The session deals with the evolution of language. Might historical linguists have something to say on this score? Linguists have been treating language scientifically for years. A few are now trying to treat it as a selection process in particular.

Organized by: David Hull

William Croft, University of Manchester
"The Darwinization of Linguistics"

The field of linguistics was essentially historical in the 19th century and, in fact, provided an inspiration for Darwin in developing his theory of evolution. However, 20th century linguistics has not kept up with 20th century evolutionary theory. Instead, the century has been dominated by the structuralist movement, which focuses exclusively on the morphology of language and retains an essentialist approach to the analysis of language. In the latter half of the century, a movement arose that took a populational approach to language -- variational sociolinguistics. This movement focused its attention on selection (propagation) but did not deal with mutation (innovation) and all but ignored subsequent developments in biology. The goal of my paper is to redress these omissions.

Thorbjorn Knudsen, University of Southern Denmark
"Why Tacit Knowledge Protects the Evolutionary Potential of Human Language"

The present article introduces the theory of cultural evolution as a possible basis for treating language as a selection process. Cultural evolution is Lamarckian and involves the social transmission of explicit concepts by communication in processes involving choice, imposition, imitation and so forth. A possible complementary Darwinian principle operating in the social realm is defined in terms of social transmission of the tacit knowledge that underlies explicit concepts. According to this principle, termed Local Emulative Selection, some forms of tacit knowledge are not adapted (those which cannot be reached by consciousness) by their carrier. I then identify a problem of establishing a base-line for adaptation that plagues any form of Lamarckian selection. This base-line problem implies that the evolutionary potential decreases as the possibility of adaptation increases. In consequence, the social transmission of tacit knowledge, which cannot be reached by consciousness, protects the evolutionary potential associated with any form of social evolution. According to this argument, tacit knowledge helps protect the evolutionary potential of human language.

Gregory Radick, Leeds University
"Darwin on Language and Selection"

How did Darwin himself view the relationship between language and selection? Concentrating on his accounts in The Descent of Man (1871, 1874) and The Expression of the Emotions in Man and Animals (1872), I shall look in particular at two aspects of the language-selection relationship: first, the role of selection in shaping the power of language, broadly construed, in humans and animals; second, the extent to which changes within particular languages were due to a selection-like process. In both cases, it turns out, Darwin attributed rather little to selection, for reasons that point back to a number of often overlooked features of his own explanatory and theoretical commitments.

Salikoko Mufwene, University of Chicago
"Competition and Selection in Language Evolution"

The approach I propose in my book The Ecology of Language Evolution originates in my research on the development of creole vernaculars, perhaps the quintessence of language evolution under contact conditions. The target language, usually identified as the lexifier, has been restructured so differently during its appropriation by non-native speakers that the outcome is considered a separate language. My research underscores structural diversity within the lexifier itself (like variation within a species that evolves into a new state) and more diversity in the contact feature pool that also included features from substrate languages, those vernaculars previously spoken by those appropriating the lexifier. During the development of creoles every speaker's mind constituted an arena where various features competed for the same function. Much of the relevant research should focus on identifying principles that regulate how features are selected into the emergent vernaculars. The research should cover both those cases in which the competition is resolved by the selection of one alternative and those where it fails to be resolved. (The choice of a particular language as the target/lexifier is itself among the competition-and-selection phenomena to explain.) Invoking the traditional markedness model in linguistics, I have argued that it works best when augmented with an ecological dimension that articulates the ecology of language evolution. This ecology determines the markedness values of competing alternatives for particular functions and can help us identify those factors that regulate selection by determining markedness values. The feature pool notion is evidently patterned on that of gene pool in biology, except that linguistics already provides a paradigm that can articulate some details of the selection process. I also argue that it is critical to analogize languages to biological species and it is from epidemiology that studies of language evolution can derive the most adequate inspirations.

Zeynep Tufecki, University of Texas at Austin
"From Linguistics to Evolutionary Psychology: Methodological and Ontological Arguments Against Extrapolation"

The fundamental question posed by evolutionary psychology today is whether the human mind consists mainly of modular adaptations resulting from selective forces, or whether it is more like a general-purpose machine with a capacity for varied context-dependent actions. For evolutionary psychology, contemporary linguistics provides the hardest evidence. This paper examines the evidentiary status of Chomskian linguistics and, more importantly, whether it provides support for the stronger claims of evolutionary psychology.
Chomskian linguistics makes two striking claims: (i) there are universal features of languages; and (ii) linguistic ability is universally acquired by children with a poverty of stimulus, with no formal training and negligible negative feedback. However, at the very most, universality is limited to some very general linguistic features. Moreover, it is at best a necessary but not a sufficient condition for modularity or innateness: it neither requires nor guarantees domain specificity . The poverty-of-stimulus argument, which is supposed to provide the basis for linguistic innateness (along with universality) has itself been questioned on empirical grounds.
The second claim examined is whether there is remotely enough similarity between linguistic and other mental processes to make similar claims about innateness and adaptiveness for them. Humans grow up infused in formal and informal social training with ubiquitous feedback. Thus, there is no analog of the poverty-of-stimulus argument for these mental/social features. Furthermore, there is ample empirical evidence to question the existence of non-trivial universal phenomena in these categories. Moreover, establishing some similarity across disparate cultures without explicit randomization does not remotely show universality by descent: using the terminology of evolutionary biology, it establishes analogy, not homology.

David Hull, Northwestern University
"Does Selection Add Anything to Linguistics?"

Linguists worked out a theory of language change before Darwin published his Origin of Species. In fact, Darwin used the work in historical linguistics to support his own theory of evolution. But from then until quite recently, linguists and evolutionary biologists have gone their separate ways. However, in spite of so little interdisciplinary work, scholars in the two fields have developed very similar systems with the same strengths and weaknesses. For example, both invented cladograms. In my paper I will explore similarities and differences between the two disciplines. In particular, selection plays a major role in biological evolution. Does it play any role at all in language change?