The evaluation of 'deviant articulation' (that of young children, speech defective persons, aphasics, second-language learners) has usually consisted of two activities: transcription of the speech being evaluated, and comparison of the transcription against some 'standard' set of 'target' sounds. Any transcription is a description of a speaker's articulation in terms of the auditory and perceptual capacities of a transcriber. This paper presents a method for the transcription and quantitative analysis of deviant articulation. The method is based on (1) analphabetic transcription using binary articulatory categories , and (2) analysis of the transcripion in terms of the perceptual performance of the transcriber as measured by the Theory of Signal Detectability (Tanner and Birdsall, 1958). This paper comprises a report in "Development of Language Functions, A Research Program-Project (Study C: The Development of Speech and Sound Specificity in Children)." (DO)