Department of General Linguistics, University of Stellenbosch,
Stellenbosch, South Africa
rpb@maties.sun.ac.za

Theories of the evolution of human language express by their very nature claims of a historical sort: claims about why, when, where or how language emerged and/or developed in some distant past. An essential feature of these claims is that they are made in the absence of sufficient historical evidence about the evolutionary events, biological processes, physical forces, environmental pressures, kinds of (pre)linguistic entities and so on involved in the evolution of language. The paucity of this historical evidence – i.e., evidence derived from data contained in natural or man-made records of these evolutionary events, etc. – is generally seen as one of the most formidable obstacles to serious work on language evolution.

In modern work on the evolution of language, various strategies have been adopted for remedying the problem of evidential paucity and, in Pinker and Bloom's (1990, p. 727) words, for countering the "[s]kepticism about the possibility of saying anything of scientific value about language evolution". A number of these strategies involve the use of new sources of data that are believed to bear on the truth of claims about language evolution. A first source comprises data of a historical sort, including paleoneurological data. For instance, in work such as that by Wilkins and Wakefield (1995), data about impressions on the interior surface of fossil skulls – or corresponding bumps or ridges on the exterior surface of endocasts – are brought to bear on claims about the emergence of (certain "preconditions" for) the language faculty. A second source comprises data of a nonhistorical sort. For example, Pinker and Bloom (1990) attempt to bring, amongst others, data about the reliance on lexical association by right-handed people with a family history of left-handedness to bear on the truth of claims expressed in their selectionist account of the evolution of the language faculty.

Both historical data of the sort used by Wilkins and Wakefield and nonhistorical data of the kind drawn on by Pinker and Bloom are data about entities belonging to ontological domains that are distinct from the ontological domain to which the language faculty belongs. As a cognitive entity, an ancestral language faculty is separated by a distinct ontological gap from both fossil skulls and the processing behaviour of people living at present. Being the different kind of thing they are, fossilized (fragments of) skulls simply do not contain direct information about cognitive faculties. Similarly, the processing behaviour of presently living right-handers with a family history of left-handedness does not offer in any direct way information about the evolution of a mental faculty that is claimed to have come into existence in a distant past.

To be able to bring data about fossil skulls or data about the processing behaviour of right-handers with a family history of left-handedness to bear on the truth of claims about the evolution of the language faculty, the ontological gap in question has to be bridged in a proper way. If this can not be done, data about the former entities can not be relevant to the truth of the latter claims. What are in essence required, are warrants licensing the various inferential jumps from paleontological or behavioural data to the truth of claims about the evolution of the language faculty. This gives rise to the question of what the source of the required inference licences might be.

Developing ideas outlined in (Botha, 1998, 1999), the present paper identifies bridge theories as such a source of inference licences. A bridge theory is characterized as a structured set of empirical assumptions which systematically interrelate properties of entities of one ontological domain with properties of entities of another ontological domain. To elucidate and appraise the role of bridge theories in modern work on language evolution, the paper focuses on the way in which data about fossil skulls or endocasts have been brought to bear on the truth of claims about the evolutionary emergence of language as a mental faculty. The peer discussion in Behavioral and Brain Sciences (Vol. 18, 1995) of Wilkins and Wakefield's target article on the emergence of the neural preconditions for language is used for illustrating the following points:

1. The constituent assumptions of the required bridge theories are often not stated clearly or justified explicitly.
2. The various licensing assumptions are not always "parcelled out" into different bridge theories but are lumped together into a single holdall theory.
3. Some of the required bridge theories – or their constituent assumptions – are often quite controversial.

If the problem of evidential paucity is to be solved, the paper concludes, it is imperative that flaws such as (1) - (3) are eliminated from bridge theories.

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