Some Thoughts on Jasanoff, “Science and Reason in the Public Sphere”: Symmetry and Realism

Yesterday afternoon I had the pleasure of hearing Sheila Jasanoff give a talk as part of York University’s STS Program’s Seminar Series, in which she spoke on the subject of “Science and Reason in the Public Sphere.” For good reason is she one of the more prominent scholars in STS. I have seen Jasanoff speak once before and have read a variety of her works, and overall, she strikes me as a careful, measured, and eminently reasonable scholar.

However, I find myself at odds with various aspects of her work, especially the broader ways she conceptualizes science, its public role, and “our” relationship to it as STS scholars, and a lot of these issues were prominent in her talk. What follows is not meant to be a rigorous critique, just a general reflection, or perhaps, that convenient way of deflecting criticism and meandering around ideas, a “provocation,” about some of the broader problems that I think surround these topics. Indeed, none of these issues are uniquely specific to Jasanoff, they are broad issues that permeate STS. Mostly, they are just things that I found myself thinking about while listening to her speak, which is testament to the fact that she gave an excellent talk.

Two issues that came up were the classic problems of symmetrical analysis, and realism as an expectation from science. These are very old issues that have been discussed ad nauseam elsewhere. Most scholars are firmly entrenched in their positions on them, so it’s unlikely anything said here will change any minds, but discussing them will at least help frame what follows in later posts and perhaps introduce someone somewhere to some new ideas. 

In a previous post about Michael Shermer’s perceived “liberal war on science,” I noted that he adopts a “sociology of error” – an asymmetrical form of analysis – in trying to attempt to find psychological explanations for why it is that 20% of Democrats in the US don’t believe in anthropogenic global warming, while assuming that assent to the AGW theory by the other 80% is somehow indicative of their rationality. In other words, Shermer assumes that the AGW theories are true, and that this fact about the world somehow constitutes an explanation of why people believe it to be true. Or, perhaps slightly more fairly, Shermer assumes that such a theory is the best available explanation for observed global warming, and that the 80% of Democrats who believe in the theory do so because they are capable of making scientifically valid judgements about climate change and recognize it to be the best theory available. There are philosophical issues involved here, but even disregarding those, it is quite clear that the 80% of Democrats who believe in climate change don’t do so because they are qualified experts in the field of climate science. Shermer would argue that climate change sceptics disbelieve in AGW because of “social” or “political” factors, but it is also the case that people who believe in AGW do so because of “social” or “political” factors.

Science and Technology Studies (STS) scholars are trained to look at scientific disputes “symmetrically” – that is, they do not simply try to account for the social factors accounting for false or irrational beliefs, but also those accounting for beliefs that are held to be true or rational. So, in the above example, they wouldn’t only try to explain why people don’t believe in AGW, but they would also try to explain why people do believe in AGW. However, this has caused a great deal of consternation in trying to figure out what normative role STS scholars can have, since it is supposed that we are not allowed to accept that scientific facts “really are” correct or rational; rather, they are the result of social negotiations among scientists. In other words, scientific facts are “social constructs,” not representations of reality.

[Edit: This is a perhaps a silly construal of what is meant by social construction. Sorry for repeating it. Without going too much into it, few thinkers hold that representations of reality are in contrast to social constructs. More reasonably, one version of a social constructivist position holds that representations of reality are social constructs, in that there are many ways we can represent reality, and moreover, there are particular (social) reasons that we should prioritize representing certain realities over others. Very few, if any, scholars think that science does not represent reality in more or less accurate ways. They mostly challenge that such representations are incomparable, or uniquely superior to others, though the details of such comparisons need to be flushed out with specific examples. Sometimes certain representations do appear unparalleled and especially accurate. There is no implication of epistemological equivalence here between all possible representations of some phenomenon. However, this is all normatively and philosophically speaking, that is, in the realm of discussion about how the world “really is.” Returning to methodological impartiality, the supposition is that STS scholars should not accept the truthfulness, or the accuracy of some representation, as an explanation of why that representation comes to be accepted].

The philosophical basis for this view is that realism or correspondence theories of truth are wrong. (Long editorial note: While there are certainly valid philosophical objections to these notions, it is worth pointing out in the name of reflexivity how uncritically these proclamations are bandied about. Very rarely have I encountered thorough-going critiques of realism or correspondence theories that are fully sensitive to the range of philosophical debates on these matters (though they exist). Graduate students are taught to rehearse these idioms, though when pushed, they are barely, if at all, able to articulate why it is that realism or correspondence theories fail, without merely begging the question on social construction. Such accepted wisdom is as egregious as any disciplinary indoctrination that STS scholars find amongst scientists/philosophers of science.)

In any case, because of this philosophical underpinning, STS scholars often uncomfortably find themselves trying to explain why it is that anyone should listen to scientists if what they give us aren’t actually facts about the world. One major avenue is to simply evade the question and proclaim that STS scholars are not in the normativity business – they make no claims about whether scientists are doing good or bad science, they are simply offering sociological analyses. Others find themselves wanting to say that we should listen to what certain scientists have to say, but not because they have facts about the world, but rather because they are a trustworthy epistemic group (a criteria that can be satisfied a bunch of ways), or some other reason that can’t be said to rest on realism.

Jasanoff positions herself somewhere among this second group. The notion of a constitution is prevalent in her views on public science. She supposes that we need something like a constitutional expert governance in which the rights and responsibilities of expert institutions with great public import and authority are clearly outlined. In her talk, she contended that STS scholars (and it was unclear if she intended for this to apply to the general public as well) should not base their trust in scientists on traditional notions of scientific accountability – how well they meet Mertonian-type norms, how transparent their work is, how solid the evidence is, etc. – but instead on the degree to which they offer constitutional accountability. Thus, first and foremost STS scholars should be doing symmetrical analyses and asking questions about what would make the authority of science justifiable, that is, consistent with democratic and constitutional ideals.

After her talk a question was raised concerning her position, which went something like this: if we, as STS scholars, notice in our studies that there is a particular institution or regulatory body or group of scientists, that seem to be colluding with say industry forces to minimize the risk of some technical undertaking (the specific example used was about the Nuclear industry), and this can be shown because the rest of the scientific literature – that is to say, the facts – shows that this group is misleading people, what’s wrong with saying so? What’s wrong with saying these scientists are getting the facts wrong and it’s demonstrably because they are influenced by ulterior motives (i.e. “politics”)?

Firstly, Jasanoff responded by saying that it’s troublesome (and asymmetrical) to talk of a particular scientific claim being influenced by politics, since all science is political. This is a common refrain in STS, but one I often find frustrating. Its tautologous nature serves to conflate all senses in which science can be political, and thus stifles important discussions that could be based on further distinctions. By stating that all science is political, or similarly, that all knowledge is socially constructed, one is trying to further demolish the asymmetrical “realist” view of scientific knowledge. One cannot differentiate, so the symmetrical argument goes, between good and bad science by claiming that the former escapes political influence (and thus gives us real knowledge about the world), while the latter is implicated in it. STS shows us that all science is political, so there need be some other criteria of good and bad science.

Yes, but all science is political in different ways. Science can be political in the sense that some scientists try to have their work be empirically verifiable, robustly supported by relevant data, consistently and thoroughly analysed with accepted methods and tools, and presented as transparently as possible, because if one does not abide by these social norms their work will be dismissed and they will allowed to participate in a particular scientific cultural group. Or science can be political in the sense that some scientists take money from tobacco companies or chemical companies to produce work that does not meet scientific norms in an attempt to mislead the public and politicians about the dangers of cigarette smoke or pesticides. Both are political but in drastically different senses. The latter sense is typically what is understood in common parlance when someone says that science is political. While I completely concede to the underlying theoretical point of pointing out that all science is political, it tends to demolish the distinction between being excellently political and being political in a terrible sense.

Beyond this, she re-emphasized the point about realism – as STS scholars “we” all know that realism is untenable, so we shouldn’t say things like one group is mislead by politics or economic incentives and it is the facts that tell us so. Instead, she said we could couch this critique in our goal of constitutional accountability – if one group out of many is not conforming to accepted standards, then this is all we need to point to this inconsistency and question the degree to which this might affect accountability. It would be something analogous to noticing that a judiciary body does not have accepted standards or that there is a powerful group of lawmakers that are scorned by other groups of lawmakers. This might cause us to be apprehensive about the legislative body’s capacity to guarantee their constitutional obligations. (Or something like this. It was an extemporaneous answer and I am writing from memory. I hope I am not misrepresenting her position. In any case, she’s not the only person to position themselves in such a way, so hopefully it will suffice for the sake of argument).

But there is a persistent problem with this middle position that allows analysts to be normative but does not require them to trust scientists because we believe them to give us “facts” about the world. It’s all well and good to say that there are a slew of social factors that establish some group of experts as trustworthy and others as untrustworthy. We might even say that the most important social factor is how well a group of expert follows the established epistemic norms of their field. But in this case, there is a lingering question about the precise details of those epistemic norms. For many, what constitutes the epistemic norms is far more important than whether a group of experts faithfully follows them. Indeed, one should have little trust in an expert group with dubious epistemic norms, no matter how consistently they are applied.

A consensus is an important and noteworthy thing amongst a group of experts. It should indeed instil a great deal of faith in the knowledge for which it stands. However, though most climate sceptics (for example) make the point almost wholly rhetorically, they are right to insist that a consensus is not a sufficient condition for truth (or whatever less philosophically offensive term you want to use here). The possibility exists, however remote, that the vast majority of experts in a field have colluded to manufacture a consensus about a false theory. So in this case one needs to resort to questions about methods and rules of inquiry.

The more cynical part of me wants to suggest that STS scholars have dug themselves into a theoretical hole. They have spent a great deal of effort to demonstrate that the way science works is not really how scientists (and certain philosophers) think it works, scientific truth really isn’t what scientists (and certain philosophers) think it is, and what makes scientific expertise valuable is not what scientists (and certain philosophers) think makes it valuable. So when bluntly and forcefully asked, “But why should I believe anything scientists have to say if what they give us isn’t accurate knowledge about the world?” an appealing position is to evade the question altogether, or say something about the trustworthiness of certain social networks. They simply do not want to have discussions about epistemic norms per se.

When pushed, however, a third and very popular response presents itself: No one denies that scientists give us useful knowledge about the world, just they don’t give us truth, and that science is not the only way of knowing about the world, and it shouldn’t be afforded as much cultural authority as it has since it’s mostly rhetorical (on that note, here’s a topic for a later post: Are we all just instrumentalists?). But if the discussion reaches this point, we’ve already unwittingly entered the realm of epistemic norms. What is it about scientific knowledge that makes it so useful? Might it have something to do with the principles by which it is constructed? And might there be better and worse ways of constructing useful scientific knowledge?

To have these discussions one need not subscribe to realism or correspondence theories or belief in “facts” or whatever other strawman one wants to say scientific knowledge cannot be held to rest on. And I am perfectly happy to have these discussions. I see absolutely no reason to adopt some sort of meta-impartiality according to which, in virtue of my training as an STS scholar to be able to do symmetrical analyses of the sociological aspects of science, I am not allowed to make normative judgements about the epistemic norms used by scientists. Frankly, I am surprised that a meta-impartial, anormative position is seriously entertained. And let me be clear, I’m not talking about making normative judgements about science well outside my purview as an STS scholar, like when talking at the bar or something. Impartiality depends on the purposes of your study. To believe it to be an institutional requirement of being an STS scholar, or that a study doesn’t qualify as an STS study otherwise, is far too restrictive of a disciplinary standard.

Perhaps one might reply that I simply don’t have the qualifications to make such judgements, that scientific norms are too varied and specific to their context to be able to apply any general philosophical reasoning to these issues. Well, I would say it depends on the scale of the question. Am I able to make judgements about the code used for one climate model as opposed to another? Probably not. Am I able to note that one group relies on systematic evidence and detailed study while the other relies on specious reasoning and rhetoric? Arguably yes. Of course, the most technical debates are typically not so black and white, but it is abundantly clear that at many levels one is perfectly capable of making reasoned judgements about the epistemic merits of scientific knowledge claims. And here we find the reasons for many STS scholars’ reluctance to have normative discussions: it might turn out that we should believe certain scientific knowledge claims for essentially the reasons scientists say we should.


  1. Heheh. I really quite like this post. I think I agree, overall, with your conclusion–that there is the perhaps slightly-distasteful possibility that STS scholars “should believe certain scientific knowledge claims for essentially the reasons scientists say we should.” My (partial) response however is that the vast majority of knowledge or facts that humans interact with are “scientific” and political in ways that vindicate STS scholars. That is, I think the important bit of science is testing your ideas against “reality”. This is an imperfect practice in numerous ways, but resembles Jasanoff’s procedural point in that I think testing anything against “reality” entails engaging dialectically with other people. Other people may interpret the “reality” differently, but where it is possible for an idea to be put into practice repeatedly (or otherwise “tested”) it seems to me viable to say that a truly dialectical process where groups observe and compare their perceptions of reality should lead towards “truth”, all things being equal, because in the long-run ideas that truly bump up against reality will (maybe?) show this in their failure (socially and/or instrumentally). Now, this needs wagonsloads of caveats to be properly STS-y 🙂 but I guess my overall point is twofold:
    a) The “process” of science, while neither perfect nor monolithic, (should) necessitate accountability to both other people and to observations, and this seems like it will, in the long-run and more often than chance, lead to “science” (theories, beliefs, ideas) that better align with external reality than their predecessor beliefs. (This becomes tautological if one doesn’t assume the existence of an objective reality, but I do, even if objective reality is something viewed through a glass, darkly, from across the room, by partially blind people.)
    b) The process of living as a human involves belief in any number of things because they are socially/culturally/politically expedient and lack easily interpretable feedback contradicting them, and/or we lack the processing power and time to check all of them using the process I outline above. I use the example of relationships in my classes–most of us would consider our relationships with our families as one of, if not *the* most important things in our lives. Yet even the majority of scientists do not demand that all their child-rearing or spouse-communicating techniques be backed up by rigorous, peer-reviewed published scientific inquiry. It would be, among other things, impractical. Almost all of us instead rely on anecdotes, our own imperfect memories, inductive reasoning, and “intuition” to steer us through these so-important relationships, with, at best, some sprinkling of “scientifically validated” techniques gleaned from secondhand news reports, popular books, or very occasionally, peer-reviewed literature. Yet we seem to do ok. Clearly, it is not true that we can or do only act on scientifically proven ideas to make important decisions (who demands empirical tests on the likelihood of a successful marriage before proposing?), and we act on information coming to us “symmetrically”–a physicist doesn’t insist on getting a PhD in psychology before thinking about how they will raise their children but rather rely on information social cues (and contact with reality) tells them is “true.”

    I’ve already written too much, but suffice it to say–great blog! 🙂

    1. Thanks for the comment. Sorry it took so long to respond, I’ve been away from blogging for a bit.

      I agree with the point about science being thoroughly political, but I also suggest that science is political in such variable ways, that it is often empty to criticize expectations for political neutrality from scientists (especially insofar as this leads to doubting, in a general sense, the capacity for science to give us very useful and important knowledge about the world), since what is being expected is a very particular kind of political neutrality.

      When I say that we should believe scientists claims about climate change, for example, precisely for the reasons they offer, I’m sure it’s clear that I don’t mean that we should believe them in virtue of the fact that they are scientists, and science is the best, and we should always believe scientific facts. The reasons climate scientists offer are complex, but based on evidence, certain methodologies, interpretations of data, background theories, etc. And it is ultimately those reasons upon which beliefs about climate change are founded. Of course, the challenge for STS scholars here is often the same as the challenge for the broader public in the face of scientific controversies – they simply don’t have enough knowledge to make expert judgments about competing scientific knowledge claims. But even beyond this, many STS scholars seem uncomfortable saying things like we should believe in climate change because the scientific evidence and theories are convincing, mainly because of leftover arguments about symmetry and impartiality and such. Obviously, some take these largely methodological and philosophical issues more seriously than others. I lose no sleep over proclaiming that certain scientific knowledge claims have been compromised by nefarious interests.

      As for the relationship example, I am certainly not arguing for some neo-scientism. Some philosophers and STS scholars would be heavily critical at any suggestion that rationality in general and rigorous scientific research have any structural similarities, but I’ll make the suggestion nonetheless. Different questions require different levels of rigour, but I would suggest that even in things like child-rearing, we are making decisions based on evidence and reasoning (at least purported evidence, in the form of suggestions from relatives). And somethings clearly don’t work. Heh, now that I am thinking about it, the child-raising issue is very provocative, especially insofar as there is so much easily dismissed scientistic advice out there (e.g. such and such study says playing classical music while hanging your baby upside down will increase their IQ…). And given that it’s so provocative, it’s hard to say anything sufficient here. I’ll merely speculate that there are indeed better and worse ways of raising children, and creating relationships, but these things are notoriously difficult to study (except really obvious things, like, should you let your kids play with lead paint or not). That people engage in relationships and raise their children in such variable ways, most of which seem to work just fine, I think says more about the flexibility of people (children in particular), than about whether or not grounding our child-raising and relationship-forming behaviours in rigorous scientific research would lead to better children and relationships.

  2. “That people engage in relationships and raise their children in such variable ways, most of which seem to work just fine, I think says more about the flexibility of people (children in particular), than about whether or not grounding our child-raising and relationship-forming behaviours in rigorous scientific research would lead to better children and relationships.”

    Eh, I’m not sure about this. It seems to me that the lines to judge by how they “work just fine” are hard to draw precisely, and direct causality is difficult to concretely establish, such that it is hard to know exactly how to improve and what led to what desirable or undesirable outcome. In this way, they’re a lot like my bailiwick of study, ecosystems, and I think the comparisons are similar. There are, after all, likely many good ways to manage an ecosystem, just like there are many to raise a child, but the outcomes from some appear clearly undesirable.

    And I think we largely agree; where you say, for example ” even in things like child-rearing, we are making decisions based on evidence and reasoning” — I agree completely. My point, as someone who has been embroiled in some “good science/bad science” debates is that there’s been steady–seemingly increasing–critiques of certain things as “bad science”. (I think you’ve dealt with Keith Kloor in your blog, but perhaps I’m misremembering). There’s also a tendency in natural scientists I work with to shake their heads at the public’s irrationality and ignorance, as well as at the “bad science” among people who, for example, criticize GMOs. There’s been an unhelpful conflation of “conclusions I disagree with” with “bad science”, and this has bled into the old arguments on whether scientists are/should be “political”. My issue is that so very few natural scientists seem to accept that scientists are or should be (or can’t avoid being) political, even in the quotidian senses rather than the nefarious sense, and they certainly seem to me to reject the implications of this. For example, one implication of it to me is that, if we are all “political” in various ways, it is better to be transparent and disclose than to presume a posture of apolitical-ness, especially since (in my field) such apolitical stances can lead to horrible misanthropic results, ignorance of real governance and power issues, and reliance on the faulty “linear model” of policy production. In short, I guess that I have been using the relationships example as one way to show that we all, not just non-scientists, use a lot of intuition, induction, and hear-say as evidence for really important things, because we cannot do otherwise. While I think promoting “rational skepticism” and more formally scientific thinking among the general public is a worthy goal, I find promoting humanism and an understanding of some of the elements and implications of “symmetry” among scientists to be as important, if not more so.

    1. I think the relationship example is engaging and provocative, and I need to think about it more. I do like the complex-system analogy, and I think that captures the heart of what I was trying to say with that comment about human adaptability.

      In saying that humans are so flexible that we can raise children and engage in relationships in innumerable ways, I was trying to point to the inescapability and conspicuousness of the social constructedness of these things (much more conspicuous, arguably, than the social constructedness of science). So yes, you’re right, it’s precisely that so many things “work just fine,” or rather, what “works just fine” is too a social and cultural construct, that makes it difficult to say anything definitive about whether “scientific” child-rearing is “better.” Child psychology makes a great STS case study, because the research questions and priorities are so disparate based on where the research is being done. Some cultures are preoccupied with IQ and intelligence, others discipline, others empathy, others curiosity, and so on.

      I would wager that there is much “traditional” child-rearing advice that certainly does not have the intended effect (and even more “scientific” child-rearing advice that does not have the intended effect). We can’t know for sure, however, because, again, these systems are irreducibly complex, and there are so many confounding variables. Even trying to determine whether exposure to some chemical in children has some behavioural effect is notoriously difficult because to control for the host of other possible social and environmental variables is overwhelming (I remember reading a recent study finding a correlation between marijuana use and decreasing IQs in teenagers, and thought how difficult it would be to establish causality, since teenagers who smoke pot typically engage in many other activities that would correlate to decreasing IQs – and the study did not control for most of these). So how could we possibly establish the causal efficacy of things like piano lessons, soccer practice, curfews, video-games, “educational” baby toys, painting your child’s room a certain colour, etc.?

      But, I would also wager that so much of the advice that you get from mother-in-laws, and friends, and all other people who are eager to give their input about how to raise your children, is inconsequential. Perhaps in part because it actually makes no difference one way or another, or perhaps, even if mildly detrimental to a child’s “development,” children are remarkably resilient and adaptable. And of course, this a lot of good advice, like “eat your vegetables” and, “colouring is fun,” that doesn’t need a foundation in (formally) scientific ways of thinking, which is your point. But then again, “eat your vegetables” is consistent with scientific knowledge (though I am unfamiliar with the state of research on the funness of colouring).

      Finally, I think your example is really provocative, precisely because there seems to be an increase in desire for “scientific” (certainly, scientistic) advice for child-rearing, if the amount of books on the subject and the number of child-psychologists on talk-shows is any indication. I wouldn’t be surprised if there is already an STS case study on the “scientization” of raising children.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s