Abstract :
The commentator observes that the classification used in the above article (ibid., vol.23, no.3, p.64-74, Mar. 1990) is arbitrary and that truly significant properties of computer architectures, like topologies and communication methods, are missing entirely. The author responds that to the extent a taxonomic scheme represents a particular viewpoint, all such schemes have an element of arbitrariness. He also argues that the overall purpose of a taxonomic scheme is to provide a unified conceptual framework by means of which objects can be classified in a succinct and elegant manner. The scheme may not be able to address all aspects of the objects of interest. Rather than extend it arbitrarily, inevitably making it cumbersome, one may retain it intact and construct independent schemes for other aspects.<>
