Children who say they remember past lives: the research

What the research actually is

The research on children's past-life memories at the University of Virginia is a real, decades-long project, not a single experiment. Since the psychiatrist Ian Stevenson founded the university's Division of Perceptual Studies, researchers there have collected more than 2,500 documented cases of young children, most between ages two and five, who describe specific memories of another person's life and death. The case files are real and carefully kept. What is disputed is not whether children say these things, but whether the pattern points to anything beyond ordinary explanation.

How a case gets built

The division's current director, the child psychiatrist Jim Tucker, continues files Stevenson started compiling in 1961; Stevenson led the division until his death in 2007. Tucker meets families himself, shows children arrays of photographs, and checks their specific statements against birth records, death records and other paperwork tied to the life a child describes. Field notes from each case are coded across more than 200 variables and added to a database that now holds over 2,500 cases gathered over 50 years, most of them from outside the United States.

What the pattern looks like

By the division's own account, children's statements typically begin between ages two and five and fade out by around age seven, and about 30 percent of cases include a birthmark or birth defect that researchers say corresponds to a wound on the person the child describes, checked where possible against medical records. In Tucker's own case series, about 70 percent of the previously identified individuals died a violent or unexpected death, near 20 percent of children carry a matching birthmark, and the typical gap between that death and the child's birth is about 16 months.

Two documented examples

In one widely reported case, a four-year-old Oklahoma boy named Ryan Hammons described a past life in Hollywood, including a white house, a swimming pool and three sons. His mother, Cyndi, recognized him pointing at his own face in a 1930s film still and identified the man as Martin Martyn, a Broadway dancer turned talent agent; researchers confirmed that an address and a name Ryan had given, 825 North Roxbury Drive and a "Senator Five," matched Martyn's actual life and his acquaintance Senator Irving Ives. A separate case in the division's files involves James Leininger, whose family said he described being a World War II pilot shot down in battle.

Where critics push back hardest

The sharpest published examination of an American case comes from Benjamin Radford, deputy editor of Skeptical Inquirer, who examined a Virginia boy the magazine called Andrew Lucas. Radford found that Andrew's mother had shown him photographs of six soldiers and asked, "Were these your friends?" before he answered, "I was friends with them, a lot," a pattern Radford reads as a child echoing an adult's cue rather than recalling anything. He also noted that a television crew, not a neutral researcher, had helped connect Andrew to the Marine his family settled on.

Radford points to cryptomnesia, the documented tendency for people, especially under hypnosis or suggestive questioning, to produce sincere, detailed accounts of events they never lived through, and cites the discredited 1950s "Bridey Murphy" case as an earlier example of the same effect. He argues a fair test would show a child mixed photographs, including strangers, and ask him to supply names unprompted, rather than confirm names an adult had already supplied.

Named objections, and the researchers' replies

Other critics have pressed different points for decades. The philosopher Paul Edwards argued reincarnation cannot be squared with a brain-based account of consciousness. Leonard Angel accused Stevenson of "backwards reasoning" when he matched birthmarks to old wounds. Keith Augustine and Champe Ransom both pointed to parents who already believed in reincarnation shaping children's stories through leading questions, and Stevenson's interviews, conducted over one and a half to four days with the help of local interpreters, could plausibly have let that kind of bias in.

Stevenson's supporters reply that the strongest cases had statements written down and witnessed before anyone matched the child to a specific deceased person, which limits how much coaching or memory distortion could explain. Of roughly 1,600 cases Stevenson examined for the possibility of deception, he found deception in only seven, according to a summary of the critical literature published by the Society for Psychical Research.

Belief, culture and the shape of the evidence

Cases have turned up on every inhabited continent, including in the United States and Europe, but researchers have long noted they surface more often where a culture, such as Hindu communities in India or Buddhist communities in Sri Lanka, Thailand and Myanmar, already treats reincarnation as part of the natural order. That pattern alone does not settle whether the accounts are real memories, invented ones or something in between. It is one reason Tucker and Stevenson both sought out cases in the United States specifically, to test whether the pattern holds up outside cultures that expect it.

Is it real, and where the research stands now

Researchers in the field tend to be careful when asked directly whether this is proof of anything. Tucker has said he understands "the leap it takes to conclude there is something beyond what we can see and touch," and frames his own role as documentation rather than proof. The biologist Michael Levin of Tufts University has called Tucker's methods "first-rate research" while still saying that current science can neither confirm nor rule out what the strongest cases mean.

Stevenson himself put the disagreement in terms that still hold: "The wish not to believe can influence as strongly as the wish to believe." Stated plainly, the record comes down to this: the documentation is real, researchers and critics still disagree about what explains it, and no single case has satisfied independent skeptics working from ordinary explanations like family suggestion, leading questions or cryptomnesia. Real or Legend stamps this record disputed for that reason.

Questions people ask