4

Growing Up North: Exploring the Archaeology

of Childhood in the Thule and Dorset Cultures

of Arctic Canada

Robert W. Park

University of Waterloo

ABSTRACT

In late prehistoric times the Canadian Arctic was occupied sequentially by the peoples known to archaeologists as
the Dorset and the Thule cultures, the latter being the direct cultural and biological ancestors of the Inuit who
live there today. For archaeologists interested in exploring aspects of childhood, the Thule culture has three very
desirable characteristics: potentially magnificent preservation due to the effects of permafrost; a complex and varied
material culture; and, from their Inuit descendants, a rich and detailed body of ethnographic information that can
be drawn upon for analogy. With these advantages it is possible to identify a wide range of Thule items, especially
miniature versions of implements, specifically associated with children. These can be studied to explore the roles of
children in Thule society. Intriguingly, it is far more difficult to identify items associated with children of the Dorset
culture, despite their material culture’s being equally complex and their sites being often almost as well preserved.
This chapter summarizes the results of research into childhood among the Thule and shows how the experience of
childhood for Dorset children may have been somewhat different from that experienced by their Thule successors.

Keywords: children, Inuit, Thule, Dorset, Arctic, miniatures

T

his chapter summarizes some of the research that I have
been carrying out into childhood among the precontact

inhabitants of the Canadian Arctic. However, before getting
to that I would like first to briefly explain what brought
me to do this research. My introduction to the archaeol-
ogy of childhood came about quite by accident. In thinking
back over all the years of my training as an archaeologist I
can remember just three occasions when children were even
mentioned by my teachers or field directors. The only oc-
casion when the topic came up in the context of a course
involved looking at grave goods buried with children and
then comparing them with grave goods buried with adults
of the same society. I was taught that if some children were
buried with grave goods as rich as those found with adults,
one could conclude that they came from a society in which
high rank was inherited. The other two times I can recall
children being mentioned occurred when I was a field assis-

Archeological Papers of the American Anthropological Association, Vol. 15, pp. 53–64, ISSN 1551-823X.

C



2006 by the American Anthro-

pological Association. All rights reserved. Permissions to photocopy or reproduce article content via www.ucpress.edu/journals/rights.htm.

tant in the excavation of Iroquoian sites in southern Ontario.
On one occasion a supervisor made the casual observation
that he thought the reason so much of the pottery we found
was so thoroughly fragmented was because of children—his
idea was that large fragments of broken pots would have
been fun targets for young boys throwing stones. And fi-
nally, back in the lab during the analysis of pottery from
those sites, I was taught that some pots were identified by
the rather odd designation “juvenile” pots—they were con-
sidered to be young girls’ first attempts at manufacturing
pots. So in each of the contexts in which children were men-
tioned, children and childhood were never the focus of the
archaeologist’s interest. Instead, in the first case, children’s
graves were a means to learn about the (adult) political orga-
nization of a society; in the second case children were seen as
a site-formation process affecting the material culture pro-
duced by adults; and in the third case children were invoked

54

Robert W. Park

as a means of separating out seemingly aberrant potsherds
so that they would not confuse the stylistic and other kinds
of analyses that were being applied to the remaining “adult”
potsherds.

With such a lack of exposure to even thinking about

children as a possible topic of interest in the archaeological
record, it is hardly surprising that when I was approached a
few years ago to write and present a conference paper on the
archaeology of childhood in the Arctic, which had become
the geographic focus of my research, I initially declined. For-
tunately I eventually reconsidered and agreed to do the paper.
It was well received, so I worked up and published a version
(Park 1998). Shortly afterward I was asked to write a shorter
version for a popular archaeology magazine. I was flattered
by the request and complied (Park 1999). However, when I
received the galleys for that second article I was taken aback
to discover that the editor had inserted into my introduction
the sentence “Childhood is a magical time.” I had not written
anything like that and in subsequent bargaining with the ed-
itor the best I could manage was to get that addition altered
to “Childhood can be a magical time.”

I have recounted these anecdotes to make a point that

to me lies at the heart of any archaeology of childhood: that
editor, despite having read both the scholarly and popular
versions of my paper, implicitly but fervently believed that
we already understand exactly what childhood is all about.
Even more significantly, that editor also believed that the par-
ticular experience of childhood with which we think we are
familiar is a human universal. That inference is of course
incorrect (e.g., Aries 1962; Hirschfeld 2002:613). Ethno-
graphic research demonstrates that childhood can be defined
differently from culture to culture, and therefore the expe-
rience of being a child undoubtedly varies as well. And if
synchronic variation is possible, then so of course is varia-
tion over time. Hence there is the need for an archaeology
of childhood, in which at least one goal should be to explore
how and why childhood varied. I am not sure why my teach-
ers and field directors were not interested in childhood as
a topic of archaeological research—my guess is they either
shared that editor’s belief or they felt childhood was one of
those things beyond the reach of archaeology. Whatever their
reasons, I believe we should try to ensure that archaeologists
cannot get as far into their careers as I did without learning
that childhood is an important, interesting, and accessible
topic for archaeology!

Childhood and Arctic Prehistory

Archaeological research has shown that the first in-

habitants of Arctic Canada and Greenland, collectively re-
ferred to as Paleo-Eskimos, migrated into that region from

Alaska more than four thousand years ago (McGhee 1996;
Maxwell 1985). The adaptability and resourcefulness of
the Paleo-Eskimos in that daunting region is documented
by their survival there for approximately three millennia.
The last of the Paleo-Eskimos belonged to a culture known
as Dorset. Approximately one thousand years ago another
group of Alaskan migrants arrived in Arctic Canada: the
people known to archaeologists as Thule. There is no con-
vincing evidence for acculturation between the Dorset and
the arriving Thule and it is conceivable that the Dorset had
mostly or completely died out prior to the arrival of the Thule
(Park 1993, 2000). Whether or not the Dorset had died out,
it was the Thule who gave rise to the Inuit who inhabit the
region today—from Alaska to Greenland the diverse Inuit
groups who greeted Europeans when the latter eventually
entered those regions were the direct biological and cul-
tural descendants of the Thule (e.g., Collins 1937; Dumond
1987; McCullough 1989; McGhee 1972; Mathiassen 1927b;
Morrison 1983; Taylor 1979). Thus, from an archaeological
perspective the archaeological sites in the Canadian Arc-
tic and Greenland derive from two chronologically distinct
cultural traditions that had developed separately from one
another for at least three thousand years. Further, despite
the fact that the Dorset and Thule inhabited the same re-
gion and thus must have faced many of the same ecological
and social problems, the material culture produced by these
two cultures is both easily distinguishable and often very
well preserved as a result of the cold environment. Under
those circumstances an obvious question is, can we iden-
tify any evidence in the archaeological record of the Thule
and Dorset cultures for the kind of childhood documented in
the extraordinarily rich body of ethnographic data available
concerning the Inuit descendants of the Thule? To answer
that question this chapter will first summarize some of the
ethnographic data that might be expected to show up in or
have a bearing upon the archaeological record of childhood
among these cultures (for a more detailed review of these
ethnographic data, see Park 1998).

The Concept of Childhood

In traditional Inuit worldviews a child was not seen as an

entirely new or unique individual. A newborn infant would
be given the name of someone who had died recently and
it was believed that the child would gain, along with the
name itself, a “name spirit” that would imbue him or her
with some of the attributes of the previous owner or owners
of that name. Lee Guemple (1979, 1988) has written about
Inuit conceptions of childhood based on his research among
the Qiqiqtamiut Inuit of the Belcher Islands, but his inter-
pretations seem consistent with ethnographic observations

Childhood in the Thule and Dorset Cultures

55

throughout the Inuit world. In the Belcher Islands a child re-
ceived the spirit and its name when four days old. Therefore,

From four days after birth a child possesses its entire life
potential in capsule form; and the business of socialisa-
tion becomes one of assisting the new member (who is
really an old member) to realise the potential of his or
her pre-established identity. It follows that there are not
children in the self-formulated world of the Qiqiqtamiut,
at least not in the Western “empty vessel” sense of what
it means to be a child. [Guemple 1988:135]

This Inuit view of children and childhood is evident

in most ethnographic accounts and will become important
in evaluating the archaeological record of childhood in the
Dorset and Thule cultures.

Stages of Childhood

Early childhood in traditional Inuit society was a time

of close and constant interaction between a child and his or
her immediate family. Jean Briggs states that “babies and
small children are the affectional center of the family; their
every need is attended to with immediate solicitude, and
during most of their waking moments they are cuddled and
played with by any and all adults and older children who
may be present” (Briggs 1974:266). Describing this stage
of childhood in southern Baffin Island in the 1880s, Franz
Boas noted that

Young children are always carried in their mothers’
hoods, but when about a year and a half old they are
allowed to play on the bed, and are only carried by
their mothers when they get too mischievous. When the
mother is engaged in any hard work they are carried by
the young girls. [Boas 1888:565–566]

Children were, however, quite free to investigate their

immediate surroundings. Guemple observes that

Infants and young children are allowed to explore their
environments to the limits of their physical capabilities
and with minimal interference from adults. Thus if the
child picks up a hazardous object, parents generally leave
it to explore the dangers on its own. The child is presumed
to know what it is doing even if it is incapable of execut-
ing its designs because of physical limitations. [Guemple
1988:137]

Once children became a little older they enjoyed almost

complete freedom to go and do as they wished, especially
during the summer (Briggs 1970:165–166; Honigmann
and Honigmann 1953:43). Describing Copper Inuit life on
Victoria Island, Diamond Jenness writes that “generally
speaking, boys and girls grow up like wild plants, without
much care or attention from the time they can run about
till they approach puberty” (Jenness 1922:169–170).

Describing this stage in life among the Caribou Inuit, Kaj
Birket-Smith writes that

It is true that one does not notice Eskimo upbringing
much. No parent will ever beat his child . . . The children
are allowed to do practically as they like. Naturally they
are not angels and, from the time when they become
old enough to be naughty and until they understand that
they ought to be good—between about three and seven
years—they are sometimes bad; but on the whole they
are astonishingly obedient and well brought up. [Birket-
Smith 1929:288]

The generally “good” behavior of Inuit children

appears not to have come without considerable effort and
stress. Teasing and shaming helped eliminate inappropriate
conduct among children (Briggs 1970; Guemple 1979:44).
Conversely, socially valued accomplishments would be
publicly marked with positive comments, especially when
a child performed a task successfully for the very first time
(Guemple 1979:45).

The education of children seems to have taken

place mostly by observation (Honigmann and Honigmann
1953:39–40), entirely consistent with the Inuit conception
of children described above. Guemple notes that

Children are presumed to be socially whole and complete
from shortly after birth and so are believed to require not
so much to be taught as to be guided and directed by
adults. What Europeans would undertake to invoke in
children by “teaching” Inuit attempt to accomplish by
drawing out that which the child is believed already to
know. [Guemple 1988:134]

Elsewhere, he writes,

If the Inuit adult attitude toward children includes the
view that they are nothing less than diminutive adults
with all the potentialities of adults, this view can only be
sustained because there is no great burden of knowledge
to acquire that is not an ongoing part of the everyday
life of everyone in the community. Children do not
need to be “taught” how to perform tasks which they
can experience through observation and participation.
[Guemple 1979:50]

The relative freedom from responsibilities enjoyed by

children in Inuit society eventually came to an end. Boas
states that “when about twelve years old they begin to help
their parents, the girls sewing and preparing skins, the boys
accompanying their fathers in hunting expeditions” (Boas
1888:566). Guemple (1979:44) suggests that children start
performing these tasks considerably earlier, commencing by
the age of five or six. Briggs (1975:177) notes that there
can be considerable differences between groups in the age
at which children begin to perform the tasks they will be
responsible for as adults, but that children 10 years old can
be expected to perform chores requiring some skill.

56

Robert W. Park

Childhood Activities

Much of childhood in Inuit society was devoted to fun

and games and some activities that were just that: games.
These included various ball games, running races, hide-and-
seek, a version of tag called “Wolf and Raven” among the
Copper Inuit and “Caribou and Wolves” among the Caribou
Inuit, and even competitions to see who could keep quiet
the longest (Birket-Smith 1929:291–292; Boas 1888:570;
Jenness 1922:218–219). According to Boas, children also
enjoyed telling one another fables and singing short songs.
“Comic songs making fun of any person are great favorites,”
he notes (Boas 1888:572). Of course, most of these same
pastimes were indulged in by adults as well. One diver-
sion that seems to have been done just by children was
the use of noisemakers, one called a bull roarer and the
other known as “the buzz” (Birket-Smith 1929:289–290;
Jenness 1922:220). Spinning tops were also made (Nelson
1983 [1899]:341). A few items have been found archaeolog-
ically that have been identified as bull roarers and spinning
tops, and rounded pieces of wood are often identified as
“balls” (e.g., McCartney 1977:406), but otherwise all these
kinds of activities are likely to be invisible archaeologically.
However, another class of childhood activities is potentially
far more visible. Diamond Jenness captures this most suc-
cinctly when he writes, “One of their favourite pastimes is
to carry out, in miniature, some of the duties they will have
to perform when they grow up” (Jenness 1922:170). For my
purposes it is convenient to break down these “miniature ac-
tivities” into three broad categories: playing house, playing
with dolls, and playing at hunting.

Playing house was a common activity for all children.

Jenness states that “both boys and girls play at building
snow houses. In summer, with only pebbles to work with,
they simply lay out the ground plans, but in winter they
borrow their parents’ snow knives and make complete
houses on a miniature scale” (Jenness 1922:219). Ethno-
graphers and archaeologists have reported finding such
small house foundations constructed of pebbles from Arctic
Canada and Greenland (e.g., Birket-Smith 1929:289; David
1995:390–391; Hawkes 1916:122). Playing house also took
place within the “real” house. Jenness observes that “little
girls often have tiny lamps in the corners of their huts
over which they will cook some meat to share with their
playmates” (Jenness 1922:170).

A specific kind of play apparently limited to girls (if

the ethnographers are to be believed) was making and play-
ing with dolls. According to Birket-Smith, “dolls are the
favourite plaything of little girls” (Birket-Smith 1945:213).
Boas states that dolls “are made the same way by all tribes, a
wooden body being clothed with scraps of deerskin cut in the

same way as the clothing of men” (Boas 1888:571). Jenness
provides a little more detail concerning the manufacture of
dolls and offers a rationale for why they were made this way:
“Girls make dolls out of scraps of skin, and clothe them like
real men and women. Their mothers encourage them, for it
is in this way that they learn to sew and cut out patterns”
(Jenness 1922:219).

Children also played at hunting both land mammals and

sea mammals. Caribou were the most important land mam-
mal to most Inuit societies, and one common way to hunt
them was with bow and arrow. Jenness states that “both
boys and girls learn to stalk game by accompanying their
elders on hunting excursions” (Jenness 1922:170), and dur-
ing the summer he observed children “setting up rows of
stones and turf, injukhuit, as for a caribou drive, and digging
shallow pits, tallut, from which they launched their shafts at
imaginary deer” (Jenness 1922:219). Sea mammals, includ-
ing seals, walrus, and various species of whales, were also
of great importance to most Inuit societies. Birket-Smith
notes that “boys sometimes have a miniature harpoon head,
ikiort´ınguaq, according to what I have been told used in the
manner that the boy thrusts it into the seal killed and helps
to pull it ashore” (Birket-Smith 1924:420). Boas also pro-
vides a detailed description of play hunting, specifically the
hunting of ringed seals through their winter breathing holes:

Boys play hunting seals. Each of them has a small har-
poon and a number of pieces of seal-skin with many
holes. Each piece of skin represents a seal. Each of the
boys also has a hip-bone of the seal. Then one boy moves
the piece of skin which represents a seal under the hole
in the hip-bone, which latter represents the blowing-hole
in the ice. While moving the piece of skin about under
the bone, the boys blow like seals. Whoever catches with
the little harpoon the piece of skin in one of the holes
retains it, and the boy who catches the last of the pieces
of skin goes on in turn with his seals. The little harpoons
are made by the fathers of the boys, the pieces of skin are
prepared by the mothers. [Boas 1901:111]

In light of the archaeological data outlined below, it may

be significant that the seal-hunting game described by Boas
appears to have been limited to boys whereas Jenness says
that both boys and girls participated in the caribou-hunting
game.

Miniature Material Culture

What is so interesting and promising archaeologically

about these kinds of childhood activities—playing house,
playing with dolls, and playing at hunting—is that they
are clearly associated with an extensive miniature material
culture. In addition to those already discussed, miniature

Childhood in the Thule and Dorset Cultures

57

versions of the following items are specifically mentioned
in ethnographic accounts as having been used as toys:
sledges, kayaks, umiaks, cooking pots, snow knives, and
sleeping platform mattresses (Birket-Smith 1945:214; Boas
1888:571). While miniature items of all these types and more
have been found archaeologically at sites of the Thule cul-
ture, unfortunately we cannot automatically assume that all
these miniatures were toys and thus associated with chil-
dren. On the basis of ethnographic accounts, miniature ver-
sions of full-sized items were made or used in two other dis-
tinct contexts: as grave offerings and as the paraphernalia of
shamans.

According to Knud Rasmussen, among the Utkuhikhal-

ingmiut, “a dead man’s property usually descends to his rel-
atives, if it is worth anything. Instead of his own possessions
they put into the grave small copies of the things they have
inherited, these miniatures being carved in wood” (Knud
Rasmussen 1931:507). Among the Copper Inuit he notes that

if among the deceased’s weapons there are especially
valuable objects they may be substituted by less costly
ones. Some implements may also be replaced by minia-
ture imitations carved in wood. . . . Real articles such as
ulos or cooking pots are never laid by a woman’s grave,
miniatures being used instead. [Rasmussen 1932:46]

For the Netsilik he notes that

it is true that at the “funeral” itself no particular respect
is shown the body, and yet the patch of soil on which it
is laid seems to be looked upon as a kind of holy spot, to
which they make offerings of meat or miniature imple-
ments such as weapons, kayaks and the like. [Rasmussen
1931:263]

Thus, from an archaeological perspective, miniature

items associated with burials are likely not toys, although
Boas (1888:613) does state that a child’s toys would be buried
with him or her, and both Oswalt (1952:71) and Birket-Smith
(1924:419) report children’s graves containing dolls, pre-
sumably the possessions of the children buried there.

Miniature items were also included among the parapher-

nalia used by shamans (e.g., Burch 1988:102). According to
Rasmussen, among the Iglulingmiut:

As soon as a young man has become a shaman, he must
have a special shaman’s belt as a sign of his dignity. This
consists of a strip of hide to which are attached many
fringes of caribou skin, and these are fastened on by all
the people he knows, as many as he can get; to the fringes
are added small carvings, human figures made of bone,
fishes, harpoons; all these must be gifts, and the givers
then believe that the shaman’s helping spirits will always
be able to recognize them by their gifts, and will never
do them any harm. [Rasmussen 1929:114]

To see an example of such a shaman’s belt, see Balikci

(1984:Fig. 18).

While shamans’ miniatures certainly might be found

in houses and in midden deposits, it seems likely that they
should not be particularly common nor found in every house
since likely only a few people in a community would have
been shamans. Further, in light of what was said above about
grave offerings, it seems plausible that those miniatures used
by shamans would have been buried with them or they would
have been inherited and used by other shamans—that is, in
comparison with children’s toys they might be expected to
be more subject to curation. Thus, in light of all the available
ethnographic data, it is plausible to assume that most of
the miniature implements and human figurines (i.e., dolls)
found in contexts other than graves were toys and thus were
associated with children.

Miniatures in Arctic Archaeology

In a 1927 monograph Therkel Mathiassen pioneered the

use of small size as a criterion for identifying toys among
Thule culture archaeological finds. In that report he de-
scribed together full-sized and miniature implements of the
same type but he also included a separate heading for toys,
beneath which he wrote: “In the foregoing, when describing
the various types of implements, various miniature objects
have already been mentioned as having presumably been
used as toys: harpoon heads, arrow shafts, baleen bow, snow
knife and lamps” (Mathiassen 1927a:75). Most subsequent
site reports for Thule sites have similarly identified miniature
implements and human figurines as toys.

In a paper published in 1998 (Park 1998) I attempted

to use the data reported in all those site reports, plus data
from my own excavations, in order to learn about child-
hood in the Thule culture. In that paper I argued that these
data might be used to learn about the kinds of activities in
which Thule children participated and how much they par-
ticipated in them. Further, if children performed in minia-
ture the tasks of adults, then it might be reasonable to ex-
pect some sort of correspondence between the material cul-
ture associated with the activities carried out by adults and
the material culture associated with the activities engaged
in by children. To evaluate that simple hypothesis I as-
sembled the data presented in Table 4.1 on the abundance
of various full-sized and miniature artifacts from 31 sites
(adapted from Park 1998:Table 1). It presents information
on 9,753 artifacts, including 369 miniatures. All were ex-
cavated from winter house ruins; none came from graves.
For each functional type (e.g., “harpoon head”) the table
contrasts the frequency of miniatures with the frequency of

58

Robert W. Park

Table 4.1. Quantities and relative abundances of miniature and full-sized artifact types from 30 Thule culture sites in Canada and
Greenland

Full-sized

Miniature

n

% of full-sized

n

% of miniatures

Harpooning
Harpoon head

474

5.1

16

4.3

Harpoon

13

0.1

10

2.7

Dart head

13

0.1

5

1.4

Harpoon ice pick

130

1.4

2

0.5

Harpoon socket piece

100

1.1

2

0.5

7.8%

9.5%

Archery
Bow

96

1.0

29

7.9

Arrow

4

0.0

19

5.1

Arrowhead

330

3.5

9

2.4

Arrow shaft

209

2.2

3

0.8

6.8%

16.3%

Fishing
Fish spear side prong

134

1.4

4

1.1

Fish spear

25

0.3

3

0.8

1.7%

1.9%

Transportation
Sled slat

55

0.6

17

4.6

Kayak

39

0.4

12

3.3

Paddle

7

0.1

10

2.7

Boat misc.

0

0.0

6

1.6

Whip handle

8

0.1

5

1.4

Umiak

10

0.1

4

1.1

Sled runner

43

0.5

3

0.8

1.7%

15.4%

Other men’s activities
Man’s knife

426

4.5

22

6.0

Sling handle

45

0.5

11

3.0

Throwing board

8

0.1

10

2.7

Lance

139

1.5

7

1.9

Snow knife

210

2.2

5

1.4

Bird dart

36

0.4

2

0.5

9.2%

15.4%

Women’s and household activities
Lamp

160

1.7

23

6.2

Cooking pot

465

5.0

13

3.5

Ulu (woman’s knife)

262

2.8

5

1.4

Dish/bowl/bucket

106

1.1

5

1.4

Snow beater

67

0.7

2

0.5

Snow shovel

25

0.3

1

0.3

Platform mattress

21

0.2

1

0.3

Needle case

7

0.1

1

0.3

11.9%

13.8%

Other
Doll/figurine

N/A

99

26.8

Drum

13

0.1

1

0.3

Parka

8

0.1

1

0.3

Miscellaneous

5,696

60.7

1

0.3

60.9%

27.6%

9,384

100.0

369

100.0

Childhood in the Thule and Dorset Cultures

59

full-sized examples. The relative abundance of each minia-
ture type is then expressed as a percentage of all the miniature
artifacts; similarly, the relative abundance of each full-sized
type is expressed as a percentage of all the full-sized arti-
facts. For example, 16 miniature harpoon heads were found
at the 31 sites, forming 4.3 percent of all the miniatures
found, whereas 474 full-sized harpoon heads were recov-
ered, forming 5.1 percent of all the full-sized artifacts that
were excavated. In this case the relative abundances of minia-
ture and full-sized harpoon heads are very similar, but the
use of the frequencies of individual functional types (such
as harpoon heads) has some disadvantages. It has long been
recognized that comparisons of the frequencies of individ-
ual Thule artifact types are complicated by the very large
number of types:

The statistical techniques of artifact comparison devel-
oped for use in other archaeological areas are based on sit-
uations in which one recovers large numbers of artifacts
which can be grouped into a few classes. In the Thule sit-
uation, we have a large number of functional or stylistic
classes, with generally very small numbers of artifacts
in each class. Any attempt to measure statistical resem-
blance between such assemblages is highly influenced by
sampling error, and the resulting correlations are proba-
bly not very useful. [Taylor and McGhee 1979:115]

Therefore, a consideration for the research was identi-

fying artifact categories that could reveal information about
childhood activities and that would contain large enough
numbers of items that sampling error could be reduced to an
acceptable level.

The approach adopted was to base comparisons on

larger functional classes that better reflect activity sets based
on ethnographic information. Seven broad classes were cho-
sen, of which three deal with distinctive techniques of ob-
taining food: “Harpooning,” “Archery,” and “Fishing.” A
“Transportation” class is followed by two classes containing
diverse items whose various uses were conventionally (but
by no means universally) divided by gender: “Other men’s
activities” and “Women’s and household activities.” Lastly,
an “Other” class contains dolls as well as a large number of
full-sized types for which miniature versions have not been
found in the archaeological record; these are included un-
der the “Miscellaneous” category. Dolls are included in this
class because they presumably were not used by adult women
and because there is no corresponding full-sized item. Their
apparent importance as toys is, however, shown by the fact
that they represent over one-quarter of all miniatures. De-
spite the large percentage of full-sized items falling into the
“Miscellaneous” category (i.e., over 60 percent), they were
included in the calculations in order not to artificially skew
the results—in other words, to compare all full-sized items

with all miniature items. However, the basic proportions re-
main similar if the entire “Other” category is excluded.

In all but the “Other” class, miniatures are at least

slightly more abundant than their full-sized counterparts,
but in three classes (Harpooning, Fishing, and Women’s and
household activities) the differences are less than two per-
cent and therefore presumably insignificant. For example,
miniature harpoon parts form 9.5 percent of the miniature as-
semblage whereas full-sized harpoon parts form 7.8 percent
of the full-sized assemblage. Given all the possible com-
plicating factors, this is remarkably close and completely
consistent with children playing at their future adult roles.
However, the disparity between the frequency of full-sized
and miniature items associated with the Archery class is
much more substantial. This class forms just 6.8 percent
of the full-sized assemblage yet fully 16.3 percent of the
miniature assemblage. It is very tempting at this point to
return to Jenness’s (1922:170) statement that both boys and
girls were given bows and arrows and taught to use them by
their fathers. Archery appears to have been overwhelmingly
a male activity among adults so the disparity that we see be-
tween the full-sized and miniature assemblages may reflect
the fact that, among adults, only men owned and used bows
whereas both boys and girls had bows. Further, if I am not
reading too much into quotes by Birket-Smith (1924:420)
and Boas (1901:111) describing boys playing at seal hunt-
ing, only boys engaged in this activity, so the similarity in
proportions of full-sized and miniature harpoon parts would
likewise be consistent with the ethnographic record.

Turning to the “Other men’s activities” class, miniatures

are somewhat more abundant than full-sized items. Much or
all of this disparity may be attributable to two hunting im-
plements: sling handles and throwing boards. These too are
implements that, like bows and arrows, girls accompany-
ing their fathers on hunting expeditions might have used,
although no specific references to this were found in the
ethnographic literature. The Transportation class exhibits
the largest disparity, forming more than 15 percent of the
miniatures and less than two percent of the full-sized assem-
blage. A probable explanation for the overrepresentation of
miniatures in this class compared with their full-sized equiv-
alents is that some of the full-sized items, such as boat parts
and sled parts, are considerably larger than all the other “full-
sized” types used in this analysis and may simply be too large
to be expected in the winter house ruins from which these
data were drawn.

The first five classes in Table 4.1 are associated with

activities that conventionally were part of the role of men
in Inuit society, so it may appear that “boys’ toys” are over-
represented in the entire miniature assemblage. However, if
dolls are added to the “Women’s and household activities”

60

Robert W. Park

miniatures, then miniatures specifically associated with
girls’ activities represent over 40 percent of all miniatures.
And having demonstrated that girls appear to have used
some of the hunting equipment—at least bows and arrows
and perhaps sling handles and throwing boards—it appears
that miniature material culture may have been approximately
evenly distributed between boys and girls.

These archaeological data, presented in greater detail

in my previous publication (Park 1998), appear to docu-
ment the wide range of the miniature material culture used
by Thule children, as well as its patterned relationship to
the “full-sized” material culture of Thule adults. They also
suggest that the Inuit practice of treating children “simply
as small adults” (Guemple 1988:137) was characteristic of
their Thule ancestors as well and that Thule children did in-
deed carry out, in miniature, the tasks they would come to
perform when they grew up. Thus, even though Bugarin (this
volume) rightly questions the automatic assumption some-
times made by researchers that child behavior can be taken
as a miniature model of adult behavior, in the case of Inuit
and Thule culture the ethnographic and archaeological data
support such an interpretation.

Size Counts More Than Ever

At the conclusion of my research there were two di-

rections in which I hoped eventually to proceed. The first
was based on the fundamental dichotomy that I had drawn
between “miniature” and “full-sized” implements. In em-
ploying that simple distinction I had been constrained by the
fact that many of the site reports from which I had assembled
my data had employed that distinction. Far fewer than half of
the 9,753 artifacts in my database were actually illustrated or
described with measurements in the site reports. Therefore,
I was forced to accept all those different scholars’ catego-
rizations of artifacts as either functional (i.e., full-sized) or
“miniature”—I could not develop my own independent cri-
teria to do that. However, I was aware that the concepts of
“miniature” and “toy” probably encompassed several differ-
ent functionalities. The very smallest “miniatures” are likely
playthings that “functioned” only in a child’s imagination.
Some of them would probably have been used with dolls
while playing house and so forth. But many larger “minia-
tures” were undoubtedly quite functional in an important
fashion. These would be items simply scaled down in size
enough so that children could use them. For example, the
miniature harpoons whose use is described in the quotation
by Boas appear to have been functional—that is, they worked
for harpooning things. What made them toys was the fact that
they were used in a game to harpoon small perforated bits

of sealskin. Similarly, the bows and arrows described in one
of the Jenness quotes were toys used by children to shoot
at “imaginary deer” (Jenness 1922:219), but clearly those
toys “functioned” enough for the arrows to be launched.
And finally, as also described by Jenness (1922:170), chil-
dren’s “fathers make bows and arrows for them suited to
their strength.” Those bows and arrows, while smaller than
the ones used by adults, clearly cannot be described as toys
although they were associated with children.

Thus my simple miniature/full-sized dichotomy was al-

most certainly obscuring a great deal of interesting vari-
ability in the material culture associated with Thule chil-
dren, variability that had the potential of revealing much
more about the various stages of childhood outlined above.
Clearly, what needed to be done was to examine and measure
all the miniature artifacts and all the full-sized artifacts of the
same types in order to be able to document and quantify that
variability. Doing so would be an ambitious project, how-
ever, because the actual artifact assemblages are dispersed
in several museums across Canada, the United States, and
in Europe.

Dorset Childhood

But there was another avenue of this research that I also

wished to pursue: applying the same approach to the mate-
rial culture of the Dorset people. They also produced minia-
ture versions of items, but Dorset assemblages contain pro-
portionately many fewer miniaturized implements with the
seeming exception of one artifact type: miniature harpoon
heads. The rest of their miniatures include a much wider
range of carvings of animals and humans than is seen in
Thule. The Dorset miniatures have for many years been inter-
preted predominantly as the paraphernalia of shamans (e.g.,
McGhee 1976, 1987; Sutherland 1993:322; Ta¸con 1983;
Taylor and Swinton 1967). Although intuitively I find that
conclusion compelling for many of these items, it raises a
question that is very pertinent in the present context: where
are the toys? Did Dorset children have a similar and equally
extensive miniature material culture that is simply invisible
within assemblages containing the many miniatures used by
shamans, or did Dorset children have different kinds of toys
or no toys at all? If they had different kinds of toys or no toys
at all, then the experience of childhood for Dorset children
must have been somewhat different from that of their Thule
successors.

Because I was already carrying out research into other

aspects of Dorset culture I decided to pursue the avenue
of childhood in that culture rather than attempt to track
down and visit all the Thule assemblages in their dispersed

Childhood in the Thule and Dorset Cultures

61

Figure 4.1. Histogram of harpoon head length of 357 harpoon
heads from Dorset culture sites.

repositories. One of my students at the time, Pauline
Mousseau, had become fascinated with the miniature har-
poon heads we had excavated at two Dorset sites so we con-
centrated on just that artifact type (Park and Mousseau 2003).
We assembled measurements on as many complete or es-
sentially complete full-sized and miniature Dorset harpoon
heads as we could from published site reports and from some
unpublished data: 357 harpoon heads in all. Minimally, we
hoped that doing so would allow us to determine whether the
harpoons heads classified in the literature as “miniature” and
the ones classified (implicitly) as “full-sized” really could be
separated in a nonarbitrary fashion into those different cat-
egories. We attempted this based on the assumption that if
full-sized harpoon heads had to be at least some minimum
size in order to function successfully, this might manifest
itself on a histogram of harpoon head length as a bimodal
distribution. If so, such a finding would allow us to focus on
the miniatures and attempt to determine whether some were
toys. Then we could proceed to do for those items what I had
not been able to do for the Thule miniatures: see whether the
different sizes of “toys” could be associated with different
kinds of childhood activities.

Figure 4.1 presents a histogram of the lengths of all 357

harpoon heads used in our study. Alas, it displays no apparent
bimodal distribution and, if our sample were not somewhat
biased toward harpoon heads classified by their excavators
as “miniature” (Park and Mousseau 2003:262), the his-
togram might conform even more closely to a normal curve.
Dorset harpoon heads are subdivided by archaeologists into
a number of different types and, on the rationale that differ-
ent types might have different functions and therefore sizes
and that analyzing them all together as in Figure 4.1 might
be obscuring size differences within types, we performed
the same kind of analysis several times while restricting
ourselves to harpoon heads of a single type. Again, however,
we found very little evidence of bimodal distribution in
harpoon head length (Park and Mousseau 2003:265–266).

We next explored a factor that we termed potential func-

tionality, that is, a harpoon head’s ability to function as part
of a complete harpoon in that it can be mounted on a fore-
shaft, it can pierce the skin of an animal, and it can secure the
harpooned creature by means of the harpoon line. Leaving
aside all question of size, a potentially functional harpoon
head must exhibit three basic attributes: it must have a line-
hole or some other means of affixing the harpoon line; it
must be self-bladed (i.e., have a sharp tip) or have an end-
blade slot into which a sharp end-blade can be inserted; and
it must have either a closed or an open socket as a means of
being affixed onto the harpoon foreshaft. By these criteria,
several of the “miniature” harpoon heads in our sample were
not potentially functional because they possessed an inad-
equate socket or lacked one completely and thus could not
have been affixed onto a harpoon foreshaft. But the socket-
less examples we could examine were otherwise complete
and often finely manufactured so there is no reason to infer
that they were unfinished; we therefore concluded that they
were never intended to be affixed onto a harpoon. These are
the only harpoon heads for which it seems possible to as-
sert with confidence that they are either toys or shamans’
paraphernalia. If they are toys then they were used in imag-
inary activities—without a functioning socket they could
not have been used in games such as the one described by
Boas. However, all the rest of the very small harpoon heads
could have been used as part of small but complete har-
poons, and several harpoon foreshafts small enough to be
used with the smallest harpoon heads have been found. The
fact that there is no clear size distinction between these small
but potentially functional harpoon heads and the “full-sized”
and therefore functional harpoon heads (i.e., harpoon heads
that would have been used to hunt seals or walruses) means
that it is presently impossible to isolate with any confidence
a subset of the Dorset harpoon heads that would contain only
children’s implements or shaman’s paraphernalia, let alone
identify a subset made up solely of items associated with
children.

This was a somewhat disappointing finding in the

present context, especially since the smooth transition in size
from the smallest to the largest Dorset harpoon heads might
be consistent with their having been manufactured in increas-
ing sizes as children grew up. But that is by no means the
only possible explanation for the size of these implements.
For functional reasons the smallest harpoon heads would not
have been used to hunt seals or larger sea mammals—they
are simply too small to permit a strong enough harpoon line
to be attached or too fragile to stand the force of restraining
a large animal without the harpoon head breaking. But there
is no reason such smaller harpoon heads could not have been
used by adults to harpoon smaller creatures. The existence

62

Robert W. Park

of robust foreshafts with small tangs indicates that some of
these very small harpoon heads could well have been used
with adequately sized harpoons to hunt small creatures, es-
pecially fish but perhaps birds as well.

So with respect to childhood-associated material cul-

ture, the Dorset remain as enigmatic as before, at least in
comparison with the Thule. There is actually one more com-
plicating factor that should be mentioned. Dorset sites are
older than Thule sites and, because of that and some archi-
tectural differences between the houses of the two cultures,
artifact preservation at Dorset sites is often less complete
than at Thule sites (although there are a number of excep-
tions to this rule). This is most significant for items man-
ufactured of wood because some of the Thule miniatures
most convincingly identified as toys, including miniature
harpoon heads, are manufactured of that material. Almost
all the small Dorset harpoon heads discussed here are man-
ufactured of harder materials: bone, antler, or ivory. If there
were substantial numbers of Dorset miniature harpoon heads
made of wood, few of them have survived. But that, of course,
introduces another kind of data that should be systemati-
cally recorded for the Thule miniatures—material—to see if
there are consistent differences in the material from which
full-sized and miniature implements of the same type were
manufactured.

Discussion

The archaeological data from the Arctic with which I

am working exhibit a number of obvious advantages for ar-
chaeological research into childhood. The combination of
excellent preservation along with a technologically complex
material culture and a rich and detailed body of ethnographic
information allows for kinds of analyses that are often go-
ing to be far more difficult to apply to archaeological data
from some other prehistoric contexts. It is possible to make
a strong case for interpretations concerning childhood in the
Thule culture precisely because they are consistent with pre-
dictions based on the ethnography of their Inuit descendants.
Conversely, our present understanding of childhood and its
material culture in Dorset is poor at least in part because
we do not have comparable ethnographic data. It may be
that childhood in the Dorset culture was very similar to that
in Thule, at least with respect to its material culture, and
the similarity is simply obscured because of miniature items
of shaman’s paraphernalia being more prevalent in Dorset
culture. Conversely, it may be that childhood was different
in Dorset culture in the sense that Dorset children did not
spend as much time playing or practicing with miniature im-
plements. Perhaps they played with the much wider range of

small carvings known from Dorset, including animals and
many fabulous creatures. Or perhaps Dorset children did not
have a material culture of their own at all.

The research presented here is obviously incomplete. It

has only scratched the surface of the potential of these data
for learning about the experience of childhood in these cul-
tures, but it does suggest a number of avenues of research
that have great potential. I definitely intend to continue ex-
ploring these questions, but there is a lot of work that needs
to be done. For example, even with all the advantages I have
listed for these Arctic archaeological data, it seems clear
that one cannot a priori identify the subset of a society’s
entire material culture that would have been associated with
children—that can only be done by exploring the entire range
of material culture and identifying culture-specific criteria
that would allow one to do so. I have begun exploring size
and potential functionality for these Dorset and Thule cul-
ture artifacts but there are obviously more criteria that could
be investigated. It is my hope that readers of this volume will
be inspired to share some of that work, adding new questions
and new approaches to archaeological research into child-
hood, including in the Arctic.

Acknowledgments

I would like to thank Patricia Smith and William Fox

for first encouraging me to think about the archaeology of
childhood in the Arctic by inviting me to present a paper
on that topic back in 1997. Similarly, I would like to thank
Jane Baxter, organizer of the 2001 American Anthropologi-
cal Association session in which I presented a portion of this
paper, for convincing me that I had something more to say
on the topic.

References

Aries, Philippe

1962 Centuries of Childhood. New York: Vintage Books.

Balikci, Asen

1984 Netsilik. In Arctic. David Damas, ed. Pp. 415–

430. Handbook of North American Indians, vol. 5,
William C. Sturtevant, general editor. Washington,
DC: Smithsonian Institution.

Birket-Smith, Kaj

1924 Ethnography of the Egedesminde District with As-

pects of the General Culture of West Greenland.
Meddelelser om Grønland, vol. 66.

1929 The Caribou Eskimos: Material and Social Life and

Their Cultural Position: Descriptive Part. Report

Childhood in the Thule and Dorset Cultures

63

of the Fifth Thule Expedition 1921–24, vol. 5(1).
Copenhagen: Gyldendalske Boghandel, Nordisk
Forlag.

1945 Ethnographical Collections from the Northwest

Passage. Report of the Fifth Thule Expedition
1921–24, vol. 6(2). Copenhagen: Gyldendalske
Boghandel, Nordisk Forlag.

Boas, Franz

1888 The Central Eskimo. Annual Report of the Bureau

of American Ethnology 6:399–669.

1901 The Eskimo of Baffin Land and Hudson Bay. Bul-

letin of the American Museum of Natural History
15, Part 1:1–370.

Briggs, Jean L.

1970 Never in Anger: Portrait of an Eskimo Family.

Cambridge: Harvard University Press.

1974 Eskimo Women: Makers of Men. In Many

Sisters: Women in Cross-Cultural Perspective.
Carolyn J. Matthiasson, ed. Pp. 261–304. New
York: Free Press.

1975 The Origins of Nonviolence: Aggression in Two

Canadian Eskimo Groups. In The Psychoanalytic
Study of Society, vol. 6. Warner Muensterberger
and Aaron H. Esman, eds. Pp. 134–203. New York:
International Universities Press.

Burch, Ernest S., Jr.

1988 The Eskimos. Norman: University of Oklahoma

Press.

Collins, Henry B.

1937 Archaeology of St. Lawrence Island, Alaska.

Smithsonian Miscellaneous Collections, vol. 96.
Washington, DC.

David, Robert G.

1995 A Survey of Eskimo Sites from Mesters Vig

to Kuhn Ø, Northeast Greenland. Polar Record
31(179):389–398.

Dumond, Don E.

1987 The Eskimos and Aleuts. 2nd edition. London:

Thames and Hudson.

Guemple, Lee

1979 Inuit Socialization: A Study of Children as Social

Actors in an Eskimo Community. In Childhood and
Adolescence in Canada. Karigoudar Ishwaran, ed.
Pp. 39–53. Toronto: McGraw-Hill Ryerson.

1988 Teaching Social Relations to Inuit Children. In

Hunters and Gatherers 2—Property, Power and Ide-
ology. Tim Ingold, David Riches, and James Wood-
burn, eds. Pp. 131–149. Oxford: Berg.

Hawkes, Ernest W.

1916 The Labrador Eskimo. Geological Survey Mem-

oir 19. Anthropological Series No. 14. Ottawa:
Canadian Department of Mines.

Hirschfeld, Lawrence A.

2002 Why Don’t Anthropologists Like Children?

American Anthropologist 104:611–627.

Honigmann, Irma, and John Honigmann

1953 Child Rearing Patterns among the Great Whale

River Eskimo. Anthropological Papers of the Uni-
versity of Alaska 2(1):31–50.

Jenness, Diamond

1922 The Life of the Copper Eskimos. Report of the

Canadian Arctic Expedition, 1913–18, vol. 12(A).
Ottawa.

McCartney, Allen P.

1977 Thule Eskimo Prehistory along Northwestern

Hudson Bay. Mercury Series Paper No. 70. Ottawa:
Canadian Museum of Civilization, Archaeological
Survey of Canada.

McCullough, Karen M.

1989 The Ruin Islanders: Early Thule Culture Pioneers

in the Eastern High Arctic. Mercury Series Paper
No. 141. Ottawa: Canadian Museum of Civiliza-
tion, Archaeological Survey of Canada.

McGhee, Robert

1972 Copper Eskimo Prehistory. Publications in Archae-

ology No. 2. Ottawa: National Museum of Man.

1976 Differential Artistic Productivity in the Es-

kimo Cultural Tradition. Current Anthropology
17(2):203–220.

1987 Prehistoric Arctic Peoples and Their Art. American

Review of Canadian Studies 17(1):5–14.

1996 Ancient People of the Arctic. Vancouver: UBC

Press.

Mathiassen, Therkel

1927a Archaeology of the Central Eskimos: Descriptive

Part. Report of the Fifth Thule Expedition 1921–24,
vol. 4(1). Copenhagen: Gyldendalske Boghandel,
Nordisk Forlag.

64

Robert W. Park

1927b Archaeology of the Central Eskimos: The Thule

Culture and Its Position within the Eskimo Cul-
ture. Report of the Fifth Thule Expedition 1921–24,
vol. 4(2). Copenhagen: Gyldendalske Boghandel,
Nordisk Forlag.

Maxwell, Moreau S.

1985 Prehistory of the Eastern Arctic. Orlando: Aca-

demic Press.

Morrison, David A.

1983 Thule Culture in Western Coronation Gulf, N.W.T.

Mercury Series Paper No. 116. Ottawa: Canadian
Museum of Civilization, Archaeological Survey of
Canada.

Nelson, Edward William

1983 [1899]

The

Eskimo

about

Bering

Strait.

Washington, DC: Smithsonian Institution Press.

Oswalt, Wendell H.

1952 The Archaeology of Hooper Bay Village, Alaska.

Anthropological Papers of the University of Alaska
1(1):47–91.

Park, Robert W.

1993 The Dorset-Thule Succession in Arctic North

America: Assessing Claims for Culture Contact.
American Antiquity 58(2):203–234.

1998 Size Counts: The Miniature Archaeology of

Childhood in Inuit Societies. Antiquity 72:269–
281.

1999 The Archaeology of Childhood: Toys Tell Tales of

Growing Up in the Arctic. Discovering Archaeol-
ogy 1(2):81–88.

2000 The Dorset-Thule Succession Revisited. In Iden-

tities and Cultural Contacts in the Arctic. Martin
Appelt, Joel Berglund, and Hans Christian Gulløv,
eds. Pp. 192–205. Danish Polar Center Publication
No. 8. Copenhagen: The Danish National Museum
& Danish Polar Center.

Park, Robert W., and Pauline M. Mousseau

2003 How Small Is Too Small? Dorset Culture “Minia-

ture” Harpoon Heads. Canadian Journal of Archae-
ology 27(2):258–272.

Rasmussen, Knud

1929 Intellectual Culture of the Iglulik Eskimos. Report

of the Fifth Thule Expedition 1921–24, vol. 7(1).
Copenhagen: Gyldendalske Boghandel, Nordisk
Forlag.

1931 The Netsilik Eskimos: Social Life and Spiritual

Culture. Report of the Fifth Thule Expedition
1921–24, vol. 8(1–2). Copenhagen: Gyldendalske
Boghandel, Nordisk Forlag.

1932 Intellectual Culture of the Copper Eskimos. Re-

port of the Fifth Thule Expedition 1921–24, vol. 9.
Copenhagen: Gyldendalske Boghandel, Nordisk
Forlag.

Sutherland, Patricia D.

1993 The History of Inuit Culture. In In the Shadow of

the Sun: Perspectives on Contemporary Native Art.
Pp. 313–332. Mercury Series Paper No. 124. Hull:
Canadian Museum of Civilization, Canadian Eth-
nology Service.

Ta¸con, Paul S. C.

1983 An Analysis of Dorset Art in Relation to Prehistoric

Culture Stress. ´Etudes/Inuit/Studies 7(1):41–65.

Taylor, William E., Jr.

1979 Preface. In Thule Eskimo Culture: An Anthro-

pological Retrospective. Allen P. McCartney, ed.
Pp. iv–v. Mercury Series Paper No. 88. Ottawa:
Canadian Museum of Civilization, Archaeological
Survey of Canada.

Taylor, William E., Jr., and Robert McGhee

1979 Archaeological Material from Creswell Bay,

N.W.T., Canada. Mercury Series Paper No. 85.
Ottawa: Canadian Museum of Civilization, Archae-
ological Survey of Canada.

Taylor, William E., Jr., and George Swinton

1967 Prehistoric Dorset Art. The Beaver Outfit 298:32–

47.