A discussion of pre-historic (i.e., preliterate) language looks at the processes of affixation and inflection in the context of two conflicting theories on the complexity of those languages. The traditional view holds that the grammar used by early Indo-Europeans was at least as complex and abstract as that of any modern educated and literate speaker. The other perspective is that the early peoples are unlikely to have been very different in language and thought from present-day non-literate speakers and very different from contemporary literate language users. The two theories are outlined and discussed. It is concluded that given the absence of direct evidence for pre-historical stages and changes in Indo-European languages, the conflict will have to be resolved through: (1) an appraisal of the validity of the traditional technique of historical linguistics; (2) appraisal of the validity of the studies and descriptions of non-literate users of contemporary languages; and (3) evaluation of the latter theory's assertion that characterizations of contemporary individuals and societies are proper data for the understanding of early man. (MSE)