Loveless Lost: The Case of the Oldest Doe Identified Using Forensic Genetic Genealogy

In the summer of 1979, a family was searching for artifacts in Buffalo Cave, a dark lava tube outside of the small and remote town of Dubois, Idaho. Approximately 200 feet from the opening of the cave, the family discovered a burlap sack containing a mummified, clothed, and dismembered headless torso. The Clark County Sheriff’s Department responded and searched the cave for the rest of the body to no avail. The body was taken to the local jail, where the coroner concluded that the remains could not have been in the cave for more than 10 years due to pungent odor and preserved skin. When the remains were sent for examination by renowned Smithsonian anthropologist Dr. Douglas Ubelaker, he concluded that the victim was a White male, at least 40 years old at the time of death, dismembered below the elbows, above the knees, and above the shoulders. There was no trauma to the remains aside from the dismemberment marks, so no cause of death was determined.



In early spring of 1991, another discovery was made after a family searching the cave for artifacts stumbled upon a mummified hand exposed in the soil. After the sheriff’s office was alerted, a subsequent search revealed burlap sacks containing the arms and legs in a shallow burial. Systematic searches and probing deep into the cave by anthropologists found no further evidence. No head was ever recovered even in later searches that included student groups and cadaver dogs. No fingerprint or CODIS matches were ever made, and the case remained cold until 2019. Without a narrowed postmortem interval or timeline of the crime, the victim remained unidentified until anthropologists coordinated with the local Sheriff to bring the case to the attention of the DNA Doe Project (DDP).


Nearly 103 years since his death, Joseph Henry Loveless, a Wild West outlaw and bootlegger with numerous aliases was identified. A team of volunteers had been assembled to work on the case and 15 weeks after the autosomal profile was first uploaded to GEDmatch, the DDP contacted the sheriff’s office with a potential match (later confirmed with a direct to consumer ancestry kit via the closest living relative, an 87 year old grandchild). In the weeks leading up to the identification, historical documents related to Loveless’ colored past were discovered, including a Wanted poster which detailed clothing that he was wearing during his last jailbreak. The clothing description matched the shirt and pants found on the body in the cave, leading law enforcement to believe that Loveless was killed shortly after his final jailbreak in 1916.


Loveless’ identification is the oldest open Doe case to be closed by law enforcement using genetic genealogy. During their presentation at ISHI 31, Margaret Press and Anthony Redgrave described how anthropologists’ persistence, the Sheriff’s determination, and forensic genealogists’ tenacity culminated in a successful collaboration greater than the kept secrets of a dark, remote cave. As there was not time for them to answer all of the questions that came in during the conference, we’ve compiled them here.


How important was the Y DNA work in solving the case?

Margaret Press: I believe it shaved considerable time off the research in that it allowed the team to move branches with that surname onto the paternal side of the tree. That eliminated a lot of scenarios.


At what time in your genealogy research did you know you were working on a case that was going to make history?

Anthony Redgrave: I had the notion as soon as I heard about the case that we could potentially be making history, since it had been examined by so many different organizations, including the Smithsonian and the FBI, and they were unable to determine an identity.  What really made it clear, though, was when the WATO (What Are The Odds) probability tool became more and more insistent that our John Doe had to have been born before 1900 in order to fit in to the predicted relationships with his DNA matches.  That alone would have been enough to make history, but then we discovered that he was not only murdered, but a murderer, and a bootlegger, and user of many aliases, and an escape artist.  I don’t think there will ever be another forensic genetic genealogy case with this many twists and turns ever again.  If there is, though, I hope I get to work on it.


Typically genealogists get emotionally involved with their clients…How do you feel about Henry now that you know he is a wanted criminal that murdered multiple people?

Anthony Redgrave: We had no idea what kind of person Henry was before we found his identity, just like most Doe cases.  All people are flawed, and we don’t judge the person we’re trying to identify.  It’s a little strange with Henry because the typical sensation I have when working on a Doe case is “oh no, what happened to this poor person?” and Henry’s case was no exception – who deserves to be dismembered and buried and forgotten about? So of course I had some protective feelings about him just like I would with any other Doe.  Finding out what he did was a huge shock, but it doesn’t change that much about how I feel about putting in the work.  I don’t regret it, and I value the time we spent.  I know he did terrible things.  Terrible things were done to him in retaliation.  It’s not our place to say if that was fair.  It’s our job to find out who he was.  The fact is that everyone matters, and everyone deserves to have their name back, and every family deserves to have answers if someone is missing.


Margaret Press: To quote Harry Bosch: “Everybody matters or nobody matters.” We’re neither judge nor jury when we strive to return someone’s name.


How was the composite image of Loveless created?

Anthony Redgrave: We all tried really hard to find a photograph of Henry, but we found photos of every family member except for him – brothers, sisters, parents and grandparents.  When we put together that Henry was the same person as Walt Cairns, the physical description on the wanted poster started to make sense, as some of the facial features described were features we could see in the family photos. Having previously done photo composite work to produce a forensic image for Doe cases, I could see how a plausible image of Henry could be constructed from his relatives and the description on the wanted poster.

All of the photos I used were taken with the subject turned at slightly different angles, so I had to correct them all to be facing the same direction using a 3D rendering tool.  I compared features across the different family members to see which ones carried over the most strongly and selected those portions.  After that, I selected the portion of each family member’s face that seemed to match the description on the wanted poster and pasted them together in one image.  I smoothed the pasted together pieces out using a digital paint program, and gave him his hat, sweater and overalls from stock photo sources.  A video that goes into more detail of how I did this is available here:


Did the anthropologist ever reassess their 5  year assessment after realizing the remains were close to >100 years old?

Anthony Redgrave: The PMI estimate was never really set in stone to begin with.  Dr. Douglas Ubelaker, the Smithsonian Institution anthropologist who examined the remains, was hesitant to give a postmortem interval because he had seen remains in a similar state that were proven to be over a hundred years old when the initial estimate was given as less than five years.  That being said, he had to report something, so he hesitantly gave the estimate and made it clear that it could be incredibly off.  Knowing this from the beginning helped us to have open minds about what time period to look for a person of interest.  We only had this perspective because of working closely with the anthropologists, Drs. Blatt and Michael, who brought the case to us. On most case summaries, Dr. Ubelaker’s estimate is recorded in shorthand as five years without the elaboration of his hesitancy to give one in general.  Without their input, this case could have taken quite a lot longer.


Dr. Samantha Blatt: The initial (in 1976) estimate for post-mortem interval by Dr. Ubelaker was that the remains were AT LEAST 5 years old, but could be much older. There was and is still no reason or methodological way (other than bomb or radiocarbon dating) to assert otherwise. A number of factors play into the decomposition rate and preservation of remains. In this case, the cool, dry cave  is the biggest contributor to the preservation, but even body size, time of years, etc. are factors that are not easily or always discernible from remains. There are not specific taphonomic experiments that have been done in eastern Idaho in a cave environment like that from which to estimate decay rates. Taphonomic studies are always advancing, but there are too many intrinsic and extrinsic factors to consider to narrow the age. Unfortunately, historic archaeologists were never consulted to examine the clothing and pinpoint a time period. Post-mortem interval is one of the most difficult variables to pin down even today. As the anthropologists who participated in this case are also bioarchaeologists, studying remains in the distant past, by training, the possibility of the remains being historic was never eliminated. As far as “assessing” other aspects of the remains, yes., the remains were re-analyzed by several waves of professionals at ISU and elsewhere. Conclusions about factors of the biological profile did not vary significantly (with or without the remains being fleshed).


It looked like you had many DNA matches.  What was the largest cM value in the beginning?  Did you have to do additional targeted testing to narrow your case or mostly genealogy work?

Margaret Press: Our top matches were in the 180 cM range, but with all the intermarriage in the Loveless family the totals were a bit misleading.  Add to that the complication of innumerable half relations due to LDS polygamy in the family trees of the matches and the relationships of the top matches became harder to predict.  We did no target testing in order to come up with the candidate because we arrived at a potential ID in a relatively short period of time, and it was not necessary once we were able to confirm the paternal line with the Y DNA. Once we had his name and next of kin, we sent an AncestryDNA kit to the Sheriff’s Office, which they hand-delivered to the grandson. This was how confirmation was obtained.