One of the courses I am doing this term (arguably the best course I am doing this term) is on the Philosophy of Science. I have wanted to write a post on feminism and science for a while but haven’t had the time. Most of the data of the post comes from Kathleen Okruhlik’s Gender and the Biological Sciences[1]. I thought it would make for interesting reading. (That apart I really have nothing to post about. Life is as dull as dull can get).
Traditionally the philosophy of science has placed a very limited premium on the role social and psychological factors play in influencing the scientific process. (The most revolutionary work in this vein was Thomas Kuhns’ The Structure of Scientific Revolutions[2])
The traditional view of science is of a process that aims at objectivity and achieving a better understanding of reality. In the context of discovery or theory generation how one comes up with a hypothesis is irrelevant[3]. What is important is having a strict and rigorous process of falsifiable experiments and tests. This process will naturally act as a filter to separate the good theories from the bad[4].
The problem with this view of the scientific process arises when we consider the fact (as Okruhlik points out) that in actuality science doesn’t work like this. A scientific theory is chosen as the best available explanation of a certain phenomena over other theories, not necessarily because it is the best explanation of reality per se. This makes sociological influences at the level of theory generation particularly important.
Take for example theories the aim to explain female behavior. The theories may be different from each other in many respects but if they all arise in a deeply sexist culture, they will all be contaminated by sexism. Non sexist rivals will not be generated at all. Thus having a rigorous, objective scientific process does not automatically eliminate the sexist or androcentric bias from science.
I reproduce just two of the many examples Okruhlik gives to support her case.
One example is nineteenth century craniometry’s attempt to explain inferior female intelligence by appealing to brain size. It was suggested that the true measure of intelligence lay in the proportion of brain to body mass. But this index favoured women and the hypothesis was discarded. It was then suggested that frontal lobes were the seat of intelligence and men have bigger frontal lobes (women have bigger parietal lobes). However when research pointed to parietal lobes being the seat of intelligence the data was evaluated to argue that in actuality women have smaller parietal lobes. The one underlying assumption that scientists were unwilling to give up was that women are biologically less intelligent than men.
In another example, take the case of the lab rat. An ideal healthy lab rat is by default male. Female hormones and their effects are taken simply as nuisance variables that prevent experimenters from getting at the pure, clean, stripped down essence of rat hood as presented by the male model. (This held true for human experiments as well. Until quite recently the default archetype for medical research was a (white) male). The female species is thus covered by research on males; if she is not included it is because she is not an archetypal member of her own species.
The question that arises therefore is this: Do we simply accede that science is not as rational and objective as we want or think it to be, or do we attempt to alter the process of science so that it takes into account the normative aspects of theory generation?
To achieve the latter it is important, Okruhlik argues, to recognize that there is a bias present in the social arrangements that govern the scientific process and to try and control for these. Secondly, it is important to note that the traditional rationality of the scientific process is a simple summation of individual rationalities. It is only the inclusion of diverse standpoints that can lead to a questioning of certain underlying sociological biases and pave the way for change.
It is for this reason that the feminist standpoint and critique of science is so important. It is not that this view is more holistic, intuitive, objective or nurturant. It is important because it comes from a social and political view point that is different from the one that has dominated science. That said, real change in the scientific process will be possible only when theories are developed by scientists who not only have a solid grounding in their own disciplines but who are also open and committed to questioning the sociological biases that seep into science.
[1] From Biology and Society, Canadian Journal of Philosophy, Supplementary vol. 20 (1994): 21 -42
[2] Kuhns fundamental argument is that all normal science essentially works within a given paradigm and is thus constrained by the boundaries of the very paradigm it works within
[3] The method of Conjecture and Refutation as well as the notion of ‘falsifiability’ were put forth by Karl Popper.
[4] In Popper's opinion theories are not true or false, but rather good or bad. Once a theory has been corroborated it is accepted as the ‘current’ best theory available.
Subscribe to:
Post Comments (Atom)
55 comments:
i hadn't known that kuhn was taught at an undergrad level. am impressed and happy.
I agree that there is probably a bias towards male rat subjects in medical experiments, and I agree that some pseudo-scientists once used to try to cherry pick data to prove superior male intelligence, but I'm not sure if either of these issues reflect problems with the objectivity of modern scietific methodology.
Do you really see a masculine tilt in the practice of science today? Most scientific studies are asexual, so I'm not sure what biases will be overcome by adding a feminist perspective to the process.
Instead of adding a new evil to fight an existing one, any unintended gender based bias should be rooted out, just as any other corruption is removed, to ensure the most accurate result of a scientific experiment.
this is a great summary on an interesting subject...
I think most of the scientific papers do discuss their methodology in detail, all the assumption, process of setting up experiments, scope of their conclusions, in fact it is these things which determine the worth of any scientific work. a paper on rat behavior which uses only male rats will not be of much use, it will be a "bad" theory because it will be able to predict only a small set of behavior. I think the corrective measure is inbuilt in the way science is practiced. However I do agree, things get more complicated in case of so called "soft sciences" like psychology for example.... most of evolutionary psychology for example is highly controversial and contested because of these things... flaws in their methodology. but yes I agree with you that all this shouldn't mean we should abandon science, rather take a normative approach to set things right.
I had read a small introductory book on philosophy of science some time back. It discusses the Kuhn-Popper debate too. I wanted to read more and got hold of Popper's book too but couldn't go very far.
You mention the "female species". Are they mutually exclusive from the human species?
Anyway, a great insightful post on the history of scientific research and philosophy (I desperately hope this is history).
The traditional view of science is aimed at the so called "hard" sciences, where the identity of the experminter does not matter one whit. You can see this in the language of science reports, which is dominated by the passive voice.
The process of getting a theory works like this:
You are given a set of facts.
You come up with a set of explanations for these facts. This is termed as a hypothesis.
Then you see how well the hypothesis matches up with reality, and whether it can predict other facts. This may be experimentally verified.
If it satisfies all known facts, it gets termed a theory.
This works very well in Phyics and Chemistry. It has a few problems in biology, but mostly works there as well.
Traditionally the philosophy of science has placed a very limited premium on the role social and psychological factors play in influencing the scientific process.
This is because the philosophy of science does not take into consideration the views of short lived humans, machines or any other species. The results of the experiment are required to be independent of the observer. If they are not, then the experiment is no longer repeatable and hence , it isn't scientific.
The problem is with the softer sciences like psychology or economics, where the biases of the experimenter can affect results, or allow for false conclusions to be drawn, or simply put up unrepeatable experiments.
The whole argument against "Intelligent Design" rests on this argument. There isn't any experimental verification possible to refute the theory. Evolution can be falsified, ID cannot.
As some of us put it, these aren't really science but arts [1]. Calling something a science does not make it so.
When you are referring to lab rats, or human subjects, which experiments are you actually referring to? Where was the sample chosen from?
That said, real change in the scientific process will be possible only when theories are developed by scientists who not only have a solid grounding in their own disciplines but who are also open and committed to questioning the sociological biases that seep into science.
Is this a desirable change? What breakages are we looking at? How easy is it to get controlled access to other groups? How important is the sociological factor in the experiment?
Relativistic and quantum physics are not paradigm shifts away from Newtonian physics, even though they appear to be. They work at different levels to explain anomalies in observations from what Newtonian theory predicts. Essentially, they set limits on the space where Newtonian physics is valid..
Take for example theories the aim to explain female behavior. The theories may be different from each other in many respects but if they all arise in a deeply sexist culture, they will all be contaminated by sexism. Non sexist rivals will not be generated at all. Thus having a rigorous, objective scientific process does not automatically eliminate the sexist or androcentric bias from science.
Ah, but you are looking at half the picture. Say a hypothesis is produced to explain why females are less intelligent than the males (to follow your own example). Experimental verification disproves this hypothesis (you may have to wait for the technology to evolve which will allow you to test against the measurable paramters of this theory).
If the experiments are repeatable, you will be able to conclude that females are not less intelligent that the male of the species.
At this point, someone needs to produce a hypothesis which explains experimental results. This will happen, regardless of whether the society is initially sexist or not, simply because the new hypothesis needs to explain those results.
Note that this new hypothesis may not reduce sexism, or alter society in any signifcant way. However, the reason(s) to be sexist will change.
This doesn't just apply to sexism. Scientists have been imprisoned for less. Merely claiming that the earth moves around the sun was reason enough.
However, given that keeping the sun stationary and having the earth move makes the mathematics simpler, and explains all the facts known without having to add more complicated systems does imply that the theory is a better description of reality. (This is true in Physics and Mathematics, if the equations simplify to a few terms, you are probably on the right track).
[1] This is not demeaning the subject, just clarifying that this is not a science as we define it. These arts follow part of the scientific process, but not all of it.
it's interesting that anonymouse chooses to club psychology, economics, and intelligent design together, and that too as "arts". thanks for the laugh.
@TR: You tell 'em Prof!
but arent you being just a little anachronistic??
I ran a quick google and found plenty of examples of laboratory experiments that test BOTH male and females species for complete results.
for eg. http://www.labanimals.no/Kurssite/Tromso2001/crabbeEDIT.PDF is a link to the GENTETICS OF MOUSE BEHAVIOUR and if you just conduct a preliminary reading you will see that for a comprehensive study they have tested both male and female specimens, exactly how science expects it to be.
also this link http://www.nal.usda.gov/awic/newsletters/v9n3/9n3weerd.htm which is a study of EVALUATION OF ENVIRONMENTAL ENRICHMENT FOR LABORATORY MICE tests both sexes[at the cost of repeating myself for complete reults] and in fact, once the tests showed little difference in the behaviours of the male and female rats, 'only female species were tested' for enhanced results.
there was a third link[which honestly made little sense to me] but one thing caught my eye.
'These experiments are performed to determine how estrogenic chemicals have affected vitellogenin mRNA levels in male Japanese quail livers. Vitellogenin is a biomarker known to be induced by estrogenic compounds and is found in egg producing animals. ONLY MALE ANIMALS are studied because they do not generate high levels of vitellogenin and thus give a low background on such quantitative experiments.'
Sometimes the nature of the tsts determine the specimens chosen right?
tabalua rasa, i have dark memories of Kuhn being taught as a first year class. ...
yes anonymous has intersting insights...
"If the experiments are repeatable, you will be able to conclude that females are not less intelligent that the male of the species."
say that this is irony, say that this is irony, say that this is irony....
"This doesn't just apply to sexism. Scientists have been imprisoned for less. Merely claiming that the earth moves around the sun was reason enough."
no no, anonymous. scientists in general never have been imprisoned for sexism.
"The problem is with the softer sciences like psychology or economics, where the biases of the experimenter can affect results, or allow for false conclusions to be drawn, or simply put up unrepeatable experiments."
Shame.
How was that again with the transcendental unity of apperzeption? Just saying....
"As some of us put it, these aren't really science but arts [1]. Calling something a science does not make it so."
no? Who has funded it?
TR: Yes, exactly. I wanted to rush over and give my friends in the economics department the good news, but it turns out they don't actually have any "experimenters".
Szerelem: Good one. One minor quibble though. It seems to me that the points you make don't so much speak to the lack of objectivity in the scientific process per se, but to the influence of sociological factors on a) what questions get asked and b) how the answers are applied. There's nothing wrong with the findings of the rat experiments you describe - they're objectively valid and make perfectly legitimate theoretical contributions (presumably). There's no reason to question the experiment or its findings. What does need to be questioned / kept in mind is the scope of what these findings are / are not applicable to. The experiments may well produce fascinating and scientifically valid results for men - which is a perfectly useful and legitimate scientific objective in itself. All we need to do is recognise (as we 'arts' types have been doing all along) that the applicability of the findings is limited, and then look for ways in which similar experiments can be carried out for female subjects and different / modified theories developed.
None of that changes the fact that sociological factors may have underemphasised consideration of gender as an important variable in biological research - but it does mean that we don't necessarily need scientists who design or conduct experiments that are free from sociological biases (I'm unconvinced that's even possible) - all we need is a recognition that there's a bunch of papers waiting to be written that study sex as a mediator of established relationships, and a hundred dissertation hungry post-graduates will rush to fill that opportunity.
TR: Yes we do a lot of Kuhn. Popper as well. And critiques of both. I would think it would be difficult to have a class on philosophy of science without studying both of them :)
MT: I would disagree with you when you say that it is only pseudo-scientists who have reached androcentric conclusions. Historically there have been numerous instances where science has been heavily affected by sociological biases and it is from this perspective that feminist critiques arise.
Take for example the traditional biological view of conception that features an active sperm and a demure passive egg. The view had always been of female passivity with the sperm actively pursuing the egg, surviving the hostile environment of the vagina and defeating his many rivals. Of late however (since the ‘70s) that view has changed. Biologists now view the egg as an active agent, directing the growth of microvilli (small finger-like projections on its surface) to capture and tether the sperm. The egg and sperm are seen more as partners in the process of conception. Interestingly the microvilli were discovered in the 1890’s but never considered worthy of serious scientific enquiry. Another example we discussed in class is the case of primatology. The field focused so much on the role of male hunter gatherers that the fact that females helped provide close to 80% of the nutritional value of a family was largely ignored. It is easy to underestimate the impact social biases have on the things science chooses to observe. And this is not true only of gender. Take the example of homosexuality. We now know that homosexuality is prevalent in animals (and these have been relatively recent findings). I think in the case of goats (or rams?) this percentage is somewhere between 7 -10%. Considering these animals were domesticated thousand of years ago why has there been no scientific enquiry in this direction at all? Why have these observations been made so late?
Obviously, fields like physics and chemistry are not as open to such biases as is the case with the biological sciences (and even the social sciences). And like I said the feminist critiques of science do come from a certain historical perspective. Clearly science is more gender neutral now, but even in that case what is wrong with adding feminist perspectives to the process? Doesn’t that in fact make science more robust and protect against such biases?
Alok: You mention the process that directs an individual scientists work with the assumption that if this process is rigorous enough then a ‘bad’ theory will be rejected. This is basically Poppers method and he famously stated that how a hypothesis is generated doesn’t matter as the process will weed out the bad ones. The argument Okruhlik makes in her paper and as I wrote in the post is that science as a whole is that biases that creep in at the level of hypothesis generation and these are important. The fact is that science looks for theories that best explain certain phenomena. If we have an inherent bias in all theories that are competing to explain said phenomena the biases cancel each other out. Basically, reality is never given a chance to rebut a bad theory. So we could still end up with a theory that is biased, but given all other alternatives explains a phenomenon best.
In general though, I like feminist philosopher like Okruhlik and also Helen Longino who do want to correct this. What they argue for is a diversity of perspective that leads to a diversity of possible hypothesis which hopefully leads to a greater sensitivity against bias.
Nitin: Female specimen rather.
Anonymouse: You are giving the same of view of science that is being criticized here. Re what I wrote in response to Alok above the argument made is that this process of science doesn’t necessarily guard against bias. True physics and chemistry conform better to this process. Okruhliks paper deals specifically with the biological sciences because she argues that biological causes are often cited to explain sociological phenomena. In addition critiques of biology are not as easy to dismiss as feminist critiques of the social sciences because of its position in the hierarchy of sciences.
I would repeat what I wrote in response to MT in response to your question as to whether questioning these biases is desirable or not. Re what you said about relativist and quantum physics not being paradigm shifts away from Newtonian physics, I think Kuhn actually makes a pretty solid argument that they are. Even if you don’t buy into the radical view Kuhn has of paradigm shifts (and I don’t) I think it’s a bit myopic to say that paradigms don’t affect scientific results.
Re the case of women being less intelligent than men. If such a study was conducted in say the 17th century I would argue that you would indeed reach the conclusion that women were less intelligent than men. Women didn’t have the same avenues that men did, they didn’t have access to the same standards of education or knowledge that men did. This is exactly what I mean when I say that the rigor of the scientific isn’t enough to root out biases. You would assume that women are less intelligent than men. The problem is that this is not because of biological factors but rather as a result of social conditions.
Missquoted: I should have specified that these are historical examples. Yes scientific experiments are more gender neutral now, even in the case of lab rats. But a lot of this gender neutrality has been achieved because of the feminist critiques.
Antonia: Thank you. And I agree with you about the irony bit.
Falstaff: You are right about the two points you listed there. The lab rat example is directly from Okruhliks paper actually. In this case the argument she was making was that the results were applied across the board and have been very harmful. So what you’ve said is true, in that as long as those results are applied keeping their limitations in mind its ok.
I agree with you (and I would think a lot of feminists philosophers of science who take the stand of contextual epistemology would as well) that it is impossible to be free from bias. An interesting approach to deal with that problem is to differentiate between negative objectivity and positive objectivity. Even though we might not be able to achieve negative objectivity (freedom from bias) it is worth questioning whether we can still hope to achieve positive objectivity (in that we have accuracy or some sort of accord with reality). Helen Longino makes an interesting point in arguing that having a profusion of viewpoints and retaining different view points, making science more communal can help in this regard.
i have to apologize for having been so polemical, but I just couldn't resist.
next time I engage with more arguments.
Tabula Rasa, I wasn't dignifying ID by classifying it as an art.
Though yes, I should have phrased it better. Writing things at 5 am IST is not conducive to clarity of presentation, even when it appears to be clear at the time of writing.
antonia, The "who has funded it" question should be answered in the full disclosure of the experiment. The funding issue has become more important as the impact of science on the effects of bigger corporations grows, but the funding is still merely one of the parameters you need to consider when criticising an experiment.
Szerelem, you ask the same questions for biological studies as you would for any other ones. The idea of peer review is to consider biases and bring those out into the open. The point is not to criticise with a particular agenda, but instead to look at possible ways of accounting for errors and biases.
If J. R. social scientist claims "women are less intelligent than men" because of study "foo", then you can always test "foo" and see if the results were valid, and applicable to the hypothesis.
falstaff, got any control groups to validate your hypotheses?
szerelem, Considering these animals were domesticated thousand of years ago why has there been no scientific enquiry in this direction at all? Why have these observations been made so late?
Possibly because no one was interested in knowing about it? Or to put it another way, why are these observations being made at all?
Are we discussing the scientific process here (hypothesis, experiments, peer review, validation or alternative hypotheis), or the (mis)application of science in other fields?
Antonia: Well, your arguments are always welcome!
Anonymouse: Let me go back to the example I used earlier. Say we are living in the 17th century. J.R. social scientist claims that women are less intelligent because of study ‘foo’. We go back and test study ‘foo’ a number of times and women do come out to be less intelligent. Is this because women are biologically less intelligent than men or the social avenues available to women lead to them being less intelligent than men?
The second case yes, probably earlier no one was interested in this and therefore no one noticed. But isn’t it important that we notice it and study this considering how many misconceptions are present in society about homosexuality?
You are right about peer review of course. This is just one way to move towards objectivity. It’s also important that the contents and methods of science are public and that there is no arbitrator of science. I am not saying that science should be hijacked keeping in mind a particular agenda. The case here is that someone who has a different set of beliefs or knows different facts can look at science with a new perspective and can probably point out potential problem issues that someone else might miss completely.
anonymouse:
better and better. just to clarify: at 4 pm ist, which of the two were you dignifying when you classified intelligent design with economics and psychology?
We go back and test study ‘foo’ a number of times and women do come out to be less intelligent. Is this because women are biologically less intelligent than men or the social avenues available to women lead to them being less intelligent than men?
No, you repeat the experiments, looking for other uncontrolled variables you need to control.
Alternatively, you come up with another hypothesis which explains the facts equally well, and makes predictions.
Original hypothesis:
Women are less intelligent than men.
Experiment:
Study 'foo', which measured paramters bar, baz and quux. The control group responded in way A, test group responded in way B.
Report would include other parameters which may affect result.
Peer review: Szerelem claims that the social environment has not been controlled for. Claim that intelligence is a function of the environment rather than genetics.
Now, this _should_ result in a new set of experiments being done which control the nurture factor also.
Is this because women are biologically less intelligent than men or the social avenues available to women lead to them being less intelligent than men?
Or are your metrics for intelligence wrong? Are you accounting for the nurtition factor, and controlling it? Are you taking into consideration the health of the parents when the child was concieved (regardless of gender)?
My claim here is that different viewpoints are desirable, but claiming that the basic process of science needs to be altered to accept those views is wrong. What you want to change is the society around the process, not the process itself.
WRT the homosexuality in animals thread:
But isn’t it important that we notice it and study this considering how many misconceptions are present in society about homosexuality?
Important? It's as important as any other scientific research.
The case here is that someone who has a different set of beliefs or knows different facts can look at science with a new perspective and can probably point out potential problem issues that someone else might miss completely.
My point here is that the identity of the individuals involved is not important. It could be someone with the same beliefs, looking at the same facts, but coming to a different conclusion. If identities start to matter, you simply have bad science.
Good scientists accept that they can be wrong. Bad scientists don't.
The power of a scientific theory is in the predictions it can make. The power of the scientific process is that it can invalidate theories by using facts and experimental results.
The weakness of the scientific technique is that it cannot eliminate belief in the first place.
Economics and psychology are arts, ID is theology posing as science.
To bring up another point: it's true that male specimens are preferred to females in a lot of lab experiments, and the endocrine rationale is also true. But the fact is that if I'm studying a specific tissue or organ that is not *directly* linked to the female endocrine system, I'd prefer to work with a male specimen, cos as you pointed out there are less complications that way. Once a workable hypothesis is generated, it can be tested on females to ensure it's authenticity.
In one of Ridley's books (can't remember which one) he mentions feminism in scientific philosophy and his points mirror the ones you've articulated here. While I agree with most of the major ones, I do believe that the fields affected by sexism have a less holistic view of the subject under study. But I don't know how major the actual practical impact it. I'm only talking about biological studies, here, of course.
ok, so now we've dropped the dignification, we're left with a bald assertion. and my question is: what is particularly artistic (and hence, by implied negative correlation, unscientific) about 'economics and psychology'?
@Szerlem: I see your point and agree with you more or less. There are certainly societal (and perhaps sexist) factors that can skew the direction in which a scientific study can progress. Perhaps a feminist approach can help provide the oversight to ensure that science proceeds in a balanced manner.
@TR: Economics is certainly a science (dismal though it may be), but perhaps it is a slightly different form of science than say Physics or Chemistry.
To be simplistic (unavoidable in my case), if I add chemical A to chemical B in a controlled environment, I will always get chemical C.
If I apply force X to object Y in a controlled environment, I will always get momentum Z.
However, if I raise interest rates by I, I will not always get a predictable growth in demand for my bonds, because there are way too many factors to keep track of and provide a scientifically controlled environment.
Basically, while being a science, eco is such a complicated one that it requires a certain amount of 'gut-instinct' to produce a desired measurable result. That's what makes its practitioners artistic in a way. I would take it as a compliment.
TS, the pretty graphs? :p
I meant TS... I was thinking tabula slate in my mind for some reason :p
http://socserv.mcmaster.ca/econ/ugcm/3ll3/pareto/theories
http://links.jstor.org/sici?sici=0031-8205%28196012%2921%3A2%3C159%3ASAANIE%3E2.0.CO%3B2-G&size=LARGE
http://www.frbsf.org/publications/economics/letter/2001/el2001-13.html
http://www.zmag.org/ParEcon/writings/neoclasseco.htm
a professor from my uni was one of the key testimonies in the 'creationism' debate in kansas or wherever the case first originated. he'd talked about how theories of science are deeply culturally and contextually rooted. at a time when the divide between religiona nd the state was still taking root, science endeavoured to prove the existence of God. now science is scene as the antidote to religion. and this hegemonic discourse often occludes the fact that there is no such thing is objective truth not even in science. or maybe i'm just a cynic. the truth for me is neither out there or in here.
btw how do i create links in a comment as you did for helen longino?
:)
the comments section here is just as interesting actually.
Wow! Great discussion!!
Agree with what you say. Our biases can potentially affect the set of initial hypotheses we choose from and then there is a question of funding as anonymouse says. these corporations and special interest groups with money will make sure that there are more results to show to backup one particular set of socio-political views that they are most comfortable with. the reason such loony ideas like ID get so much coverage and media debate is the same, there is money behind it.
reg economics and psychology, they are definitely not physics or chemistry. even though, at least economists, try to imitate physics and build similar mathematical models.
anonymouse:
whew! okay, now we've also dropped 'and psychology', and we're looking at a few cherry-picked articles that basically claim that economics is not a science (or not a scientific science) because it doesn't make simple deterministic predictions. is that your point then? if it is, see below. (and *please* learn to put links into your comments. you obviously have enough facility with google to teach yourself how if you were so inclined.)
mt:
this lack of determinism caused by excess complexity is probably the key reason why some people feel (in their guts, mind you, not scientifically) that economics - or any other social science - is not a real science.
three are two responses to that. one is: give it time to grow. these are very nascent endeavors -- compare how long physics has been developing (at least since the greeks) with how long economics in its present state has existed (since adam smith, maybe? 'economics' as the greeks knew it was merely home management). psychology is even newer -- freud was only a hundred years ago, and his theories stand largely discredited. newton only came along *two thousand* years after aristotle.
the second, less flip response is that just because in physics and chemistry one can isolate micro subsystems which appear to our eyes to be deterministic doesn't really make them more scientific. as you make these systems increasingly complex, you enter the exact probability land that economists inhabit. that's why you can have astrophysicists from harvard walk into i-banks (as a close friend of mine did) and claim that the nature of their work is essentially the same, just that the variables have different names; and that's also why the professor i respected most when i was an undergrad (physics major, btw) used to say something i will never forget -- "the art of physics lies in making good approximations."
renovatio:
:-D i hear you!
szerelem:
apologies for hijacking this thread -- but that's what you get for writing on such populist topics :-D
Anonymouse: I reread my post and realized that I had written or do we attempt to alter the process of science so that it takes into account the normative aspects of theory generation. What I should have written, and what I meant, is that we should attempt to deepen/widen the scope of science beyond the traditional philosophical model.
Okruhliks argument is that social biases have an impact on the very questions (hypothesis) we ask. When we test a hypothesis reality can give us only a binary answer – yes or no (true or false). We don’t even compare a test hypothesis to all possible rival hypotheses, only to its extant rivals - that is, to other hypotheses which have actually been articulated to account for phenomena in the same domain and developed to the point of being testable. Science chooses the theory that best explains the phenomena and in this case if all the hypothesis generated are biased then there is no way the bias can be eliminated.
As you yourself note there are numerous factors that could have affected the variable under observation (and there are only so many cases in which the control group scenario can be applied not to mention there’s a limit on how effective they can be) No single theory can completely account for the data that we observe. If the data aren't completely determining our theory choices, then something else must be doing this job.
Okruhlik argues that “we can't control for every possible variable in our experimental designs; so which we take into account depends on what our background theory tells us may be relevant. The process of peer review, critiques etc could point out the flaws but this cannot be counted on as a certainty. Arguably, it is not possible to have an unbiased science (if we take bias to mean having negative objectivity). We may not even be aware of our biases. Historically, there have been numerous examples of blatantly bad, biased theories that have been believed. Having diverse view points then is not just desirable, but imperative in that it enhances the rationality of the scientific community as a whole.
In the case of gender biases in science one could say that men could have done exactly the same work. The argument is that they could have, but it’s not simply about necessary or sufficient conditions. It is not a logical necessity but also no accident that the advent of certain scientific hypotheses coincided with increased political power for women and increased representation of women in the academy and scientific communities.
Revealed: Welcome! Like Falstaff wrote the findings from a single gender study would still be valid and relevant as long as the scope and applicability are kept in mind. And then of course as you said they could be expanded. And I do agree with you that a field affected by sexism isn’t holistic. I would think that the practical impact would be to reinforce whatever sociological biases we have wouldn’t it?
MT: Yes, we do more or less agree :)
Renovatio: The graphs. They do require artistic talent! (and welcome!)
Hobo: Yes, it is questionable to what extent science can be unbiased, but that doesn’t mean that we cant aim at accuracy, which is why I think Longino’s differentiation between positive and negative objectivity makes sense.
And to see how to create hyperlinks go here :)
Alok: This is what you get for asking me to post on the philosophy of science!
TR: You are more than welcome to hijack the thread. And I do agree with what you are saying especially “just because in physics and chemistry one can isolate micro subsystems which appear to our eyes to be deterministic doesn't really make them more scientific.”
I think it’s important to recognize that there is a variety in science. Some areas maybe more exact and some may not. (I mean even biology by the deterministic standard isn’t exact is it?).
What I think is important is that these fields follow a scientific process. Poppers whole argument with Freud, Marx and Adler was that there was that their theories could be applied to explain anything. This clearly isn’t true for economics and psychology. The results they come up with might not be exact or deterministic but they are falsifiable. This again goes back to what you say about people feeling that these fields are (in their gut, not scientifically) not real science.
And P.S.: Wow. This is probably the first time I have seen you get testy.
tr, You mean, double-click selection, ctrl-t and middle button paste doesn't work for you? Tsk, tsk.
Psychology
You seem to be under the impression that I treat art as inferior to science. I don't, (I work in a field which some practitioners (including me) call a craft, not even an art. Programming is an art, not engineering.).
Arts can become sciences, and the leading edge of any field of science can be an art. The leading edge of engineering will be a craft, and some crafts will become engineering. If we are arguing with different definitions, we are never going to have a rational debate^Wdiscussion.
My claim is that we don't have enough possible experimental justification yet to be able to call psychology or economics (or any other "soft science") a science. As szerelem put it, there aren't enough controls to allow for invalidation of a hypothesis.
Philosophy of science
If economics and/or psychology become sciences, I will be glad to call it psychohistory.
To put it slightly more crudely, if the individual practising the subject matters, it's an art (or craft). If the individual does not matter, then it's a science (or engineering).
szerelem,
Historically, there have been numerous examples of blatantly bad, biased theories that have been believed. Having diverse view points then is not just desirable, but imperative in that it enhances the rationality of the scientific community as a whole.
Historically, humankind has been stupid enough that I sometimes wonder how we even survived. Memetic diversity is always good. Mutations will always arise in any group of independent organisms (or the group will die out when conditions change).
On the other hand, I also believe that given enough time, there will be better hypotheses put up to explain given sets of data.
I think that if I have understood szerelem correctly, her point is that including a feminist viewpoint will improve science. My point is that adding any rational viewpoint will do as much, regardless of why that viewpoint is being proposed, or who proposes that viewpoint.
anonymouse:
now you resort to wordplay? and link to a passage in wikipedia (of all sources), where the two relevant points made are (a) one that i've already addressed, and (b) a 45 year old quote from kuhn -- hardly impressive considering my other point about the age of the discipline.
and what exactly does "enough possible experimental justification" mean? i mean come on -- "enough possible"? and you speak of rational discussion?
szerelem:
i'm very impressed by the depth of understanding you're displaying here. hats off. i'd be curious about what your plans are, post-summer. mind emailing me?
and yeah, apologies for getting testy. i normally don't bite on stuff like this -- but that classic statement clubbing psychology, economics, and intelligent design got me going :-D
anyway, much as i've enjoyed this, i have to sign off. going offline for the next five days. ciao, all.
This might bring the wrath of everyone on my head, but i basically agree with anonymouse. The best part of science is that the theories *are* tested over and over again. And as it continually evolves there is an evolution of the initial theory. And maybe the initial pool of hypotheses might be skewed but I think it would be simply skewed towards the thinking of white male caucasians, and nothing else. Women are only one subset of the population that are not equally represented in the scientific community. The conclusion that because women are underrepresented the initial pool of hypotheses is curtailed is something that I cannot fully agree with. All of us might bring our individual baggage with us to the lab, but once you're talking results I don't think any of us use our history/culture/background to come to evaluate them. That's what I meant when I said I'm not sure about the practical realities of the skewed pool. This is one of those unfortunate cases of 'It could have been' which cannot really be proved one way or another, I think.
And also, I read St. Aubyn after your review, and I'm so glad I did! Was one of the best reads of last month :)
Anonymouse: Well, the scientific process hasn’t necessarily protected against bad theories, so it does make sense to try and make the process more fool proof. Also I am not saying that only having a feminist perspective is important. In toto a diversity of perspectives is good. Feminist perspectives are just one aspect of it, so why I don’t see why the hostility to it.
TR: Wow, thank you. Though this post has basically dealt with only one issue that has been raised about science. And really, most understanding comes from discussing / arguing / debating with my prof and other people!
Revealed: Like I said before (1)We don’t even compare a test hypothesis to all possible rival hypotheses, only to its extant rivals - that is, to other hypotheses which have actually been articulated to account for phenomena in the same domain and developed to the point of being testable. Science chooses the theory that best explains the phenomena and in this case if all the hypothesis generated are biased then there is no way the bias can be eliminated. Reality doesn’t have a chance to rebut. (Theories are assumed to be tested in isolation because individual experiments follow that pattern. But the broader process looks more like a choice between theories). (2) This is why being more careful about the hypothesis we generate is important. (3)It’s questionable whether we can be free from bias at all. (This of course doesn’t mean that this is something we shouldn’t aim at). How do we know we are not biased? (4) Its not a simple question of representation but of having a diversity of view points. Men can be feminists cant they? (4) In the post I mentioned feminism specifically but any diversity of perspectives is good. (5) I disagree with your last sentence. Skewed hypothesis have led to bad theories. A model of how science should work, might not be how science actually works. How do we bride that gap? How do we make the process more robust? (6) Yes, I enjoyed Mothers Milk a lot too. :) Though I felt St. Aubyn wasn’t able to maintain his tempo and tone till the end.
tabula rasa, not wordplay. Just pointing out _my_ opinion of what the relationship between Science and Art is.
The Wikipedia article said what I wanted to quite well, and I decided to link to it rather than repost it.
As for the experimental issues:
I can repeat the same experiments Newton did to verify his conclusions. Or Galileo's. Can I repeat Adam Smith's experiments? Or any major economic ones, for that matter?
If I can't do that, there isn't enough possible experimental justification to call it a science. Becuse the experiments are simply _not repeatable_.
I am agreeing with your point that the subjects may evolve to a point where they will stand up on their own as full sciences. I am claiming that at the current stage of development, they are not.
PS: that "classic" statement was addressing two points made in the last statement of the paragraph preceding the two examples.
Anonymouse: What about string theory?
revealed: I'm not sure how what you say makes you agree with anonymouse. All of that - theories being tested repeatedly, often with new controls / enhanced scope to build further theories, the lack of contamination by individual biases of history / culture / background of the evaluation of results, rigorous statistical procedures, etc. is just as true of economics or psychology as it is of biology or chemistry.
TR / anonymouse: on the science vs. arts question, I think the real question is this: How do you scientifically study a phenomenon or set of phenomena which is constantly changing (in an 'objective' sense) and where the theories you develop themselves cause the observed behaviour to change? It's not that economics / psychology are any less rigorous about peer review / falsifiability / statistical sophistication than the 'sciences' - the difference, I suspect, is less in the method and more in the stability of the system they're studying. Understanding gravity doesn't make matter attract itself in different ways, but understanding firm competitive behaviour (and making your results public) will almost certainly change the way firms behave in the market. If we're defining 'science' to mean the study of something that is subject to absolute / deterministic patterns of behaviour unaffected by human thought, then I'd have to agree that economics and psychology are not sciences, though it would imply that a science of human behaviour is impossible by definition. If we consider something to be a science because it follows certain methodological norms (which is the view I would tend to take) then I would argue that economics and psychology are sciences.
szerelem, what about it?
http://www.math.columbia.edu/~woit/wordpress/?cat=2
http://ezinearticles.com/?String-Theory-Science-or-Theology?&id=450748
http://www.math.columbia.edu/~woit/wordpress/?p=530
Falstaff, the question you raise is whether the observer is neutral or not.
You can have experiments where the mere act of observation affects the results of the experiment. However, the experiment can be repeated consistently by different observers, implying that there is no bias from the observer.
With economics or psychology, that repeatability is missing. My reference to Asimov's psychohistory was not meant as a joke. Asimov actually considered that knowledge of psychohistory and the second foundation would affect the behaviour of people, and hence those subjects needed to be kept secret.
anonymouse: No, that's not the question I raise. Go read my comment again. The question is - what happens when the act of observation changes the behaviour of what is being observed? Or more generally, what happens when the system you're observing is objectively unstable and not governed by any underlying deterministic principles at all?
The point is that economics and psychology are sciences in the same sense that biology is a science - we can't predict with absolute certainty how a given molecule will affect a particular organism, we can only establish, with statistical significance, what the effect will be on average. This isn't because the observer is biased, it's because organisms are heterogenous and vary in unpredictable ways. Exactly the same applies to economics and psychology.
And where do you get this crap about repeatability being missing in economics or psychology? Who does psych research without controlling for observer bias? And what major psych experiment isn't replicated countless times to test whether its findings continue to hold?
Anonymouse: Well it raised interesting questions doesn’t it?
For all the work that is happening/has been done in the field not one testable prediction has been made. The theory is simply not falsifiable. Forget predictions, there is no theory – just a bunch of calculations and hunches that suggest that a theory could exist.
The one reason that physicists have clung on to string theory is because if the idea of mathematical “beauty”, but is that a good enough reason? Should physicists be constrained by what a philosopher said science should be?
There is a pretty strong case to be made that String Theory has become the dominant paradigm. String theory pretty much dominates physics departments. At the Institute for Advanced Study, for example, the director and nearly all of the particle physicists with permanent positions are string theorists. Eight of the nine MacArthur fellowships awarded to particle physicists over the past years have gone to string theorists.
Even in this case a case can be made for diversity. There’s no need for rejecting research into this area but a little more diversity in terms of the avenues open for research, in the questions that are being asked.
Falstaff: Re Psych and repeatability - you stole the words from my mouth.
szerelem, and if you bother to read the links I posted, you would note that there are some physicists calling string "theory" as theology for precisely the reasons you quote.
Having beautiful mathematics is usually a sign that you are on the right track, but that is an effect of being on the right track and not the cause.
My bone of contention with you is/was on your claim that having a "feminist" perspective is useful. My claim is that the more perspectives, the merrier (a more general claim).
If we invent artificial intelligence tomorrow, or we meet a species which did not originate on Earth, I would treat their views with exactly the same respect.
Since both of us appear to have more of less agreed on this point, with a slightly different emphasis on each side, lets end this fork of the topic.
(And please, please, do not read emotions into my posts. When I get irritated, I tend to show it pretty clearly, given that most of my communication with my peers is over textual channels.)
Quoting falstaff:
what happens when the system you're observing is objectively unstable and not governed by any underlying deterministic principles at all?
In which case, I will simply ask if the study of that system can be termed as a science?
If the act of observation changes the sytem, and it cannot be reset to it's previous state, then can you reliably repeat those observations?
anonymouse: Fair enough. I said as much in my earlier comment. I quote:
"If we're defining 'science' to mean the study of something that is subject to absolute / deterministic patterns of behaviour unaffected by human thought, then I'd have to agree that economics and psychology are not sciences, though it would imply that a science of human behaviour is impossible by definition."
Notice that by this definition, biology isn't a science either.
In fact, if what is or is not a science is defined by what you're studying rather than how you're studying it, it does seem to make the question of what is or is not a science a fairly uninteresting one, doesn't it?
Anonymouse: Not disagreeing with the articles you linked to. Just saying that the questions it raises are interesting. About the current state of the field of physics and about philosophy of science as well.
falstaff, why isn't biology a science?
Oh, and s/human thought/neutral observer/ (I would like to allow for the observer to be a machine).
szerelem, I don't see many interesting issues around string theory (That it is popular is because a lot of people think it is cool). There are a bunch of people who claim that it isn't even a theory, and hasn't produced a falsifiable hypothesis after a couple of decades, hence it can't be treated as a theory.
The scientific process is working quite well, as far as I can see.
anonymouse: Sigh. Try to keep up, will you. Because it's not deterministic - conclusions from biological studies are only true on the average - and results will vary from individual to individual. Oh, and because the system it studies isn't static but constantly evolving (potentially influenced by interventions taking place based on past knowledge), so that entirely objective observations made today may no longer be true tomorrow.
This is as true for biology as it is for economics or psychology. If, as you claim, anything that studies an objectively unstable system isn't a science, then biology isn't one. And if biology is a science because it uses the scientific method and produces results that are free of observer bias though may be provisional over time, then economics and psychology, which do the same, are sciences as well.
falstaff, you seem to miss my point.
What about the effects of the observation results on the system? Can the observation generate a feedback loop?
Is the observer a relevant entity in the study?
woot. what an interesting conversation i've stumbled into.
imho, the principle of "observers effect", is probably not relevant in macroscopic systems such as the ones being talked about here. Aharonov and Bohm, and hawthorne, with all due respect, weren't talking about studies on this plane.
just my 2 cents @ anonymouse.
Err... whats going on here?
anonymouse: No, you seem to miss mine. Biological intervention can cause systems to change as well. Or haven't you ever heard of organisms developing immunity to drugs? And the observer is just as much a relevant entity in a biological experiment as he / she is in a psychological / economic one.
@szerelem: I ironically agree with most of what you said. I think my reaction to is what differs. Yes, there is a skewed pool of scientists and hence a bias in the field. But it has it's advantages and disadvantages and I think that these biases swing back and forth. I was recently at a grad applicatn interview weekend and out of 15 candidates exactly 3 were male. All the rest female. And this was a molecular biology program. So, I do believe this is one of those phenomena that change with time (and of course the energy and conviction of society in general). But theories being tested in isolation is not just an ideal. It's the truth. I can only talk based on my experience in research and trust me, the first thing you learn is that experiments are not planned in order to get a specific result. In fact results that turn out different from the mean are the ones that are viewed as the most exciting. So from what I've seen, I don't agree with that statement of yours.
And I think Aubyn lost some steam in the end, too. But twas a good read just for the words :).
@falstaff: First, I wasn't agreeing with his definition of science and non-science. I was agreeing with the part of his statement involving the testability of hypotheses and my belief that in practice the skewing that is present is something unavoidable but not really as harmful as we believe it is.
Second, why is Biology always picked on as the weaker of the sciences, or to make a point on the variability of science? The concept that it isn't an exact science capable of standing up without bias to repeated experiments is misinformed and frankly, infuriating. There might be in exceptional individuals an anomaly that doesn't work according to general principles, but otherwise the principles hold and all organisms react in a similar way. If you're talking immunity then you mean pharmacology. Which is not really Biology. Biology is the basic study of living organims, and trust me 99 out of a 100 times organims react in the same way to a molecule injected in them, and you can definitely predict the end result.
revealed: Sure, and 95% of the time people respond to economic stimuli in predictable ways and you can predict the result.
I'm not questioning biology's status as a science at all - I'm just saying that economics and psychology are just as much sciences as biology.
Revealed: I am not disagreeing with you about the process in which experiemnts work. The question is basically about the hypothesis that are generated. The argument made is that if biases creep in at that level the steps of experimentation/observation aren't sufficient to weed it out.
@falstaff: I understood your point. Infact I suspect I empathize with your point more than you know. But it just irritates me when people highlight Biology as an inferior science (even when, as in your case, they're trying to prove the opposite).
@Szerlem: I understood your point as well :P. Maybe I'm just being particularly inarticulate. All I'm saying is when I have a set of results I formulate the best hypotheses for those results, irrespective of who I am or where I come from. One of my ex-bosses was an Italian man, and he and I when given the same set of data came to the same set of conclusions repeatedly. And based on those conclusions we formulated hypotheses. If you think that the questions asked are skewed by the individual, I'll have to point out that the questions asked are only ever skewed by funding. And funding's skewed by practical use to humans. I believe your argument makes a lot of sense when applied to some sciences, but not all. I'm shamelessly nit-picking. So feel free to ignore :)
Post a Comment