Why it is a question
The legacy line (QUESTION-301, QUESTION-299) asks how the living reach forward to people after them. Intergenerational trauma is the past reaching forward into the living. Four routes are proposed, and they make different claims about what is inherited.
-
The germline. In mice, a father’s conditioned fear of an odour shows in his offspring and grand-offspring, with methylation changes in sperm (Dias and Ressler, LIT-1746; reviewed by Bohacek and Mansuy, LIT-1738). In humans the evidence is:
- epidemiological: Överkalix (Pembrey et al., LIT-1764) and Civil War prisoners’ sons (Costa et al., LIT-1696);
- molecular, in small samples: FKBP5 methylation in Holocaust survivors’ offspring (Yehuda et al., LIT-1692).
Critics hold that germline reprogramming erases most marks in mammals, and that human studies cannot exclude genetic, uterine and cultural routes (Heard and Martienssen, LIT-1745; Horsthemke, LIT-1762; Bird, LIT-1708).
-
The womb. Prenatal famine leaves lasting methylation marks (the Dutch Hunger Winter, Heijmans et al., LIT-1751). The exposure is the person’s own, so this is intergenerational, not inheritance through untouched generations.
-
Parenting and silence. The clinical tradition (Danieli, LIT-1701; Yehuda and Lehrner, LIT-1715) traces transmission through care and a “conspiracy of silence”. Yet meta-analyses find no general second-generation harm in non-clinical samples of Holocaust survivors’ children (van IJzendoorn et al., LIT-1718), and none in the third generation (Sagi-Schwartz et al., LIT-1770).
-
Narrative and identity. A group’s shared memory of an ancestral injury becomes a marker of its identity: Volkan’s “chosen trauma” (LIT-1727), and the sociology of cultural trauma, in which carrier groups make an event an injury to the collective (Alexander et al., LIT-1679). Lehrner and Yehuda (LIT-1793) join this to the biology.
The record already holds two neighbours:
- Cultural persistence: attitudes formed under slavery persist locally by cultural transmission (THEORY-889, from LIT-1388).
- Inherited non-genetic state in bacteria: E. coli cells inherit a transient state from their mothers (THEORY-179).
CASE-098 is a rivalry that new fans inherit before they know why it matters.
How it meets the owner’s thesis
On the identity route, descendants feel as their own an injury done to people they never knew. That is the thesis’s extended self (CLAIM-491) running backward in time. Read that way, a descendant’s grief or anger over an ancestral injury is self-regarding: the injured self includes the ancestors. It is also the reverse of Pitcher’s posthumous harm (LIT-1740), where the dead are harmed by what comes after them. Here the living are harmed by what came before them. On the biological routes no self is needed: what passes is a physiological state, not an identification.
What would count as an answer
- On the routes: a human study that separates germline transmission from uterine, parenting and cultural routes. For example, children of trauma-exposed fathers conceived after the trauma and raised apart from them, or adoptees. Or evidence that no such separation is achievable.
- On the identity route: whether descendants’ distress over an ancestral trauma follows their identification with the group, measured apart from their knowledge of the events and from family transmission. If it does, the injury is carried as part of a social self, and CLAIM-492’s test applies to the past as well as the future.
Note of 2026-10-29: the germline route, read
Five works were read: the mouse study, the Civil War study and the three sceptical reviews. Together they narrow the germline route.
- What counts as inheritance (Heard and Martienssen, NOTE-1254). An effect is transgenerational only in a generation never exposed: the grandchildren of a father exposed before conception, the great-grandchildren of a pregnant mother. Survivors’ children are an intergenerational case. Two rounds of reprogramming make transmission past them unlikely in mammals. What reaches grandchildren travels, on present evidence, by the womb, by care, by habit or by story.
- What the best animal case carries (Dias and Ressler, NOTE-1281). Offspring were more sensitive to the odour their father was shocked with, with more neurons for its receptor, through IVF and cross-fostering. The authors call what passes “sensitivity”, not fear. The proposed sperm mark is modest, in groups as small as four. A tuning of perception needs no representation of the ancestor.
- What the best human case carries (Costa et al., NOTE-1255). Sons of the worst-held prisoners of war had a higher older-age mortality hazard (about 1.11); daughters did not. The authors favour an epigenetic route only by elimination and have no data on parenting. One untested line touches the identity route: return “to a grateful nation” may have given captivity a meaning that protected sons. On that reading a group’s story buffers transmitted harm as well as carrying it.
- The human test cannot yet be run (Horsthemke, NOTE-1265). A human study can exclude genetic inheritance but not ecological or cultural inheritance. The causal step “can hardly be done in humans”. Adoptee and paternal-only designs weaken the confounds without removing them. The question’s first test is, for now, the second answer it allows: no clean separation is achievable.
- The named human cases are weak (Bird, NOTE-1264). Överkalix ignored cultural transmission within families and risks multiple testing; the FKBP5 study rests on 32 people. Bird cites a larger Swedish cohort (Vågerö et al. 2018) only for failing to replicate the paternal-grandfather effects on diabetes and heart disease. That paper’s title reports such an effect on all-cause and cancer mortality, so his account is one-sided.
Two cautions about this consensus. The three sceptics are not independent. Heard reviewed Bird’s article, Bird commented on Heard’s manuscript, and Bird and Horsthemke have written together. They are one school, not three judgements. And their chief confound, families passing on habits, is itself one of the question’s routes. The social routes, parenting and identity among them, gain share by default, not by evidence; nothing read yet chooses between them.
Note of 2026-10-29: the other routes, read at second hand
Seven works were read in place of the closed ones. They correct two of the question’s labels and shift weight among the routes.
-
Two labels are wrong. Överkalix is a male-line nutritional case, not a germline trauma case (Pembrey, Saffery and Bygren, NOTE-1320).
- The exposure was a grandparent’s food supply in mid-childhood, and good supply, not poor, went with higher grandchild mortality.
- The authors leave the mechanism open, grant that culture matters and that replication is needed, and measured no family culture.
The FKBP5 study is intergenerational, not a germline test (Youssef et al., NOTE-1310). It had 32 survivors and 22 offspring against 8 and 9 controls. Methylation was higher in survivors and lower in offspring: the child does not carry the parent’s mark.
-
The proponents concede the disputed steps (Jawaid and Mansuy, NOTE-1296). In mice, sperm RNA from traumatized fathers reproduces their symptoms in offspring. The authors say it is “not clear” how effects pass beyond the first generation, and that germline dependence “is not established in humans”. Both sides now agree on the intergenerational effect in mice and on the gaps.
-
The womb route is real and narrow (Heijmans et al., NOTE-1248). Conception during the Hunger Winter went with slightly lower IGF2 methylation six decades later (0.488 against 0.515); exposure late in pregnancy did not. It is one gene, in blood, with no outcome measured, from famine, not psychological trauma.
-
The parenting route finds nothing in the general population (Fridman, NOTE-1291). The meta-analyses’ authors, through their student, restate:
- no second-generation traumatization in non-clinical, well-designed samples (32 samples, 4,418 offspring);
- none in the third generation (13 samples, N = 1,012).
Her own registry sample of survivors’ daughters showed no psychological or cortisol difference. The comparison families shared the survivors’ national memory, so the design tests family routes with the identity route held fixed.
-
The identity route passes by category, not descent (Eyerman, NOTE-1246; Volkan, NOTE-1250).
- Cultural trauma is a group’s loss of identity and meaning, which “need not necessarily be felt by everyone”. Slavery became the trauma of black Americans whether or not their ancestors were enslaved.
- Volkan’s chosen trauma is an image, possibly mythologized, called up by a present threat. But his mechanism is parents depositing it in children, so his identity route does not separate from parenting.
-
A group’s story buffers as well as carries. Fridman’s survivors found protection in building a state. Costa et al.’s “grateful nation” (NOTE-1255) is the same suggestion.
-
The nearest thing to the identity test (Hirst, Cyr and Merck, NOTE-1305). Of children nine or younger at 9/11, 89% could report a parent’s memory of where they were, and stronger identification as American went with knowing it better. It falls short of the test: it measures knowledge, not distress, the channel is the family, and the direction is unknown.
Still unread: Lehrner and Yehuda’s paper joining the biology to cultural trauma (LIT-1793). Danieli’s handbook (LIT-1701) is read at second hand in the next note.
Note of 2026-10-29: Danieli’s framework, read at second hand
Danieli, Norris and Engdahl (NOTE-1307), read in place of the handbook, test its framework on 422 adult children of survivors.
- The parenting route runs through the parent’s style. Wartime internment reached the child only through the parent’s victim style (stuck in the loss, overprotective, volatile), with small indirect effects (.05 to .06). The post-war family weighed more than the war.
- Silence is not the main channel. Numb style, which includes the family’s “conspiracy of silence”, had an effect only through mothers and dropped out of the final model. With Fridman’s thesis, that is two works that do not put silence first.
- No comparison group. The paper locates the route among survivors’ children without showing harm against Fridman’s null (NOTE-1291). The two results are compatible.
- The ancestor and the injury come apart. Children who counted their murdered grandparents as their own reported fewer impacts, and this item had the largest total effect (.21). Meanwhile “obsession with the Holocaust” is scored as part of the harm. Taking in the ancestor went with less harm; taking in the injury is part of the harm. The question’s phrase “an ancestor’s injury becomes part of the descendant’s self” covers both, and they should be kept apart.
- A group’s story as buffer, again. Settling in Israel went with milder impacts (total effect −.16). Reading that as the nation’s story protecting the family is an inference, the third time the record has met it (Costa et al., Fridman).