theory and places them in historical context.
The history of this recorded speculation has been treated exhaustively by Hewes (1973, 1976, 1996), and it is not my purpose to review it in detail-interested readers may review these works by Hewes and my own discussions (Armstrong 1999; Armstrong and Wilcox 2002). Instead, this article assesses the current scientific status ot the various aspects of the theory that visible gesture was the earliest manifestation of the human capacity for language and that speech evolved subsequently from an original visible/gestural communication system. Our real historical journey begins with the speculations of the French Enlightenment philosophes of the latter half of the eighteenth century, such as Condillac, Diderot, and Rousseau. Some of these writers were aware of the work of teachers of deaf children, such as Pereire, and there appears to have been dialogue between the philosophes and the Abbé de l'Epée, who opened the first school for deaf youth in France in the late eighteenth century, as the theories of deaf education and the origin of language evolved in tandem (Lang2OO3; Rosenfeld 2001). …