It has long seemed to me desirable to have a procedure for identifying and recording the different kinds of fossils simply by strings of objective descriptors, without imposing any taxonomic system. This avoids the necessity to evaluate degrees of relationship or similarity, and the strictures of a nomenclatural code, and uncertainty as to the extent of variability admitted by various workers into a named taxon. Computers provide the means by which long, unwieldy strings of descriptors can be handled as easily as are the binomens of a taxonomic system.
An opportunity to put this conviction into practice came when it was found necessary to determine ages in cores of deep-sea clays lacking calcareous and siliceous microfossils, but containing ichthyoliths - microscopic teeth and skeletal debris of fishes (Doyle et al., 1974 ). It was not necessary for our purposes to determine which fishes produced the ichthyoliths, but only to be able to recognize assemblages of forms that are similar in sediments of similar age.
It proved quite feasible to identify ichthyoliths by coded descriptors - strings of alphanumeric codes that described features of ichthyoliths as seen in transmitted or reflected (Tway, 1984 ) light. The strings of code are often long and ungainly, but record character-states and their variability objectively, and are readily handled by computer. A summary description of the way these coded descriptors work is provided by Doyle and Riedel (1985).