Toward Certainty, Objectivity, or Testability?

Abstract

It would seem a step designed to try the reader’s patience for me now to comment that the classification of retinal ganglion cells is not a straightforward, objective task, but is as much a product of the observer’s presuppositions as any other scientific proposition; and is likely, therefore, to provide its share of grist for the philosopher’s mill. Yet it is a fair and relevant criticism of neurobiologists (including myself) that we have been largely unaware of the problems of methodology inherent in classification, and of the substantial literature that exists on those problems. There has been little awareness, either, of the central part played by classification (whether of nerve cells, plants, animals, aphasias, or rocks) in the organization of bodies of knowledge.* Perhaps as a consequence, much of the variety and inconsistency found among neuronal classifications stem from differences between scientists in our presuppositions; differences that, though less exotic than the gulf that intrigued Foucault, are nevertheless substantial and highly influential on our work.