1Since his excellent 1997 book La mesure de l’esprit (The measurement of the spirit/mind), Olivier Martin has been known for his keen reflections on the social construction of numbers and quantification, and his sharp critical analyses of indicators and measurement. When that first book was published, a lively scholarly literature on these themes existed, stimulated by the work of thinkers such as Stephen Jay Gould, Ian Hacking, and Alain Desrosières. In that context, Olivier Martin’s contribution was an important one. Today, he offers us a broader approach, which admirably integrates this theme into a general reflection on quantification and its history, starting from the earliest examples and arriving at the systematization of quantification today in evaluation and management procedures.
2He opens the book with a set of ‘investigations’ (enquêtes) that provide insight into the history of the use of numbers and quantification in various contexts. He evokes commercial uses and the measurement of time, emphasizing the social origin of the quantification and usage of numbers in its various forms. He emphasizes the great leap into abstraction that led to measurements intended as universal or universally comparable, with the appearance of the metric system as one of the most exemplary cases. These chapters lead naturally into an exploration, familiar from Martin’s previous work, of how societies are ‘put into statistics’ (mise en statistique). They present a masterful synthesis of work carried out over the past 20 years on the history of statistics, the translation of society into numbers, and their uses as tools of government. As he points out, ‘statistics do not simply enumerate pre-existing realities: they also contribute to forging these realities [.…] As they invent themselves, societies develop the tools (both practical and intellectual) to measure, manage, and govern themselves. Statistics are incontestably among these tools’ (p. 73). We also appreciate the homage paid in this chapter to Alain Desrosières and to the role of his major work, The Politics of Large Numbers, in this renewal.
3The final three investigations, the chapters that close the first part of the book, are closely articulated: starting from a historicization of quantification in the sciences to pose the question of scales of measurement, they end with the question of ‘quantitative indicators aimed at evaluating the performance, efficiency, quality or activity of an organization, a department, or even an individual’. His reflection on scales of measurement prominently features psychophysical measurement and the clinical diagnosis of intelligence (both dealt with in depth in Martin’s first book). The chapter on evaluation narrates the strange recent history of the ‘sociology of evaluation’, which, starting from the evaluation of public bodies, processes, and strategies, spills over into the evaluation (including self-evaluation) and quantification of individuals.
4The book’s second part seeks to draw a lesson from these investigations, setting out a complex sociology of quantification. Its three chapters undoubtedly represent the best illustration of the importance of Martin’s research: they justify an approach that encompasses multiple aspects of quantification and measurement, usually analysed in different ways, to investigate the ‘reasons for quantification’ and its effects. Quantification is doubtless a ‘social fact’, as the title of Chapter 8 has it. But Martin systematizes this, moving beyond the debate on realism (a key concern of Desrosières, as Martin underlines), arguing that while ‘figures are neither “true” nor “opaque veils”’, they reflect a reality and can have strong performative effects. They express the presence of power, while also providing it with tools. Martin presents a very persuasive critical analysis of various methods of evaluation, founded on conventions, although their initiators seek to provide them with a foundation that affirms their character as scientific and realistic because quantified. But this is just one among many equally fascinating lines of argument in the book.
5This work is thus extremely stimulating. Written with great clarity and drawing on a very large literature, the book takes us on a long journey through the history of quantification and its analysis as a social fact. This important book constitutes an essential critical tool on the processes, ‘practices and techniques’, and use of quantification. It demonstrates their multiplicity, emphasizing that ‘quantification is not (necessarily) an act of scholarly knowledge’ without falling into a simplistic denunciation of the use of measurement. This work takes its place in the context of a welcome critical sociology of science. One might perhaps regret that, in analysing quantification as a ‘social fact’, Martin does not sufficiently systematize the relation between the nature of particular political and economic powers, on the one hand, and the uses and forms of quantification that lie behind them, on the other. Martin looks at both authoritarian regimes and liberal economies, showing that the role of quantification is not exclusive to any one, while highlighting the differences. But he does not clearly mark out the distinct usages and forms of quantification in these different systems.
6Likewise, the book scarcely deals with the development of new players and new forms of quantification and measurement linked to powerful economic actors in the digital sphere, from Google to Facebook, which are developing new tools at the edges of the quantifiable whose consequences remain difficult to determine. Will quantification give way to the digital processing of massive quantities of data which are no longer analysed using summary numerical indicators? The development of new practices that orient, or determine, without quantifying? And will state powers be marginalized by these economic actors as a result? Martin evokes the emergence, in the 19th century, of the idea that ‘to govern is to calculate and direct by numbers’ (p. 215), a conception that is still dominant today and that can lead to the idea that it ‘is possible to build societies by purging them of politics (i.e. of debate, confrontation, the search for compromises, and negotiated solutions), while subjecting them to quantified laws’ (p. 218). But what consequences will the rise of artificial intelligence and of new tools for processing ‘Big Data’ have for ways of governing and for the relative place of states and digital economic monopolies in all of these processes?
7That is another line of investigation, it is true, one that this work gives us great reason to enter into. This is a fine and fascinating book, filled with essential critical reflections that open up many questions and avenues for future research.