This article deals with the definitional legacy of the end of ideology debate and the problem of defming ideology. The definitional legacy of the debate is described in terms of assumptions about the characteristics and functions of ideology, and several problematic features are identified. A major argument is that certain theoretical difiiculties characterizing the debate were the result of viewing various attributes as definitional properties of ideology rather than hypothetical variables. In an effort to overcome the problematic features of the defmitional legacy ofthe debate, the article recommends the construction of a minimal defmition of ideology which converts characteristics and Functions oAen viewed as a priori elements of ideology into an extensive battery of hypothetical variables.The end of ideology debate ranks as one of the major controversies in the social sciences during the postwar decades, and the debate was waged on truly a grand scale. It stretched over a period of more than fifteen years, from the mid-1950's into the 1970's. To an unprecedented degree the debate engaged scholars on both sides of the Atlantic. Thousands of pages were written in several languages, and anthologies of contributions to the debate stand as monuments of the con&* versy (Waxman 1968, Cox 1969, Allardt & Rokkan 1970.In surveying the prolific writings of the debate, however, one is faced with the difficult question as to the importance of the debate with respect to theorizing about ideology. The purpose of t h i s article is to deal with this question and examine several of the theoretical underpinnings of the end of ideology debate. This is done by focusing on the problem of defining ideology. This problem constitutes an appropriate focus because the influence ofmany ofthe definitional assump tions of the debate stillpervade thinking on ideology and much of the everyday usage of the tenn. T h i s article is a revised version of a paper presented at the workshop 'Ideology after the End of Ideology Debate: Recent Theorizing and Empirical Applications', the ECPR Joint Sessions of Workshops, Freiburg. 1983. I would like to thank the workshop participants for their helpful criticisms and comments. In additi0n.Elia.s Berg, University ofStockholm, Barbara G. Haskel. McGill University, and Evert Vedung, University of Uppsala, have offered valuable suggestions.
111