Say what you like about the science wars, they’ve got legs. A few days after finishing a review of The One Culture, the book that was supposed to end the quarrels between scientists and sociologists/historians of science, I learned that two of the contributors to that volume have been involved in a new skirmish. The opening shot was a new book by the physicist Steven Weinberg, called To Explain The World. The riposte came from the historian of science Steven Shapin, who has reviewed the book under the uncompromising title 'Why Scientists Shouldn't Write History.' Will Thomas, who drew my attention to the review, comments that Shapin’s review is ‘unduly divisive’ and that historians ‘ought to take seriously...the objections and perspectives of scientists.’ I agree, but I would go further. The reason historians should not write scientists off is not just that the scientist's perspective is valid but also that it overlaps with the perspective of some historians. To a large extent, the division that Shapin sees between scientists and historians is better understood as a division between historians. The methodology that Shapin endorses in his review is a strong form of anti-presentism. He writes: ‘History is properly about trying to understand the world of the past in its own terms.’ Most historians would agree with some reading of that statement. Most would also agree with Shapin that the job of professional historians these days is not that of ‘judging the past by the standards of the present.’ But Shapin goes further, in two ways. Firstly, he suggests that historians cannot legitimately say that an earlier event was a precursor to a later event. The true historian will ‘thump the table’ while reading Weinberg’s book, ‘insisting that searching for anticipations and foreshadowings is both wrong and illogical—‘ahistorical’ as they’d say.’ Certainly there are historians who would say this. But many others would say that an anticipation is merely a case of a past event resembling, or having an effect upon, a later event, and that anticipations are meat and drink for any historian who aspires to narrate or analyse past events—that is, for any historian worthy of the name. Secondly, Shapin seems to say that no-one should judge the past by the standards of the present. That is, such judgements are barred not only to professional historians but also to anyone who wishes to write accurate accounts of past science. Shapin’s subtitle captures his view nicely: ‘Plato was ‘silly’. Bacon ‘overrated.’ Galileo ‘behind the times.’ The suggestion is that anyone who makes such claims has made a fundamental methodological mistake, analogous to affirming the consequent or using a telescope to prove a mathematical theorem. Again, many historians would agree with this. But what’s wrong with saying that a past theory was false or that a past scientist used an unreliable method to reach a theory? Can such statements be verified? Apparently. Do they inevitably lead the author into errors of historical fact? If you think the answer is 'yes,' I'ld like to know your reasons. Arguably, Weinberg’s error is disciplinary, not conceptual. He has used past science in a way that professional historians do not usually use it, and in doing so he has underestimated the preciousness of professional historians. Shapin’s blanket anti-presentism obscures the real error in Weinberg’s book. Here I must confess that I have not read Weinberg’s book, so I stand to be corrected. Based on Shapin’s description, however, it seems plausible that Weinberg has judged past scientists by first establishing whether their theories were true or false, and then by assuming that the true ones must be the result of sound reasoning and the false ones the result of incompetence.** Plato’s cosmology does not match our own, hence Plato was ‘silly’; Newton’s cosmology is very much like our own, so he must have been a flawless genius. The problem is not the judging, or even the judging-by-today’s-standards. The problem is judging the rationality of an individual by the truth of their theories. This common error is the reason we have the symmetry principle. In his eagerness to run rings around Weinberg, Shapin skates over the distinction between truth and reasonableness and thereby repeats Weinberg’s mistake. Shapin’s anti-presentism is tied to his defense of the autonomy and expertise of the professional historian. His target is the dogma that ‘writing history is pretty straightforward and that being a 21st-century surgeon gives you a leg up in documenting and interpreting, for example, theories of fever in the 17th century.’ There are really two dogmas here. The first is that history is a doddle. This dogma is false (though whether history is as difficult as surgery is an open question). The other dogma is that knowledge of present-day science can be useful when studying past science. This is a much more plausible dogma, and Shapin is in danger of replacing it with the opposite dogma that scientists have nothing in particular to contribute to the history of their disciplines. Shapin reaches this conclusion by analogy. ‘Modern installation artists don’t think they can produce adequate scholarly studies of Dutch Golden Age paintings, and it’s hard to find offensive linesmen parading their competence in writing the history of rugby.’ The plausibility of both analogies is due partly to the fact that installation artists, unlike historians, are not heavily involved in reading and writing argumentative prose. The analogies are suspect because scientists are heavily involved in those activities. More importantly, the analogies rely on the fact that installation art is not painting and that linesmen are not rugby players. But the question is not whether physicists can help with the history of botany, or whether ethics committees—the linesmen of biology—have something to contribute to the history of biology. The question is whether painters can help with the history of painting, rugby-players with the history of rugby-playing, and physicists with the history of physics. And it seems to me that the answer in all cases is that they can, and they do. The assumption that lies behind Shapin’s analogies is that present-day activities cannot be compared with their closest equivalents in earlier epochs. Now, we can all agree that activities have changed over time, and that the risk of anachronism is real. But how much have events changed, really? How should we weigh the threat of anachronism against the special insights that a practitioner can bring to the study of their practice? And is there really a trade-off between insight and anachronism? After all, it is possible to believe that the earth moves, or that species evolve, or anything else, without automatically attributing that belief to every past scientist. Shapin’s vision of history is skewed in other ways. On the authority of Thomas Kuhn, he reports that ‘linear and cumulative progress is a problematic notion.’ Very well – but how can Kuhn or Shapin make this claim without making judgements about whether earlier theories were better or worse than later theories, and has not Shapin foresworn all interest in making judgements about past science? Shapin is right that the notion of progress is problematic for historians of science – but is this because we have shown that science does not make progress, or because we have decided not to address the question? Shapin says that historians would ‘express bemusement at Mr. Weinberg’s insistence that science advances by rejecting teleology, even as he depicts its history as a triumphal progress from dark past to bright present.’ But is it really so absurd to find purpose in human action and not in brute nature? And is Weinberg really so wedded to ‘triumphal progress’ if he thinks that science went backwards in the Middle Ages, as Shapin reports? This historian is not bemused. I am not saying that Weinberg's book is flawless. As Shapin points out, it ignores most science apart from physics, all science after Newton, and just about everything that seventeenth-century scientists wrote about religion.** Nor am I saying that Weinberg's errors are unrelated to his eagerness to evaluate past scientists and to find anticipations in past science. What I am saying is that anticipations and evaluations do not lead inevitably to bad history, and that at least some professional historians recognise this. More generally, several of Shapin's criticisms of Weinberg reflect the fact that he is a particular type of historian, and not that he is a historian as opposed to a scientist. This conclusion raises a question that I cannot hope to answer in this post but that is too important to omit. If historians disagree about how to do history, but agree that scientists sometimes write bad history, how should historians go about improving scientists' history? One answer is that historians should set aside their internecine disputes when dealing with scientists: they should only criticise scientists for errors that the vast majority of historians would recognise as errors. The problem is that this seems to tacitly resolve those internecine disputes in favour of the more liberal historians. If Shapin stops criticising Weinberg for evaluating past scientists, will not Weinberg think that it is OK to evaluate past scientists, thereby writing history that I can accept but that Shapin cannot accept? Still, there is surely some value in identifying points on which historians agree about the historical errors of scientists. These convergences may not be the whole solution, but they are surely part of it. And in the search for historiographical common ground, I think we could do much worse than the symmetry principle that I mentioned earlier in this post and that I have discussed at length on this blog. The vast majority of historians of science would agree, I think, that we should not assume that past scientists who held true theories did so for good reasons, and that those who held false theories did so for bad reasons. Scientists who make these assumptions should be the first targets of anyone who is interested in raising the bar of popular history of science. ** These criticisms of Weinberg's book are based on the data I had at the time of writing, namely a few hints that Shapin dropped about the book in his review. Weinberg has since denied the charges. This denial does not effect the main point of this post, ie. that the division that Shapin sees between scientists and historians is better understood as a division between historians. Expand post.
Tuesday, February 24, 2015
Wednesday, February 11, 2015
How to end the science wars: a review of Harry Collins and Jay Labinger, The One Culture? A Conversation About Science, part II/II
This is the long-delayed second part of a two-part review of Harry Collins and Jay Labinger, The One Culture: A Conversation About Science (2001). In the first part I argued that we—by which I mean, roughly speaking, scientists and sociologists of science—would more easily reach agreement about science if sociologists acknowledged their past relativism and if everyone was charitable in debate. It would also help if we set aside the interesting but irrelevant question of whether the truth of a belief can (partly) explain the belief. In this post I make three other recommendations: revive the internal/external distinction, or something like it; be clear about how our visions of science differ, if we think they do differ; and beware tacit philosophy of science. Revive the internal/external distinction. Gravity waves are tiny ripples in space-time created by bodies with mass. In 1970, several experts believed that the physicist Joseph Weber had detected gravity waves in his laboratory. By 1980, no experts believed this except Weber. What changed their minds? One plausible answer is that the experts made many sincere and careful attempts to replicate Weber’s results, and these replications failed. Another answer is that there was only one such failure, and that it was carried out by a physicist who happened to have more polemical skill and social prestige than Weber. The former answer says that the experts followed the evidence, and the latter answer that they followed their most powerful colleague. For want of better terms, let’s call the former an ‘internal’ answer and the latter an ‘external’ one. I don’t know which answer is the right one in the case of Weber, but we should be mindful of the difference between the two answers, because otherwise the science wars will never end. The reason is that scientists tend to play up internal explanations of true beliefs, whereas sociologists emphasise the external ones. If we do not distinguish between these two kinds of explanation, we will not even be able to characterise this disagreement. Worse, we may end up exaggerating the extent of the disagreement. Most of the contributors to The One Culture do not make any distinction along the lines of the internal/external one. More often they distinguish between ‘natural’ and ‘social’ factors, which they too often see as synonymous with ‘scientific’ and ‘social’ ones. For example, Michael Lynch takes ‘social factors’ to refer to ‘a range of personal, circumstantial, and institutional considerations,’ and ‘natural factors’ to mean ‘objective reality, nature itself, or properties of the physical world’ (271). What is missing in Lynch’s distinction is the idea that scientists’ beliefs might result from a third kind of cause, namely experiments undertaken by scientists and the deductions made by them—the meat and drink of internal historians of science. No progress can be made if one party (sociologists) does not recognise, even conceptually, the preferred explanations of the other party (scientists). When sociologists do recognise the scientists’ preferred explanations, they tend to absorb them into the category of ‘society’ or ‘culture.’ This has the effect of obscuring points of agreement between scientists and sociologists, since scientists do not think ‘experiments and deductions’ when they see the words ‘society’ and ‘culture.’ This is not due to any deep misunderstanding on the part of the scientists. It is just not how the words ‘society’ and ‘culture’ are usually used when talking about science. The point may be illustrated by a dispute between Harry Collins and Trevor Pinch, on the one hand, and the physicist H. David Mermin on the other. Collins and Pinch wrote an account of early experimental tests of the theory of relativity in which they argued that those tests were inconclusive. They maintained that it was not these tests, but rather the ‘culture of life in the physics community’, that persuaded physicists to believe the theory. I leave the rest to Mermin:
What this means depends, of course, on what ‘culture’ is taken to include. If, as I now understand Collins and Pinch’s intent, ‘the culture of life in the physics community’ refers to the cumulative impact of all theoretical and experimental work bearing on relativity since 1905, then [their thesis] is correct. But their readers...are likely to conclude that the momentum was generated by nothing more than many years of growing more and more comfortable with the [two inconclusive tests].Be clear about other disagreements. In my previous post I said that the most visible disagreement in The One Culture (about whether the truth of a belief can explain the belief) is not a real disagreement. And in this post I have argued that the real disagreement (between internal and external explanations of true beliefs) is scarcely visible in the book. So, are there any disagreements in the book that are both real and visible? Collins thinks there are, but I’m not convinced. In a chapter near the end of the book, Collins says that sociologists prefer the ‘rough diamonds’ of science to the ‘crown jewels’ of science. He summarises his point by saying that sociology ‘reduces the quasilogical authority of science’, but really he draws several contrasts that need to be treated separately. None of them pick out a clear disagreement between scientists and sociologists. Process v product. One of these contrasts is that, whereas Collins studies the early, uncertain stages of scientific debates, scientists such as Weinberg (who seems to be the foil that Collins has in mind in this chapter) emphasise the polished theories that emerge from this process. This is a difference in research interests, not a disagreement. Skills v knowledge. Collins says that he stresses the ‘assiduousness, experience, skill and virtuosity’ of scientists, and the fact that they are ‘the kind of people whom it makes sense to trust’ because they have ‘the right kind of expertise.’ By contrast, scientists play up their ‘privileged access to reality’ and their ‘extensive store of knowledge about the way the world works.’ Whether this is a disagreement depends, as ever, on how the terms are defined. The disagreement vanishes if scientists say that it is precisely their ‘experience’ and ‘skill’ that gives them ‘privileged access to reality.’ Collins himself suggests that ‘the right kind of expertise’ might consist in making ‘experimentally or observationally based claims’ rather than ‘book-based claims.’ This claim is remarkable for its conventionality. It looks like all that quasi-sociological talk of ‘trust’ and ‘skill’ and ‘expertise’ is really just another way of saying that we should believe what scientists say about nature because they study it directly instead of relying on bookish authorities. In short, Nullis in verbia. Simple v complicated justifications. Collins says that ‘scientific procedures do not speak for themselves but have to be judged and interpreted’, and that ‘experiments and theories [are] less decisive in bringing scientific controversies to a close than uninvolved scientists and others generally think they are.’ His idea seems to be that the intellectual side of scientific debates are more complicated than scientists make out—he is saying that it is not just a matter of doing one or two experiments and drawing the obvious conclusion. The problem is that none of the scientists in the volume deny this, not even Weinberg. In fact, one of them (Mermin) turns the tables and accuses Collins of giving an insufficiently complex account of the early evidence for the theory of relativity. Social v scientific education. Collins says that the sociology of science ‘turns the public understanding of science into a matter of social education rather than scientific education.’ What he means is that non-scientists lack the time and expertise to carry out a thorough assessment of the evidence on either side of scientific debates over such things as global warming and genetically modified foods. We rely on the testimony of the experts, which means that we need to decide who are the experts on the debate in question. This is not a trivial problem, especially when the experts appear to disagree. Collins implies that scientists have trouble understanding these points. He should have another look at the following passage from the scientists Jean Bricmont and Alan Sokal:
When confronted with experts, any individual or small group of individuals is in a difficult situation. There is no way to find the time and the means to check even a small fraction of the experts’ assertions. And yet, in many practical situations we have to decide whether or not to trust their claims. How should we proceed? This is a truly interesting and difficult question … [and one where] many sociological considerations become relevant (46).Beware of tacit philosophy of science. Michael Lynch writes that the science wars are ‘a metaphysical battle fought by conscripts who have limited training in the martial arts of philosophy’ (53). Lynch says that the debate would be improved if the participants recognised that they were arguing over questions that have challenged philosophers for centuries, if not millenia. Some of the questions are metaphysical: do causes and categories really exist in nature, or are they just tools that help us to understand nature? Other questions are epistemological: what is the best way to justify scientific theories, and are our justifications strong enough to give us confidence in the existence and nature of unobservable entities like quarks and DNA? Lynch is right that these are fraught topics, and that anyone who wants to discuss them seriously should have a least a passing acquaintance with the relevant writings of trained philosophers. However, to judge from The One Culture, Lynch is mistaken if he thinks that scientists and sociologists still see these issues as a major front in the science wars. None of those issues appear in the list of ‘open questions’ that the editors provide at the end of the book (299-300). And, as I mentioned in my previous post, the only avowed relativists in the book are methodological ones. This reticence is both encouraging and vexing. It is encouraging because it suggests that, pace Lynch, scientists and sociologists are arguing over topics that they have some special competence in, rather than ineptly reproducing the debates of professional philosophers. The reticence is vexing because it may conceal more disagreements than it resolves. Collins, with his stress on 'skill' and 'expertise', often sounds like an instrumentalist, ie. someone who thinks that scientific theories are great instruments for predicting and controlling nature but who is loath to take the next step and say that the successful ones are probably true. Steven Shapin, for all his asides about the impotence of philosophy, defends a thesis that is nothing if not philosophical, namely that there is no single method that characterises all forms of science. Several contributors imply that the sociology of science has shown that science does not achieve 'certainty', a claim that is both normative and epistemological. If we make such claims then we should be clear that we are making them and that they are, at least in part, philosophical claims. This would not solve the problem of how non-philosophers can reach agreement on philosophical questions. But it would at least clarify where the disagreements lie. *** The stated aims of The One Culture were to get scientists and sociologists talking to each-other again, and to get clear about their points of agreement and disagreement. The upshot of this review is that the book achieves the former goal but has mixed success with the latter. To sum up my criticisms:
- the editors ignore the main source of agreement between the two parties, namely the fact that sociologists have retreated from their full-blooded relativism of the 1970s and 1980s. - the editors misidentify methodological relativism as a major source of disagreement. Even Harry Collins admits that the truth can help to explain a belief, and anyway the interesting question is not whether truths can explain true beliefs but whether social factors routinely play a decisive role in the formation of true beliefs. - The One Culture barely addresses the latter question (about social factors), which is not surprising given that most contributors do not acknowledge the distinction between social and intellectual factors. - Collins identifies another persistent disagreement, concerning 'crown jewels' and 'rough diamonds,' but this distinction is overdrawn, as we see when we unpack the gemmological metaphor. - some contributors hint at more substantial disagreements over such things as instrumentalism and the unity of science, but these are philosophical questions that scientists and sociologists cannot resolve on their own.The good news is that the question about the decisiveness of social factors is a question for scientists and sociologists, and there is no reason why they should not be able to answer it together—as long as they can agree to distinguish social factors from intellectual ones. Expand post.