Yesterday morning nearly 2000 historians of science gathered in a vertiginous lecture hall at the University of Manchester, UK. Hasok Chang, the keynote speaker, told them that they could benefit from studying the technical content of science. Not a very controversial claim, you might think. After all, science does have technical content, just as it has journals, military contracts, and priority disputes. The fact that the talk was controversial—and the initial reaction on twitter suggests that it was—shows just how sensitive historians of science still are to what was once called the internal/external debate. Having written about this debate before on this blog, I can’t help commenting on the talk. I agreed with much of Chang's talk, but not all of it. I think that in some respects he went too far in defending internal history of science, and that in other respects he did not go far enough. Update: the video of Chang's talk can now be viewed for free here: http://www.ichstm2013.com/blog/audio-and-video/. In the rest of this post I've inserted (in square brackets) the time of key events in the video. Here’s what I agree with: The internal/external debate is still a live one. It may take a different from than it once did, and some of us may be repelled by the very idea that the debate continues. But there is no doubt that some historians still worry that the history of science is in danger of “losing its science,” while other historians worry that it is bad history or bad politics, or both, to separate the technical content of science from its social and political aspects. Worriers of the former kind include the historian of physics Olivier Darrigol, whom Chang quoted in his talk [2:50]. If Chang cited worriers of the latter kind, I don’t remember who they were and would be grateful to anyone who could jog my memory [I could not find any such citations when I watched the video of the talk. Update to the update: Michael Weiss, watching more carefully than I, has noted the citation of Kathryn Olesko at 4:50 and the less direct citation of Kathryn Olesko and Robert Kohler at 5:30]. The internal/external distinction is coherent and useful. Chang made free use of the terms “internal,” “external,” “internalistic,” and “externalistic.” He also asserted that the internal/external distinction is not a false dichotomy. He dismissed some other distinctions, such as those between practice and theory and between the social and intellectual (of which more below). But he had no shame in insisting on the internal/external distinction as one way of dividing up past science, and of dividing up the books and articles we write about past science. Internal history of science can be good history. One of the key slides in Chang’s presentation was a list of “reasons for doing history” that included describing, understanding, using, overcoming, and appreciating the past [42:00]. Chang’s point was that internal history of science could serve all of these aims. In other slides he argued that internal history of science could serve other aims, like teaching science in schools and advancing present-day science. But he made it clear that these extra-historical goals were not the only ones that internal history could serve. In his view, as in mine, internal history of science can be a genuinely historical enterprise. Internal history of science has a bad image among professional historians of science. This was probably one of Chang’s most controversial claims, largely because of its rhetorical effect. Like anyone who claims that the pendulum has swung too far in one direction, Chang gave encouragement to those who would swing it in the opposite direction. This will alarm historians who think that the pendulum has not yet swung far enough away from internalism, or who worry that historians encouraged by Chang will swing it too far back to internalism. But as far as I can tell, Chang’s basic point is correct. Since about 1985, the best way to mark oneself as naíve and out-of-touch, at least among historians of science in the UK and US, is to write books and articles that say little or nothing about the politics, sociology, rhetoric, architecture, print culture, visual culture, etc. of science. Darrigol did not err when he complained, in a chapter that Chang cited, that internalists tend to be seen as “fossils” who cling to a discredited brand of history [2:50]. There is more to history of science than internal history. Although Chang defended internal history, he did not defend it as the exclusive mode of history of science. His point was that internal history is more worthwhile than it is often thought to be, not that no other history of science is worthwhile. Historians should be “pluralistic,” he said. “There are no enemies here, except those who are in the habit of making enemies.” (I would add that pluralism should not stop us from rejecting opinions we consider false, and that we should be able to disagree with someone without making an enemy of them). So much for my agreements. In the following, the claims in bold are the ones that I endorse and that I want to contrast with some aspects of Chang's talk. The real debate is between internalists and hybridisers. The label “externalist” is almost as widely shunned as “internalist.” I do not know any historians of science who pride themselves on ignoring the technical content of science. But I know many who pride themselves on integrating the technical content of past science with its social or political elements, and many who think that this integration is one of the main goals—perhaps the main goal—of the discipline. Consequently, the most common criticism of internal history is not that it pays attention to the technical content of science but that it does so exclusively. Far from countering this criticism, Chang appeared to endorse it. He said that the distinction between the “social” and “intellectual” factors was a false dichotomy, and invoked well-known works of hybrid history to make his point [10:00]. I don’t recall all of his examples, but they included Peter Galison’s Image and Logic and Steven Shapin and Simon Schaffer’s Leviathan and the Air Pump. These works are (rightly) celebrated, and often they are celebrated precisely because they marry the content and context of science. I fear that the chief lesson that many people will take from Chang’s talk is that we are in dire need of more works of this kind. This would make things harder, not easier, for those who wish to emulate internalist works such as Darrigol’s magisterial Electrodynamics from Ampère to Einstein. Internalists should defend their work as history of science and not as history of science. There are two ways a work can fail to be good history of science. It can fail to be good history, or it can fail to be about science. It is important to keep these two criteria separate when assessing any historical genre, whether internalist, externalist, or hybrid. Obviously, the two criteria are distinct: there is good history that is not about science, and poor history that is about science. The main reason the distinction is important is that the correct answer to the question “what good is internal history of science?” may depend on which criterion one uses. For example, one might think a) that internal history of science is on a par with non-internal history as history, but that b) it is more squarely about science than external or hybrid history. b) is bound to be more controversial than a), simply because it is a claim for the superiority of internal history rather than a claim for its parity with other kinds of history of science. Chang seemed to make the latter, provocative claim. The title of his talk, “Putting science back into the history of science,” when read alongside the talk itself, suggests that Chang equated “science” with “technical content.” This would imply that, in his view, internal history of science is indeed more squarely about science than external history of science. Perhaps Chang is right about this. But in my view it would be better for internalists not to make that claim until they have established the less provocative (but still important and controversial) claim that internal history is no worse, as history, than external or hybrid history of science. It may not even be necessary for internalists to make the provocative claim just described. Personally, I am concerned about whether I am writing good history, not about whether I am writing history about science—even when I am writing about the technical content of science. Here's another way of putting it. I am puzzled, and sometimes irritated, by those who insist that internal history of science is somehow a second-rate form of history. But I’m unbothered by those who imply that internal history of science is no more “about science” than external or hybrid history. The best way to defend internal history is by the hybridiser’s own standards. Hybrid history of science tends to be surrounded in a halo of historiographical virtues. It is said of hybrid histories that they are contextual, cultural, and causal, that they respect actor's categories, honour the symmetry principle, display the contingency of science and treat science as a construction, a product of human activity. These terms of praise are not often applied to works of internal history of science. So it is easy to get the impression that the hybridiser's virtues are beyond the reach of the internalist. The challenge for the internalist, as I see it, is to show that there is no such imbalance. Chang's approach was not quite so direct. He did list certain historiographical virtues that he thought were within the reach of history. But these virtues were not, by and large, the ones that are most commonly associated with non-internal history of science. There were two notable exceptions to this rule. Chang did mention that internalist works can show the contingency of science, and that they can be “cultural.” Chang made the latter point with an intriguing remark that may be paraphrased as follows: “It is only an anti-intellectual culture that does not consider intellectual activity to be cultural activity.” Chang's point here was not that intellectual activity is always shaped and sustained by something other than its technical content. On this view, internal history of science is not yet cultural history but can always be turned into cultural history by adding in the politics, sociology, rhetoric, etc. This may be true, but it is not what Chang meant, I think. What he meant was that works of internal history are already cultural. No additives are necessary to make them into works of cultural history. Darrigol's Electrodynamics from Ampère to Einstein is as much a cultural history as Leviathan and the Air Pump. To think otherwise is to treat intellectual activity as a special sort of human activity, one that is not on its own cultural. There is an irony here: those who reject internal history on the grounds that it treats science as “exceptional” are thereby committing the very error they claim to save us from. Endnote. Having said all this, I had better say why I think that internal history is not a second-rate form of history. The reason is simple: the standard argument against internal history, and in favour of hybrid history, is guilty of a whopping inconsistency. The usual argument is that only hybrids do justice to the inextricability of social and epistemic factors in past science (or something along those lines). The obvious reply is that there are lots of other inextricable pairs that the historian should try to knit together—like theory and experiment, or mathematics and physics, or science in France and science in Germany—the list is endless. Given that no single work can hybridize everything, it is inconsistent to single out the science-society dyad for special attention. An example might help to make the point. As far as I know, no-one has asserted that all books and articles dealing with scientific instruments must also deal with scientific theories, and that those that fail to do so do not qualify as proper history. That would be absurd. It is equally absurd to insist, as many historians of science seem to do, that all books and articles that deal with the technical content of science must also deal with the sociology, politics, rhetoric, etc. of science. Expand post.