Family sociology has gained in sophistication from developments in related disciplines. For example, research by historians of the family has put to rest, at last, the classical family of Western nostalgia (Goode, 1963). Unfortunately, this mythical family does not rest in peace, but inhabits the consciousness of the general public with a wondrous tenacity. The use of the cohort analysis in the study of social change or continuity directs attention to the changing characteristics of specific age cohorts or generations not only as they move through their life course but also among successive cohorts in a particular society. Both foci warn us against generalizations regarding parent-child relations as fixed. Intergenerational relationships must be studied in historical context, especially in terms of the linkages between family and other institutional spheres as these undergo continual change. While we may never disabuse the general public of the notion that there once was a time in which the extended family reigned supreme-and, more importantly, that mutual respect and satisfaction governed adult child/ aged parent interaction-we must, as family sociologists and practitioners, take an unromanticized view of intergenerational relations at the distal end of the life course. The question usually asked of this topic is can such bonds be strengthened? The one we propose here is why have any such bonds persisted? The modern family is characterized by choice: whom to marry, where to live, how to earn a living, how many children to bear, and, increasingly, how to conduct interpersonal relations and allocate tasks within the nuclear family. As we move from the family of obligatory ties to one of voluntary bonds, relationships outside the nuclear unit similarly lose whatever normative certainty or consistency governed them at earlier times. For example, sibling relationships today are almost completely voluntary, subject to disruption through occupational and geographic mobility, as, indeed, it might be said of marriage itself. Is this also to be the fate of parentoffspring ties in later life? There are many indicators of growing distance between generations, especially so in later life. There are also clues to enduring qualities of the parent/child bond. We shall examine the most important of these forces-centripetal and centrifugalat both the societal and familial levels. *Adapted from a more extensive essay, Parent and child in later life: Rethinking the relationship. In R. M. Lerner & G. B. Spanier (Eds.), Child influences on marital and family interaction: A life span perspective. New York: Academic, forthcoming.