The problem of syntax has people on all sides of the issue “desperately” hoping a solution will somehow appear, University of Maryland syntactician Juan Uriagereka told an Evolang-conference audience in Barcelona this afternoon. Most accounts of language origins simply ignore the rise of syntax (of the formal, Chomskyan sort), while those that glance at syntax still do not offer any hint as to how humans evolved the powerful system we enjoy.
Speaking formally, as long ago as 1962 Noam Chomsky proposed a hierarchy of machines that can be used to generate strings of symbols. The strings become more complex and variable as you move up the hierarchy. From a syntactician’s (Dr. Uriagereka preferred the term “professional”) point of view the evolution of language could be understood as an evolution up this hierarchy to the point (fairly high up the ladder) where natural languages are possible, but inquiries into language origins do not track this process.
Professionals would be more sympathetic to functional accounts of language evolution if research showed in detail how the syntax simply follows (or emerges) from the development they do trace, but so far no such demonstration has been offered. In the meanwhile the expressed hope that the syntactic structure is simply “there” sounds more desperate than proven.
Dr. Uriagereka spoke more favorably of the idea that a narrow language faculty limited to Homo sapiens was built on top of a broad language faculty that was already found in many other animals, most notably in our nearest primate relatives. Implicit in this picture of broad and narrow faculties is the idea that the lower parts of Chomsky’s hierarchy are found in the broad faculty and that the jump to the narrow faculty involved a transition to the natural language generator. The problem, of course, is that the jump is hard to account for through natural selection. This difficulty might be explained as the result of a virus that got tangled in human-lineage DNA and spread like an epidemic. Such evolution has happened in some other systems (e.g., adaptive immunity in cartilaginous fish), but Dr. Uriagereka admitted that this account too is likely to strike some as another desperate stab. (He did not address the detail that some such improbable intervention might be necessary at every step up the hierarchical pathway.)
From a linguist’s perspective we need to understand how people can “squeeze” multi-dimensional “syntactic thoughts” into words and phrases, and back again. A promising approach, Dr. Uriagereka said, comes from the investigation of ‘procedural memory,’ which oversees rule-governed behavior. Procedural memory is important for things like rhythmic activities and knot tying as well as speech, although Dr. Uriagereka suspects that these capacities depend on an ability that evolved first for speech. He noted that the FOXP2 gene appears to be expressed in procedural memory.
Comments