Annonse
The catastrophe in 1956 shook Belgium. Researchers have now provided answers to the bereaved
“It has been a national trauma in Belgium,” says researcher Magnus Deli Vigeland. The consequence of this discovery also reveals an unpleasant truth.
Published
On August 8, 1956, a fire broke out in an elevator shaft in a mine in Marcinelle, Belgium.
The carbon monoxide that spread down into the mine killed all 262 men who worked there. More than half of them were Italian guest workers.
“It has been a national trauma in Belgium,” says Magnus Deli Vigeland, who researches forensic genetics at Oslo University Hospital.
Especially because 14 of the dead miners were never identified.
Vigeland has now investigated the case with Norwegian, Swedish, and Belgian colleagues. This would lead to a new answer. And an uncomfortable truth.
Annonse
Belgian police requested help
The case took a new turn in 2021 after many years of pressure from the descendants of the victims of the mining accident.
The remains from the unmarked graves were exhumed and DNA-tested.
Belgian police managed to identify three of the bodies. Then they hit a wall.
The methods the police use to test DNA are cheap and simple. They only contain a few genetic markers that can be easily measured in a lab.
"Forensic genetics is special. You can't just create new systems at will. All DNA profiles in the world are based on these old markers,” says Vigeland.
Vigeland and his Norwegian colleagues are mathematicians and statisticians. They have developed several methods that can be used to identify people in difficult DNA cases.
That's why Belgian police asked them for help.
Only had time to work for four months
Italian Rocco Ceccomancini was hired to work in the coal mine in Marcinelle in 1956.
The 19-year-old only managed to work there for four months before the disaster on August 8.
He was never identified among the dead.
Now, 70 years later, Ceccomancini has three living family members who could be DNA-tested.
They included one female cousin and two daughters of a male cousin.
“The gold standard is to match against immediate family members, preferably parents or children. More distant relationships are more challenging,” says Vigeland.
Annonse
Had to create new methods
Vigeland points out that the easiest way to solve the mystery would have been to analyse the entire genetic code. That is, all three billion or so of the bases A, C, T, and G.
But that's expensive.
“As important as this case is in Belgium, there was no money to do this,” says Vigeland.
Fortunately, Swedish researchers wanted to help.
They ran the DNA samples through a self-developed test that measures nearly 5,000 points in the genetic material. Both from the 14 victims and from families who were missing someone after the mining accident.
But with so many peripheral family members, standard statistics did not provide a reliable enough answer as to whether they were related, says Vigeland.
The Norwegian researchers had to create new methods.
Simulated many family trees
The whiteboard and the computer became the most important tools.
To solve the puzzle, the researchers simulated what happens when DNA is passed on through a family tree.
It works like this:
DNA is packaged into 23 pairs of chromosomes.
One set from the mother, one from the father.
Before passing the chromosomes on to your own children, each chromosome pair is mixed with each other.
Not like two decks of cards that are shuffled together.
But more like picking random piles of cards from each deck, and then making a new deck.
A lot of variation in what relatives share
This rough mix can yield strange results.
Maybe a large chunk of a chromosome is completely different between you and your brother, while the same section is exactly identical to that of your second cousin.
“There’s a lot of variation in how much DNA relatives actually share,” Henrik Nordtorp says.
He is also involved in the new study and is working on a PhD in forensic genetics with Vigeland as his supervisor.
“Statistics, mathematics, and data analysis become tools to determine something precise about exactly that variation,” says Nordtorp.
A new identification
By running many simulations of how DNA may have moved through the family trees of the deceased workers, the researchers were able to provide an answer.
For each of the remaining eleven families, they calculated the probability that they were related to one of the deceased miners.
One of the answers led to a formal identification. The researchers were able to conclude that the Italian Rocco Ceccomancini was one of the 14 who were buried in an unmarked grave.
The Norwegian researchers also confirmed the three identities that the Italian police had already come up with.
The new study describes only these four individuals.
Revealed an uncomfortable truth
But what about the rest?
The answer surprised the researchers.
“What we discovered is that several of the living family members do not match any of the victims,” says Vigeland.
The consequence of this finding reveals an unpleasant truth.
Namely, that some of the deceased must have been handed over to the wrong family after the accident.
“The identification process that took place in the old days was much more uncertain than today. Back then, they didn't have DNA,” says Vigeland.
With over 250 dead, it must have been an extremely demanding situation, he says.
Most viewed
“Our analyses unfortunately suggest that mistakes were made,” says the researcher.
For some of the missing victims, there were also too few living family members to be able to conclude whether they were among the 14 in unmarked graves.
More pieces to the puzzle
The new study joins a number of historical cases that have been solved with the help of DNA, according to Kirstin Janssen, an associate professor at the Centre for Forensic Genetics at UiT The Arctic University of Norway.
“It’s very fascinating what these tools can do,” says Janssen.
She finds it very positive that the researchers behind the new study have developed new methods for identification.
“The more pieces that fit into the puzzle, the more certain the answerbecomes,” says Janssen.
Because even though the established methods for testing family relationships work quite well, there is always a small risk of being wrong.
“You might miss someone, or you could discover a family relationship that doesn’t exist,” she says.
Could be useful in other disasters
The new tools are openly available to everyone.
The UiT researcher believes they could be useful in similar types of cases in the future.
“These methods will help to identify people in the event of natural disasters, terrorism, war, or major accidents," says Janssen.
References:
Vigeland et al. Using autosomal and X-chromosomal SNPs to identify a victim of the 1956 Marcinelle mining disaster, Forensic Science International: Genetics, vol. 82, 2025.
An article about the identification from the Institute of Natural Sciences, Belgium.
———
Translated by Ingrid P. Nuse
Read the Norwegian version of this article on forskning.no
Related content:
Subscribe to our newsletter
The latest news from Science Norway, sent twice a week and completely free.
