Social science isn't "science"?

By Razib Khan | August 2, 2010 9:53 am

Update: The title is way too strong as a reflection of my opinion. I’ve added a question mark.

A friend once observed that you can’t have engineering without science, making the whole concept of “social engineering” somewhat farcical. Jim Manzi has an article in City Journal which reviews the checkered history of scientific methods as applied to humanity, What Social Science Does—and Doesn’t—Know: Our scientific ignorance of the human condition remains profound.


The criticisms of a scientific program as applied to humanity are deep, and two pronged. As Manzi notes the “causal density” of human phenomena make teasing causation from correlation very difficult. Additionally, the large scale and humanistic nature of social phenomena make them ethically and practically impossible to apply methods of scientific experimentation. This is why social scientists look for “natural experiments,” or involve extrapolation from “WEIRD” subject pools. But as Manzi notes many of the correlations themselves are highly context sensitive and not amenable to replication.

He concludes:

It is tempting to argue that we are at the beginning of an experimental revolution in social science that will ultimately lead to unimaginable discoveries. But we should be skeptical of that argument. The experimental revolution is like a huge wave that has lost power as it has moved through topics of increasing complexity. Physics was entirely transformed. Therapeutic biology had higher causal density, but it could often rely on the assumption of uniform biological response to generalize findings reliably from randomized trials. The even higher causal densities in social sciences make generalization from even properly randomized experiments hazardous. It would likely require the reduction of social science to biology to accomplish a true revolution in our understanding of human society—and that remains, as yet, beyond the grasp of science.

At the moment, it is certain that we do not have anything remotely approaching a scientific understanding of human society. And the methods of experimental social science are not close to providing one within the foreseeable future. Science may someday allow us to predict human behavior comprehensively and reliably. Until then, we need to keep stumbling forward with trial-and-error learning as best we can.

CATEGORIZED UNDER: Social Science
  • Pingback: Tweets that mention Social science isn’t “science” | Gene Expression | Discover Magazine -- Topsy.com()

  • dcwarrior

    Well, is weather prediction a science? Weather is chaotic and is fundamentally unpredictable. People are even more so.

    I agree that many times the social sciences devolve into philosophy or politics but the attempt is still there. Isn’t the “science” part of it the method, not the success?

    After all, we aren’t saying the study of “dark matter” isn’t science, even though those studying haven’t a clue as to what it is.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    dcwarrior, i think the attempt to think clearly, and analytically, about social topics, and the use of quantitative methods, are all important. the title probably overemphasizes my skepticism. though most people are unfortunately underinformed about how context sensitive a lot of social science findings are (e.g., they aren’t easily replicated, so hinging an argument on positive finding X is probably not sound practice).

  • Chris T

    Psychology is probably the closest to becoming a true science. It finally has access to tools that allow it to begin examining the physical sources of human behavior.

    Unfortunately, much of the rest of social science is advocacy in the guise of science. Even worse is that their advice is sometimes used to make federal policy (particularly in education).

  • http://www.factsandotherstubbornthings.blogspot.com Daniel Kuehn

    How exactly do you define science, Razib? Maybe as you say your title came out a little strong, but I have much the same reservations about this post as dcwarrior.

    Too often, I find physicists, chemists, and some biologists confuse “experimental science” with “science”. Social science is harder to derive conclusions from – if that’s all you’re getting at, I agree with you. But I’m not so sure the challenges faced by social science are tantamount to a refutation of its status as a science. If anything, social scientists (particularly economists) come out of the whole ordeal even more attuned to concerns about causation than more experimental scientists do.

  • onur

    If social sciences aren’t science, then what are they? Disciplines of humanities? If yes, then the distinction, the wall, between social sciences and humanities collapses and all of them are bound together under the category of humanities never using the social science category again. That would relieve the scholars of today’s social sciences, as the pressures and expectations on them to be scientific would diminish significantly. I think that would be more realistic given the apparent limitations to scientificity in today’s social sciences (as summarized in the above article). BTW, I am excluding archaeology here, as it is a science by all criteria.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    i can’t defend the title, so i’ve posted an update. thanks for holding me to account :-)

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    more specifically, i probably could defend the title, but it would be rhetoric. i probably agree more with some of the criticisms here than the full literal interp of the title.

  • Chris T

    Just rename it to “Is Social Science science?”

    The current title just looks bizarre now.

  • http://occludedsun.wordpress.com Caledonian

    Weather is chaotic and is fundamentally unpredictable. People are even more so.

    But the weather isn’t noticeably changed when we observe it. Human beings are notoriously sensitive to the observation of human beings.

    Imagine physics was forced to confront all of the paradoxes of quantum mechanics from the get-go, instead of dealing with comfortable and predictable macrostate phenomena first. That’s something like what cognitive psychology has had to cope with.

    If social sciences aren’t science, then what are they?

    There’s a word for things which aren’t science but present themselves as such: ‘pseudoscience’.

  • http://www.factsandotherstubbornthings.blogspot.com Daniel Kuehn

    This is not to be argumentative – just to (anecdotally) point out how tenuous the relationship between experimentation and science is – I’m watching some “Shark Week” while prepping some food right now – so far I’ve seen them work through some theories on three different sharks. Every single case of data collection that is reported on for each of these three theories has been of the “tag them and observe them” sort. Observational data – no experimentation to speak of.

    Funny… the Shark Week narrator keeps calling the guys conducting these observations “scientists”!

    I think it’s easy to slip into the “science as experimentation” mode if your particular science relies on experimentation. But if one starts paying closer attention to the “science” that they hear and read about, an immense amount of it is observational empiricism rather than experimental empiricsm, which is exactly what social scientist do.

    The problem is, as the City Journal article alluded to, social scientists are taking the organism with the most complex brain we’ve ever seen and observing their interaction with billions of other equally complex individuals. Instincts, preferences, feedback, information processing ability, etc. are all considerably more complicated for humans. That’s what we’re left to deal with.

    I think outsiders (btw – probably worth mentioning at this point that I’m an economist) are also probably unaware of distinctions between social sciences too. Economics and psychology are on considerably firmer footing than, say, much of anthropology, sociology, and political science. Some of this is for the same reason that physics is on firmer footing than economics – because the subject matter of sociology is a lot more complex than the subject matter of economics or psychology. But a lot of this difference is also because of ideology or philosophy is explicitly incorporated into knowledge creation in disciplines like political science and sociology in a way that it simply isn’t in economics and psychology.

    The other advantage I see in something like economics is that economists have been pretty good about distinguishing between normative and positive claims. Values and consequences for the social sciences are important. That’s fine – that’s why it is done. But it’s also something we need to be explicit about.

  • http://www.factsandotherstubbornthings.blogspot.com Daniel Kuehn

    Caledonian –
    You make awfully strong claims about social science, but when they are prefaced with something like this: “But the weather isn’t noticeably changed when we observe it. Human beings are notoriously sensitive to the observation of human beings” that clearly demonstrates a lack of understanding of how social science is actually done, I’m wondering if you’re informed enough to make those claims.

    I’ve always felt that social science is simply a branch of biology. Do you think humans are any more sensitive to being observed than any other complex organism we set ourselves to studying? Do you think we face any challenges in understanding human social organization that we don’t face studying the social organization of other complex animals?

  • http://www.mazepath.com/uncleal/ Uncle Al

    The USSR, China’s Great Leap Forward, Castro’s Cuba, Cambodia’s Khmer Rouge, Israel’s kibbutzim, Peru’s Sendero Luminoso… the US War on Poverty and federal Department of Education demonstrate preponderant successes of centrally administered social science. Economics! Social science has never been sufficiently imposed, for sufficiently long times, with adequate budgets.

  • Chris T

    The primary failing in the social sciences is their frequent failure to consider their objects of study as a physical system. Rather than study human behavior as emergent from an immensely complex biological system, they often divorce behavior and the mind from physical reality and treat them as discreet phenomena. Without considering physical constraints, one can come up with any explanation one wants without regarding how it fits in with explanations of other phenomena. Thus we get a large field of isolated and often contradictory theories without any way of falsifying them (absolutely critical for science).

    Attempts to explain behavior in terms of the underlying biology is often met with derision and hostility (ie: genetics and evolution).

    Meteorologists study weather as a physical system with limits. So, even though weather is phenomena emergent from an immensely complex system, the basic processes of it can be understood by examining the underlying physical influences and causes.

  • http://occludedsun.wordpress.com Caledonian

    that clearly demonstrates a lack of understanding of how social science is actually done

    Oh, please. We’ve known for decades that people behave fundamentally differently when they know they’re being observed than when they think they’re alone. Any attempt to scientifically study human social interactions that doesn’t rely on unobservable information-gathering techniques is screwed from the start, and the hurdles involved in getting humans to study human beings are immense.

    Scientists are supposed to approach their fields of inquiry with as few preconceptions as possible, but that can’t be done with people.

  • Tim Bartik

    I have to say that this post and Jim Manzi’s article are both disappointing. I can only speak to what they omit from my perspective as an economist; the other social sciences will have to find other defenders. First, both of you appear to know about only a small number of the many experiments that have been done within economics. I don’t know what links you allow, but I would suggest that readers consider the following: (1) the many experiments in development economics conducted by the Poverty Action Lab at MIT and the New Haven-based group Innovations in Poverty Action — Esther Duflo had a recent TED talk that highlights some of this work, and the New Yorker and Business Week have recently had prominent articles that discuss both the accomplishments and limitations of this work; (2) the many experiments over many years conducted in welfare reform, job training, and education by MDRC; (3) some recent very good experimental work in education conducted by Roland Fryer at Harvard. All of this material can easily be found by a little googling. and this is just a small sample of experimental research in economics. Second, there is some good recent work that shows that some non-experimental methods can give results similar to experiments, for example, approaches that use natural experiments and evaluate results using a regression discontinuity estimation approach. And by the way, Jim Manzi’s conclusion that it is rarely possible to change people through programs does not appear to be true. Consider, for example, the research literature on the long-term effects of preschool education, for example as covered at the website of the National Institute for Early Education Research.

    Of course, most of these experiments have to do with what would be considered microeconomics rather than macroeconomics. (I should point out that contrary to what most people think, the majority of economists work on microeconomic issues, not issues of monetary or fiscal policy. ) It’s harder to do experiments or even quasi-experiments on the macro-economy. There, the results depend more on good modeling, and it is rarely possible to fully “confirm” models, rather it is only possible to falsify models.

  • Mariana

    Uhr, yeah, you don’t wanna mess with economists.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Isn’t the “science” part of it the method, not the success?

    NO! NO! NO! It’s the success.

    The reason why people want to be scientists is that physics, chemistry, and biology have been so tremendously successful, and in particular, so much MORE successful than folk physics, folk chemistry, and folk biology. Claiming to be scientific because you use methods borrowed from the successful sciences, even though they don’t work, is just mimicry.

    If these criticisms are true, social science loses its right to trade on the prestige of more powerful sciences. There is still a lot to be said for the collection and analysis of data, even if you can’t predict the future and don’t have an adequate theory of social behavior.

    As far as that goes, alchemy’s theory and claims to scientific stature were bogus, but they knew a lot about the production of acids, bases, salts, and metals and the properties of these. They also invented glassware that is still used.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Here’s Brad DeLong: One of the dirty secrets of economics is that there is no such thing as “economic theory.” There is simply no set of bedrock principles on which one can base calculations that illuminate real-world economic outcomes.

    The prestige of economics has suffered an enormous collapse in the last few years, because a lot of the most prestigious economists claimed that an economic collapse like we’re suffering now had been made impossible: “The Great Moderation”. Not all economists disagreed, but the theoretical structure of economics is so weak that the dissidents couldn’t disprove the optimistic claims of the others.

    What I think has been lost, for the moment and maybe forever, are the grandiose claims to have finally become Scientific. What’s also been lost, however, has been the demand that other people be Scientific. The secular, empirical, disciplined study of society, even if it doesn’t cross the threshold of Science, still provides usable information. But the grand theories have lost a lot of their oomph.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    So, even though weather is phenomena emergent from an immensely complex system, the basic processes of it can be understood by examining the underlying physical influences and causes.

    The point of what people were saying above is that the results of this are disappointing in terms of prediction past a few days. Science has its prestige beause of its tremendous success, and meteorology hasn’t had these tremendous successes, though more so than the social sciences.

    The primary failing in the social sciences is their frequent failure to consider their objects of study as a physical system.

    This can be a failing, but it’s not the primary failing. It’s like the demand that biology explain everything down to the level of individual molecules or individual quarks.

  • onur

    BTW, biological anthropology has nothing to do with anthropology, rather it should be counted as a branch of biology and thus a natural science.

  • http://www.factsandotherstubbornthings.blogspot.com Daniel Kuehn

    Caledonian –
    You misunderstood my point. I was disputing “notoriously”, not the problem itself – which I go on to say is no worse than in biology or any other observational science of a complex organism.

    However – in many of the social sciences individuals aren’t directly observed anyway, rendering the very real problem you cite moot.

    John Emerson –
    Could you be a little more specific? As I understand it, the downturn has only really threatened one particular theoretical position in economics. The people you heard saying that we beat the business cycle where to a large extent those that had a political incentive to make those claims. The brush strokes you paint with are far, far too broad. I can quote some pretty kooky claims by natural scientists that had a political incentive to make those claims too. So again – could you be more specific? Nothing we’ve seen contradicts Keynesianism, monetary disequilibrium, or monetarist theories. They are problematic for RBCT and EMH theories – and hopefully will help us arbitrate between those new classical positions and other theoretical positions. You act like economics as an entire field of inquiry has been laid to waste by this downturn. Again – please be more specific if you’re going to make these sorts of claims. Otherwise you start to sound silly. This downturn, I think, is very important for arbitrating between a few competing theories. If you’re going to rely on some politico that scored points by claiming he beat the business cycle, then you’re really not talking social science.

  • onur

    I’ve always felt that social science is simply a branch of biology.

    Biologists would laugh at your statement.

  • onur

    The point of what people were saying above is that the results of this are disappointing in terms of prediction past a few days. Science has its prestige beause of its tremendous success, and meteorology hasn’t had these tremendous successes, though more so than the social sciences.

    However limited it is, meteorology is certainly a science.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Daniel Kuehn: First, if there are many schools of economics, then economics isn’t a well-theorized science. There are subfields of biology, physics, and biology, but not schools.

    Second, at the time the DOMINANT school of economics (it wasn’t a kooky individual or group) made those claims, nobody was able to knock them down. Other schools disagreed, but they couldn’t make a solid case against those claims until attempts to apply these principles collapsed.

    I go into this a little in comments the DeLong thread. Economics is a valuable area of study, but it doesn’t give unique answers to a lot of questions and for that reasons is an imperfect or incomplete science and should reduce its claims.

    As I’ve said many times, economics has all the right answers right there on the shelf, mixed in indistinguishably with all the wrong answers.

    My theory is that economics is a highly-skilled kind of policy study, or skilled advocacy sort of like law. No one says lawyers don’t know anything — they know a lot. But no one calls them scientists.

    And if there are areas of science where unique answers are given and everyone agrees on them, maybe that are is scientific.

  • J

    Which disciplines of social science are the most scientific? Microeconomics? IQ research?

  • onur

    Which disciplines of social science are the most scientific? Microeconomics? IQ research?

    As I said above, I don’t count biological anthropology as a “social science” or a branch of anthropology, instead it is a natural science as a branch of biology. So we are left with archaeology as the only scientific “social science”.

  • Chris T

    “The point of what people were saying above is that the results of this are disappointing in terms of prediction past a few days. Science has its prestige beause of its tremendous success, and meteorology hasn’t had these tremendous successes, though more so than the social sciences. ”

    The discussion was over method and results. However, the initial conception of what one is studying underpins both. A physical system has limits, no matter how complicated it is. Understanding that a system has limits goes a long way towards making a problem tractable.

    Meteorology has made progress precisely because it understands that, even though there are an immense number of variables involved, the behavior of weather is bounded by the limits of of those variables. By studying both the variables and the emergent weather phenomena, meteorology has achieved some ability to predict.

    The problem with the social sciences, is that they’ve been ignoring the underlying physical causes and therefore have failed to bound the phenomena they’re studying.

    “This can be a failing, but it’s not the primary failing. It’s like the demand that biology explain everything down to the level of individual molecules or individual quarks.”

    It doesn’t need to; it’s sufficient that it understands that they exist and are ultimate causes. By knowing the limits of the physical causes, biology can make what it studies tractable.

    By treating human behavior as though it is separate from the physical, social science has failed to define its limits and therefore failed to come up with adequate explanatory power.

    Attempting to explain human behavior while acting as though genetics doesn’t exist is like attempting to explain weather while acting as though the laws of thermodynamics do not exist.

  • http://www.rishon-rishon.com David Boxenhorn

    John, I have no doubt that Economics suffered a huge loss of prestige in your eyes, but I doubt that many actual economists ever thought that they could do what you thought they could do. Physics is great at predicting the results of controlled experiments, less good at predicting the results of real-world physical problems, like whether the WTC would fall down if hit by a plane (you have to do a lot of work to figure out the answer, and you might still get it wrong), and even less well at predicting the weather (which is just physics, after all). Predicting the economy is like predicting the weather. In contrast, economics is quite good a predicting the results of some well-defined economic policies. For example, when the minimum wage is increased, unemployment pretty reliably goes up. The economy, like the weather, simply has too many moving parts. We don’t have a good engineering for either one.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Meterology has achieved some ability to predict, but the results are enormously disappointing compared to what was hoped for, and this is not just the result of inadequate data or failure to find the right theory. What the right theory tells you is that the intractability is systematic.

    A lot of what’s being said here, by me anyway, suggests that the same systematic intractability is present in the social sciences.

    When social science studies the behavior of millions of hundreds of millions of people, the value of a genetic description of them all will be limited.

  • http://www.factsandotherstubbornthings.blogspot.com Daniel Kuehn

    “Daniel Kuehn: First, if there are many schools of economics, then economics isn’t a well-theorized science. There are subfields of biology, physics, and biology, but not schools.”

    You’re making these judgements because they sound good to you. You have absolutely no basis for this sort of claim. The reason why there are competing schools/theories (whatever you want to call it) is because of the compelxity of the data, and what was earlier called the “causal density” of the phenomenon being studied. Is it as determinant as other sciences? Of course not. But that doesn’t make it poorly theorized or unscientific.

    Reliable economic forecasts can be made considerably further into the future than reliable meteorological forecasts (I’m not sure what someone who suggested the contrary was thinking of, or what experience they have with economic forecasting). That doesn’t say anything about the scientific content of either – it’s a function of the type of complex system each science is dealing with.

    People have odd expectations with social science indeed. Nobody expects a biologist to be able to tell you if a particular pride or herd or pod will succeed or not. And yet for some reason when economies undergo depressions it is somehow a failure of the science. It doesn’t make sense. The task is to understand the underlying processes. The extent of predictive power from that understanding when looking at specific cases is dependent upon the nature of the phenomenon you’re studying.

    “No one says lawyers don’t know anything — they know a lot. But no one calls them scientists.”

    Because the knowledge production and collection of lawyers has absolutely nothing to do with the scientific method. The process of knowledge production in economics and most other social sciences does. You are making these claims apparently just because they sound right to you. You need to be more dispassionate than that. How is knowledge obtained in various disciplines? What is the content of that knowledge (is it objective, or is it subjective like art criticism, or is it perhaps normative like law?). Those are the standards you need to be looking into.

    And yes – some sciences have messier or more complex data and messier or more complex phenomenon of study.

  • gaddeswarup

    I remember an example from Bernard Cohn’s “An anthropologist among the historians and other essays”. I think that he was studying female infanticide among some groups in North India. He found that in one of the clans, the tradition is to marry girls upwards, that is to boys of higher economic status in the society. This involved huge marriage expenses and dowry which led to the fragmentation of the properties of many and eventually to female infanticide to preserve family properties. This sort of work seems to be like science to me. May be in general it is not easy to find such clear cut outcomes and topics and in general it may depend on the topic. Possibly in many situations only some heuristics and possible scenarios can be guessed. I think Daniel little has been trying to outline some of these in his blog “Understanding Society”

  • http://www.factsandotherstubbornthings.blogspot.com Daniel Kuehn

    John – listen to David.

    Think about geologists or seismologists too. Do they “lose prestige” when an earthquake strikes? Of course not. The phenomenon they study is by its very nature predictable in a different way. The task is to understand the process, utilizing the scientific method.

    Add to that list meteorologists, scores of biologists, all paleontologists. It depends upon the nature of the subject you are studying.

    What you’re doing is cargo-cult science. You’re trying to replicate the form of the analysis you like best instead of thinking of the nature of the problem at hand and the scientific approach to that problem. And yes, I am deliberately turning Fenyman on his head because ironically he was doing precisely what he accused social scientists of.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    David Boxenhorn: Economists have not been at all modest in their claims, and in particular they have sneered at historians and sociologists. I think that with he recent enormous failure (“Great Moderation” = “biggest downturn since 1929″) they will be forced to moderate their claims a great deal, and that will be a good thing.

    “For example, when the minimum wage is increased, unemployment pretty reliably goes up”: this fits economic orthodoxy and theory and was never empirical. Card and Krueger’s 199 New Jersey study called it into question. Results are inconclusive. The minimum wage wiki has a bunch of stuff.

    Daniel Kuehn: there’s a modest sense of the word “science” that I would credit to economics: “The secular, rational, empirical, disciplined study of society”. But that’s not the definition most people use. Science is the pot of gold at the end of the rainbow, and people who claim to be scientific claim dominance over non-scientists. What I’m saying is that either historians and economists are both scientists (and few claim that) or else neither is.

    And you can give reasons and excuses as long as you want to, but as long as the consumer of economics has to choose between multiple conflicting schools, with no way of choosing between them, people are rightly going to doubt economics claims to be a science. You need more unique, agreed-upon answers.

    This is all about whether economics gets the pot of gold. I say that it doesn’t, but it’s not economics’ fault. People should quit looing for the pot of gold. (And there are far more problems with economics than with meteorology).

    Did anyone read the DeLong link.

  • dcwarrior

    mind you, I think the hard scientists have a ton to learn from the social and behavioral world, even those which cannot be termed as science.

    A lot of the controversies of science have as a partial consequence the need to explain and convince lay people of extremely esoteric and complex matters (as writers and reporters do). I see a lot of disdain among the hard scientists of the need for THEM to get better at talking to, writing to and convincing lay people of their findings and views (as lawyers and PR people do).

    I see in these blogs a lot of “they need better science education” or, “why are they so irrational?” As if people who know nothing of what you do should just take your word for it? Political scientists, pollsters, political consultants and marketing consultants can give a lot of help to the hard scientists – just as the hard scientists say, “why don’t people defer to me in my own field” let me ask, in hearing the answers to that question, why don’t hard scientists defer to those people, in their own field?

    And finally, if the hard sciences, and by extension, the hard scientists, are so rational, why do they care about such trivalities as “prestige” and whether social scientists have it or not? The desire for prestige is itself a complex psychological and sociological phenomenon, that a completely rational individual might neither seek nor value. Yet the OP shines a light on the fact that, yes, many WOULD like to separate the hard sciences from and elevate them above, the soft sciences.

  • Chris T

    “When social science studies the behavior of millions of hundreds of millions of people, the value of a genetic description of them all will be limited.”

    You hardly need to for genetics to be useful. Just knowing the general bounds genetics puts on human behavior and abilities would be extremely useful. Heck, just recognizing that genetics does put bounds on human behavior would be a great start. (Instead of tentatively acknowledging it when pressed and then running from it as a fast they can.) Very little work has been done to make social science even remotely tractable. Thus we wind up with a wide range of active theories that don’t even agree on the basics.

    Macroeconomics has arguably gone in the other direction and oversimplified everything. (ie: all recessions are caused by a shortfall in demand or a limited money supply or treating individuals as sterile variables in equations.)

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Seismologists are reasonable, sensible people who don’t brag. Economists are the biggest braggarts in the university. Seismologists also don’t run the federal Reserve or the banking system and claim that their policy choices are scientific.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    Think about geologists or seismologists too. Do they “lose prestige” when an earthquake strikes? Of course not. The phenomenon they study is by its very nature predictable in a different way. The task is to understand the process, utilizing the scientific method.

    economists lost prestige because they made some macroeconomic assertions which were wrong, and, whose wrongness was very costly. e.g., “the great moderation.” many are also very skeptical of the extremes to which microeconomists took the rational actor model in terms of how robust it was in real world conditions.

    i grant that economics is very hard. i also know that outside of the mathematical sciences and mathematics economics has the greatest cognitive firepower per person at the graduate level. but economists regularly spoke with authority in a way in 2007 which in hindsight seems definitely more than what was warranted by the error bounds of their assertions. or at least the contingent nature of the truths which they were elucidating.

    i say this as a person with a deep interest in economics and economic history, and who respects economists and economists as attempting to explore a muddled topic, social phenomena, with a clear and crisp language. i hope economics gets further. but seriously, please check yourselves sometimes. the average economist is far smarter than the average biologist (going by GRE scores), but they don’t know their discipline as well because of how hard it is, and a little humility would help.

    tim bartik, i’m familiar with behavioral economics, and enjoy the results. i also blogged robust changes in behavior from pre-school programs in the 1960s 1 week ago. so i know about the predominant failures, and few successes, in this area. you’re more impressed with economics than i am, but perhaps you know to expect far less because you know how hard genuine robust results are.

    let’s not turn this into a shouting match btw. everyone here is probably pretty interested in social phenomena. my post was motivated by the tendency of people in policy debates to go to google, and find a social science finding which validates their own normative preference, not understanding that in many fields there are many contradictory results, and no consensus (sometimes the people using google know this and are operating in bad/rhetorical faith, sometimes not).

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Daniel K: As soon as economists themselves develop a more reasonable attitude toward their own work, like seismologists, I’ll concede. But they’ve been unreasonably arrogant for the last 50-70 years, and I don’t think that they’ve quit yet.

    DeLong’s link that no one read was a step in the right direction.

  • Chris T

    Heck, physicists tend to be far more circumspect when it comes to theories that have survived every conceivable test thrown at them than economists tend to be with ideas that are virtually impossible to verify.

    When you have a well known economist calling fellow economists stupid for disagreeing with him on a regular basis in a widely read newspaper, it doesn’t suggest economists are particularly humble.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    I’m being a devil’s advocate a little. In my experience the question “Is X science?” has always been asked during a pecking-order dispute, often in terms of who’s smarter (higher IQ). Mathematicians disdain physicists who disdain chemists and so on. Economists disdain sociologists and so on. The question “Is economics a science?” is usually asked either from the point of hard scientists asking whether economists should be allowed into the club, or else from the point of view of economists trying to get the whammy over sociologists and historians.

    By and large I think that it’s a mistaken game. The important thing about any area of study is how good a job it does with the material it works with, not about where it stands in the pecking order. Furthermore, the whole game involves demanding kinds of results from the proposed scientists which are not possible because of the nature of their objects of study.

    A lot of this is similar to what Daniel K is saying.

    BUT: economists have played this game way too hard, and they need to be brought down. And they aren’t to blame for not achieving the impossible, but they’re to blame for claiming to have done it. And because of the worldly power of their biz, the possibility of corruption is serious.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    propositions

    1) economists are very smart

    2) economics is very hard

    3) economics as a discipline has had very significant real-world impacts in terms of policy

    4) economists are often wrong

    the combination of #3 and #4 is the major issue combined with the arrogance which goes with #1. if economic orthodoxy is as de facto contingent as it often seems to be, there needs to be a softer touch (in an idealized/stylized situation economics can make very good predictions, the problem is that those situations may be a lot rarer than we assumed).

    the flip side of this is that unlike cultural anthropology or sociology the errors of economics can be corrected because the errors are clear and distinct. i’ve given upon those fields, but not economics. but economists do sometimes seem to operationally confuse clarity of language (formalism) with clarity of understanding. you’re still looking through the glass darkly dawgz.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Again, look at the DeLong link above. DeLong says that economics hasn’t been theorized and as a result can’t give unique answers. By “hasn’t been theorized” he means “has been theorized to many different ways”. DeLong is a fully accredited mainstream economist, though most don’t agree with him.

    One heterodox economist, Daniel Davies, says that within academic economics, encumbered by a lot of other stuff, there’s a nice little science of management trying to get out.

  • DK

    A vote for John Emerson: Yes, it is a success that matters. There is nothing god-given in the scientific method. Scientific method is only used because it delivers. If a field uses scientific method and it never works, that filed is not science. By definition. Social sciences as we have now are not sciences. They *probably* can be sciences – but only if they shed all the pretense that currently makes 99.99% of their output a pure unadulterated pseudoscience. Not likely to happen – this trough feeds too many.

  • gaddeswarup

    This seems to be a problem. It is difficult to trust economists but we cannot do without economics. Here is a comment from energy forecaster (?) Vaclav Smil which may also apply to economics. From The Perils of Long-Range Energy Forecasting: Reflections on Looking Far Ahead via Farnam Street:
    “…a new century will make little difference to our ability of making point forecasts: we will spend more time and money on playing the future game—but our predictions will continue to be wrong.

    But acknowledging these realities is not the same as advocating a complete abstention from looking far ahead. There is a fundamental difference between decisions that are good only if a particular prediction turns out to be correct—and the ones that are good for a range of alternative futures: scenarios, rather than point forecasts, are thus much more valuable, both from heuristic and from practical points of view. As the future is inherently unpredictable, it is the decision analysis or contingency planning under a range of alternative scenarios that should be pursued most diligently. Techniques comprising this approach range from narrative and normative scenarios to Monte Carlo simulations and to stochastic programming.”

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    DK, I wasn’t saying that they could be sciences but that they probably shouldn’t expect to. The reason is the difficulty of the material, as others have said.

    In various areas of study science has had tremendous successes, and social scientists started hoping that they too could have such tremendous successes. But they were, by and large, overoptimistic.

    The question about economic, sociology, anthropology, etc. shouldn’t be “How scientific is it?” but “How good is it?” And it’s not true that if your knowledge isn’t scientific it’s nothing at all. Whether or not eco is a science, someone still has to make economic policy. What should they know?

    Razib, I’m not so sure that the errors of economics are as easily corrected as that. Sometimes they are.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    Razib, I’m not so sure that the errors of economics are as easily corrected as that. Sometimes they are.

    fair enough. but let’s observe that we’re talking about economics, and not sociology or anthropology. part of this is about relevance. economics has a lot of real-world impact in terms of influence in the halls of power. but part of it is clarity. cultural anthropology and sociology are by and large not even science today, and don’t make a pretense to be such (more so cult anthropology than sociology). at least in the USA.

  • Markk

    Science is about models, the tools to make and understand models of the physical world (which includes humans) and the social activity needed to make the models known and used by enough people to give them some robustness and understanding of their areas of applicability. What are the models of social science? I don’t mean things like Quantum Mechanics – huge overarching things. Even physicists don’t really use those most of the time. I mean the rules of thumb turned into mathematical relations that are interlinked with each other. Unique answers are NOT required. We might have a model that things could go one of two different ways, but not several others. Or that says, hey, 3/4ths of the time we don’t have a clue but 1/4th of the time something we can predict will happen when certain things are in place.

    I think these do exist in “social sciences” but they aren’t very robust. Crowd models and traffic models DO work more than they don’t. The people movement models used by architects work more than they fail. I think there are many more that have a significant probability of describing reality more than they fail. I think these models aren’t the point often in descriptions of social sciences, so perhaps their limitations and actual successes aren’t out there.

    It is science to understand that because of chaotic interactions one CANNOT get a high probability prediction, and to understand what kind of group statistics might be applicable under those conditions. That is one of the things that distinguish science – understanding its limits in precision and area of applicability. So I would answer YES to the question mark. But what that yes consists of hasn’t been well delimited to me.

  • Chris T

    “There is nothing god-given in the scientific method.”

    Right, but you need good method to achieve success. Before even method, though, you need to define the problem. The social sciences have never really defined what they’re studying and are flailing about because of it.

    They generate a lot of information, but have no way of interpreting it.

  • Pingback: Tweets that mention Social science isn’t “science”? | Gene Expression | Discover Magazine -- Topsy.com()

  • DK

    John Emerson: “The question about economic, sociology, anthropology, etc. shouldn’t be “How scientific is it?” but “How good is it?”

    Yes, in general of course. But the specific question in the title of the post is exactly that: “how scientific is it?”. And the simple answer is “not at all”. Because we know empirically that scientific method, when applied to appropriate subject appropriately, gives solid knowledge. Ergo, “social sciences” are either not using scientific method correctly or their subjects are not appropriate for the application of the scientific method. In any case, it means that these sciences are not sciences at all.

    “And it’s not true that if your knowledge isn’t scientific it’s nothing at all. Whether or not eco is a science, someone still has to make economic policy. What should they know?”

    100% agree. Never meant to imply otherwise. Science is only a tool. And tools have their limits.
    Philosophy is not a science. I don’t think anyone is saying it shouldn’t exist.

  • Doug1

    This post from Steve Sailer get’s at what’s most wrong with Manzi’s article incisively:

    http://isteve.blogspot.com/2010/08/jim-manzis-new-article.html

  • http://backreaction.blogspot.com/ Bee

    Interesting. I like to say we have to finish the scientific revolution, since it did not extend to the natural sciences. If experiments are possible, that’s certainly part of it. But more general than that, it means stopping to apply arguments from the social sciences in a trial and error fashion. Because that’s what we’ve been doing in politics and economics for centuries. Unfortunately, meanwhile the consequences of our errors are becoming increasingly costly and devastating. Time to use a scientific method.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    “Because we know empirically that scientific method, when applied to appropriate subject appropriately, gives solid knowledge.”

    That’s not really empirical, that’s more like a definition, and it’s circular because you put ion the “appropriate subject” caveat. In other words, if you have solid knowledge, you know that the subject was appropriate and that the scientifically method was applied appropriately.

    You still can have secular, empirical, rational, hypothetical-deductive thinking about social things. You just won’t get results with the power clarity, and certainty of physics. (Even in physics, Heisenberg said that quantum physics was easy, but turbulence in fluids was impossible).

    My concern has been, first, to get away from the pecking-order talk. More successful scientists believe that if they did social science they’d do a much better job of it, and sometimes they do, but history is littered with failed efforts, and in many respects econ has been one of them. Both Samuelson’s mathematization of econ and the more recent quant invasion may have done as much hatm as good.

  • http://haquelebac.wordpress.com/2010/04/06/my-fossil-railroad/ John Emerson

    Bee, the gist of what several of us have said is that, because of the difficulty of the material, that might not be possible. People have been trying for centuries. It’s not like it’s a new idea.

    one more time I recommend the Brad DeLong link

  • Jim Manzi

    Thanks to all you commenters. I can’t tell you how fascinating I find this conversation to be. I touch on the various methodological issues in the article, but about 2/3 of the upcoming book is devoted to trying to put forward a rigorous but practical framework for this topic.

    In reference to the debate on this thread, and in very shortest form, I am convinced that science is defined by method, but that the only justification for its method(s) is, ultimately, the creation of predictive rules that enable increased mastery of our physical environment.

    In the book I propose and defend the propositions that: (1) while experimentation is a subset of science, science without controlled experiments is just sloppy philosophy (and I devote a chapter to considering the superficially falsifying cases of sciences like astrophysics or historical geology); (2) in general, controlled social experiments require randomization; (3) there have been some randomized experiments that have created tiny islands of truly scientific knowledge of human society, in the specific sense of identifying reliably some non-obvious social programs that improve human flourishing; (4) the application of these experimental methods is growing, should be encouraged, and is likely to create some improvements to this knowledge; but, (5) it is very unlikely that social science will in the foreseeable future create anything approaching a comprehensive understanding of society, and that our political institutions should be predicated on continuing, profound ignorance of what effects social interventions will actually have.

  • bob sykes

    “A friend once observed that you can’t have engineering without science …”

    Actually, we had engineering for a few thousand years before we had any science. Often the flow went the other way, eg thermodynamics after steam engines.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    bob, people often make a distinction between technology and engineering. but hey, who cares? you’re bob sykes, and your powers of perception are so incredible that we must bow down to your assertions.

  • Pingback: “Don’t Trust One-Offs” « Around The Sphere()

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »