DOUBLE
REFRACTION
Looking twice at the history of science

Sunday, May 20, 2012

What's wrong with what historians of science do now?

version française -----------------------------------------------
According to Will's picture, there are fairly serious problems with the current practices of historians of science, practices I summarised in the previous post. The problems are not self-evident. As a comment-writer asked recently on EWP, “How does the idea of historians seeking “the invisible” differ from what, I think, we all hope to produce in our scholarship, that is, the new?” Likewise, it is not obvious what is wrong with writing case studies, focusing on the sociological aspects of science, and trying to enlighten the public. In this post I'll try to say what WT thinks is wrong with each of these practices. But first, a note and a clarification.

The note is this is a long post, for the simple reason that WT has lots of criticisms. My premise is that it is easier to read a single 3,000 word post than to read the 3 years of EWP posts that this post is designed to distill.

The clarification is that WT is not always critical. He thinks that historians made genuine gains in the 1970s, '80s, and '90s; the problem was the failure to build on those gains, and to forget the work on which the gains themselves were built. A 2008 post set the tone: “I can see how such things were refreshing given the state of the historiography 20-30 years ago, but why today?”

Even today, there are brilliant exceptions: “The best history-writing of today is much better than the histories produced by [various past schools in the history of science].” But too many of us are engaged in “cargo cult historiography,” re-living the past in form but not in substance. Here is a final quote to drive the point home: “The richness, honesty, and critical depth of many of the debates in the social studies of science in the late ’70s and early ’80s continues to surprise me, since their full contours were not very well preserved in later rehearsals.”

WT has more to say about the road to the current state of the history of science, much of which I hope to summarise in the next post but one. For now, the thing to clarify is that most of WTs criticisms are aimed at the over-extension or mis-application of practices that were once healthy. Now for the criticisms themselves.


IDEOLOGY

Some important topics are not hidden by any ideology. Sure, all historians want to produce “the new.” In this sense we all want to reveal something that was previously invisible. And if something was invisible before, presumably there was something that caused it to be invisible. And making it invisible usually involves identifying this cause. So far, so good.

BUT, the cause need not be an ideology of science, ie. an intellectual prejudice put about by scientists to serve their interests and to cope with strains resulting from their failure to perform their social function. It might be an intellectual prejudice put about by non-scientists (eg. by historians of science). Or it might not be an intellectual prejudice at all. For example, it might be the failure of an archivist to open a crucial carton because it was too tightly packaged and the assistant was away sick that afternoon.

Or it might have simply been ignored by scholars, as the topic of bureaucracy is ignored, because it has not been hidden by ideology. Some things are simply under-documented, not covered up. There is evidence for this in the success of historiography from below, ie. history-writing that treats actor's histories as descriptive rather than merely polemical. In short, there is more novelty in history than that which is hidden by ideology.

I agree with this, but would insist on the following distinction. On the one hand, there is the question of “visibility”, ie. whether or not a topic has been hidden by a prejudice put about by scientists. On the other hand, there is the question of the “depth” of understanding that is achieved in studying a topic, with data-collection and time-line-creation being fairly shallow and explanation, analysis and interpretation being deeper.

These two questions are distinct. That is, it is not the case that studies dealing with invisible topics are ipso facto deep, nor that studies dealing with visible topics are ipso facto shallow. The danger is that we forget the former: that is, that we imagine that studies of highly visible topics can only ever be exercises in mechanical fact-collecting rather than deep analysis, and hence that they are inherently less interesting to historians than the topics with low visibility.

Some important topics are not evils due to ideology. There is also more novelty in history than that which is a product of the evils of ideology. Historians focus on evils such as scientist's over-confidence in their predictions (think bankers) or the abuse of the authority of science by non-scientists (think pharmaceutical companies). But sometimes scientists are not over-confident in their predictions, and sometimes the public is not convinced by the epistemic aura of scientists. Current practice leaves no room for such episodes.

As WT put it: “[Current] historiography is good at chronicling the uses of the authority of science and resistance to this authority, but to what extent can it capture traditions of scientific deference to public opinion in political issues, or traditions of the public not deeming scientific figures to be a source of authority relevant to their main concerns at all?”

Ideologies lead to case studies. Sometimes WT downplays the previous two criticisms. Recently he wrote that “the power and appeal of the cult of invisibility is in the fact that its central insights are more-or-less correct.” This seems to be a concession that ideology really does hide things, and that it really does cause evil. On this reading, WTs real beef is with case studies. And the main problem with our self-imposed task of making ideologies visible is that it causes us to settle for case studies.

As noted in the previous post, historians imagine that case studies are adequate to the aim of making ideologies visible; they also believe that case studies are all that historians can hope for once one accepts the “critical insight” that past science is irreducibly diverse and discontinuous. Unfortunately, case studies do not add up to good history. In the next section I give five reasons why they do not, based on my reading of EWP.


CASE STUDIES

Our case studies are incomplete. If history is a puzzle, our case studies account for only a fraction of the pieces. This is the mildest complaint about case studies, and not one that WT often expresses, if at all. WT thinks that current historiography does leave undocumented vast areas of the history of science, but in his view the solution is not to write a vastly greater number of case studies.

They are also unsynthesised. One reason case studies are not enough is that (to return to the puzzle metaphor) there is a difference between collecting the pieces in a box and assembling them on the board. The problem with the case study approach is that “you do not organise, arrange, cross-reference, and rearrange your knowledge.” The “consolidation of gains,” otherwise known as progress, is either forgotten or frowned upon.

One substitute we have for synthesis is to write “commentaries” that have a combative but otherwise ill-defined relationship with existing literature on a topic.

Another substitute is to “assemble thought-provoking images together in various thematic combinations, like they do in galleries of modern art.” This thematic approach is alarmingly easy to parody, as WT does in the following:
The role of ‘hats’ in history is not well studied.  This conference, which may result in an edited volume, aims to work toward an understanding of the importance of hats in a variety of different periods and locations.  We define the term “hat” broadly, and it could be taken to mean bonnets, helmets, or even hoods.
The problem with the thematic approach to synthesis is that our themes often do not pick out a coherent historical entity. Coherent historical entities are things like “the French Revolution” and “18C natural philosophy.” Non-entities include things like “everything that happened in January in the 18C” or “all scientists whose name begins with F.” WT thinks that the theme “representation” is a historical non-entity in this sense. That is, there are many instances of representation in the history of science, but there is nothing for a historian to say about those instances as a whole because they all had different audiences and served different functions.

There are better ways to synthesise. Historians can try to identify long-term “traditions of practice,” study the evolution of organizations and groups of people, attempt high-level syntheses of (say) pre-1600 science, or at least pay as much attention to chronological questions as they do to epistemic ones. And if we cannot do these things ourselves, we should give credit to the “workhorses”--prolific, thorough, well-organized scholars--who do it for us.

We do not have a navigable archive of our case studies. But there is another, more basic problem than the failure to synthesise. To make synthesis easier we need to collect all our case studies in one place and have some way of finding the ones that deal with particular themes or topics.

This is the scholarly equivalent of sorting puzzle pieces into basic categories and laying them roughly on the board. Each time we want a particular kind of piece--say, a blue piece for an area of sky, or a piece whose shape shows that it goes on the puzzle border--we do not want to have to search through the entire box looking for instances of that kind. If we are interested in a person or period, we want a quick way of accessing the sum of historical writing that has some concrete historical connection to that person or period.

Concreteness is crucial here. Suppose we have a historical interest in Galileo's experiments on free fall. If this is a genuinely historical interest, we would want to know about experiments on free fall and related topics among Galileo's contemporaries, and not about a nineteenth-century case study with some loose thematic connection to Galileo (say, the theme of idealized experiments). Presumably WT would like historians to create historiographical versions of his array of contemporary American physicists (ACPA), a cross-referenced data-bank of basic information about 20th century American physicists.

WT thinks that no such navigable archive currently exists for the historiography of science. Granted, we publish the occasional bibliography or handbook. But the former are unweildy and the latter only useful if the researcher knows in advance what she is looking for. In general, the discipline shows an “almost criminal neglect of creating and maintaining professional-level guides and reference materials.” The leaders of the field may well possess synthetic, high-level views of past literature, but we have no means (and apparently no interest) in making the private knowledge of leading scholars available to the masses.

This lack of wholesale archival projects is made worse by bad habits on the level of individual books and articles. In particular, footnotes tend to be “an opportunistic endeavour” in which the author gestures towards a “grab-bag” of well-known works without specifying the historical relevance of those works to the book or article in which the footnote appears.

Our case studies are marred by over-generality. WT has some criticisms which suggest that the case study tradition is not just incomplete but also harmful. In other words, the pieces we have gathered are not just unassembled but unsuitable for assembly. One of these criticisms arises from the fact that case studies in the history of science aim to serve as instances of some very general theme or thesis. This leads to at least four problems.

Firstly, it means that although our case studies contain useful information, this information is hard to access because the authors feel they have bigger fish to fry. The case study method “actually makes a piece of writing say less by trying to say too much.” WT has given some examples of books that would have been better if their authors had not tried so hard to be Wittgenstein, or had offered more than “window dressing on a basic Latourian sociological point.”

Another problem with over-generality is that nothing new is said. “No new ideas can easily emerge from the Gallery [of practices] because the Gallery was selected in the first place through the terms of a tacit narrative or according to words seemingly drawn out of a hat, the properties of which are already understood.” At most the case study acts as a positive or negative instance of a general thesis, but this does nothing except increase or decrease our confidence in the thesis. It does not tell as much about the past.

A third problem is what cognitive psychologists call “confirmation bias.” This term does not appear on EWP, but the idea does. Confirmation bias is the mental habit of attending to data that confirm a pet theory and ignoring data that are inconsistent with the theory or are indifferent to it. For example, we are guilty of confirmation bias if we “take away a conclusion about the relationship between science and the Cold War because we have selected events indicative of our preconceived understanding of what the relationship between 'science' and the 'Cold War' was like.” One reason for synthetic history and for navigable archives is to overcome sampling problems like this one.

But even if the sampling problem could be overcome, a fourth problem would remain. “Even if a reading of the Gallery manages to escape [confirmation bias] it is still possible to abstract an idea or practice from the Gallery that makes little sense removed from a system of ideas of which it was initially a part.” Ironically, this tendency to yank ideas from their contexts is a standard criticism of the old history that we are supposed to have escaped from.

Our case studies suffer from perspectivism. Another reason why our pieces are unsuitable for assembly is that the pieces themselves have not been formed in a co-ordinated way. Different authors deal with different topics from different perspectives: there is “a fracturing of perspectives between museology, sociology, critique, epistemology, and historical explanation.” New approaches appear from art historians, literary historians, feminist historians, sociologists, “and on and on.” Even individual authors sometimes “veer disorientingly back-and-forth between actors, themes, and time periods.” And this perspectivism has exaggerated the degree of diversity and discontinuity between different people, events and traditions, and even between different aspects of the same people, events and traditions.

To be sure, any attempt at scholarly synthesis involves tweaking some of the elements to be synthesised. But today the problem is exacerbated the “Rashomon Posture” that excuses or even celebrates historiographical chaos on the grounds that the past really is chaotic, or that its order is beyond the reach of historians. Even Michel Foucault would have disapproved of “the simple accumulation of interpretations” now practiced by historians of science.

I may have put words into WT's mouth here by presenting an argument that is supposed to cast doubt on our intuitions about the disorderliness of the past. The argument is that some of the apparent chaos may just be an artefact of the different styles, tools, and interests of the different historians who have studied the past. Still, this argument seems to be in the spirit of WT's thoughts on the topic. It is also a nice inversion of one part of Will's Picture mentioned above. According to historians, the diversity and discontinuity of the past is part of the justification of the case study method. I am suggesting that the diversity and discontinuity of the past is partly an artefact of the case study method.


ANTI-PHILOSOPHY

Philosophy is a fact of history. “Anti-philosophy” in the history of science is the preference for explaining the beliefs and actions of scientists in social or cultural terms rather than in terms of such things as data, evidence, and sincere argument. The obvious problem with this preference is that scientists do engage in sincere argument, and do revise their beliefs in light of data and evidence. Historians of science should be able to describe “things like being legitimately converted from one position to another by evidence and logic, which, we must presume, has indeed happened from time to time in the history of science.” To presume otherwise is just absurd.

Philosophy is a core fact for the history of science. Sometimes WT goes further, suggesting not only that philosophical explanations have a place in the history of science, but that they are at the centre of the discipline. The last President of the History of Science Society tried to “bring back a balance of interests [to the field] – science at the core, along with plenty of room for social history, economic history, political history, environmental history, and so many other histories.” WT seems to endorse this idea that “science”--in the sense of data, evidence, and sincere argument--should be at the “core” of the discipline.


PUBLIC ENLIGHTENMENT

Our insights are not ours alone. Our project of enlightening the general public assumes that there are a lot of unenlightened souls out there. But this may not be the case. Our historiographical amnesia (discussed in my next post) means that we exaggerate the prejudices of past scientists and historians, and of past actors in general. As for people in the present, we may be mistaken about their level of ignorance if WT is right that we sometimes ignore other science professionals such as journalists and museum curators.

The resulting complacency does nothing to improve our own writing: “it’s my fear that to not even mention the existence of these professionals … will only serve to sink our discipline further into delusion about the quality and relevance of our work.”

We may also exaggerate the prejudices of the public because we have found those prejudices in academic books, and imagine that the public has read the same books. Here is WTs version of this idea: “Is 'everyone' supposed to have read [the now-unfashionable sociologist] Robert Merton? Or did Merton (or maybe those pesky textbook history boxes) penetrate the 'public' imagination in a way that newer science studies people have yet to do, but if we only present enough department seminars we are sure to?”

Our insights are not really insights. The things we call insights are either false, trivial, or incoherent. Take Thomas Kuhn's idea that past science is strewn with data and theories that were neglected not because they were wrong but because they were inconsistent with the dominant theory of their period. WT cautions that this view can “overemphasize the extent to which science naturally encloses itself within silos, thus requiring some external intervention to free it from its myopia.”

Then there is the “banal” idea that “science is not context-independent.” As for incoherence, this can be found “by descending into the usual sociological hell of illustrative case examples, labyrinthine jargon, and funny diagrams (here, things like the 'Periodic Table of Expertises').”

Falsity and triviality are combined in "post-modern equivocation", a term that appears in this book by John Zammito. Post-modern equivocation is the habit of making ambiguous claims that are true but trivial on one reading and non-trivial but false on the other reading. By conflating the two readings, or by jumping from one to the other at convenient moments, post-modernist thinkers appear to say things that are both true and non-trivial when in fact they are always either false or trivial.

Our insights are too general to be useful. Suppose that our insights really were ours, and that they really were insights. Even then, they would be too vague to be of much use in real-life situations such as policy debates. At most they can serve as “cogent reminders” that do not help our audience or our own discipline. “Only novel and pointed contributions to public discourses will elevate the profession in the public eye.” Again, complacency threatens: “blanket warnings against dogmatic views of science (or blanket advocacy for science) as grounded in the examples of historical incidents will mainly serve to heighten a professional sense of self-righteousness.”

Our outreach is poor anyway. Historians sometimes write and speak as if lots of non-historians are reading and listening. Sometimes we are right about this, but more often we are deluded. Note the jibe about “department seminars” five paragraphs ago. Note also WTs puzzlement at reviews in obscure scholarly journals that attack popular science books. (But note this response to WTs puzzlement.)

Outreach conceals our real gains. Blanket warnings not only encourage complacency in the profession but conceal the good work we do (WT does not deny that good work exists). In our rush for public relevance, “we have played down our more specific accomplishments in favor of highlighting a few key insights that can be directed outward to the public.”

Outreach is not history. WT ends his recent four-post summary with “a warning that the imperative of engagement with broader audiences can — though it need not — displace other scholarly norms at the core of professional identity.” The suggestion is that our over-enthusiasm to engage with the general public is partly responsible for the things historians of science now do, and hence for the problems that they now face.

***

If these problems are as serious as WT makes out, why do they persist? See my next post for an attempt at an answer.

1 comment:

  1. goedkoop nike air max schoenen , een combinatie van elegante stijl en geavanceerde technologie, een verscheidenheid aan stijlen van goedkoop roshe, de aanwijzer loopt tussen uw exclusieve smaakstijl.

    ReplyDelete