Bryant targets a view that she calls ‘objectivism’, or ‘metaphysical realism’, which she attributes to Kripke, and to Putnam during his scientific materialist phase in the 1970s. She accepts Hartry Field’s three-point summary of the position. Objectivists hold that the world consists of some fixed set of mind-independent objects, that there can only be one true and complete description of the way this world is, and that truth involves some sort of correspondence between our signs and the things and classes of things which exist in nature (38). As she sees it, the objectivist considers scientific research to be a kind of applied special metaphysics aimed at aligning our referring terms and terms for classes, relations and functions with the objects, classes relations and functions that constitute the order in the world. Once the alignment is complete, the truths available in this perfect world-representation will include de re necessary truths, recognized empirically, and revelatory of the essence of things. Natural kinds will be distinguished by distinct essences that determine distinct behaviours, and the truths about the natural world will be accessible once the single complete set of those essences is specified.

If this position sounds Leibnizian, Bryant’s view is best described as a pluralistic version of Locke. She notes that Locke believed that the real essences of things determined their behaviour, but that investigators could at most construct a nominal essence for them (20). Medin and Ortony, two psychologists of categorization whose work Bryant admires, maintain that when establishing categories we take certain properties, salient to our explanatory purposes, as essential to the members of the category (62-3). This tendency relies on a common belief (heuristically valuable, but unjustifiable) that things have essences that make them what they are. Other characteristics of things are considered accidental, or as common enough and accessible enough to permit an initial categorization of an individual, but as inessential. The properties that are taken to be essential to a kind are selected because they are useful in explaining some of the regularities marking the behaviour of members of that kind.

This view appears congenial to objectivism. However, Bryant agrees with J. V. Canfield that any necessity adhering to the link between essential properties and the kind with which they are associated is purely de dicto, an effect of the linguist activity undertaken (92-7). Just as the same object can play different roles in different activities, intended for different purposes, so a category term can change its role and its necessary connections to other terms as the use to which it is to be put changes. We can detect a commonality of meaning through some set of these changes whenever there is a stability of meaning through some set of these changes whenever there is a stability of content in the terms employed. A sufficient invariance in links between a term and its associates will establish a level of conceptual stability for it.

Connecting the necessity involved in the relation between natural kind terms and their essential properties to the linguistic and cognitive activities of the users opens up the possibility that there is more than one way to classify a field of study into natural kinds. Bryant maintains that, while there is a world independent of theorizers about it, how that world is organized depends on the cognitive purposes of the theorizers (a matter of epistemology) as well as on the characteristics of the world (a matter of ‘metaphysics’). Different cognitive aims do, and should lead to different categorizations of nature.

These conclusions fit a number of phenomena well: coexisting incompatible systems of classification in the life sciences, distinct, incompatible theories for distinct purposes in the physical sciences, continuity in meaning despite theoretical change in the history of the sciences. It appeals to anyone attracted to ontological pluralism. Still they are not quite full Putnamian internal realism. Putnam appears to think that the world has no structure apart from human construction. The best theory of the world possible will still be a human creation that has succeeded by human standards of verification.

On the other hand, Bryant apparently holds that the structure(s) of the world depend on human selection. Objects in some field of study are classified into natural kinds, on the basis of properties taken to be essential to membership in that kind, but the kinds and their connections to essential properties are determined partly by the theoretical aims of the classifiers. Kinds and essences, therefore, depend on what could be loosely called an epistemological frame, it is not the one under development by the classifier; they are a part of the material employed in that development. While kinds and essences are parts of an epistemological frame, the properties definitive of them are prior to it.

If there is a basic level of classification, there must be frame-independent properties employed to get characterization under way. Bryant seems to concede this, claiming commitment to a ‘mind-independent world populated by real entities and real properties’ (17). It is classification of these objects that requires a view point. The extreme complexity of the world demands it: no single way of organizing information about any domain could describe its complexity, or account for the regularities to be found in it (111-16). This leaves an undiscussed metaphysical puzzle. How could the world manifest regularities which, taken together, are too complex in principle to admit a single unified account?

-Thomas Mathien, University of Toronto

TO REQUEST A CATALOGUE:
M-GH2-01
285 Madison Avenue
Madison, New Jersey 07940
Phone (973) 443-8564
Fax (973) 443-8364
fdupress@fdu.edu