Children’s Past-Life Claims Show Shorter Gaps After Sudden Deaths

For decades, a research unit at the University of Virginia has gathered accounts of children who say they remember the lives of people now dead, then turned each case into rows of coded numbers. A new analysis of 1,678 case records finds a link between how suddenly the earlier person died and how soon the child was born. Saeedeh Sadeghi and colleagues at the University of Virginia School of Medicine do not conclude that reincarnation is real. They add that the results cannot be taken as evidence for or against any metaphysical reading, and that more research is needed.

Few corners of research draw as much side-eye as children’s past-life claims, and much of the quarrel concerns what the children said, which is hard to check decades on. So the authors measured something else: the gap, in months, between the earlier person’s death and the child’s birth, which they call the intermission. Hardly settled ground.

The records come from a university archive of 2,254 cases, and in 1,678 of them investigators decided that one named dead person fit the child’s statements. Trained assistants coded each case against a list of 208 variables, among them how the earlier person died. Each death also got an unexpectedness score, from 0 for a death foreseen more than a month ahead to 4 for a sudden death nobody saw coming. Unnatural deaths (accidents, violence, suicide) made up 57% of the cases with this information, and 68% of deaths were rated completely unexpected.

Here’s the awkward part: the unit that collected, investigated and coded these cases is also the one that analyzed them. So how do you test an archive you built yourself? The authors went hunting for the dull explanations first, asking whether the link grew stronger exactly where bias should be strongest.

Twelve Months Against Twenty-One

A median is the middle value when cases are lined up by size, which stops a few extreme cases from steering the result. Among 1,075 cases with both scores on file, the median gap ran from 21 months after the most expected deaths to 12 months after the most sudden ones. That is about 43% shorter, in an archive where the median gap is 16 months and the longest, at 150 years, did not get a vote.

An analysis in 1988 of 326 cases had already reported shorter gaps after violent deaths. For Sadeghi, who led the new analysis, the larger archive brought a surprise: “To our surprise, the relationship was still there when we tested it in a much larger number of cases, and the social and psychological factors we examined did not explain it away.”

Five Tests for a Dull Explanation

The authors then ran five bias checks, built on one logic: a mundane explanation should work hardest where it has the most room. Four checks covered how well the death was documented, how accurately the child described it, whether the families already knew each other, and the difference in social standing between them. Across those four the link held except among 332 loosely documented cases, where it was weaker, and no check changed its strength by more than the authors could tell apart from chance. The fifth check asked whether a family’s religion teaches reincarnation. It did alter the link, but against a bias story: clear in the 255 families outside such religions, only marginal in the 626 inside them.

The effect is small: sudden death accounts for only about 2% to 3% of the variation in gaps, a figure the authors call modest. Nor is it a clean replication of the 1988 work, since the archive holds most or all of those older cases.

A Careful Paper, a Bolder Release

The university’s press release says the team ruled out sources of bias. The journal article is more careful: its checks gave little support to bias accounts, but the data are observational, and overlooked social or psychological factors could still be at work. Religion also stood in for belief (a rough proxy), so the families the release calls nonbelievers are really those outside faiths that teach reincarnation.

In plain terms, the work hands skeptics a specific pattern to explain and hands believers nothing they can fairly call proof. The authors name two routes forward: fuller social and psychological models, or broader frameworks in which consciousness is not entirely reducible to known biology. They also want contemporary cases, where details are written down before anyone has identified the earlier person. Until then, the archive can show a pattern but not its cause.

Marieta Pehlivanova, one of the paper’s two senior authors, sees the gap in understanding as an invitation: “The mechanisms underlying these findings are currently unclear, but that gives us even more reason to keep looking, not to look away.” Over to the skeptics.

FAQ