IT WAS a tough life in Roman Britain and it seems that hard-pressed natives
killed some of their newborn children to reduce the strain on family resources.
But the first detailed genetic study of infant remains from the period suggests
that perhaps the victims weren’t just baby girls.
Simon Mays from the English Heritage Centre for Archaeology in Portsmouth and
Marina Faerman from Hebrew University in Jerusalem studied 31 infant skeletons
found at two English sites where they believed Britons practised infanticide.
They wanted to test the old assumption that it was mostly baby girls who were
killed.
Previous researchers believed infanticide took place because they found an
abnormally large number of children from graves in Roman Britain who died about
40 weeks after conception—the average age of a newborn child. And
archaeologists also found a disproportionate number of adult male skeletons
throughout ancient British sites.
Advertisement
But Mays and Faerman uncovered a different picture. They measured the length
of the babies’ bones, and determined they had died between 38 and 41 weeks. But
by examining their DNA for fragments of the X and Y chromosomes, they identified
the gender of 13 of the children: 9 of the victims were boys and only 4 were
girls.
Their finding agrees with a study from 1998 of Roman remains found in
Ashkelon, Israel, where more boys were found than girls. At the time researchers
thought this was an anomaly because the site was near a brothel, and perhaps
female children were being kept as future workers.
Mays says his study is too small to say whether more boys were sacrificed
than girls in Roman Britain. But if the babies were indeed murdered then
obviously it wasn’t just girls who were victims.
Andrew Chamberlain of the University of Sheffield interprets the results
differently. “Infanticide probably went on, but there’s no guarantee that these
infants are representative of this.” Instead, he thinks the mix of genders found
by Mays may indicate that these were natural deaths.
- More at: Journal of Archaeological Science (vol 28, p 555)