Identifying Constructions of Childhood: Beyond Adjectives to the Promise of Verbs


Chris Borst, McGill University

Reconceptualist ECE enjoins us to use “the image of the child” as our evaluative lens in children’s services. Doing so depends on having a rich schema of the constructions of childhood we will encounter, but those available to us remain limited. To propose and pilot methods for the identification of a rich (i.e., multidimensional) schema of constructions of childhood. Recent work on constructions of childhood, associated with the so-called “new wave” and “ontological turn”, has focused on processes of construction, rather than on the enumeration of constructions . But solutions to the latter problem remain unsatisfactory. The standard solution appears to be that provided in Smith’s 2014 The Government of Childhood and Lancy’s 2008 The Anthropology of Childhood . Yet, these offer a schema (good, bad, potentially useful) that would have been familiar at the outset of the 19th century. Malaguzzi claimed there were “hundreds” of images, but gave us only a Manichean clash of “the image we need” and “the image we don’t”, subsequent work deeming the latter to be the version of human capital theory associated with Heckman and colleagues, decorated with references to epigenetics and neuroscience, and the former to be a critical-democratic discourse that (despite invocations of Foucault, Bourdieu, Deleuze, Haraway, and Braidotti) would have been familiar to Dewey. Pacini-Ketchabaw et al.’s 2015 Journeys offers only ad hoc lists whose items don’t all even clearly differ. A potentially more fruitful route is sketched in works such as Heywood’s 2001 A History of Childhood and Hendrick’s 2003 Child Welfare , which suggest small sets of binary oppositions – multidimensional models – as ways of organizing historical variation. This chimes with work in cross-cultural psychology and psychological anthropology, parenting studies, the study of political ideologies, and multiple currents in linguistics, as well as Bourdieu’s work on the “space of lifestyles”. The electronic texts in the Project Gutenberg repository, most pre-dating 1927 for copyright reasons, were downloaded and passages pertaining to children were computationally extracted from the over 54,000 texts in English. Various methods were applied on a pilot basis to samples from this corpus, including both the hand-coding of passages and computational methods. The familiar images from Smith, Lancy, Malaguzzi, et al., emerged clearly from the adjectives used about children and young people, which were predominantly about positive-negative ethical and aesthetic contrasts, saving only a heavy emphasis on their smallness and some mention of their weaknesses and strengths. A much richer picture, however, was suggested by the verbs – the things children and young people do and have done to them. Nouns (other than nominalized verbs) and adverbs seemed to convey unexpectedly little information. That good/bad/potentially useful conveys something about how children are conceived is clear, if not necessarily anything unique to them. Such evaluations are encoded in adjectives. The promise, however, lies in looking beyond adjectives and scaling up an analysis of the semantic space suggested by the verbs used about children. A rich empirical schema of constructions of children can potentially be generated computationally from the verbs and, to a lesser degree, the adjectives used in discourse about children in a larger and more contemporary corpus. The most effective way to generate this semantic space is still being piloted.

This paper will be presented at the following session: