Impossible Engineering: Technology and Territoriality on the Canal du Midi (review)

Impossible Engineering is an ambitious and stimulating book. It asks important questions, though the answers it proposes are not entirely convincing. As in her previous work on the gardens of Versailles, Chandra Mukerji crosses intellectual and disciplinary boundaries with incredible ease, mobilizing a vast array of scholarship to tackle historical cases in a new way.

The book deals with the construction of the Canal du Midi, one of the most famous technological enterprises of seventeenth-century France, a case that has already attracted a significant amount of scholarship. Mukerji succeeds in taking a new approach to the subject, moving away from the classical narrative of Pierre-Paul Riquet as the genius (and to some extent, martyr) who constructed the canal all by himself. Instead, she stresses the collective dimension of this work and tries to give it a new meaning in social, political, and technological terms, thus engaging with an important question in the history and sociology of technology, namely the nature of collective technical action.

If the Canal du Midi was “impossible engineering,” according to Mukerji it is mainly because nobody at the time had the necessary skills and knowledge to build it. It had to be the work of a “collective intelligence,” one that did not pre-exist. As Mukerji brilliantly shows, the canal resulted from the interplay of the entrepreneur Riquet (a tax farmer with no previous knowledge of engineering), various learned experts and engineers (some of them sent by Jean-Baptiste Colbert, the minister of treasury, in order to control this work), and numerous local artisans and workers, whose skills proved essential in the process of construction and the necessary innovations it required. One of the most interesting features of Mukerji’s book in that respect is her attention to the work of Pyrenean peasant women, who brought specific hydraulic techniques to the project.

This collaboration is presented by Mukerji as “distributed cognition,” a concept she borrows from Edwin Hutchins. When cognition is distributed, the problem of communication arises: how could a varied cast of actors find a “common ground,” given the huge social and cultural divides between them? Mukerji’s answer is disconcerting: they were all Romans in one way or another, so that “they all knew some elements of Roman engineering.” The elites knew the classics, and the locals “unknowingly reproduced elements of classical culture” (p. 11). They were thus Romans because they never stopped using Roman techniques.

This argument is highly problematic, because none of these Romans is actually Roman—or if you want, they are so differently Roman that they do not have much in common. On the one hand we have the classical culture of the elite and their imperial dreams of France as a New Rome. Seventeenth-century France, of course, was not Rome, and the elite vision of a New Rome depended on an invented Rome of the past. Diana Davis’s Resurrecting the Granary of Rome (2007), an environmental history of nineteenth-century French-colonial North Africa, shows very nicely, for instance, how the classical political rhetoric of being like (and in many ways better than) the Romans can have extremely little to do with the actual Rome.

On the other hand, we have an array of artisans and workers, ranging down the social scale to non-literate peasant women. To argue that peasants and artisans were Romans because they unknowingly reproduced a “tacit knowledge” of Roman technical traditions is to assume a quite naïve perspective on the history of technology: a few similar features do not make a technology the same as it was hundreds of years ago, and to elide the two erases vital transformations (a process that is integral to the very concept of tacit knowledge transmission Mukerji uses). As we know through the subtle critique of Maurice Daumas in his Le cheval de César (1985, 1991), riding a horse from Rome to Paris, though apparently a similar enterprise...