This essay began at a Berkley Center community luncheon when Becky Yang Hsu—associate professor of sociology at Georgetown—presented her research on China, on the grave-sweeping, the offerings, and the difference between an ancestor and a hungry ghost. The comments I offered from across the table drew on the collective-memory writing coming out of Europe, Danièle Hervieu-Léger and Grace Davie above all. The connection has stayed with me since, and it is worth setting down with more care than a luncheon allows. What does the European memory literature actually contribute to the sociology of religion, and does the Chinese case bear it out? The short answer is that the framework is useful—and that in China it can be watched in practice: families keep their dead present by remembering them, and lose them when they stop.
Sociologists measured European religion by Sunday attendance and survey professions of belief, and by those measures most of Western Europe looked rapidly de-Christianized after the 1970s. Yet the same societies kept producing patterns that did not fit. People who told pollsters they did not believe in God baptized their children anyway; churches filled at Christmas and at moments of public grief. Whatever was persisting, the surveys had no name for it, and Hervieu-Léger and Davie set out to supply one.
Let us begin with Hervieu-Léger’s argument in La Religion pour mémoire (1993): the French sociologist sets aside the question of what religion asserts and fixes instead on how it is handed down, defining it as belonging to a lineage of believers stretched across time. That redefinition explains why someone who does not believe a word about transubstantiation can still feel that giving up the rite would cost something, and why the cost is social rather than theological. Davie found the same logic at work in the institution itself. Her vicarious religion is religion performed by an active minority on behalf of a much larger inactive majority—one that not only tolerates the arrangement but approves of it and expects it to continue. The active few keep the church running as a public utility for a public that rarely walks through its doors.
Paul Connerton came to the question from the study of memory itself: How Societies Remember (1989) distinguishes inscribed memory—texts, archives, monuments—from incorporated memory, carried in gesture and repeated bodily practice, and shows how a gesture repeated without comment can carry a tradition when stated belief has worn thin. On their account, what looked like disappearance was a change of form.
The deeper contribution is methodological. Once one accepts Hervieu-Léger’s definition, the unit of analysis is no longer the individual respondent at a survey moment but the chain itself—grandparents, parents, children, the rites that bind them, the points at which it frays or holds. That kind of question wants ethnographic and historical work—and Hsu’s grave-sweeping research is exactly that.
The framework has not gone unchallenged. To be sure, the demographic arithmetic belongs to Steve Bruce and David Voas, and on its own terms it is hard to dispute: one generation’s nominal Anglicans become the next generation’s nones, belief eroding at nearly the rate of attendance, so that the chain looks to be breaking rather than changing shape. Charles Taylor and the spiritual-revolution writers press from another direction—the framework looks at what is fading rather than what is taking its place, a milieu of subjective-life spirituality whose participants have no chain because they want none—and for Meredith McGuire the trouble is the lineage itself, too stable a container for the improvised, syncretic religion her interviewees actually practice. Each objection holds part of the ground. The memory literature survives them all the same, having broken the habit of thinking the question could be settled from survey data alone.
The Chinese distinction between ancestors and hungry ghosts makes the work of collective memory unusually literal. An ancestor, in the orthodox Han scheme, is a dead person still being remembered—fed, addressed, given tea and incense, summoned at Qingming, set a place at New Year. Let the line of remembrance lapse and that person becomes a hungry ghost; the descendants decide which he is by keeping up the rites or letting them go. What Hervieu-Léger had to describe as a structural metaphor is here a literal calendar of obligations, performed in the open by named people on fixed days.
Hervieu-Léger’s framework is more useful here than Davie’s. Her claim—that what makes a thing religious is its function as a lignée croyante (believing lineage), binding the believer to forebears behind and descendants ahead—reads almost like an ethnography of Chinese ancestral practice written in French sociological prose. Han veneration rests on exactly that premise: the living person is one link in a chain that must not break. While it holds, ancestors remain hun, settled in their tablets, present at meals. When it breaks—a line dies out, a family scatters, the offerings simply lapse—the ancestor becomes guǐ, the wandering hungry ghost whom strangers must feed because his own people no longer will. The settled ancestor and the hungry ghost are the same dead person at two points on the chain, before and after the people responsible for him stopped showing up.
The European literature has its own literary anchor. At the pivot of their arguments, both Hervieu-Léger and Davie reach for a line from Paul Bourget’s 1902 novel L’Étape—“Je ne peux pas vivre sans mes morts” (I cannot live without my dead). It is spoken by Jean Monneron, a young man raised in a thoroughly secular Republican household who finds his way back to Catholicism once he sees that he cannot make sense of himself without the dead behind him. That recognition—that belonging rests on lineage rather than on doctrine—is just what the European sociologists are after, and in Monneron’s mouth it has a directness the academic restatements lose.
Davie’s vicarious religion adds something Hervieu-Léger does not, and the pattern she found in Anglican England fits contemporary Chinese practice closely. Anna Sun’s fieldwork in Shanghai, Beijing, and Hong Kong keeps turning it up: it is the elderly grandmother, or the one designated cousin, or a hired Daoist specialist who keeps the ancestral observances going while everyone else treats them as cultural background. Davie’s image of the church as a bank of memory carries over—temple, tomb, and ancestral hall are the deposit boxes. Confucian filial piety supplies the moral pressure Christian Europe has largely lost, which is part of why the chain has held. Sun’s Harvard Divinity Bulletin essay is titled “Turning Ghosts into Ancestors”; the passage from guǐ back to ancestor is exactly what the rites exist to accomplish.
Connerton is indispensable here, because Chinese ancestral practice runs almost entirely on incorporated memory. The kowtow, the burning of joss paper, the sweeping of the grave, the rice bowl with its chopsticks set deliberately upright: the tradition survives because it lives in bodies. It is one reason the Cultural Revolution could not root the practices out—you can burn a lineage book and padlock an ancestral hall, but a habitus passed from grandmother to grandchild over a rice bowl is much harder to reach.
The Chinese case puts its heaviest pressure on an assumption buried in the European literature—that the chain of memory is, in the end, a metaphor for how a tradition persists in the habits and affections of the living. In the Chinese scheme it is closer to ontology. The ancestor is not merely remembered; he persists, in some real sense, because he is remembered, and he subsists on what the descendants set out for him. Hervieu-Léger reached toward this when she called modernity an amnesiac condition, but the Chinese tradition says it far more bluntly than she let herself. A society that forgets its dead does not just mislay a tradition; it manufactures a population of unattended spirits who must then be fed by strangers—which is how the seventh-month Ghost festival became a municipal undertaking across East and Southeast Asia.
The contrast sharpens if one sets the Bourget line back in front of it. When Monneron says he cannot live without his dead, the claim is psychological: lacking his ancestors he has no settled identity, and Catholicism is what holds the line for him. Said in a Chinese ritual register the same words turn practical—he cannot live without his dead because the dead must be fed, unfed dead grow hungry, and hungry dead make trouble for the living. The words carry over unchanged while the meaning does not, and that change—from the dead as the ground of identity to the dead as dependents who must be maintained—is where the most interesting comparative work in the sociology of religion is now opening up.
Hsu’s grave-sweeping research is where these debates can be joined. In the Chinese ancestral case the chain is visible in a way it rarely is in Europe and its breaking has a consequence everyone can see—the hungry ghost—so whether a lineage holds, frays, or breaks can be read off practice rather than inferred from a survey return, something close to a natural experiment for deciding between Hervieu-Léger and her critics. None of this settles the quarrel, but it tells sociologists where to look.
