Claire Lehmann describes some of the work of contrarian social psychologist Lee Jussim:
It appears that descriptive stereotypes are a crutch to lean on when we have no other information about a person. When we gain additional insights into people, these stereotypes are no longer useful. And there is now a body of evidence to suggest that stereotypes are not as fixed, unchangeable and inflexible as they’ve historically been portrayed to be.
So, it would appear that “we” make use of stereotypes exactly as I make use of stereotypes (or as I try to), as crude first approximations, rather than as the last word. Bigotry, it has always seemed to me, means not having no prejudices, but rather having prejudices which you are unwilling to alter, when faced with circumstances which do not fit your prejudices. And it seems that “we” think like that also.
Good to know. My thanks to Bishop Hill for telling me about this piece.
Having written all of the above (apart from the thanks to Bishop Hill), I see that the previous posting here is also about stereotyping.
I read Stereotype Accuracy: Toward Appreciating Group Differences, of which Lee Jussim is one of the co-authors, about ten years ago. It was impressive, but I could see why it hadn’t been picked up in the press. Aside from stomping on PC prejudices, it was EXTREMELY hard going. Even an unbiased journalist wants an easy life.
Incidentally “When we gain additional insights into people, these stereotypes are no longer useful” strikes me as misleading. To the extent that a stereotype is accurate, ie describes characteristics of members of a group with reasonable statistical accuracy, meeting a member of that group is likely
to reinforce the stereotype as being useful. Of course, since a stereotype is only going to be accurate statistically, and won’t cover all sorts of information about members of the group that aren’t included in the stereotype, then obviously what you specifically discover about a person from an individual encounter is going to overtrump contrary predictions from the stereotype. But the stereotype will remain useful nonetheless, both about other members of the group, and about those characteristics of the particular individual you have met, on which your individual encounter has not yet thrown any further light.
Being sniffy about stereotypes is just another aspect of lefty aversion to science – an aspect of their idea that humans are set apart from other parts of nature, and are not susceptible to analysis using normal scientific rules. We all – even lefties – have stereotypes regarding cars, dogs and so on. Not all cars are exactly the same, and not all dogs are the same. But a good solid stereotype about cars or dogs is helpful the next time you come across such things.
Hey – maybe that’s the problem ! Maybe it’s just very hard to write about stereotypes in snappy readable prose. I may just go and tack that on to my stereotype of “stereotype.”
Not having initial assumptions is stupid (Are you really going to treat a child the same way you treat an adult?) Not changing those assumptions in the face of contradictory evidence is also stupid.
Good post.
Sterotypes are vital – one can wait to see what each individual lion will do (perhaps this individual lion has a horror of eating meat), one can only judge on the basis of past experience (our own and that of other people) concerning lions.
“Prejudice is latent wisdom” – the refusal to “profile” (for fear of cries of “racism” or “Islamophobia”) cripples American (and other) security.
However, there are often individual exceptions to a general rule – and the refuseal to accept the existence of exceptions is bigotry.
Even a general rule may turn out to be wrong.
One’s own past experience (and the experience of others) could be wrong – a biased sample, or a series of misunderstandings.
A stereotype is the verbal equivalent of a statistical summary. It is as true or false as the statistical numbers that its words imply. It says plenty about the politicised rubbish that is almost all of social science that most social scientists need to be told – and refuse to hear – that people use verbal as well as numerical statistical summaries and that when they meet real exemplars they usually recognise that these real individuals need not be on the average line of the stereotype.
“Social scientists are SWJs” is a stereotype. “Far more social Scientists are SWJs than are not” is a more precisely-phrased stereotype. “The CVs of social scientists show that the ratio of SWJs to others in the field is 10:1, leading to a much larger ratio in the published and popularised researched” is a statistic from the article linked by the post. No wonder they tell you to ignore stereotypes.
“Incidentally “When we gain additional insights into people, these stereotypes are no longer useful” strikes me as misleading. To the extent that a stereotype is accurate, ie describes characteristics of members of a group with reasonable statistical accuracy, meeting a member of that group is likely to reinforce the stereotype as being useful.”
True. Otherwise the stereotype is not such.
Cristina : Otherwise the stereotype is not such
“not such” = not a stereotype ?
“not such” = not accurate ?
“not such” = not useful ?
A stereotype is still a stereotype even if it is neither accurate nor useful. Though you’d expect such a stereotype to die off after a while.
The social, economical and psychological “sciences” are the offspring of the Enlightenment. Hence the leftist bias.
Lee Moore, not a stereotype 🙂
Rather than “stereotype”, I prefer to use the word “archetype”. It says something similar, but it doesn’t carry as much baggage. And yes – it doesn’t take that long to see how any particular individual/lion/car/dog differs from the archetype, for good or for ill.
A little word substitution can often defuse knee-jerk reactions. Once upon a time there was an advertising campaign that said “Discriminating shoppers prefer ….” But then the word “discriminating” got a very bad reputation. So when I was using this in a story, I said “Discerning shoppers prefer….” Same idea, 99-44/100% less knee-jerk.
Call it stereotype, archetype, or something else – nobody can long survive treating everything in the world ab initio.
May i suggest that we should have a Popperian attitude to stereotypes?
A stereotype makes predictions about how people behave, otherwise what’s the point of having one?
But if it makes predictions, then a stereotype is falsifiable.
Our attitude then becomes crucial: if we look _only_ at the evidence supporting the stereotype, it becomes effectively unfalsifiable, and pernicious.
If otoh we make an effort to look at evidence that a given individual does not fit the stereotype, then the stereotype can be a useful starting point; not _true_ mind you: just _useful_.
If we keep meeting people who do not fit a given stereotype, then at a certain point we’ll naturally decide that the stereotype was “wrong”, or more precisely useless; or even worse than useless.
Something like that happened to me when i moved to Denmark: my stereotype of Danes changed pretty drastically. My opinion of the Danes did not become better or worse, mind you: just different. And of course i still realize that many Danes do not fit my stereotype, and very few fit it well.
So true.
Thinking about politics, economics, and the human mind like a scientist yields political science, Keynesianism, and psychology.
Thinking about politics, economics, and the human mind like a philosopher yields theology, Austrian economics, and religion.
1. The difficulty with Popper and stereotypes is that it takes a lot more than one contrary observation to falsify a proposition that is only claimed to hold “usually.” Four heads in a row does not falsify the proposition that the coin is unbiased.
2. A stereotype can be wrong but useful – “Avoid pythons as they’re usually highly venomous”
3. Or right but useless : “Men like women with big boobs”
(OK this has ceased to be entirely useless since docs got into the big boob manufacturing business – but it was utterly useless for the first 200,000 years it has been known.)
It struck me a long time ago, that stereotypes are simply a form of information processing, utilized when there is a limited amount of information available. It would be stupid and indeed on occasion self-destructive to say to oneself “I’ll amass more information before I process it and draw conclusions”. You might be waiting for ever. One has to use what one has got even if the conclusions are temporarily imperfect.
And without stereotypes there would be no language. They underly every lexeme
in our vocabulary.
Shlomo Maistre, what is the result of thinking about politics, economics, and the human mind like a scientist philosopher or vice versa?
Snorri Godhi, you may, of course, but it’s a futile exercise. The stereotype is such because it includes the main features of the group it describes. If it falsifiable, then it’s not a stereotype.
To what extent a stereotype is a personal and/or a social construction is a curious matter.
Heh. What do you get when you mix water and sewage? Sewage.
So in answer to your question, thinking like a scientist philosopher or philosopher scientist yields what thinking like a scientist yields.
Cristina : The stereotype is such because it includes the main features of the group it describes. If it falsifiable, then it’s not a stereotype.
My pedantry alarm is activated by this. A stereotype is such because it alleges certain features of the group identified. If it asserts them sufficiently precisely, it is falsifiable – we can observe reality and compare it with the allegation. (Though as noted by me above, we have to work much harder on our falsification where, as with stereotypes, we’re dealing with “usually.”) But once we’ve done enough comparing with reality, if we find that the alleged features are in fact usually absent rather than usually present, we may conclude that the stereotype is inaccurate or wrong. But it’s still a stereotype.
A stereotype is like a theory in this respect. A theory that gravitational force is inversely proportional to the distance between two masses is wrong. But it’s still a theory.
Shlomo Maistre, Arcesilaus would have answered like you. It’s the same “logical bravery” noted by Kant in reference to all the ancient philosophies.
Cristina… marry me.
Futile requests aside – that indeed would have been Arcesilaus’s response, but what’s yours, Cristina? Do you concur with my perspective?
Lee Moore, that’s a very interesting comment. Can your personal observation falsify a stereotype? How?
Lee Moore,
The key difference is falsify vs contradict. An individual instance of something may contradict a stereotype, but to falsify a stereotype there must be a sufficiently high share of many observations contradicting the theory of the stereotype to render the theory void of sufficient explanatory power to qualify as a stereotype.
In other words there may be individuals that contradict the stereotype that Jews are not particularly skilled at basketball, but to render that stereotype false a sufficiently high share of an adequately high number of observations must consistently violate that stereotype.
For my sins I do agree with you, Shlomo Maistre 🙂
Cristina : Can your personal observation falsify a stereotype? How?
Yes. By observing that experience conflicts with prediction sufficiently frequently as to satisfy whatever statistical measure one is using. Same way as scientists falsify scientific theories.
Thanks to Lee Moore for replying as i would have, had i not been asleep.
The confusion seems to be that Cristina did not distinguish between stereotypes and true stereotypes; and not between falsifiable and false, either.
WRT Shlomo’s latest comment, i admit that i was unclear about another distinction: between falsifying the predictions about a single instance, i.e. when the stereotype does not fit an individual, and falsifying the predictions about most instances, i.e. about the group. Only the 2nd can be said to “falsify the stereotype”. I am not sure that “contradiction” is the appropriate word for the 1st, though.
Thinking about economics like a mathematician yields left-wing economic philosophies.
Fiscal leftists fail to realize that economics is a behavioral science. Just look at all the schemes they concoct. They never take into account that changing the formulas changes the incentives. For instance, nobody on the left predicted that the combination of low interest rates and easy mortgage credit would spur the house-flipping fad. When Frank-Dodd passed the folks who hate big banks getting bigger did’t predict that the law would encourage more bank acquisitions, as the cost of Frank-Dodd compliance was effectively a regressive tax on financial institutions. When Seattle hiked its minimum wage in order to wean minimum-wage earners off of welfare, the idiots didn’t figure out that workers would ask for fewer hours so they could earn less and thus keep their benefits.
Lee Moore, sorry to tell you this, but you cannot falsify a stereotype more than you can change any particular component of said stereotype. At most you can only change your opinion about the thing described by the stereotype or about the stereotype itself.
Snorri Godhi, I’m really lucky because I have you to untangle the web of my mind.
Sorry Cristina I simply have no idea what you’re saying.
Lee Moore, you can change your opinion about a group described with a stereotype if your personal observations do not correspond with what the stereotype made you to expect from this group.
You can change your opinion about this stereotype and declare it useless, if that is your conclusion after your observations.
You cannot, however, change (falsify) the stereotype itself because it is not a personal construct, but a collective one. Incidentally, this is one of the few instances where vox populi approaches vox Dei
Clear now?
OK, I think I understand what you’re saying. But I will beg to differ, in spades.
Stereotype = “Men are tidier than women”
My personal observation = this stereotype is wrong
But I conclude that the stereotype is wrong, based on the belief that my observations are representative of everybody’s observations (or the observations that anybody would make if he was observing.) If other people’s observations, as reported to me, differ from my own observations, then (if I believe their reports) I will reject my conclusion that the stereotype is wrong. I am not rejecting the stereotype because my observations are in some way privileged, but simply because those are the observations that I am aware of. If you make your observations available to me, then I will take them into account.
The idea that a stereotype is is a “collective construct” not a “personal construct” is misconceived. A stereotype is a proposition about objective truth in the world. Whether I am the only person in the world aware of the proposition, or one of millions who is aware of it, makes no difference – it is an assertion about the facts. And as such it may be falsified by comparing the assertion with observations.
This is no different from any scientific experiment. The theory is compared with experience, and lives or dies accordingly. Whether I am the one doing the experiment may have an effect on whether I believe the experiment has been competently done, but assuming I accept that it has, it makes no difference who does the experiment.
So whether the stereotype “men are tidier than women” is right or wrong is nothing to do with how many people assert it or believe it to be true. It’s a question of whether or not it is true.
Thinking about particular sterotypes, certainly I think the stay-at-home Anglo view of Americans is incorrect (brash, loud and garish with over-ostentatious displays of wealth) having lived and worked with Americans over the years.
I would also question the Nigerian stereotype (as a bunch of greedy liars and scammers) on the same basis as the American one, but with the preface that there is no way in hell I am going to live in Port Harcourt or some similar Nigerian shit hole to experience the place first hand.
In truth, doesn’t the stereotype of foreigners tend to reflect the sub-groups that are most often caricatured rather than any form of real representation (e.g. the brash American tourist stands beside the gun-toting yokel from the Ozarks or the Adirondacks)
To characterise the French by experience with Parisian waiters or rebellious farmers doesn’t in any way represent the majority of the French, nor even a substantial minority.
Lee Moore:
” But I will beg to differ, in spades.” 🙂
” But I conclude that the stereotype is wrong, based on the belief that my observations are representative of everybody’s observations (or the observations that anybody would make if he was observing.)”
That’s a gratuitous and misleading assumption. Nonetheless, you are still changing your opinion, no mine for example, about the specific stereotype under discussion.
” Whether I am the only person in the world aware of the proposition, or one of millions who is aware of it, makes no difference ”
Yes, it makes all the difference because a non-shared ” proposition about objective truth in the world” is not a stereotype. It’s just your idea about that objective truth. Being it personal, you can adequate it to reality with as many observations, validations or refutations as you see fit.
John Galt: indeed, national stereotypes are almost invariably wrong, in my experience. If not plain fantasies, they are at least decades out of date; and not just 2 or 3 decades, either.
That is, of course, in itself a stereotype; or a meta-stereotype, if you wish:
Every national stereotype is wrong.
(So much for Cristina’s Vox Populi, Vox Dei.)
As with every stereotype, there are exceptions to this meta-stereotype.
Note that this does not apply to different ethnic groups living within the same country: they interact with each other sufficiently often for me to think (possibly wrongly) that their stereotypes of each other are reasonably accurate. After all, it is in their own interest to be able to figure out who they are dealing with.
That’s a gratuitous and misleading assumption.
Why so ? If I observe that leaves tend to fall off trees in the fall, rather than in the spring, I conclude that others will observe the same thing. If they report otherwise, then obviously I will want to investigate further. But when a scientist performs an experiment he expects that any other scientist will get the same result. If other scientists report different results then he investigates further. Why would you expect different observers to observe different facts about the world ?
“Nonetheless, you are still changing your opinion, no mine for example, about the specific stereotype under discussion.”
Sure. But as well as changing my opinion, I am concluding – if you have kept to the original opinion, that your opinion is wrong. Your opinion of the validity of the stereotype does not remain right until you realise it is wrong. It’s wrong if it’s wrong, even if you don’t realise it.
” Whether I am the only person in the world aware of the proposition, or one of millions who is aware of it, makes no difference ”
“Yes, it makes all the difference because a non-shared ” proposition about objective truth in the world” is not a stereotype. It’s just your idea about that objective truth. Being it personal, you can adequate it to reality with as many observations, validations or refutations as you see fit.”
So you’re saying that a stereotype is a proposition about the world that is widely asserted, and the term cannot be used of propositions about the world that are narrowly asserted. I disagree, but you can use the term as you wish. The essential point is a proposition about the objective facts, may be true or false – and may be falsified by a comparison between the objective facts and the assertion. There is no difference in kind between such propositions that are widely or narrowly asserted, whether you choose to call the former stereotypes and the latter something else.
Lee Moore, I’d suggest that you review the concept under discussion again. Just to be sure we are talking about the same thing, which seems not to be the case.
Lee Moore,
I can’t speak for my virtual Samizdatawife (just married!) but I suspect that this is the crux of the disagreement between you two folks.
Cristina mentioned that a stereotype is a collective construct – I agree and not just because she is my Samizdatawife (should we just go with blogwife? I don’t want to piss off Perry de Havilland/the powers that be but I don’t know of any other blogs at which you comment sadly). Therefore, even if I believed that Jews in general are excellent at basketball I could not deny that the stereotype indicates otherwise. I might hold steadfastly to my belief that Jews are excellent at basketball, but just because it is my belief does not make it a stereotype!
So, yes, stereotypes are propositions, but more specifically they are widely held propositions about the world.
Alan Henderson,
Is economics really a behavioral science? Can human action be studied “scientifically” as one studies the cosmos or organic matter?
On my grumpy days I don’t really accept the notion of behavioral science at all. All the sciences are, as Mr Rutherford informs, either physics or stamp collecting.
Shlomo Maistre, LOL
“On my grumpy days I don’t really accept the notion of behavioral science at all.”
Good because they are not. Trained in real and fake sciences, I can attest the veracity of that assertion. Every time the despair with behavioral science threatened to overwhelm me, the philosophy saved me.
Well, I think that behavioural science is correct! Whenever I reward someone for something they have done, I get more of that behaviour. I spell words the British way because I am rewarded with a feeling of independence from American Cultural hegemony, for instance.
And a lot of trouble is caused by words. ‘Falsifiable’ should be replaced by ‘testable’, since falsifiable carries the implication that it is false, whereas ‘testable’ implies you will test if something is true or false.
And stereotyped people seem to be averages, so one or two deviant examples won’t destroy a stereotype.
Cristina, how many philosophers would it take to change a light bulb?
A. None! As Marx said, the point of philosophy is to change the world, not lightbulbs!
Shlomo:
But how widely? national stereotypes change from one country to another. Most infamously, not every country has the same stereotype of Jews. Just checked wikipedia and Pitcairn Island has a population of 56 people. Would you say that there is such a thing as the Pitcairn stereotype of Jews? what if 50 people emigrate? would there still be a Pitcairn stereotype of Jews if it is shared by just 6 people? what if 5 more people emigrate? at that point the Pitcairn stereotype is the opinion of a single person, just as Lee was saying.
At a guess, you have never done research in the “hard” sciences, otherwise you would know that it is much more messy than you think … though admittedly not as messy as behavioral research.
Having said that, take note of what Nicholas wrote:
Indeed, i have just as much faith in the Law of Effect as in Newton’s laws of motion…and more than in Newton’s law of gravitation.
Nicholas:
No it shouldn’t! OK, sometime it’s convenient to have fewer syllables, but see below.
…but only to the unwary.
…which is why it is not used: because you cannot test whether a scientific theory is true. Predictions from the theory, of course, can and should be testable.
As Snorri says, there is not A stereotype of Jews, but several, often contradictory. In deference to Cristina’s linguistic choice I will refer to widely asserted propositions about a group as “stereotypes” and narrowly, or individually, asserted propositions about a group as “fizzlewicks.”
Having got the language sorted out, we can set about the analysis. Both stereotypes and fizzlewicks are usually, but not always, about human groups, often to do with allegedly typical aspects of behaviour. And both stereotypes and fizzlewicks are usually statistical propositions – ie they assert that “usually Xs are Y” rather than “Xs are always Y.”
So when it comes to falsification (which concept, incidentally, does not imply falsity, just the possibility of demonstrating falsity if it is present, by contradiction with observation) stereotypes and fizzlewicks require statistical falsification – ie a number of contradictions are required to demonstrate the probable absence of the claimed relation – a single contradiction is insufficient.
But whether we are dealing with a stereotype or a fizzlewick, falsification is achieved in the same way. If a stereotype disagrees with reality, it is false, regardless of how widely held it is, and regardless of whether believers still cling to it in the face of contrary evidence. It remains a stereotype, of course, but it is a false one. Likewise with fizzlewicks. We are all familiar with false fizzlewicks that are clung to with great tenacity in the face of overwhelming contrary evidence (a female acquaintance’s fizzlewick about the general truthfulness and faithfulness of her husband springs instantly to mind. I am very confident that she is the only person in the world who harbours this particular fizzlewick.)
In the face of the identical usage and logical structure of stereotypes and fizzlewicks, the question may arise as to why we should bother to use different terms for each. But linguistic usage is an unfathomable mystery. Perhaps we do it just for fun.
But perhaps those that wish to do it, do it for a reason associated with a stereotype about stereotypes – that stereotypes are typically unfavourable (ie critical of the group in question) and false. Thus a stereotype might be particularly harmful – widely held false unfavourable propositions about a group are going to be more unfairly harmful to members of that group than equivalent fizzlewicks. But as the book I mentioned up top explains this stereotype of stereotypes – that they are usually false and unfavourable turns out to be false. Most stereotypes are favourable and true (see book for details.)
Lee’s discussion of stereotypes and “fizzlewicks” (be careful, Lee: you’ll create a new word if you’re not cautious – the times are favourable to it 🙂 ) reminds me of the words of a US government report of some decades back. The report stated that Japanese-American women should not be deemed a minority group “because there are too few of them to constitute a minority group”.
I think it’s better to speak of “his own personal stereotypes” or “a stereotype not widely held” than to make too much use of Christina’s emphasis on the numbers who credit a given stereotype. The _connotations_ of the word ‘stereotype’ may suggest sizeable numbers holding it, just as the _connotations_ of the word ‘minority’ clearly suggested the same thing to whatever bureaucrat wrote that report, but a word has a literal meaning as well as its connotations. Since I see stereotype as meaning just “verbalised statistic”, I naturally see it as a thing that can belong to a few before it is held by many. “97% of Calais camp migrants are male” was presumably known to some one or few who surveyed the camp before it became a headline that I recall seeing on Breitbart London and do not recall seeing on the Guardian online. “They’re almost all male, you know – they’re economic migrants, not refugees” is now heard from enough people to meet anyone’s definition of stereotype, and denied or evaded by enough others to make a rival stereotype, but presumably it was at one time held by few. Meanwhile, at any given moment, the sex ratio of those in the Calais migrant camp is in principle observable, so one of these stereotypes is in principle falsifiable.
(Unless of course, the holder’s principles were learnt recently at one of those Universities where it is a thoughtcrime to regard sex as something determinable by so crude a method as looking. And indeed, following David Cameron’s announcement that tents and other items would be given to “women and children” in the camps, is it conceivable that some men behaved in a way we could call “self-identifying as women” – in a certain sense. 🙂 )
Oy.
Snorri Godhi,
Is 100 a lot of hamburgers? Sure. Is 10 a lot? Maybe. Is 3 a lot? Nope. Where is the magical line? I don’t know. Does this mean that there is no meaningful definition of “a lot” of hamburgers? Nope!
You have no idea how messy/not messy I think the hard sciences are. But you are correct that the hard sciences are not as messy as behavioral “sciences”/research.
Okay, thanks for the information. Economics still isn’t real science!
Here is, IMHO, a relevant and well-argued post from the Maverick Philosopher: Profiling, Prejudice, and Discrimination. It contains, for example:
Best regards
Nicholas
” Whenever I reward someone for something they have done, I get more of that behaviour.”
Do you need some pompous dimwit to tell you that? I seriously doubt it.
” Cristina, how many philosophers would it take to change a light bulb?
A. None! As Marx said, the point of philosophy is to change the world, not lightbulbs!”
LOL
Lee Moore
” In deference to Cristina’s linguistic choice I will refer to widely asserted propositions about a group as “stereotypes” and narrowly, or individually, asserted propositions about a group as “fizzlewicks.””
Thank you, Lee, but there is no need. It’s not my personal choice, but the proper definition given by the behavioral ” sciences” (plural where I studied that) A stereotype is a belief held by a group, not an individual. How many? Nobody says.
” a single contradiction is insufficient.”
Ergo you cannot falsify a stereotype with your personal observation.
” But linguistic usage is an unfathomable mystery”
Alas, it’s another science.
” But perhaps those that wish to do it, do it for a reason associated with a stereotype about stereotypes – that stereotypes are typically unfavourable (ie critical of the group in question) and false. Thus a stereotype might be particularly harmful – widely held false unfavourable propositions about a group are going to be more unfairly harmful to members of that group than equivalent fizzlewicks. But as the book I mentioned up top explains this stereotype of stereotypes – that they are usually false and unfavourable turns out to be false. Most stereotypes are favourable and true (see book for details.)”
True (vox populi approaches vox Dei and all that)
” a single contradiction is insufficient.”
Ergo you cannot falsify a stereotype with your personal observation.
I’m sorry, Cristina, but your ergo simply doesn’t ergo. It’s got nothing to do with who is doing the observing, but whether the observing – by whoever – contradicts the stereotype. What is insufficient is a single contradiction. A thousand contradictions out of a thousand observations by a single observer will be quite sufficient to falsify pretty much any stereotype.
If the English (of long ago) believed that Chinese people have tails, a single traveller making thousands of observations of untailed Chinese, and none of tailed Chinese, can falsify the stereotype of the Chinese as a tailed people. Even if the rest of the population of England has never seen a Chinese person, and continues to believe in the stereotype, despite the traveller’s report.
Sorry, but you’re all getting caught up in the truth or falsity of stereotypes, but that’s a side issue. (Personally, I agree that they are usually true; as to whether they are usually “favorable”, I’m not so sure.) What matters is not whether a stereotype is accurate, but whether it is believed by a sufficiently large group of people. (Insert your own definition of “sufficiently large”; it’s unspecified.) And once that critical mass of believers is achieved it’s extremely difficult to overcome, objective falsification notwithstanding. One traveler reporting that Chinese don’t have tails isn’t going to disabuse the masses of that notion.
I agree with Brian that stereotypes provide “crude first approximations”, and are quite useful as such because they generally have some basis in reality (or they wouldn’t have arisin in the first place). And I like Niall’s description of them as “the verbal equivalent of a statistical summary”, because any statistical analysis admits of the possibility (likelihood) of outliers (which is why we developed the concept of standard deviations).
Shlomo is correct that economics isn’t real science. It belongs in the Behavioral Sciences department, not the Math Department. It wasn’t always the case, but for the last century or so economics has suffered from physics envy.
Laird are you of the view that human behavior can be studied scientifically? If so can you explain? Political science for instance strikes me as a misnomer.
If the Chinese people don’t have tails, why do they have tailors? Got you there!
Science is the study of any natural phenomenon, including human behavior, which may help us predict what shape that phenomenon may take in the future. The degree of this predictability varies widely depending on the phenomenon studied. The degree of variance is the greatest when we compare behavior of inanimate objects (the study of physics) with that of living creatures, especially sentient ones such as humans. That great degree of variance in itself does not make the study of human behavior and the attempts to predict it (by means of stereotyping, among others) useless or pointless. And, while behavioral sciences may not yield themselves well to the Popperian Scientific Method, saying that human behavior may not be studied scientifically in the general sense of the word is throwing the baby out with the tub water.
The term ‘science’ is derived from the Latin ‘scire’ to know, hence ‘scio’ is ‘I know’.
What has become what we now call physics and its branches was once termed ‘natural philosophy’. I dare say that most people these days regard ‘science’ as something done by people wearing lab coats because of the experimental nature of it (most astronomy etc. excluded).
Broadly, human behaviour can be studied by reason, and making deductions about what is logical about human behaviour (the inescapable search for a maximisation of utility as a taxis), i.e. by looking at what people consciously do, we see what they do to maximise utility, even if the choice is constrained by fear or force.
Or human behaviour can be studied by recording outcomes, looking at what people have done, and trying to conclude from that what the motive was.
Shlomo, I agree with what Alisa and Mr Ed said. Human behavior most certainly can be studied scientifically*; the adverb refers to a process, not an outcome. My point was simply that economics is the study of applied human behavior, not physics, and as such is not subject to rigorous quantification.
* And as Alexander Pope taught us, “the proper study of mankind is man.”
I confess myself puzzled by the notion that human behaviour cannot be studied scientifically – sufficiently so, that I assume that such folk have some rather particular notion of what scientific study entails. Perhaps it would be helpful if such folk – if they be listening – could indicate whether they think that animal behaviour can be studied scientifically. Is the problem specifically with humans, or more generally with biological objects ?
Lee Moore writes:
Over many years, I worked on Automatic Speech Recognition (ASR) – so by machine. As part of that work, I frequently had contact with applied psychologists working on human perception of speech and other sounds. The work they did showed quite definitely that there is scientific work in that field, and all fields associated with how the brain and associated sensors (ears, eyes, etc) work. This is even though there is variability of response between different people, different occasions and different though related measurements. So there is quite a lot of experimental measurement that is variable, though within bounds that allow useful conclusions. Animal behaviour can also be studied in similar ways, though the opinions of the subjects is usually more difficult to determine and is subject to greater quantisation of the different responses.
There are two specific aspects of the issues with experiments on human brains and related functioning.
Firstly, there is the problem with large groupings (as usually found with economics and often found with sociology). Experiments cannot be recreated with adequate repeatability, in the same way as with physics and chemistry and even with smaller-scale biological and applied psychological experiments.
Secondly, the subjects of experiments on humans can often be equally as ‘intelligent’ as are those studying them. Humans may well anticipate the expected results and deliver results that vary from them – this is particularly the case where experiments are repeated on the same subject (either in exactly the same way or in a way that can be related back to the earlier experiment). Thus the assumption of “all other things being equal” often does not apply, and this much more so that with experiments on animals.
How this all relates back to Brian’s original post on stereotypes is, I think, getting rather tenuous.
Best regards
I think this is the crux of the matter.
Good to see an emerging consensus that i can agree with, in all comments beginning with Alisa @10:44 today. (There were some more before that.)
WRT physics envy in psychology and the social sciences, i recommend (to those who can think of nothing better to do) reading the essay: Newton’s effect on scientific standards, by Imre Lakatos. The thrust of the essay is that Newton has had a two-sided effect: his scientific work has had a positive effect on the hard sciences (duh!); while his professed methodology has had a negative effect on the “soft” sciences.
The root of the problem (and of course i am oversimplifying) is that Newton felt the need to legitimize his results by a spurious epistemology, which has little to do with the methods he actually employed: the epistemology was designed a posteriori to legitimize his results. (Whewell said pretty much the same thing about a century before Lakatos; though at the time of Whewell, Newton had not yet exerted a pernicious effect on the social sciences.)
The reason Newton felt the need for a spurious epistemology, is that the Cartesian ideas generally accepted at the time de-legitimized Newton’s results.
Imre Lakatos is an under-appreciated intellectual giant. Of course he was standing on the shoulders of other giants, most notably his mentor Karl Popper; though they had a falling out, unresolved by the time of Lakatos’ early death.
Alisa,
I obviously was not arguing that the study of human behavior is pointless or useless.
To call the study of human behavior scientific is to degrade the term science by nullifying its distinguishing characteristics – chief among them the rigorous quantification, as Laird puts it, that is so aptly applied to such fields as physics, chemistry, etc.
I’m not arguing that we should not study human behavior; I’m arguing that to call the study of human behavior science not only plays into the hands of leftists who want to quantify such things but is also simply not true and, thus, leads to such things as Keynesian economics.
Austrian economics is not science; it is philosophy (in the modern sense of the term).
Snorri,
Personally, I’m glad to see an emerging consensus that I can disagree with. I mean that seriously. Being a retrograde, stubborn contrarian gets me up in the morning!
Lee Moore,
I can say that I personally do not believe that the study of the behavior of sentient creatures is science. There are multiple ways I could explain why but ultimately at the end of the day I must confess that I suspect the reason comes down to my stance on the mind-body problem. A creature that is alive is not a mechanical robot as Hobbes suggested. It cannot be studied as one studies why the rain falls or how quickly moons circle planets.
But man fools himself.
” Personally, I’m glad to see an emerging consensus that I can disagree with. I mean that seriously. Being a retrograde, stubborn contrarian gets me up in the morning!” LOL
” A creature that is alive is not a mechanical robot as Hobbes suggested. It cannot be studied as one studies why the rain falls or how quickly moons circle planets.”
Precisely. The struggle to call ” science” the study of human behavior is a tendentious attempt to reduce us to the status of automatons.
Shlomo:
There you go! trapped into Newton’s pernicious self-serving propaganda, without a glimmer of a way out!
Much better science than Keynesian economics; though no more than marginalist micro-economics.
My contrarianism takes a different form: i compulsively look for some claim with which i can disagree; but when i find it, i am not happy: i get angry. (Less angry after adopting a low-carb diet.)
I know you didn’t, and I am not arguing – I simply quoted your question to Laird because it got me thinking further (and aloud).
The importance of quantification depends on the application of the scientific results in question.
When talking about physics, such applications are usually referred to as engineering, and any engineer will tell you that applied scientific quantification needs to be only as rigorous as the client demands.
The same is true when it comes to behavioral sciences, (we don’t usually call applications thereof engineering, although there is such a thing as social engineering). Here, as with physical engineering, quantification depends on the demands of the “client”. An individual seeking to study human nature and apply that knowledge in his private life and interactions with other individuals, is one such client of behavioral sciences, and he is not likely to demand quantification nearly as rigorous as someone who wants to buy a car or have a bridge built.
Another sort of client for behavioral sciences, especially economics, are governments and various institutions they control (formally and informally). With these clients, rigorous quantification does come into play – or, to be more precise, and far more importantly, the appearance of rigorous quantification (appearance, because we all seem to agree that actual quantification of human behavior, including human interactions we call economics, cannot be nearly as rigorous as that of physics).
I couldn’t agree more. But neither is Keynesianism – the only difference being that the Austrians seem to be much more honest about not being scientists. However, I could still refer to psychology and sociology as behavioral sciences, without much cringing. Applications thereof is a different matter again – see ‘clients’ above.
” An individual seeking to study human nature and apply that knowledge in his private life and interactions with other individuals, is one such client of behavioral sciences”
The above description encompasses every human being, including those unaware of the existence of behavioral sciences.
” However, I could still refer to psychology and sociology as behavioral sciences, without much cringing”
Why?
No it does not, as not all individuals are interested in behavioral sciences.
Because while the quantification employed by psychologists and sociologists, if any, is unrigorous compared to those employed by physicists, they don’t tend to present appearances to the contrary – unlike economists.
Snorri,
I suppose this might be true if I held Newtonian physics entirely correct while free of any philosophical underpinning; sadly, the depth of my thinking is rather deeper, I’m afraid. I am a big fan of self-serving propaganda in general, though.
Alisa appears to disagree with your perspective on this, which reminds me of the old quip that a family may be unhappy in many ways but happy in just one. Truth, of course, is singular: it is always aligned with purity and unity.
Perhaps a reasonable definition of science is a field of study where one can make a hypothesis and test that hypothesis through a controlled experiment – this is being able to make a falsifiable prediction and being able to test it while controlling for all confounding variables. I must note that I have never seen an economist, political scientist, or sociologist test a theory on humans while controlling for all confounding variables – if for the woefully simple reason that man is incapable of understanding most (let alone all!) of the (confounding) variables that renders all human decisions intrinsically unique at the moments they are made. This reality makes forming testable theories regarding anything involving human nature a futile endeavor.
Not many political scientists predicted that World War One would break out when and how it did. Not many political scientists predicted that the Cold War would end when and how it did. Not many political scientists predicted the rise of ISIS.
What theories were specifically supported by those political scientists that did predict these events (I assume there must be some, right?)? How have we successfully applied their theories? And how can we test their theories in controlled experiments without excessive confounding variables?
Alisa,
This is the modern delusion that one can engineer society as one engineers a watch. They are plainly different things.
Just because some people can make changes to society based on data and analyze the results does not make it a form of engineering.
So let me see if I understand your argument.
Businesses claim to sell (A) which they contend has value because of (B). Governments and other institutions buy (A). Therefore, there must be value in (A). Therefore, (B) is true.
Not convinced.
Lots of people and businesses sell bull*hit.
If the entire healthcare system were privatized and if also everyone could purchase any sort of drug for any reason at any time without the sign-off of a psychiatrist then how many psychiatrists would be out of jobs?
But psychoanalysis must be science because, um, people say it is and they spend big bucks to get it.
Maybe the problem is that in our Enlightened “scientific age” we forget that a field of study may be valuable even though it is not scientific? Maybe.
You obviously don’t, and for the life of me I can’t even see what are you trying to say, and how it relates to my comments. No offense etc.
Much fewer than you seem to imagine, but that’s just my anecdotal impression – not a scientific statement by any means.
Nope, not really – not in the sense you seem to imply. Science, as Laird had aptly mentioned, is a process, a field of study is where that process takes place, and from where it takes its object of study. Both are valuable to what I referred to as “clients”. Does that help?
You said that research grounded in behavioral science is sold to clients such as governments. I was saying that I don’t see how this proves behavioral science to be science. Maybe I misunderstood.
Actually, yes, that does help me understand your perspective a little better. Thanks.
I just do not see science as a method taking place in behavioral fields that focus on human behavior like economics, political science, sociology etc.
Shlomo:
Sadly your thinking is rather shallow, i’m afraid: you fail to understand the contrast between Newtonian physics and Newtonian propaganda. (See my comment on Lakatos.)
As a rough outline this is pretty good, but the devil is in the details.
First, i note that you did not say anything about _quantitative_ predictions; and rightly so: while desirable, they are not indispensable.
Second, an experiment is not always needed to test a prediction: one can test astronomical predictions without experiments.
Third, and most important, how do you know that you have controlled for ALL confounding variables? it is a contradiction of terms to say that you know that there are no confounding variables that you don’t know about. The reality is that, in physics as in economics, one has to have a theory about which variables are relevant and which variables aren’t. If your predictions are correct, then it’s OK: neither the theory making the predictions nor the theory about which variables are relevant, are falsified.
The funny thing is, if your predictions turn out to be incorrect, then, in physics as in economics, you can always blame some variable that has not been accounted for; e.g. an as yet undiscovered planet, or dark matter, just to mention ad hoc assumptions that have actually been used. Note that the first assumption has been found to be correct, and the second is widely believed to be true.
I suppose that i could easily find examples of ad hoc assumptions to avoid falsification of Keynesian economics, if i thought that Krugman’s column is worth my time.
I also note that you are confusing making predictions with making correct predictions. Science is a process, as others have pointed out: making predictions is what matters, making correct predictions will come in due time.
Snorri : an experiment is not always needed to test a prediction: one can test astronomical predictions without experiments
Interesting thought, perhaps “observation” is a better word than “experiment.” Or, as I think I recall from a Feynman lecture it’s a matter of comparing the prediction with experience. So it’s certainly not necessary to set up an experiment, as such. It would be possible to advance an astronomical theory about what was happening in Galaxy XYZ, predicting that it would flash green and explode next Tuesday, and then one could look out of the window next Tuesday and see if it did flash green and explode. But since we’d be dealing with the speedalite and all that swaddling, actually it would all have happened millions of years ago.
So the key thing, I think, is to be able to make a prediction of something you don’t already know has happened, from what you do already know, or can deduce. And then test the prediction against the observation. Even if the event to be observed has already happened. It doesn’t have to be a “new” event, it just has to be “new” to the predictor.
That’s why gathering a huge pile of data and then making pretty patterns from it – a la global warming modelling – doesn’t count as science. Because your test “observations” are already in the data you knew when you made the model – you specifically made the model to accommodate them. But it would be perfectly OK, scientifically, for Scientist C to come up with a theory and predictions to test against historical observations already possessed by Scientist D, but which Scientist C hasn’t seen when making his predictions.
Alisa, do you know a single individual not “seeking to study human nature and apply that knowledge in his private life and interactions with other individuals”?
“But it would be perfectly OK, scientifically, for Scientist C to come up with a theory and predictions to test against historical observations already possessed by Scientist D, but which Scientist C hasn’t seen when making his predictions.”
Lee Moore, do you think that possible in the study of human behavior?
Possible, certainly. Perhaps less likely in these web-enabled days where information is more readily available and shared than before. But I can certainly imagine Scientist C suggesting that, say, men with allele B of some gene were 2.5 times as likely to be rapists as men with allele A. And it later being discovered that Scientist D happened – for some other purpose – previously to have compiled a database of rapists’ DNA in the City of Baltimore.
If you’re asking whether Scientist C should always and necessarily be believed when he claims never to have seen Scientist D’s collection of data, before that data later miraculously supports Scientist C’s theory, then (a) the answer is no and (b) you have an untrusting suspicious mind (which is a good thing to bring to any scientific enquiry)
Fair point. I did miss that.
Shlomo: glad to see that you are willing to concede a point, when you see it’s fair.
Lee: you might be aware that Einstein did not know that Brownian motion had already been observed, when he predicted it.
Also, of course, Einstein was not aware that general relativity would predict the anomalous perihelion of Mercury, when he conceived it. This latter case is of particular interest, because astronomers had been aware for a long time of the anomalous perihelion of Mercury, but had been unable to use induction to obtain general relativity from the data. The message seems to be that the hypothetico-deductive method succeeds where the strictly inductive method fails.
Touché 🙂
Cristina: probably none, but how is this relevant to the discussion?
Oh, dear.
Alisa, this way:
C: ” The above description encompasses every human being, including those unaware of the existence of behavioral sciences.”
A: ” No it does not, as not all individuals are interested in behavioral sciences.”
I know that, I’m asking why did you bring that up in the first place?
Alisa, Dec 9, 11:02 pm
“An individual seeking to study human nature and apply that knowledge in his private life and interactions with other individuals, is one such client of behavioral sciences, and he is not likely to demand quantification nearly as rigorous as someone who wants to buy a car or have a bridge built.”
…and?
Alisa, may I suggest you to read the complete thread? I think it could be easier that way.
Cristina, I read the complete thread, and made a comment to which you replied with the comment I am wondering about. Feel free to elucidate what brought your comment about in reply to mine – or not. And either way, have a good night.