Weingarten and Chisholm (2009) are to be congratulated on a fascinating and stimulating article. However, it suffers from a lack of any real definition of “religion.” This might not matter, except for three things. First, most (if not all) of the evolution-of-religion literature that Weingarten and Chisholm are addressing shares this lack of definition. Second, “religion” is a Western concept, and most societies worldwide (especially the small-scale, traditional societies) do not have an equivalent word or concept. Third, there are major problems with most of the working definitions that seem to be implied in the literature. If religion is “belief in supernaturals,” we have not only the well-known problem of belief in Santa Claus, ghosts, Mickey Mouse, and other dubious divinities (see Atran 2002; Winkelman and Baker 2008) but also the problem that social scientists believe in and routinely attribute agency to unreal or undemonstrable entities such as neoliberalism, the elite class, the modernist program, the colonial mentality, and, for that matter, Society itself. Is social science a religion? Conversely, most societies worship the spirits or life force of perfectly real things: mountains, trees, bears. We may not agree that a bear spirit does all the things that the worshipers say it does, but we cannot deny that bears are alive and real. What, then, is a “supernatural”? Is it anything more than something you believe in and I don’t? If religion is defined as deeply held but unprovable foundational beliefs, we have a similar problem. How do we separate religion from communism, microeconomic theory, human rights, or moral philosophy? In fact, atheism would be a religion, being a thoroughly unprovable set of beliefs about the divine. If religion is intense personal experience of mysticism, awe, reverence, and the like, we have to deal with the singular lack of such exalted emotion in the average congregation member; much religious practice is merely boring, as much to the devout as to anthropological observers. William James (1903) long ago pointed out that even within very small homogeneous groups there are enormous variations in “religious” emotionality, and Radin (1957) confirmed this for small-scale traditional societies. We are left with Durkheim’s (1995 [1912]) definition of religion as a collective representation of community that holds a group together through engaging emotion and then motivates people to follow their social codes. This, again, could apply to communism, or high school spirit, or political partisanship, or nationalist fervor (cf. Anderson 1991). So can other definitions of religion as social practice or social bonding. Different experts highlight different aspects of religion as foundational and thus wind up talking past each other. It seems that the only way to define religion is to say it is an occasionally institutionalized set of social conventions that unite all the above in various combinations. If religion is indeed that complex, looking for “the God gene” (Hamer 2005), or any simple evolutionary explanation, is hopeless. We have to explain how each component evolved and then how they have come to be assembled in the various ways ethnographers have documented. In the thousands of smallscale traditional societies that lack a concept of religion or any differentiation of “supernatural” from “natural” beings, we will have to consider all these various things separately and make no assumptions about their being united in one institution set over against other cultural institutions. I am duly convinced by Weingarten and Chisholm (2009) that attachment theory is one place to begin understanding religion. But we need a great deal more in order to explain the origins of awe, or fear of ghost poison, or mysticism, or the Storm God.