Australian Journal of Cultural Studies
Vol. 3 No. 2, December 1985

Representing and Intervening

Ian Saunders

Ian Hacking, Representing and Intervening, Cambridge: Cambridge University Press, 1983.

For the student of semiotics and cultural forms the chief interest in the philosophy of science over the last two decades has been in its dealings with the nature of meaning and interpretation. 1962 saw the publication of Thomas Kuhn's epoch marking The Structure of Scientific Revolutions, and since then the battles have been fought over the implications of accepting a characterisation of scientific theory as in some way actively structuring experimental phenomena, a characterisation akin to and often overlapping with the linguistic a prior ism that thinks of language as generating and organising our 'world'. Ian Hacking wants to change all that in Representing and Intervening, debunking under the banner of history the supposed centrality of theory in favour of the myriad experimental and technological interventions that, he says, ought to constitute our sense of science. He wants to move from puzzles of rationality to problems of reality, from representing to intervening.

Despite the constraints of the book's form ('Introductory topics', more on which below), Hacking achieves a good deal towards that shift through the telling accumulation of historical detail. Theory oriented philosophers and theory oriented textbooks have simply re written history to support the view that theory comes first, be it as a conscious problem hypothesis (Popper), a novel source of prediction (Lakatos), or as a subconscious theory matrix (Hanson; Feyerabend). So, he relates that in 1965 the radioastronomers Arno Penzias and R.W. Wilson, while trying to locate specific energy sources in space, found an unexpected background temperature reading of 3°K. It was only after that discovery that a different group of scientists circulated the hypothesis that if the universe had originated in a Big Bang then there would be a uniform residual temperature throughout space, and happily the independent experimental work 'meshed beautifully with what would otherwise have been mere speculation' (59). Memory is short, though, and by 1979 a textbook, F.M. Bradley's The Electromagnetic Spectrum, can retell the history as if Penzias and Wilson had begun with the notion of testing that hypothesis about the Big Bang, whereas they were in fact unaware of it. Observation becomes, in respect, theory dominated.


History does matter, Hacking insists, and the philosopher's cavalier attitude to it won't do. Imre Lakatos in particular is amusingly revealed as being remarkably inventive with the historical examples he produces to support his version of theory centrality, but he is not alone. Hacking's own view turns out to be rather Wittgensteinian: there is no one thing that always 'comes first', rather it might be on occasion a hunch, a technological innovation, an experimental tinkering, an unexpected result (like the 3°K background reading), ... or even a theory.

That is not to deny that we need background presuppositions and habits of mind to make any observation whatsoever, but, he maintains, to call these full-blown theories is to stretch the notion an intolerable extent. If it is to be of any use we need to be able to distinguish between (say) a theory of perception and the loose bundle of linguistic habits and submerged assumptions behind a sentence like 'I can see a brown table over there'.

Indeed it is the concentration on observation sentences like this one at the expense of detailed history that above all raises Hacking's ire. Questions of language, meaning, and all that meta-talk take us to the dizzying exchange between materialist and positivist, where the ultimate response can only be the careless 'Talk as you will' of Rorty's depressing conversationalist. It is not that, for example, van Fraassen is mistaken about empirical adequacy, it is just that the issue could not be decided one way or the other, so little can be gained from tracing the rhetorical spiral. 'Realism and anti-realism scurry about, trying to latch on to something in the nature of representation that will vanquish the other. There is nothing there' (145). Realism, anti-realism, or any other ism will not issue as the conclusion to a sophisticated syllogism on the metaphysics of rationality or the nature of interpretation—that, Hacking says, gets things in precisely the wrong order. Use electrons to do something, test violations of parity in weak neutral current interactions, say (Hacking despairs at Galileo, Newton, and the well encrusted Standard Example), and you are committed to pragmatic realism about electrons. An armchair scientist-philosopher might well believe that observation is produced by theory, or that atomic or microscopic entities can be subsumed into instrumentalism, but one's view changes when you actually do anything. Microinjecting a living cell, for example:

We see the tiny glass needle—a tool that we have ourselves hand crafted under the microscope—jerk through the cell wall. We see the liquid oozing out of the end of the needle as we gently turn the micrometer screw on a large, thoroughly macroscopic


plunger. Blast! Inept as I am, I have just burst the cell wall ... ( 190) .

Pragmatic realism has little to do with empirical adequacy or explanatory power of the hypothesised entities; you become a realist when you can use them.

Of course it is something of a shock to be told that observation can come before theory, or experiment before language, so it is as well to note that Hacking is more than well acquainted with recent work in the philosophy of language, and equally with structuralism and post structuralism. Indeed, the work of Michel Foucault seems to provide at times an implicit motivation (Hacking has written on Foucault elsewhere). There is the hinted view that there are deep, global sets of possibilities that determine how we think, and that these change over long periods of time. However, this is not developed past a casual reference to the oddity of sixteenth century science. More importantly, although perhaps not entirely consistent with that notion of global form, there is the belief that the complex experimental, technological and social make-up of a period cannot be reduced to any representing theory of structure. So in Foucault's case it finally makes no sense to analyse out a causal or explanatory structure of intellectual and political power because both cause and explanation exist only insofar the power is reified as the surface of events. One loosely related argument Hacking gives has already been described, namely that the role of theory in discovery has been overestimated, and that one is committed to realism through the ability to fiddle with electrons or cell structures or whatever, rather than as a result of any metaphysical epiphany. Theories of representation merely 'scurry about'.

However, there is another, less persuasive, and to my mind more Foucauldian suggestion that colours the rhetoric of the text, that is that intervention dominates representation. Here a troublesome crux in Foucault's project—that it becomes impossible to characterise what would be involved in understanding or contesting the identifying descriptions he makes—surfaces in Hacking's book in the form of a paradoxical sense that he argues against representationalist philosophy of science just as he is representing both the philosophy and the science. Now Hacking doesn't claim to be hunched over the microscope rather than his typewriter (although maybe he would like to be), but even if he were it is hard to see what an understanding of what he was up to could consist in, or why he might choose to call himself a pragmatic realist, if the only or most important step in the game was intervention.


The better account, if a less exciting one, is the one he illustrates with Bacon's experimenting, speculating bee: both qualities are important. That the larger claim seems to stalk the text without being met head on is perhaps in part the consequence of the 'topics' style of the book. Hacking moves across several centuries of philosophy and experimental history with familiarity and a fluent, lively shrewdness (the sections on Putnam, Lakatos, microscopy and subatomic re search are notably fine), but the invitation to pick and choose combined with a systematic distrust of universalising theory allows him to entertain both versions of intervention, while only defending the more modest. The philosophical problem of the relation between intervention and representation, while given a well deserved shot in the arm of history, is in the end abandoned in favour of a sense of the irreducible variousness of things:

God did not write a Book of Nature of the sort that the old Europeans imagined. He wrote a Borgesian library, each book of which is as brief as possible, yet each book of which is inconsistent with every other.

Ian Saunders teaches at Griffith University.

New: 7 December, 1997 | Now: 27 April, 2015