In various tones, in divers contexts, to one purpose or another, with greater or less frequency according to the times, we have heard bruited-about the concept of class interest. Or, at all events, we hear the term. But what is the concept? What does the term mean? From the confidence with which the words are enunciated, there might seem to be no reasonable doubt. Class interestsare the interests of a class. Perhaps so, but I fear that tautology does not enlighten us. Self-evidence, I fear, may prove a cloak for unintelligence and unintelligibility. And yet, so far as I am aware, no one has seriously and fundamentally attempted to show what the concept of class interest is or, more precisely, what the term does or should or even might signify, much less to examine the adequacy of the numerous possible interpretations.' This, therefore, is the task that I here set myself. I shall try my best to discover a tenable interpretation; what the attempt will come to in the end, we shall see at the end. It is the Marxists, of course, or those under their close influence, who employ the term most frequently and insistently. But others employ it, too, either in self-explication or in social analysis; and we are likely to find it uttered in some expedient place by almost anyone who does not consistently deny the reality of socioeconomic classes. The use of the term is, therefore, familiar to us at least in a vague way, but in a way sufficiently definite, I believe, for a pretty general agreement on the field of application. And to this familiarity, to determine whether some particular notion is an interpretation of the term, I shall often make an implicit appeal. But if there is any difference between the Marxist and the non-Marxist applications, I shall take the Marxist one as decisive, for the Marxist application possesses the weightiest claim, both practically and theoretically, to the first importance. Among the users of the term there is no disagreement-or in logical consistency could be none-on the basic presumption that socioeconomic classes do exist. And this presumption I shall not call into question; it provides the immediate context of the inquiry, and to question it would be to render the inquiry pointless from the start. Naturally, if the concept of a class were invalidated in a relevant respect, the concept of a class interest would be invalidated too. However, the concept of a class presents itself a very difficult and thus far unresolved problem;2 and since it might prove impossible to discover a tenable notion of class interest even if the meaning of the other term were settled, I think I am justified in examining the subsidiary problem on its own merits. It does seem that different interpretations of class interests would probably best cohere in some sense or in various senses with different interpretations of class itself. Still, our concern lies, not (except incidentally) with the notion of class, but rather with the notion of interests as attributed to the presupposed classes. I choose here the word "attributed"; I might have chosen "imputed." And