On Certain Practical Inadequacies of Distinctive Feature Systems

Abstract

Existing distinctive feature systems have several practical shortcomings. Although the deficiencies do not necessarily have important consequences for theoretical linguistics, they constitute a considerable obstacle to diagnosing and treating speech disorders. Nevertheless, distinctive features are being used and advocated increasingly in speech research and have even been suggested as a metric for measuring intelligibility distortion. A tentative set of language-specific features based on articulation is suggested as a replacement for universalist features, which are based partially on marginal or contrived acoustic properties and are motivated by a concern for optimum notational economy.