In thinking about the results reported here from the 2000/01 NJPS, it is crit ical to have a clear-eyed view of how the social landscape in which American Jews find themselves has changed in the post-World War II period. Some of the papers (Phillips and Fishman; Phillips and Keiner) argue that a soft version of rational choice can help to shed light on such key indicators of the contempo rary Jewish situation as intermarriage, and I agree that it can. But a choice-based framework is useful only insofar as one has a firm conceptual grip on the range of options faced by minority individuals, along with their perceived benefits, risks, and costs. If that configuration has changed substantially over time, past research may not give useful guidance, indeed may be fundamentally incom mensurate with the findings from new data. The applicability of past models is, quite appropriately, called into question by some of the papers, most notably, Phillips and Keiner. The concept of a social boundary provides a useful way of specifying some of the major changes that have taken place in the last six decades or so, as well as of identifying the value of the American Jewish case for comparative studies. By a boundary, I mean a social distinction that individuals make in their everyday lives and that shapes their actions and mental orientations towards others; it is typically embedded in a variety of social and cultural differences between groups that give a boundary concrete significance (so that those on one side think of those on the other, are not like us because ...). When we discuss ethnici ty, the kinds of boundaries we invoke have the character that Weber attributed to the ethnic group: namely, that they are rooted in a belief in com mon descent?i.e., in a shared history based on a common point of origin in the past, which may be real or putative. Admittedly, this subjective belief in a shared history may be felt more by one side of a boundary than the other and thus, to a greater or lesser extent, the boundary may be imposed, as is typically the case with race-like distinctions.