Abstract
Neologistic jargon aphasia has resisted clear and unambiguous treatment since the syndrome has been recognized. The puzzles that remain in the analysis of this syndrome are due in large measure to the elusive native of the neologism itself. Where does it come from? Or better yet, where could it come from? The present paper attempts to demonstrate that several sources for the production of neologisms are plausible, but focuses upon two possibilities only. The first possible source is some sort of random generator, which is capable of producing syllable strings to serve as surrogate lexical items when the patient cannot access target lexical representations. The second possible source is the mechanism of phonemic paraphasia, which operates on an accessed lexical representation in such a way that it transforms its segmental structure variously. Two types of neologism are distinguished: the ‘target-related’ neologism, where the target has not been rendered unrecognizable, and the ‘abstruse’ neologism, where the target is not recognizable. The target-related neologism is rather uncontroversial; it stems from phonemic paraphasia. The abstruse neologism, on the other hand, is the truly recalcitrant item, since its source is much more recondite. The paper explores some possible reasons for why we might need a random generator in normality, under the assumption that new mechanisms are not created by brain lesions. In addition, the processes of phonemic paraphasia are explored in terms of syllabic structure constraints, and several principles are suggested to account for the nature of certain kinds of phonemic paraphasias–specifically the doublet-creating error.