Terror management theory (TMT) is a very impressive achievement. It is rooted in plausible biological and psychological assumptions and serves a range of integrative functions, both vertically in terms of resurrecting Freud, Rank and Maslow, and horizontally in terms of integrating self-esteem, just-world, cognitive dissonance, social identity and impression management theory and research. Further, at level of parsimony and integration it really cannot be faulted; as a descriptive framework it buys a great deal with a relatively small number of assumptions. For example, it is able to keep essence of cognitive dissonance-the importance of consistency motive as well as finding a place for self-esteem and self-presentational motives by relating all of them to the more basic goal of maintaining integrity of cultural anxiety buffer. The experimental research establishing specificity of salience of mortality concerns relative to other aversive experiences or failures is also impressive. Given all of these positive features of integrative power, parsimony specificity of mortality salience effects, and unique predictions, what are limitations of terror management model? At a metatheoretical level, perhaps basic issue is need for a need-type formulation. At heart, all need-type models involve a type of tension-reduction dynamic. In this case we seek to avoid facing our mortality. In effect we have a psychological version of second law of thermodynamics-the entropy principle. Everything eventually comes down to avoiding terror that recognizing death would bring. I find this somewhat ironic, because terrorists all over world seem to seek death readily to destroy evil nonbelievers. Indeed, continued existence of outgroup seems a much greater threat than their own death as they strap explosives to their bodies for a higher religious cause. Similarly, mass media describes events everyday in which people risk their lives daily for petty reasons, in which guns are pulled at slightest provocations. Indeed, I see a world where for many people preserving life-their own or other peoples'-has no meaning. The present theory would seem obliged to account for such behaviors. Let me leave pop sociology, however, and point to scholarly critiques regarding motive-type explanations, Perhaps most famous is that of British philosopher Peters (1958), who systematically reinterprets Freudian type-motivational assumptions from a rule-based model. For example, there is nothing motivational about assuming that primary and secondary processes involve different rules of information processing. In this context, we invoke motivational concepts when rule following breaks down. Moreover, although one would not know it from Pyszczynski, Greenberg, and Solomon's description of social identity theory, current model as elaborated by Turner (1987) is a nonmotivational formulation fueled by metacontrast and assumptions about role of shared social categories. Within context of this model, for example, social influence is interpreted in terms of attempts to maximize ingroup similarity, with members attempting to resemble prototypic ingroup position. Moreover, Turner derived from this model a condition where social identity increases even when one's ingroup fails; thus at issue is no simple self-enhancement through ingroup success. At level of specific claims, I am somewhat uncomfortable with authors' interpretation of Swann's self-verification effects as support for dissonance theory. For example, as elaborated by Brehm and Cohen (1962), dissonance is more than consistency seeking. It involves adding parameters of choice and commitment to any attitude-discrepant position, which is not true of Swann paradigm. Indeed, in Swann's view an information-processing-overload avoidance explanation is favored over a motivational interpretation. At a more basic level, I have following kinds of concerns. The concept of a cultural worldview anxiety buffer strikes me as rather vague conceptually-is this a schema of values, beliefs, attitudes? Can it be measured, manipulated, primed? Can this assumption be