Archaeologia Islandica - 01.01.2006, Qupperneq 63

First steps towards an archaeology of children in Iceland
cases recording large numbers of parents
exposing their children.
Another line of argument which
Clover follows concerns polygny. She
proposes that early medieval Scandi-
navian society practised polygyny and
points out the correlation between socie-
ties that practise polygyny and those that
practise female infanticide (Clover 1988:
170-2). The existence of polygny in early
medieval Scandinavia is also is open to
question because, again, it is supported
by later literary evidence. Certainly for
early Icelandic society it is hard to imag-
ine that there was a desire to reduce the
number of women who might survive to
reach child-bearing age and then actually
bear children and help sustain the new
colony. In this context Clover cites many
examples from other cultures where boys
are seen as more valuable than girls but
this runs counter to the needs of a society
only recently having migrated to a new
and relatively hostile environment such as
Iceland. All healthy workers were surely
valuable; and women’s deaths in child-
birth should have made them more valu-
able not less. It seems possible that in the
very earliest decades of Iceland’s coloni-
sation there were more men than women
but it seems unnecessary to suppose that
this imbalance was not rectified by, say,
AD 1000.
If we turn to the archaeology,
adult burial populations have also been
used to support the view that selective
female infanticide took place in early
medieval Scandinavia. In many cemeteries
fewer adult females than males have been
found. Yet the data ought simply not to be
used in this way. There are well-known
and continuing difficulties in determin-
ing the biological sex of adult skeletons
with the problem that the sex of many
skeletons cannot be determined. Probably
more important is the tendency to overes-
timate the number of males where a sex
determination is made. Osteologists know
and accept that this is the case; techniques
are not yet sophisticated enough for us to
make better assessments of biological sex
for archaeological skeletal material (e.g.
Walker 2005). Even the stronger propo-
nents of selective female infanticide rec-
ognise the limitations for their case based
on the archaeological evidence (Wicker
1998: 210-16).
The strongest noted imbalance
between genders was solely based on an
analysis of Norwegian burials accompa-
nied by grave goods (rather than of all
burials). It assumed that the grave goods
associated with the buried individual can
be easily and directly related to the bio-
logical sex of the person buried (Dom-
masnes 1979: 98-103; Clover 1988:
165). Ratios of three male style burials
to one female style burial (and indeed
much higher) have been detected on this
basis. The usefulness of this information
for demographic patterns is limited for
the reasons already mentioned: it is only
a measure of the apparent gender of part
of the population and not their biological
sex; in reality, gender may have been far
more complex than a simple binary divi-
sion between all males and all females
(Welinder 1998). And, in fact, where oste-
ological analysis has taken place, the ratio
of males to females can be much closer
to 1:1. The classic study of Danish graves
gave a male:female figure 1.16:1 and a
more recent Swedish one gave exactly
1:1 (Sellevold et al 1984:178; Bolin
2004:176). It might also be argued that
cremation burial, a rite which occurs in
variable frequencies all over Scandinavia
in the early middle ages and in which gen-
61