And a Little Child Shall Lead Them

Posted on April 21, 2013


I have to admit that I was a little mystified when Archaeology  online recently publicized research that used errors in working a single stone tool to propose that its less-skilled maker was probably a child.

Not that I think the idea is implausible.

Errors in production have been a staple of the archaeology of childhood for a long time now, with studies of both stone tools and pottery pointing out that learning implies developing skill, so that learners should be detected through products made with less skill.

Over a decade ago, Kathryn Kamp (in 2001) and Patricia Crown (in 1999) proposed that children learning to make pottery would have produced characteristic products that showed less skill and/or more error. They both drew on general theories of cognitive development, stressing that children would only be able to carry out particular steps in production, or indeed, conceive of a production sequence, at specific ages. Crown’s study assessed the likely age of the maker by examining how designs were drawn, relying on cognitive studies showing children develop skills for different steps in drawing at different ages.

Nor have such analyses been limited to pottery.

In 2006, John Shea published “Child’s Play: Reflections on the Invisibility of Children in the Paleolithic Record” in Evolutionary Anthropology. In this article, Shea provided an exceptionally clear model of what we might expect the evidence of children learning to work stone tools would look like.

Among other things, he noted that “learners are profligate knappers”.  In his experience teaching students, they produced twice the debris that he, as a more accomplished crafter, did. This led him to observe that a large proportion of ancient lithic assemblages might actually testify to the presence of beginners:

learners’ knapping byproducts could actually outnumber those of competent adult knappers in some assemblages.

Shea also observed that learners produced the most variable assemblages; he suggested that “stone tools made or used by children are likely to be relatively small, to fit the small hands of their makers or users”; and that learners would likely have been restricted to local, easily acquired, low cost materials.

Many of Shea’s suggestions are exemplified in a lovely article published in 2008 in the Journal of Archaeological Method and Theory. Called “Playing With Flint: Tracing a Child’s Imitation of Adult Work in a Lithic Assemblage”, the article presents a case study from a Neolithic site in Sweden, based on theories of learning as a social activity.

The author, Anders Högberg, detects spatial evidence of a single highly skilled craft worker, producing a highly stylized tool type, surrounded by debris from lithic working that was significantly more varied, “the unsystematic production of flakes”. The skilled producer used high quality raw material, while the “unsystematic” work was on low quality material. His conclusion is that a skilled adult worked on a formal tool, observed and imitated by a child who moved around more, and who, while able to produce worked flakes, did not do so to the systematic ends of the worker being imitated.

So what surprised me wasn’t that an archaeologist had proposed that a child had made a less-than-perfect stone tool; it was that Archaeology apparently thought this was news, when in the research world I inhabit, it is decidedly not.

My reaction to studies like this is somewhat different. Since we have long used evidence of less skilled craft production as a way to infer the presence of children, we have also faced the fact that the equation is neither simple nor automatic.

We debate whether we should assign the role of learner automatically, or primarily, to children. One of the things I like about Kathryn Kamp’s decade-old article on Sinagua pottery making is that, while she reviews the literature on children’s motor skills and cognitive capacities, when it comes to talking about the archaeological evidence, she asks how we can see “beginners”. Not children– beginners.

Then there is the problem of identifying error. Kamp provides cautious guidelines, noting that while beginners probably often do make errors and produce “imperfect products”, there are reasons why an experienced crafter might produce a less-than-perfect product. This point is echoed by others.

What makes the study by Anders Högberg so useful is that it mobilizes a number of different kinds of evidence within a framework that acknowledges all these issues, to conclude convincingly that the less systematic crafting at the site was likely done by a child.

His use of the word “unsystematic” avoids a problem in interpreting some craft products as showing “errors”. It is only when he considers whether this could be an adult that the word “error” even comes up:

The trials and errors of a skilled toolmaker would exhibit flake material which revealed a purpose of making an artefact recognisable as a typological form.

This makes the definition of a skilled craft worker clearer: someone who works with a purpose of making something systematically is fully embedded in a craft.

Högberg then is able to contrast children– not just beginning learners– with skilled adults by emphasizing play as characteristic of the cognitive approach of a child: unsystematic by definition, even though based on imitation.

I do not want to assume that the Archaeology magazine news item fully represents the study it mentions. So I do not conclude that the author, Sigrid Alræk Dugstad of Stavanger University, is uninformed by this preceding literature.

Indeed, much of the leadership in the archaeology of childhood came from Scandinavian archaeologists following in the footsteps of Grete Lillehammer, who published an important article on archaeology of childhood in 1989. Högberg cites this tradition of research in his article, and Lillehammer herself wrote the introduction for a collection of papers, where Dugstad apparently made the argument that is now getting attention.

This edited volume, published in 2010 by the archaeology museum at the University of Stavanger in Norway, is called Socialisation: Recent Research on Childhood and Children in the Past. It was based on a 2008 conference of the Society for the Study of Childhood in the Past, which also publishes an entire journal devoted to the topic, Childhood in the Past.

In the conference program, Dugstad’s paper appears with the title “How abortive artefacts can be informative archaeological research objects”, as part of a session dedicated to children as stone tool makers. The abstract for her conference paper touches on the points made in the other studies mentioned here– play, learning, and the importance of looking at stone tool products in context:

Teaching skills and passing on knowledge were very important, and we can assume that much of the transfer of knowledge happened relatively early in life, either through play or more structured apprenticement. Technological processes, tools, waste and retouched pieces as well as their context and associations can give opportunities to get closer to individuals, e.g. children.

So how do we get from this research, clearly part of a rich, decades-long tradition that has developed strong grounds for inferring children’s actions, to the flat-footed pronouncement that “Stone Age Kids Learned by Doing”?

The Archaeology brief note was based on a press release from the University of Stavanger. The Archaeology note is so brief that it never makes clear that the press release cites a source, an article, titled “Early child caught knapping: A novice early Mesolithic flintknapper in southwestern Norway”.

Frustratingly, though, the university press release does not give a full citation of the publication venue. But despite the changed title, this almost certainly refers to the paper in the 2010 conference volume Socialisation.

The press release also refers to a masters’ thesis incorporating this work, presumably Dugstad’s 2007 thesis “Hushold og teknologi: en studie av tidlig preboreale lokaliteter” (Household and Technology: A Study of Early Preboreal Localities) listed in the university’s online catalogue.

I have not been able to obtain a copy of the conference volume to see the full paper included there. What the press release suggests is that the study was firmly rooted in, and fully cognizant of, the wider body of work about children, crafting, and learning in the past. Dugstad is cited as saying that

a succession of failed strokes, terminating in many hinge and step fractures, indicates that axe was made by a novice flintknapper, probably a child.

Where Dugstad’s argument appears to differ from the other examples cited here is in taking a single, less well shaped artifact as a kind of signature of a novice, who she infers would have been a child:

the axe has probably not been produced by an adult. Errors are too numerous and striking to have been performed by a skilled and experienced flintknapper. This is probably a child’s work.

Other researchers have urged more caution, taking a comparative approach that sketches out contrasts between more and less well made products within a single social setting. Dugstad is quoted as suggesting that a more experienced craft worker would have had the skill necessary to correct errors, rescuing the work in progress, rather than repeatedly trying to fix one error and making another, leading to the discard of the tool in question:

one can see that the axe was made by a person with poorly developed theoretical knowledge and motoric skills. Given the numerous and characteristic failed strokes, it is also probable that the beginner had not received any form of direct instructions on how to proceed in manufacturing the tool.

The key phrase here is “poorly developed theoretical knowledge and motor skills”.  These comments are not included in the Archaeology news item.

Consideration of the relatively undeveloped motor skills and cognitive challenges of children as learners has been central to the arguments about children learning crafts that have been published over the last decade.

Theories of learning do more than point us toward analyses of errors in production. In a paper published in the book The Social Experience of Childhood in Ancient Mesoamerica, Jeanne Lopiparo suggested that adults in the Ulúa Valley of Honduras between 500 and 1000 AD made molds to help beginning learners– specifically children– produce competent examples of figurines. As I described her conclusions in Ancient Bodies, Ancient Lives,

molding served as a technology that allowed anyone, regardless of skill, to produce an image that was recognizable and that represented the contributions of different people to the social group. Lopiparo suggested that children would have been especially important participants in the crafting of figurines, learning how to be members of their society through their participation in the shaping of these conventionalized images.

The archaeology of childhood crafting is probably one of the best examples of how to think through identifying human actors in convincing ways using material traces. I just wish that Archaeology online had used its wide reach to help people understand just how rich and well developed this literature actually is.