A person wanting to use evolutionary theory to support something like more men in nursing or more women in science would have to explain how one should expect these career demographics to impact the relevant fitness variables and why.
That isn't what I was arguing. I was arguing against the idea that evolutionary theory was somehow a justification for gender roles. Just because there is a reason why things are the way they are doesn't make it a good reason, nor does it mean we should continue doing it in an appeal to tradition.
I have never -- not once -- heard anyone coherently explain why any particular human field of endeavor should be expected to have the same demographic balance as it would if its members were randomly sampled from the entire population, evolutionarily or otherwise.
I don't think any reasonable person expects every field to be perfectly divided into a representative sample of gender norms. What skewed statistics show is that there is a bias towards a group, and it's up to us to identify the 'why' of the basis.
To use StairC's firefighter example, looking for smoke is a perfectly acceptable first step. It's identifying a potential problem. Then we would obviously search for why there is smoke. If it's because someone is having a licensed bonfire, that's fine. But if we get back reports from everyone who visited the site with the smoke that a house was also 'hot' and 'there was a lot of flickering red light', we might draw a reasonable conclusion that there is some sort of fire going on in the house.
Now look, not every field has to be completely equally representative. That's not the issue. The issue is that it's clear there are still barriers in place to people choosing a field or profession they'd otherwise want to choose. If my wife can introduce herself to a patient every day as 'Dr. Mrs. Jay13x', wear a white coat every day that has 'Dr. Mrs. Jay13x' stitched into it, and be the one to arrive every time the patient asks to speak to their doctor, and still regularly get mistaken for a nurse by patients of all demographics, there is clearly some sort of bias going on against the perception that my wife could be a doctor. If every single female doctor she knows has the same stories, the area has some sort of bias. If this is something that gets reported across the country, that's even more of an indication something is up.
To then say 'well most women choose to be nurses not doctors' would completely miss the social and cultural issues that led to that point. Nursing and Teaching are female dominated fields because they were one of the few professions open to women just 60 years ago. The domination of those fields is a relic of a time where women were thought to be intellectually or professionally inferior, and often a cheaper alternative to hiring men. Cultural norms haven't progressed as rapidly as civil rights laws, and so there are still cultures that discourage women, either implicitly (through passive discrimination or sexism) or explicitly (as in they still have 'old boy's clubs') in a lot of fields. Now, I'm not saying those fields have to have an equal representation, I'm saying that we can't ignore that there is still progress to be made and that in a lot of fields the choices people make are still informed by old social norms.
But with if society itself is in itself an evolutionary construct, and its functions created for a purpose?
And what if in a rapidly advancing society technologically allowing genders to fill equivalent roles as compared to low tech societies, failure to adapt results in the death or decay of the system? From our point of view, that'd be bad for the most part, right?
The idea that gender roles evolved to be as they are today doesn't justify gender roles as they are today, it just explains them, much like the occasional vestigial tail. The reason equity in the work place matters, rather than people 'just leaving it up to choice' is because outdated norms inform the current situation, and cultural attitudes are a large part of what's pushing people away. It's hard for men to get into nursing when cultural attitudes look down on it, and it's hard for women to get into STEM for the same reason.
A person wanting to use evolutionary theory to support something like more men in nursing or more women in science would have to explain how one should expect these career demographics to impact the relevant fitness variables and why.
That isn't what I was arguing. I was arguing against the idea that evolutionary theory was somehow a justification for gender roles. Just because there is a reason why things are the way they are doesn't make it a good reason, nor does it mean we should continue doing it in an appeal to tradition.
Your comments imply that whether gender roles are "good" or "bad" is context dependent, i.e. gender roles might have been "good" in the past but are now "outdated" due to changing circumstances. If so, this implies that the purpose of gender roles is to maximize some kind of parameter in society, and that this parameter has changed over time.
Comments like these make me think we need to have a meta-discussion about what's actually being debated:
1. What makes a gender role "good" or "bad"? Is it a utilitarian maximization issue (i.e. let's maximize the total happiness and efficiency of society)? Is it instead about maximizing individual rights, even if doing so does not necessarily benefit society at large (i.e. men and women are intrinsically entitled to have equal rights even in a situation where this equality somehow harms society)? If it's the former (utilitarianism) then we might decide that certain gender roles were "good" in the past but are "bad" today. If it's about the latter (human rights), then our gender norms shouldn't change because of technology or cultural attitudes.
2. Are gender roles "outdated" in the specific context of modern western society, or were gender roles a bad thing in the past too? In other words, did society's gender-based expectations only recently become out-of-whack because of rapid changes in technology and communication, or have gender norms lagged behind the times for hundreds or thousands of years?
3. It sounds like you think "a rapidly advancing society" can dictate that it's time to do away with certain gender norms. Could the reverse happen? In other words, could future advancements in society favor the creation of new gender norms instead? Would it be morally OK to impose new gender norms on people?
Your comments imply that whether gender roles are "good" or "bad" is context dependent, i.e. gender roles might have been "good" in the past but are now "outdated" due to changing circumstances. If so, this implies that the purpose of gender roles is to maximize some kind of parameter in society, and that this parameter has changed over time.
More like 'correct' or 'incorrect'. Let's preface this by saying that many gender norms are based on incorrect stereotypes in terms of what women are capable of doing, which were reinforced by the gender norms in a self-fulfilling loop. If you treat them as inferior, you're going to perceive them as inferior and create a feedback loop.
The 'good' and 'bad' doesn't really enter into it, at least in terms of what I'm arguing.
1. What makes a gender role "good" or "bad"? Is it a utilitarian maximization issue (i.e. let's maximize the total happiness and efficiency of society)? Is it instead about maximizing individual rights, even if doing so does not necessarily benefit society at large (i.e. men and women are intrinsically entitled to have equal rights even in a situation where this equality somehow harms society)? If it's the former (utilitarianism) then we might decide that certain gender roles were "good" in the past but are "bad" today. If it's about the latter (human rights), then our gender norms shouldn't change because of technology or cultural attitudes.
I suppose I would fall under maximizing individual rights based on what I've said so far. My issue is that the fundamental arguments behind a lot of these discussions are flawed. As I stated in my previous post, women don't just 'choose' Nursing and Teaching, they're largely driven there because of cultural norms, while they're perfectly capable of most professions. Just like any civil rights issue, even when the legal barriers are gone the social and cultural barriers still exist.
2. Are gender roles "outdated" in the specific context of modern western society, or were gender roles a bad thing in the past too? In other words, did society's gender-based expectations only recently become out-of-whack because of rapid changes in technology and communication, or have gender norms lagged behind the times for hundreds or thousands of years?
That's a good point. Don't look too deeply at the evolutionary standpoint of my argument, I was being a bit hyperbolic while I dismissed the idea that evolution justifies norms. Basically, when people are considered property for thousands of years, it can be hard to break out of old established patterns.
In terms of 'lagging behind', the issue is what I just mentioned, that legal barriers have finally been removed after thousands of years, but social and cultural barriers are sluggish in following suit.
3. It sounds like you think "a rapidly advancing society" can dictate that it's time to do away with certain gender norms. Could the reverse happen? In other words, could future advancements in society favor the creation of new gender norms instead? Would it be morally OK to impose new gender norms on people?
Again, it's the fact that the legal requirements for women to conform to certain roles have been removed while the other barriers have not, or have been slow to change.
And honestly, yeah, tech advancements could change what we consider traditional gender roles quite a bit, but it would really depend on the tech, wouldn't it?
Nursing and Teaching are female dominated fields because they were one of the few professions open to women just 60 years ago. The domination of those fields is a relic of a time where women were thought to be intellectually or professionally inferior, and often a cheaper alternative to hiring men.
At the risk of further derailment: I doubt that was the reasoning as much as agricultural society undergoing industrialization thinking of women as physically inferior or capable of hard labor. Before efficient birth control and the concept of workplace safety there was also a serious incentive to not put women into work in factories handling dangerous chemicals, which could easily damage not only the mother but also the child. I would also argue that the concept of women being intellectually inferior is contradictory to the claim that women were enlisted as teachers, and that it most likely has more to do with the fact that writing, reading and languages were considered appropriate hobbies for women for long before they became teachers, and they were therefore simply suitable for that job. To back up my claims I would note that women are the predominant gender amongst teachers in nearly all cultures globally - many of which think of teachers as very valuable people, and in some cases as a well-paying job. As a long-term teacher, which was a state-funded position, my grandmother received higher pension than I will likely receive wages in the first 30 years of my career.
Quote from Valanarch »
I recognize that it wasn't arbitrary before. However, due to the existence of birth control, it now has lost its purpose and become arbitrary.
The claim that birth control somehow nullifies gender roles directly assumes that no women will ever want to become pregnant, or that employers will not consider the possibility that their future employees might want to. Simply looking at statistics show that a considerably larger percentage of women than men are pregnant at any given moment.
Quote from Valanarch »
If anyone was allowed to feel the emotions they wanted, have the roles they wanted, be in the groups they wanted, wear what they wanted, be called what they wanted, and do whatever they wanted without having to take gender into consideration, how would people feel dysphoric?
Ah, the magical world where there is no conflict amongst anyone, everyone gets to be in the groups they want despite the groups not wanting them to be there, everyone willingly cooperating with people and calling them what they want without actually knowing what it is, and people doing whatever they want and this somehow not rubbing other people the wrong way - who might choose to punch them in the face because that is what they wanted. Indeed, you have described something that is either an utopia or pure anarchy, and I am not actually sure which. Either way, I am certain that the state you described is unattainable for as long as two people exist - and that has nothing to do with gender or gender roles but with people being people.
And as has been shown earlier in the thread, people would still feel dysphoria - because they would identify with a different gender than what their sex matches. There is also the fact that people that are free to be attracted to certain people do not necessarily receive the same attraction back.
Quote from Valanarch »
If anyone could be allowed to feel what we now associate with femininity and masculinity without it being considered abnormal for those characteristics and feelings to be felt by someone of a different sex, there wouldn't be dysphoria.
It has nothing to do with the feelings, but the expression of said feelings. In fact I recall a research that showed that not only men do have emotions in response to multiple forms of stimulus, those emotions are stronger than those of women - at least if neurological scanning is correct. Men simply are expected to have much stronger inhibition mechanisms when it comes to the display of said emotions.
Quote from Crashing00 »
If it did, we could measure the consequences of our failure to maintain this balance in terms of observable failure of our society to continue itself.
The United States with a large hispanic population most likely cannot, but Japan and Finland certainly can. In contrast many developing countries display great population growth. Amazingly enough the conclusion to be drawn is that high technology levels and highly educated female workforce both significantly reduce birth rate. The explanation can be derived from evolutionary psychology and game theory, notably the handicap principle and prisoner's dilemma. Countries with high technology level and educated women allow women for more time to choose a desirable mate, which elevates the degree of handicapping men need to do in order to demonstrate their fitness. It is a prisoner's dilemma situation where competing males systematically betray each other in order to have a higher chance of mating - where they would all have higher number of offspring if they all decided to limit their earnings and education level to a certain standard, or just to live in a third world craphole.
Of course, the ability to draw conclusions like this is why some people vehemently oppose evolutionary biology. Another fan favorite is the fact that evolutionary biology can explain rape as an action that in some cases increases the fitness of the rapist. The naturalistic fallacy is of course the problem here, and the predisposition of some people to think that anything that can be explained by nature is good. I partially blame the modern day marketing and the "all natural"-strategies.
Private Mod Note
():
Rollback Post to RevisionRollBack
The Sage is occupied with the unspoken
and acts without effort.
Teaching without verbosity,
producing without possessing,
creating without regard to result,
claiming nothing,
the Sage has nothing to lose.
Your comments imply that whether gender roles are "good" or "bad" is context dependent, i.e. gender roles might have been "good" in the past but are now "outdated" due to changing circumstances. If so, this implies that the purpose of gender roles is to maximize some kind of parameter in society, and that this parameter has changed over time.
More like 'correct' or 'incorrect'. Let's preface this by saying that many gender norms are based on incorrect stereotypes in terms of what women are capable of doing, which were reinforced by the gender norms in a self-fulfilling loop. If you treat them as inferior, you're going to perceive them as inferior and create a feedback loop.
The 'good' and 'bad' doesn't really enter into it, at least in terms of what I'm arguing.
I think it really does come down to "good" and "bad," but let's ask some more questions and find out:
Let's say society has the following widespread gender norm: Men are more likely to be successful engineers then women. As you mentioned, the mere fact that this norm exists is self-fulfilling. It will become a true statement by virtue of being a widespread norm, since engineering firms will want to hire men over women, believing the men more likely to be successful. Accordingly, this gender norm is "correct" in that it correctly reflects the state of reality. So when you say "correct," and "incorrect," I think you mean something other than "correctly reflects the state of reality."
So I think what you really mean is something like "correctly reflects biological attributes of men and women that are not tied to cultural norms." But there are two issues with using "correct" in this way:
(1) You're not avoiding the whole good/bad thing. You're just kicking the can one step down the road. What you're saying, implicitly, is that a gender norm is good if it meets your definition of "correct," and bad if it meets your definition of "incorrect." But you haven't said why this makes the norm good or bad. The why is very important, because that's really the core of what's being debated here. Is it a utilitarian reason? Is it something to do with natural rights? Is it some other reason altogether?
(2) Some gender norms don't really fit into the "correct" and "incorrect" mold in a coherent way. What about this norm: "Women ought to wear dresses and men ought not to wear dresses." Is this norm "correct" or "incorrect?"
1. What makes a gender role "good" or "bad"? Is it a utilitarian maximization issue (i.e. let's maximize the total happiness and efficiency of society)? Is it instead about maximizing individual rights, even if doing so does not necessarily benefit society at large (i.e. men and women are intrinsically entitled to have equal rights even in a situation where this equality somehow harms society)? If it's the former (utilitarianism) then we might decide that certain gender roles were "good" in the past but are "bad" today. If it's about the latter (human rights), then our gender norms shouldn't change because of technology or cultural attitudes.
I suppose I would fall under maximizing individual rights based on what I've said so far. My issue is that the fundamental arguments behind a lot of these discussions are flawed. As I stated in my previous post, women don't just 'choose' Nursing and Teaching, they're largely driven there because of cultural norms, while they're perfectly capable of most professions. Just like any civil rights issue, even when the legal barriers are gone the social and cultural barriers still exist.
If it's about individual rights, then would you agree that it doesn't make sense to call gender norms "outdated?" Presumably you would say that women have been entitled to the same human rights as men since the dawn of humankind. So if the problem with gender norms is that they restrict human rights in unequal ways, then these norms aren't outdated; instead they've always been "bad," "incorrect," oppressive or whatever term you want to use. It's not that society has progressed too fast. Even if society was not progressing at all we would need to do away with unequal norms.
2. Are gender roles "outdated" in the specific context of modern western society, or were gender roles a bad thing in the past too? In other words, did society's gender-based expectations only recently become out-of-whack because of rapid changes in technology and communication, or have gender norms lagged behind the times for hundreds or thousands of years?
That's a good point. Don't look too deeply at the evolutionary standpoint of my argument, I was being a bit hyperbolic while I dismissed the idea that evolution justifies norms. Basically, when people are considered property for thousands of years, it can be hard to break out of old established patterns.
In terms of 'lagging behind', the issue is what I just mentioned, that legal barriers have finally been removed after thousands of years, but social and cultural barriers are sluggish in following suit.
Am I correct that you think the "old established patterns" were always wrong? It was wrong to treat women unequally 2000 years ago and it's just as wrong today? Or was there a time when those old norms were good (or at least acceptable) but that time has now passed?
3. It sounds like you think "a rapidly advancing society" can dictate that it's time to do away with certain gender norms. Could the reverse happen? In other words, could future advancements in society favor the creation of new gender norms instead? Would it be morally OK to impose new gender norms on people?
Again, it's the fact that the legal requirements for women to conform to certain roles have been removed while the other barriers have not, or have been slow to change.
And honestly, yeah, tech advancements could change what we consider traditional gender roles quite a bit, but it would really depend on the tech, wouldn't it?
The reason I'm asking this question is because I want to get in your head and understand what you're really arguing for. Specifically, what's the goal with abolishing gender norms? Is it to do away with all gender norms and create a gender-egalitarian society (the rights-based position)? Or is it to select the most efficient set of norms for society (the utilitarian position)?
Let me illustrate what I'm talking about with an extreme example. Let's say we do a bunch of studies and determine that we could drastically reduce rape by imposing a male-only curfew. Men would not be allowed outside their houses after 10pm.
The pure utilitarian response to this norm: "First, show me the studies and prove to me that this would be a highly effective way to prevent rape. But if it turns out that you're right and we could prevent a ton of rape with this policy, then we should consider it. We should weigh the costs against the benefits."
The pure human rights response to this norm: "Any norm that imposes special burdens or restrictions on someone just by virtue of their gender or sex is wrong. It doesn't matter how much good this policy would do, it's inherently a bad policy because it's discriminatory."
So I'm trying to see which side you come down on (or maybe you come down on some third side I'm not thinking about).
At the risk of further derailment: I doubt that was the reasoning as much as agricultural society undergoing industrialization thinking of women as physically inferior or capable of hard labor. Before efficient birth control and the concept of workplace safety there was also a serious incentive to not put women into work in factories handling dangerous chemicals, which could easily damage not only the mother but also the child.
Yeah... this isn't really true. Women were exposed to just as dangerous conditions during the industrial revolution as part of the textile industry. In South and Southeast asia, you see the same thing happening today.
I would also argue that the concept of women being intellectually inferior is contradictory to the claim that women were enlisted as teachers,
I didn't say it made sense. They were put into teaching because of their mother-like qualities. They were not, of course, believed to be intellectual equals of men.
and that it most likely has more to do with the fact that writing, reading and languages were considered appropriate hobbies for women for long before they became teachers, and they were therefore simply suitable for that job.
No. Up until the 19th century, teaching was man's job. I wish I could find the newspaper article from the 1830s I saw a while back that talked about how women's qualities make them better teachers, or the article that discussed to politics behind them being able to pay women much less than men, but women didn't take over teaching until the early-mid 19th century. Men still dominated the upper level positions in teaching, as well, and still do today, despite the enormous gender gap (an effect known as the glass escalator).
To back up my claims I would note that women are the predominant gender amongst teachers in nearly all cultures globally - many of which think of teachers as very valuable people, and in some cases as a well-paying job. As a long-term teacher, which was a state-funded position, my grandmother received higher pension than I will likely receive wages in the first 30 years of my career.
Well, only globally if you don't include Africa. But even then, there is the effect of globalization over the last 100 years that influenced this trend, especially when you consider the imperial influence of just a few countries in the 19th century.
(1) You're not avoiding the whole good/bad thing. You're just kicking the can one step down the road. What you're saying, implicitly, is that a gender norm is good if it meets your definition of "correct," and bad if it meets your definition of "incorrect." But you haven't said why this makes the norm good or bad. The why is very important, because that's really the core of what's being debated here. Is it a utilitarian reason? Is it something to do with natural rights? Is it some other reason altogether?
No, I'm not. I'm saying one is rooted in facts, and another isn't.
Sure, I have personal beliefs about what constitutes a 'good' or 'bad' norm, but that's irrelevant to the point I was making.
(2) Some gender norms don't really fit into the "correct" and "incorrect" mold in a coherent way. What about this norm: "Women ought to wear dresses and men ought not to wear dresses." Is this norm "correct" or "incorrect?"
Sure, and this isn't a gender norm I'm trying to decide by correct or incorrect, any more than we could ask if a tie should be formal wear. I was specifically talking about norms that inform career choice. Clothing traditions are actually a really complex topic and branch out further than the simple sexisms that pose a barrier to entry in career fields.
If it's about individual rights, then would you agree that it doesn't make sense to call gender norms "outdated?" Presumably you would say that women have been entitled to the same human rights as men since the dawn of humankind. So if the problem with gender norms is that they restrict human rights in unequal ways, then these norms aren't outdated; instead they've always been "bad," "incorrect," oppressive or whatever term you want to use. It's not that society has progressed too fast. Even if society was not progressing at all we would need to do away with unequal norms.
This is essentially moral relativism. Its easy to sit back in my comfy chair and type up how people in the olden days should have done something. But I don't face the same realities they faced, and while I like to think I know a better way of life, my great-grandchildren may look back at me as barbaric for whatever reason. I call certain norms outdated because they fail to live up to the new framework we've created. As a society, we've made a value judgment that women should be allowed an equal chance in the workplace as men. But thinking that should be the case and dismantling all social barriers are two different things.
I don't mean 'society has progressed too fast', I mean we're still catching up in the social norms department from what had been institutionalized years ago. Just because we make something legal doesn't fix all the social issues associated with it, and it can create all new ones.
Am I correct that you think the "old established patterns" were always wrong? It was wrong to treat women unequally 2000 years ago and it's just as wrong today? Or was there a time when those old norms were good (or at least acceptable) but that time has now passed?
Again, moral relativism. Of course, as a modern guy, I don't think I should have been able to purchase my wife with a cow and two sheep. But that doesn't mean the entire system was wrong then, or that our system is completely right. Systems with women being married off as property came together for a reason, just as arranged marriage came. Just because I would never want to be in an arranged marriage doesn't mean it's without it's own values. We just prioritize freedoms over the other social 'good' that institutions like that benefit.
We can detest all the horrific things the United States did in the 19th century, but we arguably wouldn't be as well off today if they hadn't been done. Does that make them 'good' or 'bad'? I think reality is a little more complicated than boiling everything down to good and bad, because it depends entirely on whose point of view you're taking, and when.
The reason I'm asking this question is because I want to get in your head and understand what you're really arguing for. Specifically, what's the goal with abolishing gender norms?
I never said abolish all gender norms. I've been pretty specific in what I disagree with.
Let me illustrate what I'm talking about with an extreme example. Let's say we do a bunch of studies and determine that we could drastically reduce rape by imposing a male-only curfew. Men would not be allowed outside their houses after 10pm.
The pure utilitarian response to this norm: "First, show me the studies and prove to me that this would be a highly effective way to prevent rape. But if it turns out that you're right and we could prevent a ton of rape with this policy, then we should consider it. We should weigh the costs against the benefits."
The pure human rights response to this norm: "Any norm that imposes special burdens or restrictions on someone just by virtue of their gender or sex is wrong. It doesn't matter how much good this policy would do, it's inherently a bad policy because it's discriminatory."
So I'm trying to see which side you come down on (or maybe you come down on some third side I'm not thinking about).
I think a 'pure' philosophical position is a dangerous one. Honestly, I'm for finding a balance between utilitarianism and the things I believe to be rights (which is another concept that's mired in moral relativism).
Look, I'm not trying to be evasive here, although it might come off that way a little bit. I think the problem is that we keep hitting on the moral relativism argument here, which can fundamentally confuse any argument. I don't believe in any kind of universal constant to morality, so when I talk about right or wrong I'm talking pretty strictly about as viewed from the present, not in any kind of objective sense. Does that clarify things for you? Would it be easier if I talked about from the subjective versus objective point of view?
I suppose, to give you a good answer, I would say abolishing specific gender stereotypes is 'good' from a modern perspective because we've expanded the idea of what freedoms apply to who.
So I think what you really mean is something like "correctly reflects biological attributes of men and women that are not tied to cultural norms."
Sure.
So my question is, why is this important? Why is "biological correctness" the goal we should be aiming for with cultural norms?
I mean, I can certainly come up with reasons, but the specific reason you choose has some pretty important implications. So I might say biological correctness is important because it's more economically efficient. Not having gender-based barriers to entering a particular profession leads to a more efficient allocation of resources and better utilization of people's talents. But this argument would imply I'm ok with gender discrimination in situations where discrimination turns out to be economically efficient for whatever reason. Not only that, it implies that we as a society should be actively looking for situations where discrimination would be economically efficient, and doing everything we can to implement this "efficient discrimination." Some people might not be ok with that, in which case they would need to select a different justification.
That's why I keep asking about this. I'm trying to figure out, for example, whether you think gender equality is important in-and-of-itself or whether you think gender equality is only important because it leads to some other good thing (like economic efficiency or utilitarian maximization or whatever).
(1) You're not avoiding the whole good/bad thing. You're just kicking the can one step down the road. What you're saying, implicitly, is that a gender norm is good if it meets your definition of "correct," and bad if it meets your definition of "incorrect." But you haven't said why this makes the norm good or bad. The why is very important, because that's really the core of what's being debated here. Is it a utilitarian reason? Is it something to do with natural rights? Is it some other reason altogether?
No, I'm not. I'm saying one is rooted in facts, and another isn't.
Sure, I have personal beliefs about what constitutes a 'good' or 'bad' norm, but that's irrelevant to the point I was making.
The statement "men tend to be more successful in engineering than women" is absolutely rooted in facts. It is 100% true that men tend to be more successful in engineering than women, and this statement has probably been true for as long as engineering has existed as a profession.
The problem is that the reason for the disparity is (probably) entirely cultural.
You're making the judgement that it's "good" for a norm to be rooted in biologically-motivated facts, and "bad" for a norm to be rooted in culturally-motivated facts. That's a good/bad value judgement, and it's central to the point you're making.
(2) Some gender norms don't really fit into the "correct" and "incorrect" mold in a coherent way. What about this norm: "Women ought to wear dresses and men ought not to wear dresses." Is this norm "correct" or "incorrect?"
Sure, and this isn't a gender norm I'm trying to decide by correct or incorrect, any more than we could ask if a tie should be formal wear. I was specifically talking about norms that inform career choice. Clothing traditions are actually a really complex topic and branch out further than the simple sexisms that pose a barrier to entry in career fields.
Should a gender norm ever inform career choice? Under what circumstances?
If it's about individual rights, then would you agree that it doesn't make sense to call gender norms "outdated?" Presumably you would say that women have been entitled to the same human rights as men since the dawn of humankind. So if the problem with gender norms is that they restrict human rights in unequal ways, then these norms aren't outdated; instead they've always been "bad," "incorrect," oppressive or whatever term you want to use. It's not that society has progressed too fast. Even if society was not progressing at all we would need to do away with unequal norms.
This is essentially moral relativism. Its easy to sit back in my comfy chair and type up how people in the olden days should have done something. But I don't face the same realities they faced, and while I like to think I know a better way of life, my great-grandchildren may look back at me as barbaric for whatever reason. I call certain norms outdated because they fail to live up to the new framework we've created. As a society, we've made a value judgment that women should be allowed an equal chance in the workplace as men. But thinking that should be the case and dismantling all social barriers are two different things.
I don't mean 'society has progressed too fast', I mean we're still catching up in the social norms department from what had been institutionalized years ago. Just because we make something legal doesn't fix all the social issues associated with it, and it can create all new ones.
I have no problem with you taking a moral relativist stance, but you keep waffling between relativism and absolutism. If you're a moral relativist, then why do you have a problem with social barriers to women's success in the workplace? To a moral relativist, the fact that society allows those barriers to exist means that they're, by definition, morally acceptable, right?
Systems with women being married off as property came together for a reason, just as arranged marriage came. Just because I would never want to be in an arranged marriage doesn't mean it's without it's own values. We just prioritize freedoms over the other social 'good' that institutions like that benefit.
Yes, the old systems "came together for a reason." And the discrimination that currently exists in society exists for a reason as well. But as you said just a little earlier:
Just because there is a reason why things are the way they are doesn't make it a good reason, nor does it mean we should continue doing it in an appeal to tradition.
Based on this statement, it shouldn't matter that things "came together for a reason." What we should be asking is whether they came together for a good reason. But when I ask you how we decide whether something is a good reason or not, you retreat to moral relativism, and claim "I think reality is a little more complicated than boiling everything down to good and bad, because it depends entirely on whose point of view you're taking, and when."
You can't have it both ways. If we care about whether something is a "good reason," then we need a definition of "good reason" that we can apply. If, on the other hand, you're a relativist and you reject the idea of "good reason" and "bad reason," then there's nothing wrong with sexist social norms as long as they exist for any reason at all.
The reason I'm asking this question is because I want to get in your head and understand what you're really arguing for. Specifically, what's the goal with abolishing gender norms?
I never said abolish all gender norms. I've been pretty specific in what I disagree with.
I never said "abolish all gender norms" either. To be clear, what I meant was: "what's your goal or purpose in seeking to abolish the specific gender norms you want to abolish?"
I don't believe in any kind of universal constant to morality, so when I talk about right or wrong I'm talking pretty strictly about as viewed from the present, not in any kind of objective sense. Does that clarify things for you?
As viewed from the present, society has certain sexist gender norms in place. If you're a moral relativist, by what authority do you pronounce these norms to be bad? If the moral standards of our day permit these sexist norms to continue to exist, then they're fine, aren't they?
Recognizing something is a symptom of a serious problem does not mean you’re privileging the diagnosis of the result. Let’s check out the conversation.
Firefighter 1: How should we look for dangerous fires?
Firefighter 2: How about we look for smoke?
...
Policeman 1: I found a dead body.
Policeman 2: Hah! Stairc is a murderer. I knew it all along!
Policeman 1: How did you arrive at that?
Policeman 2: Well, if Stairc were a murderer, we'd find dead bodies, right? And we found a dead body. WARNING SIGN!
Policeman 1: That seems like a stretch. After all, we'd see the same "warning sign" if you or I were the murderer.
Policeman 2: Uh, Stairc being a murderer has a symptom of finding dead bodies. Other stuff does too, sure, I'm not saying that all dead bodies result from Stairc committing murder - but we should certainly arrest Stairc first whenever we see a dead body...
Policeman 1: Stop privileging the Stairc-did-it hypothesis!
Why does Fireman 2 sound more reasonable than Policeman 2? Well, the answer has to do with basic Bayesian epistemics that you would have learned about if you had read the article I linked the first time around.
In order for X to be properly considered a "warning sign" for phenomenon Y, at least two things have to be true:
1) Whenever Y obtains, we expect X to obtain. "With sufficiently high probability, where there's fire, there's smoke."
2) Whenever Y doesn't obtain, we expect X not to obtain. "With sufficiently high probability, where there's no fire, there's no smoke."
In other words, we must obtain control of both the Type I and Type II error rates of our hypothesis test.
What we both seem to agree on is that looking for racism(/sexism/etc.) through the proxy variable of group outcomes satisfies (1) -- if there were systemic racism, you might well expect to see unequal outcomes. "With sufficiently high probability, where there's racism, there are unequal outcomes."
Where we appear to disagree is (2). Because you believe unequal outcomes are the "smoke" for racism's "fire," you must believe (2) is the case: "With sufficiently high probability, where there is no racism, there are no unequal outcomes." I, however, do not believe (2) is the case, nor will I grant it as a hypothesis.
Now remember: the entire point of this conversation is that I've never heard a coherent argument for (2). Your attempt at a coherent argument is to point out that the unequal outcome statistics are an adequate warning sign of racism. But that depends on (2) being true!
This is a circular argument.
I’m also rather uninterested in this line of argument, because it’s going to devolve into a cluster of definitions and nitpicking when you’ve already agreed on the main point.
You may be uninterested, but this is the whole thing. I intend to press the point so that it's clear to the audience what you're doing wrong.
Clearly I’m using the words “culture” and “institution” in a much broader sense than you are - and likely a much looser sense - because there really isn’t anything that would fall outside those definitions as I’m dealing with them.
Right, you're redefining words to have meanings that are incomprehensible to a fluent English speaker. You're trading one fallacy (non-exhaustive case analysis) for another. (equivocation)
The problem is that your incoherence here bubbles up through the rest of the argument. Every time you write "culture" an ordinary English speaker thinks you mean something like "the social beliefs and practices of humankind" but you actually mean "anything not biological." I don't need to tell you why this is a recipe for disaster. If you need a word for a new concept, make up one that isn't already freighted with meaning. Because I'm in a lighthearted mood I suggest "Stairculture," though you could just go with "non-biological."
Considering your own loose use of the word “coherent” I think you can appreciate that.
However loose my usage of the word "coherent" may have appeared to you previously, I hope it is now clear that it applies in spades.
Why not?
I believe I know exactly why not, but you should tell us. You haven’t justified this statement sir.
I don't know that I can add any more reasoning to the statement. It's almost tautological. If a process is random, then you expect a statistical test of randomness, when applied to that process, to show that the outcomes of that process appear random with high probability. If a process is not random, then you don't expect that. What more can I say? There's no missing reasoning there.
We expect CEOs to be experienced in business, educated, good at communication and public speaking and a whole host of other skills. Naturally their selection wouldn't be random in terms of these characteristics.
I concur that we can expect a high degree of non-randomness across fields in some characteristics (high IQ for scientists, say). The question, again, is why we would expect randomness in any other characteristics. There is no coin-flipping or dice-rolling going on in a typical CEO/scientist/athlete selection process. Where do you expect the randomness to come from?
I’d be delighted to take a look at this flaw in my reasoning here. I even gave you the sample argument I would have responded with. Please show me how that counter-argument I gave applies to the argument I’ve advanced here.
My objection was directed at the original argument I responded to -- which, to refresh your memory, was this: (I've cut out some fluff but those interested can just scroll up the thread for more context)
Quote from Stairc »
Men are disproportionately represented in high business offices compared to the population at large. In fact, I can narrow it down further to 2 possibilities.
1) It's just chance.
2) There's a reason for it.
...
And, since stastical improbabilities are indeed improbable, most assume either option 2 or 3 is at work.
The analogy that I hoped would get you to understand my objection without us having to go over material I'm sure you already know goes something like this:
Religious argument:
1) The cause of the universe is either natural or supernatural.
2) It is improbable that the cause of the universe is natural. (Because of fine-tuning or some other such nonsense.)
3) Therefore, it is most probable that the cause of the universe is supernatural.
Argument about racism:
1) The cause of unequal outcomes is either biology or culture.
2) It is improbable that the cause is biology. (Maybe? I actually asked you directly what you were alleging was improbable and you didn't answer.)
3) Therefore, it is most probable that the cause of unequal outcomes is culture.
The same flaws are present in both these arguments. Namely, why should I believe the dichotomy in (1) and why should I believe the assignment of probability in (2)?
Repeatedly stating that I'm wrong isn't useful.
Don't portray me dishonestly here. In every case where I've said you were wrong, I've also said why. In the few cases I abbreviated the reason, it's because I thought you would be able to recognize the mistakes by reference to religious arguments you've been in where you've seen the mistakes made by others.
Crashing, I try my best to avoid people setting up strawmen to such an extent. Whatever position you're arguing with isn't mine. I believe we are either utterly failing to communicate, or that you are not having this discussion in good faith. I prefer to believe the former.
In either case, I don't have a continued interest in discussing this topic with you.
In either case, I don't have a continued interest in discussing this topic with you.
At the beginning of this discussion, I was looking for a reason why we should expect random-seeming demographic outcomes. I asked that because in the absence of such an expectation, we have no way of knowing whether or not a non-random outcome is indicative of a problem, because the type I error rate is totally unbounded. It could be like using smoke alarms to detect fires on a planet where the ambient atmosphere is mostly smoke. Your alarm is constantly ringing but there's no actionable phenomenon taking place.
You chose to respond negatively, so I assumed you disagreed, and that could well be the straw man you speak of. Ultimately, though, at the end of our exchange, I don't believe I've received any such reasoning from you, so the matter stands where it was. Either you never disagreed with me in the first place, or we've failed to communicate as you suggest. Thanks for trying, though.
Private Mod Note
():
Rollback Post to RevisionRollBack
A limit of time is fixed for thee
Which if thou dost not use for clearing away the clouds from thy mind
It will go and thou wilt go, never to return.
Crashing, I try my best to avoid people setting up strawmen to such an extent. Whatever position you're arguing with isn't mine. I believe we are either utterly failing to communicate, or that you are not having this discussion in good faith. I prefer to believe the former.
In either case, I don't have a continued interest in discussing this topic with you.
I think you should reread Crashing's post. It seemed to me like a pretty direct refutation of the arguments you've been making.
I'm not as savvy with statistics as Crashing. This is how I'm simplifying his argument for my own understanding:
Example 1: I only have one person who texts me on a regular basis, named Joe. If my phone gets a text, it makes sense for me to assume it probably came from Joe. It's possible it came from someone else, but its reasonable for me to guess it came from the person I normally get texts from.
Example 2: I have 1000 people that text me on an equally regular basis. If my phone gets a text,it would not be rational of me to guess it came from Joe. He's one of a thousand people it could have come from. Sure, it's possible the text came from Joe, but there's no good reason for me to suspect it came from him specifically.
If someone sees gender inequality and says "we should investigate whether that inequality came from sexism," they're secretly making an affirmative claim. The affirmative claim is that we live in a world where sexism behaves more like Joe in example 1 than Joe in example 2. In other words, you're're making the claim that sexism is one of a fairly small number of possible causes of gender inequality. If sexism is one of thousands or millions of equiprobable causes of inequality (like in example 2), then it's not rational to nominate sexism as one of the prime suspects in the investigation.
I'm not as savvy with statistics as Crashing. This is how I'm simplifying his argument for my own understanding:
I'll respond to your own understanding then.
Example 1: I only have one person who texts me on a regular basis, named Joe. If my phone gets a text, it makes sense for me to assume it probably came from Joe. It's possible it came from someone else, but its reasonable for me to guess it came from the person I normally get texts from.
Example 2: I have 1000 people that text me on an equally regular basis. If my phone gets a text,it would not be rational of me to guess it came from Joe. He's one of a thousand people it could have come from. Sure, it's possible the text came from Joe, but there's no good reason for me to suspect it came from him and not someone else.
If someone sees gender inequality and says "we should investigate whether that inequality came from sexism," they're secretly making an affirmative claim. The affirmative claim is that we live in a world where sexism behaves more like Joe in example 1 than Joe in example 2. In other words, you're're making the claim that sexism is one of a fairly small number of possible causes of gender inequality. If sexism is one of thousands or millions of equiprobable causes of inequality (like in example 2), then it's not rational to nominate sexism as one of the prime suspects.
Crashing himself said that he'd advise people to get lumps checked for cancer, while also asserting that he recognizes cancer is actually a less likely explanation. If Crashing can do it, so can a civil rights activist.
Instead of altering my fireman example, engage with it. You job is to spot dangerous fires. You start by making a list of signs that are present when there are dangerous fires. You keep an eye out for those things.
Then someone reports a situation to you that has one or more of these signs. In point of fact, a fire alarm has gone off at the local college. At this point, you do *not* say to yourself, "A fire alarm is one sign of a dangerous fire, therefore this must be a dangerous fire". You say, "It might be a false alarm. However, to be safe, let's check it out." Then the fire department rolls out and establishes whether there's a fire or not.
The fire fighters are not concluding that the fire alarm represents a dangerous fire ahead of time. However, paying attention to fire alarms and checking them out is an efficient way to keep an eye out for things that *might* be dangerous fires. You should then conduct an unbiased investigation and draw what seems to be the most reasonable conclusion from the evidence.
Likewise, if you hear that a teacher is grading all his black students lower than all his white students - that's the smoke. Now you check to see if there's fire. You don't presume the teacher's guilt, you should examine the work and see if it's justified. You should examine the scores of the black kids compared to the white kids leading up to this and see if this represents a sudden change within this one classroom or whether it's more likely a fair assessment of the students' abilities. You should consider many factors before calling someone a racist. You should not, obviously, assume guilt without investigation.
TLDR
If someone sees gender inequality and says "we should investigate whether that inequality came from sexism," they're secretly making an affirmative claim. The affirmative claim is that we live in a world where sexism behaves more like Joe in example 1 than Joe in example 2.
Then by saying someone should investigate whether a lump might be cancerous, you secretly make an affirmative claim. The affirmative claim that we live in a world where cancer behaves more like Joe in example 1 than Joe in example 2 (i.e. is actually the MORE likely culprit, not even equally probable).
In order for X to be properly considered a "warning sign" for phenomenon Y, at least two things have to be true:
1) Whenever Y obtains, we expect X to obtain. "With sufficiently high probability, where there's fire, there's smoke."
2) Whenever Y doesn't obtain, we expect X not to obtain. "With sufficiently high probability, where there's no fire, there's no smoke."
In other words, we must obtain control of both the Type I and Type II error rates of our hypothesis test.
Is this not too narrow? Take Stairc's cancer/tumor example and plug it into (2) above. "With sufficiently high probability, where there's no cancer, there's no tumor." Quite often people will have a tumor that isn't cancerous, but you agreed earlier that a tumor is a warning sign of cancer. Either you are taking a very loose definition of 'sufficiently high probability', or (2) isn't always necessary for something to be a warning sign. And if (2) isn't necessary in the tumor scenario, why not? Is it 'because the consequences of a "tumor" outcome are so severe as to considerably overweight that option in the final calculation of expected value'? If someone were to convince you that the consequences of sexism in the workplace are severe, would you relent in making them prove that sexism in the workplace satisfies (2) above?
Taking a step back, the whole 'privileging the hypothesis' concept seems a bit misapplied here. The hypothesis that the skewed distribution of sexes across professions and positions is due in part to sexism would only be privileged if we had no reason to select that hypothesis over any other, right? Well, do we agree that some fields were predominated by women and some by men throughout large swaths of human history for reasons that were in large part sexist? If so, can we not consider this our prior distribution? Because this is how I see it. The Bayesian prior here, based on history, is 'a skewed distribution of sexes among professions is in large part due to sexism', which means that we need a whole lot of credible experience proving the opposite before the posterior doesn't include sexism as one of the major reasons for an unequal distribution of sexes. Certainly not the only reason, and certainly not a given, but a legitimate (and not privileged) option as one of the probable reasons for the skewed distribution of sexes among professions. To put it another way, if we agree about the historical sexism bit, then the situation is much more akin to bitterroot's Example 1 than Example 2.
Having gone through all that, your response here confuses me a bit:
Quote from Crashing00 » »
Quote from Stairc » »
There's no inherent reason why a field needs to have an equal distribution based on the population. It's not important in itself, but it's a warning sign that may indicate a problem or prejudice in the culture.
To say that it could indicate prejudice is to say nothing at all. People's life trajectories correlate with trillions of variables; any given statistical analysis could be responding to any subset of these correlations.
If I understand you correctly, you're saying that there are any number of reasons a person enters a particular profession, and to say that prejudice could be one of them is meaningless because there is no reason to consider that a significant reason over all the others - ie Stairc is privileging the prejudice hypothesis. Do I have this right, and if not would you clarify what you mean? The reason I ask is I'm trying to understand the crux of what you are arguing.
That's why I keep asking about this. I'm trying to figure out, for example, whether you think gender equality is important in-and-of-itself or whether you think gender equality is only important because it leads to some other good thing (like economic efficiency or utilitarian maximization or whatever).
Sure, I think equality is important in and of itself, because I'm product of today's changing cultural norms. Is it necessarily the most efficient, effective or 'best' for humanity's longevity? Maybe not, I don't know. I do believe it's a good thing because of both the utilitarian value and from an equality standpoint, and perhaps more importantly because if I was a woman, I'd want those options available to me.
The statement "men tend to be more successful in engineering than women" is absolutely rooted in facts. It is 100% true that men tend to be more successful in engineering than women, and this statement has probably been true for as long as engineering has existed as a profession.
The problem is that the reason for the disparity is (probably) entirely cultural.
And that's exactly the point I was making. Although you chose to phrase the 'norm' in such a way that it agreed with your point. If you phrased it as "Men tend to be more successful in engineering than women because they're better at the STEM fields", it wouldn't be rooted in facts, would it? I suspect you think I was trying to make a different point than I was.
The only point I was making was that, as above, the old cultural attitudes that associated capabilities with sex were simply incorrect factually, much like 19th century pseudo-science as to the capability of 'lesser races', and that the current state of things is due, in part, to those norms. Yes, I have personal beliefs that are more complex, but I was simply responding to a point.
I have no problem with you taking a moral relativist stance, but you keep waffling between relativism and absolutism. If you're a moral relativist, then why do you have a problem with social barriers to women's success in the workplace? To a moral relativist, the fact that society allows those barriers to exist means that they're, by definition, morally acceptable, right?
Moral relativism isn't amorality, as you seem to imply. It's simply the recognition that morality is subjective. I can both acknowledge that my beliefs are simply a product of my times AND hold those beliefs as having value. You keep asking question that probe both my personal and intellectual beliefs, which don't have to be the same thing, hence 'waffling'. I can view something as bad today without believing it was necessary 'bad' then. It depends entirely on the perspective of the group you're viewing things through.
But, since you want clarity, the bottom line is I believe old gender norms to be bad, at least in terms of restricting women from anything but domesticity.
Likewise, if you hear that a teacher is grading all his black students lower than all his white students - that's the smoke. Now you check to see if there's fire. You don't presume the teacher's guilt, you should examine the work and see if it's justified. You should examine the scores of the black kids compared to the white kids leading up to this and see if this represents a sudden change within this one classroom or whether it's more likely a fair assessment of the students' abilities. You should consider many factors before calling someone a racist. You should not, obviously, assume guilt without investigation.
I think that's fine, but I think when people use these kinds of purely statistical examples they tend to miss context.
The smoke isn't a white teacher scoring black students lower. The smoke is a white teacher scoring black students lower in culture with a history of racial discrimination and inequity. The framing of the situation changes the logical solutions.
Now, I get the point people try to make with these, in that correlation doesn't equal causation. I'm fully behind that.
Now, if you hand the teacher equal tests with white-sounding names or black-sounding names, and the black-sounding names score lower, you know you've probably got a problem.
And we've done that for women in STEM. There IS a bias, on a lot of different levels, for a lot of reasons, against women in certain fields. And it's not just 'sexism' in the classical sense, because women are just as likely to do it as men.
Is this not too narrow? Take Stairc's cancer/tumor example and plug it into (2) above. "With sufficiently high probability, where there's no cancer, there's no tumor." Quite often people will have a tumor that isn't cancerous, but you agreed earlier that a tumor is a warning sign of cancer. Either you are taking a very loose definition of 'sufficiently high probability', or (2) isn't always necessary for something to be a warning sign. And if (2) isn't necessary in the tumor scenario, why not? Is it 'because the consequences of a "tumor" outcome are so severe as to considerably overweight that option in the final calculation of expected value'? If someone were to convince you that the consequences of sexism in the workplace are severe, would you relent in making them prove that sexism in the workplace satisfies (2) above?
Look, we've got three or more different analogies on the table, all of which cash out differently upon evaluation. Each analogy is freighted with a bunch of additional factors, specific to the particular analogy, which cause us to evaluate them differently:
Analogy #1: 'lump/cancer'. Most people, myself included, would say getting the lump checked is probably a good idea.
Analogy #2: 'smoke/fire'. If we think about this one carefully, I think we find it to be somewhat different than lump/tumor. I myself have been involved in 200-some-odd events of fire detection systems going off in my building. Of those 200, only two represented actionable events, and even in those cases anyone with even a passing familiarity with the layout of the building would not have been in any real danger. It's gotten to the point where I don't even get out of my chair when the smoke alarm goes off unless building security comes by and makes me do so. Most other people in my building feel the same way.
This is quite a bit different than lump/tumor. I'd get every lump checked. I wouldn't panic at every smoke alarm.
Analogy #3: 'dead body/Stairc killed someone'. Most of us would agree that it's simply unreasonable to interrogate Stairc every time a dead body is found.
The question is, which of these buckets if any does 'unequal outcomes/X-ism' fall into? Is it like 'lump/cancer'? 'smoke/fire'? 'dead body/Stairc did it'?
You and Stairc are both asking me to treat it as if it were more like 'lump/cancer' than 'dead body/Stairc did it'. The problem is, I don't see any reason to believe that. That reasoning is the very thing I've been asking for since the beginning.
It's not enough merely to point out that the consequences of workplace sexism are bad. The consequences of letting Stairc keep killing people are bad too. What you need to get a handle on is the Type I error rate. How often does this so-called warning sign signal X-ism and how often is it detecting other things? We all agree that letting Stairc continue a killing spree is bad, if he's on one. The problem is that our test to figure out whether he's on one has such terrible flaws that we can't rationally act on it.
Taking a step back, the whole 'privileging the hypothesis' concept seems a bit misapplied here. The hypothesis that the skewed distribution of sexes across professions and positions is due in part to sexism would only be privileged if we had no reason to select that hypothesis over any other, right? Well, do we agree that some fields were predominated by women and some by men throughout large swaths of human history for reasons that were in large part sexist? If so, can we not consider this our prior distribution? Because this is how I see it. The Bayesian prior here, based on history, is 'a skewed distribution of sexes among professions is in large part due to sexism'
Firstly, in Bayesian reasoning, "prior" doesn't mean "historical." It means something closer akin to "hypothetical" -- it's what you're assuming to be true in advance of examining the evidence. That's why in statistical hypothesis testing, one purposefully chooses the prior to be hostile to the effect one is testing for. In other words, if we were looking for sexism, we would always choose a prior with no sexism.
Secondly, that something was historically highly weighted in a prior does not necessarily means it merits equal weighting when designing a prior today. Despite history, today I'm not going to put very much of my prior probability mass on, say, the statement "fire is made of phlogiston."
Thirdly, choosing a prior favorable to your preferred conclusion is exactly what privileging the hypothesis is!
which means that we need a whole lot of credible experience proving the opposite before the posterior doesn't include sexism as one of the major reasons for an unequal distribution of sexes.
And indeed, the reason I would be suspicious of someone choosing a prior straight out of the 19th century is because in the intervening years there has been an unending stream of evidence and experience of less systemic discrimination throughout Western societies. Starting from an historical prior and updating on recent history should cause everyone's posteriors to shift radically away from X-ism.
If I understand you correctly, you're saying that there are any number of reasons a person enters a particular profession, and to say that prejudice could be one of them is meaningless because there is no reason to consider that a significant reason over all the others - ie Stairc is privileging the prejudice hypothesis. Do I have this right, and if not would you clarify what you mean? The reason I ask is I'm trying to understand the crux of what you are arguing.
My position is not affirmative. Rather, I allege a gap in the argument of those who say that observing non-random-seeming demographic outcomes in a field is an adequate sign of systemic discrimination. The gap is that nobody has explained why the outcomes should appear random. As far as I can tell, nobody here has closed that gap.
The closest we've gotten is to make some analogies where in some cases you'd act even if that gap weren't closed. My argument there is that there are other cases where you wouldn't. It is then the burden of the person on the affirmative position to show that we are not in one of the cases where you wouldn't. Further, all of these analogies are different from the matter under study in significant ways, so one should be suspicious of reasoning from those analogies by default.
which means that we need a whole lot of credible experience proving the opposite before the posterior doesn't include sexism as one of the major reasons for an unequal distribution of sexes.
And indeed, the reason I would be suspicious of someone choosing a prior straight out of the 19th century is because in the intervening years there has been an unending stream of evidence and experience of less systemic discrimination throughout Western societies. Starting from an historical prior and updating on recent history should cause everyone's posteriors to shift radically away from X-ism.
So history is to be ignored, and for instance we shouldn't look to 19th century policies to find the reason for extreme poverty of American Indians? Well, we're not actively rounding them up and slaughtering them anymore, so it can't possible be that.
Look, overt discrimination has decreased because it's actively illegal, but as I just posted there is still an inherent bias against female candidates, even if it isn't as overt as 'oh she's a woman so she's not as good or she'll get pregnant in a year'. The effects of hundreds or thousands of years on our cultural aren't going to disappear overnight.
I allege a gap in the argument of those who say that observing non-random-seeming demographic outcomes in a field is an adequate sign of systemic discrimination. The gap is that nobody has explained why the outcomes should appear random. As far as I can tell, nobody here has closed that gap.
And, as I mentioned in a previous post, they shouldn't be completely 'random' (I assume by random you mean a true random distribution which would, in theory, match the population at large with a sufficiently sized sample). You're taking it to an extreme here that no rational person would argue. No field will ever have a perfect distribution, for a lot of reasons that go beyond one characteristic.
You're also implying that the only thing people have to go on is the statistic itself (hence the gap you mention), which isn't true. We know there are problems with women in STEM (and other fields) because it's constantly being reported on. That doesn't mean it's overt sexism (or racism, or whatever x-ism) or that it's the sole (or even a major) factor, or that anyone in particular is to blame for the skewed demographics - just that it exists.
How did you get that out of what I said? If anything, I am advocating not ignoring history -- specifically, the past 100 or so years. Freezing our Bayesian priors as we would have set them in 1920 is what I would call ignoring history, and taking history into account by updating on the intervening 100 years should cause our posteriors to be significantly less weighted towards the various -isms.
and for instance we shouldn't look to 19th century policies to find the reason for extreme poverty of American Indians? Well, we're not actively rounding them up and slaughtering them anymore, so it can't possible be that.
I must point out that I never said anything about American Indians. This is a new topic you are introducing, and all the words you've said here you've literally just put into my mouth. You didn't even bother to ask my position before presuming whatever it is you're presuming here. I don't know what explains the poverty of American Indians, I take no affirmative position on the matter, and my response to anyone taking the affirmative position that the poverty of American Indians today is caused by historical policies would be "prove it."
And, as I mentioned in a previous post, they shouldn't be completely 'random' (I assume by random you mean a true random distribution which would, in theory, match the population at large with a sufficiently sized sample). You're taking it to an extreme here that no rational person would argue. No field will ever have a perfect distribution, for a lot of reasons that go beyond one characteristic.
Yes, to be clear, by "random" I mean "appears as if it had been randomly sampled from the population at large." And I'm not the one taking it to the extreme! If someone does a study of a subpopulation, it results in a distribution that doesn't look like a random sample, and that person concludes that this is an indicator of discrimination, it's that person who's "taking it to an extreme." It's they who are relying on the hypothesis I'm questioning here.
You're also implying that the only thing people have to go on is the statistic itself (hence the gap you mention), which isn't true. We know there are problems with women in STEM (and other fields) because it's constantly being reported on. That doesn't mean it's overt sexism (or racism, or whatever x-ism) or that it's the sole (or even a major) factor, or that anyone in particular is to blame for the skewed demographics - just that it exists.
The question isn't whether the effect being measured exists, though. If you show me a study that says black students are on average doing worse than white students in school, I'll believe that. The question is whether the existence of that particular effect is strong enough evidence of something we'd recognize as discrimination.
And all the studies purporting to go further than just measuring the gross effect, including the ones you posted, are highly contentious. Particularly those about gender gaps. For every study alleging a gender gap caused by discrimination there is another study purporting to explain it in terms of something that isn't discrimination. I'm not saying either side has the right of it, necessarily, but the waters are muddy enough where the issue isn't going to be resolved by the studies.
Look, overt discrimination has decreased because it's actively illegal, but as I just posted there is still an inherent bias against female candidates, even if it isn't as overt as 'oh she's a woman so she's not as good or she'll get pregnant in a year'. The effects of hundreds or thousands of years on our cultural aren't going to disappear overnight.
Again, we're looking for the cause of these patterns. Detecting the patterns is one thing; tracing them back to whatever cause you're alleging is another. It's the tracing-back part where all of my objections lie. The fact that overt discrimination has been greatly reduced is relevant to the 'tracing back' part, because if we start with a high-discrimination prior and update on overt discrimination being reduced, we end up with a lower-discrimination posterior.
And that was the point I was making to jlpitt -- when we actually update the 1920s prior to the 2020s prior, we're looking at a situation where it's much harder to start from an hypothesis of discrimination.
I must point out that I never said anything about American Indians. This is a new topic you are introducing, and all the words you've said here you've literally just put into my mouth. You didn't even bother to ask my position before presuming whatever it is you're presuming here. I don't know what explains the poverty of American Indians, I take no affirmative position on the matter, and my response to anyone taking the affirmative position that the poverty of American Indians today is caused by historical policies would be "prove it."
Right, I know you didn't mention it - just like no one was actually discussing the best way of diagnosing a fire. It was just a better analogy than fire, since you didn't like some of those (and it IS directly attributable, but that's a whole other topic, you're right).
The question isn't whether the effect being measured exists, though. If you show me a study that says black students are on average doing worse than white students in school, I'll believe that. The question is whether the existence of that particular effect is strong enough evidence of something we'd recognize as discrimination.
I think this is where we are misunderstanding each other here. It doesn't have to be discrimination. It can just be a cultural bias, with no one intentionally discriminating against anyone.
And all the studies purporting to go further than just measuring the gross effect, including the ones you posted, are highly contentious. Particularly those about gender gaps. For every study alleging a gender gap caused by discrimination there is another study purporting to explain it in terms of something that isn't discrimination. I'm not saying either side has the right of it, necessarily, but the waters are muddy enough where the issue isn't going to be resolved by the studies.
Well, yeah. The pay gap isn't because of active discrimination, it's because women aren't as assertive in pay negotiations as men are. Because they aren't socialized to be assertive. It's also part of the reason the 'glass escalator' exists with men in women dominated fields. And there are a bunch of other differences, revolving around child care and maternity leave, etc. I would think very rarely would you find that there is active discrimination.
Although I'd be interested to read a study that places the gender gap's causal factors on something other than a cultural norm.
Again, we're looking for the cause of these patterns. Detecting the patterns is one thing; tracing them back to whatever cause you're alleging is another. It's the tracing-back part where all of my objections lie. The fact that overt discrimination has been greatly reduced is relevant to the 'tracing back' part, because if we start with a high-discrimination prior and update on overt discrimination being reduced, we end up with a lower-discrimination posterior.
And that was the point I was making to jlpitt -- when we actually update the 1920s prior to the 2020s prior, we're looking at a situation where it's much harder to start from an hypothesis of discrimination.
Just in case we're still crossing wires here after what I just said, define 'discrimination' for me, in your words.
The pay gap isn't because of active discrimination, it's because women aren't as assertive in pay negotiations as men are. Because they aren't socialized to be assertive
I'll come back and address your other arguments when I have time a little later, but I want to zero in on this statement for one second.
I don't think it's anywhere close to proven or accepted that this statement explains the pay gap. There are many other reasons proposed for the pay gap, even among academic feminist sociologists.
But let's pretend for a second that the pay gap is 100% attributable to women being less assertive in pay negotiations. Why would you assume that this is entirely cultural? The effect of testosterone levels on things like aggressiveness, assertiveness, and the like is extremely well-studied. And anecdotally, you can read accounts of trans men who started taking testosterone and experienced big changes in their personalities, including increases in aggressive and assertive behavior.
But let's pretend for a second that the pay gap is 100% attributable to women being less assertive in pay negotiations. Why would you assume that this is entirely cultural? The effect of testosterone levels on things like aggressiveness, assertiveness, and the like is extremely well-studied. And anecdotally, you can read accounts of trans men who started taking testosterone and experienced big changes in their personalities, including increases in aggressive and assertive behavior.
Sure - I note that there are a lot of reasons why there is a pay gap. That's just one of the most generally accepted ones. Usually the pay gap statistics are misleading because they aren't 1:1 comparisons between job types. Just so we're clear, when talking about the pay gap, I'm looking at the 6.6% figure, which is the one actually adjusted for a real comparison.
Now, why would I assume that it's cultural? Maybe it isn't. It could just as easily be explained by women's role as caregivers in the family, and on average spend an extra hour a day caring for their family than their male counterparts, and that cuts into their actual working time over the course of the year. But... wouldn't that be a cultural issue? Why are women spending more time taking care of their families than their husbands or SOs?
But let's pretend for a second that the pay gap is 100% attributable to women being less assertive in pay negotiations. Why would you assume that this is entirely cultural? The effect of testosterone levels on things like aggressiveness, assertiveness, and the like is extremely well-studied. And anecdotally, you can read accounts of trans men who started taking testosterone and experienced big changes in their personalities, including increases in aggressive and assertive behavior.
Sure - I note that there are a lot of reasons why there is a pay gap. That's just one of the most generally accepted ones. Usually the pay gap statistics are misleading because they aren't 1:1 comparisons between job types. Just so we're clear, when talking about the pay gap, I'm looking at the 6.6% figure, which is the one actually adjusted for a real comparison.
Now, why would I assume that it's cultural? Maybe it isn't. It could just as easily be explained by women's role as caregivers in the family, and on average spend an extra hour a day caring for their family than their male counterparts, and that cuts into their actual working time over the course of the year. But... wouldn't that be a cultural issue? Why are women spending more time taking care of their families than their husbands or SOs?
1. Are testosterone levels cultural? If you make the claim that the 6.6% number derives from culture, you need to show that it could not be explained by "non-cultural" sources.
2. I have no idea whether womens' role as primary caregiver in the family is cultural or not. I will say I'm aware of no human culture, ever, in which women have not been primary caregivers. I will also note that females are primary caregivers in most (all?) non-human primates. This at least suggests the difference might not be cultural.
1. Are testosterone levels cultural? If you make the claim that the 6.6% number derives from culture, you need to show that it could not be explained by "non-cultural" sources.
Do testosterone levels actually contribute to what you're talking about, though? It's an explanation, but not necessarily a good one. Females are the more aggressive sex in many mammals, and it's usually the social disposition of the group that affects aggressiveness.
Plus, some studies have suggested that testosterone isn't necessarily the 'aggressive' horomone we've previously believed it to be.
2. I have no idea whether womens' role as primary caregiver in the family is cultural or not. I will say I'm aware of no human culture, ever, in which women have not been primary caregivers. I will also note that females are primary caregivers in most (all?) non-human primates. This at least suggests the difference might not be cultural.
It's most, not all.
And, to be fair, it doesn't matter all that much when I've already posted evidence that women get offered less money when equally qualified, even before negotiations begin. Since both men and women are guilty of it, it implies something beyond biology.
1. Are testosterone levels cultural? If you make the claim that the 6.6% number derives from culture, you need to show that it could not be explained by "non-cultural" sources.
Do testosterone levels actually contribute to what you're talking about, though? It's an explanation, but not necessarily a good one. Females are the more aggressive sex in many mammals, and it's usually the social disposition of the group that affects aggressiveness.
Plus, some studies have suggested that testosterone isn't necessarily the 'aggressive' horomone we've previously believed it to be.
You, Jay13x, are taking the affirmative position that the gender pay gap is explained by "culture" (which you seem to define, roughly, as anything that is not an innate biological attribute of humans).
I, bitterroot, am not taking any affirmative position. I am completely agnostic as to whether the pay gap even exists, and if it does exist, whether it is caused by biology, culture, or a combination of the two.
To convince me that I should believe you rather than someone else, you have to prove your case. That means you have to demonstrate that biology is not the explanation for the pay gap.
I offered testotsterone as an example of something that might conceivably explain the gap. You need to prove that testosterone cannot explain the gap, not merely offer reasons why testosterone might not explain the gap. (I certainly acknowledge that it might not explain the gap, this is why I'm agnostic here. But you have not eliminated the possibility that it might explain the gap).
Moreover, there are other possible biological bases for these differences that are not testosterone. You need to explain why we can dismiss these as well.
2. I have no idea whether womens' role as primary caregiver in the family is cultural or not. I will say I'm aware of no human culture, ever, in which women have not been primary caregivers. I will also note that females are primary caregivers in most (all?) non-human primates. This at least suggests the difference might not be cultural.
It's most, not all.
And, to be fair, it doesn't matter all that much when I've already posted evidence that women get offered less money when equally qualified, even before negotiations begin. Since both men and women are guilty of it, it implies something beyond biology.
Let's assume for a moment it is true that women are offered less "even before negotiations begin."
Womens' (typical) role as primary caregiver might conceivably be the reason that women are offered less - employers (correctly?) expect women to work fewer hours. You have not shown that the tendency of women to act in a caregiving role is determined by culture as opposed to biology. Therefore you have not demonstrated that the tendency of employers to offer women less money is determined by culture as opposed to biology.
It's not enough merely to point out that the consequences of workplace sexism are bad. The consequences of letting Stairc keep killing people are bad too. What you need to get a handle on is the Type I error rate. How often does this so-called warning sign signal X-ism and how often is it detecting other things? We all agree that letting Stairc continue a killing spree is bad, if he's on one. The problem is that our test to figure out whether he's on one has such terrible flaws that we can't rationally act on it.
Why does lump/cancer get the pass if not because of the potential severity of consequences? This was the reasoning you offered in comment #101. What I’m trying to find out is, at what point can something be considered a warning sign without satisfying (2)? Is there any sort of general guidelines we can use?
Firstly, in Bayesian reasoning, "prior" doesn't mean "historical." It means something closer akin to "hypothetical" -- it's what you're assuming to be true in advance of examining the evidence.
Right, but hypotheticals are influenced by the past. I think we agree here, since you cite the last 100 years as important for how much mass to assign to sexism in our prior.
That's why in statistical hypothesis testing, one purposefully chooses the prior to be hostile to the effect one is testing for. In other words, if we were looking for sexism, we would always choose a prior with no sexism.
OK, let me first of all say that all I know about Bayesian estimation and credibility I learned studying for exams a little while back, and I haven’t formally applied it since. However, I don’t remember this being the case.
As an example, let’s look at fire alarms in your building. If you were to construct a prior right now about the likelihood of an alarm being false, would it look roughly like a binomial distribution with P(false alarm) = .99? Then, after you find out whether the alarm was false or not, you take the outcome and assign some amount of credibility to it, and end up with a posterior. If it turns out there really was a fire, you update your p to something like .90, and then next time there’s a fire this new distribution becomes your prior. Is that not the gist of it?
The point being, I understand the prior to be the best model of actual behavior we can come up with, independent of any hypothesis testing that may occur later. I'm not 'looking for sexism', I'm looking for an accurate model and then seeing whether sexism carries meaningful weight in that model.
Secondly, that something was historically highly weighted in a prior does not necessarily means it merits equal weighting when designing a prior today.
Re-reading my post, it does indeed look like I was advocating taking a prior straight out of the 19th century, but that’s not what I meant to do. What I mean to say is that sexism was so prevalent and entrenched in the workplace up through the 19th century that the last 100 years of progress have not made this information irrelevant. We have made remarkable progress, but we have not cured sexism.
And indeed, the reason I would be suspicious of someone choosing a prior straight out of the 19th century is because in the intervening years there has been an unending stream of evidence and experience of less systemic discrimination throughout Western societies. Starting from an historical prior and updating on recent history should cause everyone's posteriors to shift radically away from X-ism.
---
Freezing our Bayesian priors as we would have set them in 1920 is what I would call ignoring history, and taking history into account by updating on the intervening 100 years should cause our posteriors to be significantly less weighted towards the various -isms.
---
if we start with a high-discrimination prior and update on overt discrimination being reduced, we end up with a lower-discrimination posterior.
---
And that was the point I was making to jlpitt -- when we actually update the 1920s prior to the 2020s prior, we're looking at a situation where it's much harder to start from an hypothesis of discrimination.
I completely agree that any prior we construct today is much less weighted toward sexism than it would have been 100, or even 60 years ago. However, there is still mass associated with sexism that is weightier than most of the other trillions of possible explanations for the disparity of sexes we observe today. This is all I’m really trying to say.
Rather, I allege a gap in the argument of those who say that observing non-random-seeming demographic outcomes in a field is an adequate sign of systemic discrimination.
If this is what you’re arguing, then I think we might agree. But you seem to be saying more than this. The way you responded to Stairc, you seem to be arguing against viewing skewed demographics as a possible sign of sexism, not just an adequate one. It’s this stance, and not your quoted stance, that I am arguing against. I’m really trying to be careful about putting words in your mouth here, but I don’t know how else to interpret this:
There's no inherent reason why a field needs to have an equal distribution based on the population. It's not important in itself, but it's a warning sign that may indicate a problem or prejudice in the culture.
To say that it could indicate prejudice is to say nothing at all. People's life trajectories correlate with trillions of variables; any given statistical analysis could be responding to any subset of these correlations.
Please let me know if I really am misinterpreting you.
Private Mod Note
():
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
I don't think any reasonable person expects every field to be perfectly divided into a representative sample of gender norms. What skewed statistics show is that there is a bias towards a group, and it's up to us to identify the 'why' of the basis.
To use StairC's firefighter example, looking for smoke is a perfectly acceptable first step. It's identifying a potential problem. Then we would obviously search for why there is smoke. If it's because someone is having a licensed bonfire, that's fine. But if we get back reports from everyone who visited the site with the smoke that a house was also 'hot' and 'there was a lot of flickering red light', we might draw a reasonable conclusion that there is some sort of fire going on in the house.
Now look, not every field has to be completely equally representative. That's not the issue. The issue is that it's clear there are still barriers in place to people choosing a field or profession they'd otherwise want to choose. If my wife can introduce herself to a patient every day as 'Dr. Mrs. Jay13x', wear a white coat every day that has 'Dr. Mrs. Jay13x' stitched into it, and be the one to arrive every time the patient asks to speak to their doctor, and still regularly get mistaken for a nurse by patients of all demographics, there is clearly some sort of bias going on against the perception that my wife could be a doctor. If every single female doctor she knows has the same stories, the area has some sort of bias. If this is something that gets reported across the country, that's even more of an indication something is up.
To then say 'well most women choose to be nurses not doctors' would completely miss the social and cultural issues that led to that point. Nursing and Teaching are female dominated fields because they were one of the few professions open to women just 60 years ago. The domination of those fields is a relic of a time where women were thought to be intellectually or professionally inferior, and often a cheaper alternative to hiring men. Cultural norms haven't progressed as rapidly as civil rights laws, and so there are still cultures that discourage women, either implicitly (through passive discrimination or sexism) or explicitly (as in they still have 'old boy's clubs') in a lot of fields. Now, I'm not saying those fields have to have an equal representation, I'm saying that we can't ignore that there is still progress to be made and that in a lot of fields the choices people make are still informed by old social norms.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
Your comments imply that whether gender roles are "good" or "bad" is context dependent, i.e. gender roles might have been "good" in the past but are now "outdated" due to changing circumstances. If so, this implies that the purpose of gender roles is to maximize some kind of parameter in society, and that this parameter has changed over time.
Comments like these make me think we need to have a meta-discussion about what's actually being debated:
1. What makes a gender role "good" or "bad"? Is it a utilitarian maximization issue (i.e. let's maximize the total happiness and efficiency of society)? Is it instead about maximizing individual rights, even if doing so does not necessarily benefit society at large (i.e. men and women are intrinsically entitled to have equal rights even in a situation where this equality somehow harms society)? If it's the former (utilitarianism) then we might decide that certain gender roles were "good" in the past but are "bad" today. If it's about the latter (human rights), then our gender norms shouldn't change because of technology or cultural attitudes.
2. Are gender roles "outdated" in the specific context of modern western society, or were gender roles a bad thing in the past too? In other words, did society's gender-based expectations only recently become out-of-whack because of rapid changes in technology and communication, or have gender norms lagged behind the times for hundreds or thousands of years?
3. It sounds like you think "a rapidly advancing society" can dictate that it's time to do away with certain gender norms. Could the reverse happen? In other words, could future advancements in society favor the creation of new gender norms instead? Would it be morally OK to impose new gender norms on people?
The 'good' and 'bad' doesn't really enter into it, at least in terms of what I'm arguing.
I suppose I would fall under maximizing individual rights based on what I've said so far. My issue is that the fundamental arguments behind a lot of these discussions are flawed. As I stated in my previous post, women don't just 'choose' Nursing and Teaching, they're largely driven there because of cultural norms, while they're perfectly capable of most professions. Just like any civil rights issue, even when the legal barriers are gone the social and cultural barriers still exist.
That's a good point. Don't look too deeply at the evolutionary standpoint of my argument, I was being a bit hyperbolic while I dismissed the idea that evolution justifies norms. Basically, when people are considered property for thousands of years, it can be hard to break out of old established patterns.
In terms of 'lagging behind', the issue is what I just mentioned, that legal barriers have finally been removed after thousands of years, but social and cultural barriers are sluggish in following suit.
Again, it's the fact that the legal requirements for women to conform to certain roles have been removed while the other barriers have not, or have been slow to change.
And honestly, yeah, tech advancements could change what we consider traditional gender roles quite a bit, but it would really depend on the tech, wouldn't it?
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
At the risk of further derailment: I doubt that was the reasoning as much as agricultural society undergoing industrialization thinking of women as physically inferior or capable of hard labor. Before efficient birth control and the concept of workplace safety there was also a serious incentive to not put women into work in factories handling dangerous chemicals, which could easily damage not only the mother but also the child. I would also argue that the concept of women being intellectually inferior is contradictory to the claim that women were enlisted as teachers, and that it most likely has more to do with the fact that writing, reading and languages were considered appropriate hobbies for women for long before they became teachers, and they were therefore simply suitable for that job. To back up my claims I would note that women are the predominant gender amongst teachers in nearly all cultures globally - many of which think of teachers as very valuable people, and in some cases as a well-paying job. As a long-term teacher, which was a state-funded position, my grandmother received higher pension than I will likely receive wages in the first 30 years of my career.
The claim that birth control somehow nullifies gender roles directly assumes that no women will ever want to become pregnant, or that employers will not consider the possibility that their future employees might want to. Simply looking at statistics show that a considerably larger percentage of women than men are pregnant at any given moment.
Ah, the magical world where there is no conflict amongst anyone, everyone gets to be in the groups they want despite the groups not wanting them to be there, everyone willingly cooperating with people and calling them what they want without actually knowing what it is, and people doing whatever they want and this somehow not rubbing other people the wrong way - who might choose to punch them in the face because that is what they wanted. Indeed, you have described something that is either an utopia or pure anarchy, and I am not actually sure which. Either way, I am certain that the state you described is unattainable for as long as two people exist - and that has nothing to do with gender or gender roles but with people being people.
And as has been shown earlier in the thread, people would still feel dysphoria - because they would identify with a different gender than what their sex matches. There is also the fact that people that are free to be attracted to certain people do not necessarily receive the same attraction back.
It has nothing to do with the feelings, but the expression of said feelings. In fact I recall a research that showed that not only men do have emotions in response to multiple forms of stimulus, those emotions are stronger than those of women - at least if neurological scanning is correct. Men simply are expected to have much stronger inhibition mechanisms when it comes to the display of said emotions.
The United States with a large hispanic population most likely cannot, but Japan and Finland certainly can. In contrast many developing countries display great population growth. Amazingly enough the conclusion to be drawn is that high technology levels and highly educated female workforce both significantly reduce birth rate. The explanation can be derived from evolutionary psychology and game theory, notably the handicap principle and prisoner's dilemma. Countries with high technology level and educated women allow women for more time to choose a desirable mate, which elevates the degree of handicapping men need to do in order to demonstrate their fitness. It is a prisoner's dilemma situation where competing males systematically betray each other in order to have a higher chance of mating - where they would all have higher number of offspring if they all decided to limit their earnings and education level to a certain standard, or just to live in a third world craphole.
Of course, the ability to draw conclusions like this is why some people vehemently oppose evolutionary biology. Another fan favorite is the fact that evolutionary biology can explain rape as an action that in some cases increases the fitness of the rapist. The naturalistic fallacy is of course the problem here, and the predisposition of some people to think that anything that can be explained by nature is good. I partially blame the modern day marketing and the "all natural"-strategies.
and acts without effort.
Teaching without verbosity,
producing without possessing,
creating without regard to result,
claiming nothing,
the Sage has nothing to lose.
I think it really does come down to "good" and "bad," but let's ask some more questions and find out:
Let's say society has the following widespread gender norm: Men are more likely to be successful engineers then women. As you mentioned, the mere fact that this norm exists is self-fulfilling. It will become a true statement by virtue of being a widespread norm, since engineering firms will want to hire men over women, believing the men more likely to be successful. Accordingly, this gender norm is "correct" in that it correctly reflects the state of reality. So when you say "correct," and "incorrect," I think you mean something other than "correctly reflects the state of reality."
So I think what you really mean is something like "correctly reflects biological attributes of men and women that are not tied to cultural norms." But there are two issues with using "correct" in this way:
(1) You're not avoiding the whole good/bad thing. You're just kicking the can one step down the road. What you're saying, implicitly, is that a gender norm is good if it meets your definition of "correct," and bad if it meets your definition of "incorrect." But you haven't said why this makes the norm good or bad. The why is very important, because that's really the core of what's being debated here. Is it a utilitarian reason? Is it something to do with natural rights? Is it some other reason altogether?
(2) Some gender norms don't really fit into the "correct" and "incorrect" mold in a coherent way. What about this norm: "Women ought to wear dresses and men ought not to wear dresses." Is this norm "correct" or "incorrect?"
If it's about individual rights, then would you agree that it doesn't make sense to call gender norms "outdated?" Presumably you would say that women have been entitled to the same human rights as men since the dawn of humankind. So if the problem with gender norms is that they restrict human rights in unequal ways, then these norms aren't outdated; instead they've always been "bad," "incorrect," oppressive or whatever term you want to use. It's not that society has progressed too fast. Even if society was not progressing at all we would need to do away with unequal norms.
Is that a fair summary of your beliefs?
Am I correct that you think the "old established patterns" were always wrong? It was wrong to treat women unequally 2000 years ago and it's just as wrong today? Or was there a time when those old norms were good (or at least acceptable) but that time has now passed?
The reason I'm asking this question is because I want to get in your head and understand what you're really arguing for. Specifically, what's the goal with abolishing gender norms? Is it to do away with all gender norms and create a gender-egalitarian society (the rights-based position)? Or is it to select the most efficient set of norms for society (the utilitarian position)?
Let me illustrate what I'm talking about with an extreme example. Let's say we do a bunch of studies and determine that we could drastically reduce rape by imposing a male-only curfew. Men would not be allowed outside their houses after 10pm.
The pure utilitarian response to this norm: "First, show me the studies and prove to me that this would be a highly effective way to prevent rape. But if it turns out that you're right and we could prevent a ton of rape with this policy, then we should consider it. We should weigh the costs against the benefits."
The pure human rights response to this norm: "Any norm that imposes special burdens or restrictions on someone just by virtue of their gender or sex is wrong. It doesn't matter how much good this policy would do, it's inherently a bad policy because it's discriminatory."
So I'm trying to see which side you come down on (or maybe you come down on some third side I'm not thinking about).
I didn't say it made sense. They were put into teaching because of their mother-like qualities. They were not, of course, believed to be intellectual equals of men.
No. Up until the 19th century, teaching was man's job. I wish I could find the newspaper article from the 1830s I saw a while back that talked about how women's qualities make them better teachers, or the article that discussed to politics behind them being able to pay women much less than men, but women didn't take over teaching until the early-mid 19th century. Men still dominated the upper level positions in teaching, as well, and still do today, despite the enormous gender gap (an effect known as the glass escalator).
Well, only globally if you don't include Africa. But even then, there is the effect of globalization over the last 100 years that influenced this trend, especially when you consider the imperial influence of just a few countries in the 19th century.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
No, I'm not. I'm saying one is rooted in facts, and another isn't.
Sure, I have personal beliefs about what constitutes a 'good' or 'bad' norm, but that's irrelevant to the point I was making.
Sure, and this isn't a gender norm I'm trying to decide by correct or incorrect, any more than we could ask if a tie should be formal wear. I was specifically talking about norms that inform career choice. Clothing traditions are actually a really complex topic and branch out further than the simple sexisms that pose a barrier to entry in career fields.
This is essentially moral relativism. Its easy to sit back in my comfy chair and type up how people in the olden days should have done something. But I don't face the same realities they faced, and while I like to think I know a better way of life, my great-grandchildren may look back at me as barbaric for whatever reason. I call certain norms outdated because they fail to live up to the new framework we've created. As a society, we've made a value judgment that women should be allowed an equal chance in the workplace as men. But thinking that should be the case and dismantling all social barriers are two different things.
I don't mean 'society has progressed too fast', I mean we're still catching up in the social norms department from what had been institutionalized years ago. Just because we make something legal doesn't fix all the social issues associated with it, and it can create all new ones.
Again, moral relativism. Of course, as a modern guy, I don't think I should have been able to purchase my wife with a cow and two sheep. But that doesn't mean the entire system was wrong then, or that our system is completely right. Systems with women being married off as property came together for a reason, just as arranged marriage came. Just because I would never want to be in an arranged marriage doesn't mean it's without it's own values. We just prioritize freedoms over the other social 'good' that institutions like that benefit.
We can detest all the horrific things the United States did in the 19th century, but we arguably wouldn't be as well off today if they hadn't been done. Does that make them 'good' or 'bad'? I think reality is a little more complicated than boiling everything down to good and bad, because it depends entirely on whose point of view you're taking, and when.
I never said abolish all gender norms. I've been pretty specific in what I disagree with.
I think a 'pure' philosophical position is a dangerous one. Honestly, I'm for finding a balance between utilitarianism and the things I believe to be rights (which is another concept that's mired in moral relativism).
Look, I'm not trying to be evasive here, although it might come off that way a little bit. I think the problem is that we keep hitting on the moral relativism argument here, which can fundamentally confuse any argument. I don't believe in any kind of universal constant to morality, so when I talk about right or wrong I'm talking pretty strictly about as viewed from the present, not in any kind of objective sense. Does that clarify things for you? Would it be easier if I talked about from the subjective versus objective point of view?
I suppose, to give you a good answer, I would say abolishing specific gender stereotypes is 'good' from a modern perspective because we've expanded the idea of what freedoms apply to who.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
So my question is, why is this important? Why is "biological correctness" the goal we should be aiming for with cultural norms?
I mean, I can certainly come up with reasons, but the specific reason you choose has some pretty important implications. So I might say biological correctness is important because it's more economically efficient. Not having gender-based barriers to entering a particular profession leads to a more efficient allocation of resources and better utilization of people's talents. But this argument would imply I'm ok with gender discrimination in situations where discrimination turns out to be economically efficient for whatever reason. Not only that, it implies that we as a society should be actively looking for situations where discrimination would be economically efficient, and doing everything we can to implement this "efficient discrimination." Some people might not be ok with that, in which case they would need to select a different justification.
That's why I keep asking about this. I'm trying to figure out, for example, whether you think gender equality is important in-and-of-itself or whether you think gender equality is only important because it leads to some other good thing (like economic efficiency or utilitarian maximization or whatever).
The statement "men tend to be more successful in engineering than women" is absolutely rooted in facts. It is 100% true that men tend to be more successful in engineering than women, and this statement has probably been true for as long as engineering has existed as a profession.
The problem is that the reason for the disparity is (probably) entirely cultural.
You're making the judgement that it's "good" for a norm to be rooted in biologically-motivated facts, and "bad" for a norm to be rooted in culturally-motivated facts. That's a good/bad value judgement, and it's central to the point you're making.
Should a gender norm ever inform career choice? Under what circumstances?
I have no problem with you taking a moral relativist stance, but you keep waffling between relativism and absolutism. If you're a moral relativist, then why do you have a problem with social barriers to women's success in the workplace? To a moral relativist, the fact that society allows those barriers to exist means that they're, by definition, morally acceptable, right?
Yes, the old systems "came together for a reason." And the discrimination that currently exists in society exists for a reason as well. But as you said just a little earlier:
Based on this statement, it shouldn't matter that things "came together for a reason." What we should be asking is whether they came together for a good reason. But when I ask you how we decide whether something is a good reason or not, you retreat to moral relativism, and claim "I think reality is a little more complicated than boiling everything down to good and bad, because it depends entirely on whose point of view you're taking, and when."
You can't have it both ways. If we care about whether something is a "good reason," then we need a definition of "good reason" that we can apply. If, on the other hand, you're a relativist and you reject the idea of "good reason" and "bad reason," then there's nothing wrong with sexist social norms as long as they exist for any reason at all.
I never said "abolish all gender norms" either. To be clear, what I meant was: "what's your goal or purpose in seeking to abolish the specific gender norms you want to abolish?"
As viewed from the present, society has certain sexist gender norms in place. If you're a moral relativist, by what authority do you pronounce these norms to be bad? If the moral standards of our day permit these sexist norms to continue to exist, then they're fine, aren't they?
Policeman 1: I found a dead body.
Policeman 2: Hah! Stairc is a murderer. I knew it all along!
Policeman 1: How did you arrive at that?
Policeman 2: Well, if Stairc were a murderer, we'd find dead bodies, right? And we found a dead body. WARNING SIGN!
Policeman 1: That seems like a stretch. After all, we'd see the same "warning sign" if you or I were the murderer.
Policeman 2: Uh, Stairc being a murderer has a symptom of finding dead bodies. Other stuff does too, sure, I'm not saying that all dead bodies result from Stairc committing murder - but we should certainly arrest Stairc first whenever we see a dead body...
Policeman 1: Stop privileging the Stairc-did-it hypothesis!
Why does Fireman 2 sound more reasonable than Policeman 2? Well, the answer has to do with basic Bayesian epistemics that you would have learned about if you had read the article I linked the first time around.
In order for X to be properly considered a "warning sign" for phenomenon Y, at least two things have to be true:
1) Whenever Y obtains, we expect X to obtain. "With sufficiently high probability, where there's fire, there's smoke."
2) Whenever Y doesn't obtain, we expect X not to obtain. "With sufficiently high probability, where there's no fire, there's no smoke."
In other words, we must obtain control of both the Type I and Type II error rates of our hypothesis test.
What we both seem to agree on is that looking for racism(/sexism/etc.) through the proxy variable of group outcomes satisfies (1) -- if there were systemic racism, you might well expect to see unequal outcomes. "With sufficiently high probability, where there's racism, there are unequal outcomes."
Where we appear to disagree is (2). Because you believe unequal outcomes are the "smoke" for racism's "fire," you must believe (2) is the case: "With sufficiently high probability, where there is no racism, there are no unequal outcomes." I, however, do not believe (2) is the case, nor will I grant it as a hypothesis.
Now remember: the entire point of this conversation is that I've never heard a coherent argument for (2). Your attempt at a coherent argument is to point out that the unequal outcome statistics are an adequate warning sign of racism. But that depends on (2) being true!
This is a circular argument.
You may be uninterested, but this is the whole thing. I intend to press the point so that it's clear to the audience what you're doing wrong.
Right, you're redefining words to have meanings that are incomprehensible to a fluent English speaker. You're trading one fallacy (non-exhaustive case analysis) for another. (equivocation)
The problem is that your incoherence here bubbles up through the rest of the argument. Every time you write "culture" an ordinary English speaker thinks you mean something like "the social beliefs and practices of humankind" but you actually mean "anything not biological." I don't need to tell you why this is a recipe for disaster. If you need a word for a new concept, make up one that isn't already freighted with meaning. Because I'm in a lighthearted mood I suggest "Stairculture," though you could just go with "non-biological."
However loose my usage of the word "coherent" may have appeared to you previously, I hope it is now clear that it applies in spades.
I don't know that I can add any more reasoning to the statement. It's almost tautological. If a process is random, then you expect a statistical test of randomness, when applied to that process, to show that the outcomes of that process appear random with high probability. If a process is not random, then you don't expect that. What more can I say? There's no missing reasoning there.
I concur that we can expect a high degree of non-randomness across fields in some characteristics (high IQ for scientists, say). The question, again, is why we would expect randomness in any other characteristics. There is no coin-flipping or dice-rolling going on in a typical CEO/scientist/athlete selection process. Where do you expect the randomness to come from?
My objection was directed at the original argument I responded to -- which, to refresh your memory, was this: (I've cut out some fluff but those interested can just scroll up the thread for more context)
The analogy that I hoped would get you to understand my objection without us having to go over material I'm sure you already know goes something like this:
Religious argument:
1) The cause of the universe is either natural or supernatural.
2) It is improbable that the cause of the universe is natural. (Because of fine-tuning or some other such nonsense.)
3) Therefore, it is most probable that the cause of the universe is supernatural.
Argument about racism:
1) The cause of unequal outcomes is either biology or culture.
2) It is improbable that the cause is biology. (Maybe? I actually asked you directly what you were alleging was improbable and you didn't answer.)
3) Therefore, it is most probable that the cause of unequal outcomes is culture.
The same flaws are present in both these arguments. Namely, why should I believe the dichotomy in (1) and why should I believe the assignment of probability in (2)?
Don't portray me dishonestly here. In every case where I've said you were wrong, I've also said why. In the few cases I abbreviated the reason, it's because I thought you would be able to recognize the mistakes by reference to religious arguments you've been in where you've seen the mistakes made by others.
Which if thou dost not use for clearing away the clouds from thy mind
It will go and thou wilt go, never to return.
In either case, I don't have a continued interest in discussing this topic with you.
Remaking Magic - A Podcast for those that love MTG and Game Design
The Dungeon Master's Guide - A Podcast for those that love RPGs and Game Design
Sig-Heroes of the Plane
At the beginning of this discussion, I was looking for a reason why we should expect random-seeming demographic outcomes. I asked that because in the absence of such an expectation, we have no way of knowing whether or not a non-random outcome is indicative of a problem, because the type I error rate is totally unbounded. It could be like using smoke alarms to detect fires on a planet where the ambient atmosphere is mostly smoke. Your alarm is constantly ringing but there's no actionable phenomenon taking place.
You chose to respond negatively, so I assumed you disagreed, and that could well be the straw man you speak of. Ultimately, though, at the end of our exchange, I don't believe I've received any such reasoning from you, so the matter stands where it was. Either you never disagreed with me in the first place, or we've failed to communicate as you suggest. Thanks for trying, though.
Which if thou dost not use for clearing away the clouds from thy mind
It will go and thou wilt go, never to return.
I think you should reread Crashing's post. It seemed to me like a pretty direct refutation of the arguments you've been making.
I'm not as savvy with statistics as Crashing. This is how I'm simplifying his argument for my own understanding:
Example 1: I only have one person who texts me on a regular basis, named Joe. If my phone gets a text, it makes sense for me to assume it probably came from Joe. It's possible it came from someone else, but its reasonable for me to guess it came from the person I normally get texts from.
Example 2: I have 1000 people that text me on an equally regular basis. If my phone gets a text,it would not be rational of me to guess it came from Joe. He's one of a thousand people it could have come from. Sure, it's possible the text came from Joe, but there's no good reason for me to suspect it came from him specifically.
If someone sees gender inequality and says "we should investigate whether that inequality came from sexism," they're secretly making an affirmative claim. The affirmative claim is that we live in a world where sexism behaves more like Joe in example 1 than Joe in example 2. In other words, you're're making the claim that sexism is one of a fairly small number of possible causes of gender inequality. If sexism is one of thousands or millions of equiprobable causes of inequality (like in example 2), then it's not rational to nominate sexism as one of the prime suspects in the investigation.
I'll respond to your own understanding then.
Crashing himself said that he'd advise people to get lumps checked for cancer, while also asserting that he recognizes cancer is actually a less likely explanation. If Crashing can do it, so can a civil rights activist.
Instead of altering my fireman example, engage with it. You job is to spot dangerous fires. You start by making a list of signs that are present when there are dangerous fires. You keep an eye out for those things.
Then someone reports a situation to you that has one or more of these signs. In point of fact, a fire alarm has gone off at the local college. At this point, you do *not* say to yourself, "A fire alarm is one sign of a dangerous fire, therefore this must be a dangerous fire". You say, "It might be a false alarm. However, to be safe, let's check it out." Then the fire department rolls out and establishes whether there's a fire or not.
The fire fighters are not concluding that the fire alarm represents a dangerous fire ahead of time. However, paying attention to fire alarms and checking them out is an efficient way to keep an eye out for things that *might* be dangerous fires. You should then conduct an unbiased investigation and draw what seems to be the most reasonable conclusion from the evidence.
Likewise, if you hear that a teacher is grading all his black students lower than all his white students - that's the smoke. Now you check to see if there's fire. You don't presume the teacher's guilt, you should examine the work and see if it's justified. You should examine the scores of the black kids compared to the white kids leading up to this and see if this represents a sudden change within this one classroom or whether it's more likely a fair assessment of the students' abilities. You should consider many factors before calling someone a racist. You should not, obviously, assume guilt without investigation.
TLDR
Then by saying someone should investigate whether a lump might be cancerous, you secretly make an affirmative claim. The affirmative claim that we live in a world where cancer behaves more like Joe in example 1 than Joe in example 2 (i.e. is actually the MORE likely culprit, not even equally probable).
That doesn't hold up.
Remaking Magic - A Podcast for those that love MTG and Game Design
The Dungeon Master's Guide - A Podcast for those that love RPGs and Game Design
Sig-Heroes of the Plane
Is this not too narrow? Take Stairc's cancer/tumor example and plug it into (2) above. "With sufficiently high probability, where there's no cancer, there's no tumor." Quite often people will have a tumor that isn't cancerous, but you agreed earlier that a tumor is a warning sign of cancer. Either you are taking a very loose definition of 'sufficiently high probability', or (2) isn't always necessary for something to be a warning sign. And if (2) isn't necessary in the tumor scenario, why not? Is it 'because the consequences of a "tumor" outcome are so severe as to considerably overweight that option in the final calculation of expected value'? If someone were to convince you that the consequences of sexism in the workplace are severe, would you relent in making them prove that sexism in the workplace satisfies (2) above?
Taking a step back, the whole 'privileging the hypothesis' concept seems a bit misapplied here. The hypothesis that the skewed distribution of sexes across professions and positions is due in part to sexism would only be privileged if we had no reason to select that hypothesis over any other, right? Well, do we agree that some fields were predominated by women and some by men throughout large swaths of human history for reasons that were in large part sexist? If so, can we not consider this our prior distribution? Because this is how I see it. The Bayesian prior here, based on history, is 'a skewed distribution of sexes among professions is in large part due to sexism', which means that we need a whole lot of credible experience proving the opposite before the posterior doesn't include sexism as one of the major reasons for an unequal distribution of sexes. Certainly not the only reason, and certainly not a given, but a legitimate (and not privileged) option as one of the probable reasons for the skewed distribution of sexes among professions. To put it another way, if we agree about the historical sexism bit, then the situation is much more akin to bitterroot's Example 1 than Example 2.
Having gone through all that, your response here confuses me a bit:
If I understand you correctly, you're saying that there are any number of reasons a person enters a particular profession, and to say that prejudice could be one of them is meaningless because there is no reason to consider that a significant reason over all the others - ie Stairc is privileging the prejudice hypothesis. Do I have this right, and if not would you clarify what you mean? The reason I ask is I'm trying to understand the crux of what you are arguing.
And that's exactly the point I was making. Although you chose to phrase the 'norm' in such a way that it agreed with your point. If you phrased it as "Men tend to be more successful in engineering than women because they're better at the STEM fields", it wouldn't be rooted in facts, would it? I suspect you think I was trying to make a different point than I was.
The only point I was making was that, as above, the old cultural attitudes that associated capabilities with sex were simply incorrect factually, much like 19th century pseudo-science as to the capability of 'lesser races', and that the current state of things is due, in part, to those norms. Yes, I have personal beliefs that are more complex, but I was simply responding to a point.
Moral relativism isn't amorality, as you seem to imply. It's simply the recognition that morality is subjective. I can both acknowledge that my beliefs are simply a product of my times AND hold those beliefs as having value. You keep asking question that probe both my personal and intellectual beliefs, which don't have to be the same thing, hence 'waffling'. I can view something as bad today without believing it was necessary 'bad' then. It depends entirely on the perspective of the group you're viewing things through.
But, since you want clarity, the bottom line is I believe old gender norms to be bad, at least in terms of restricting women from anything but domesticity.
I think that's fine, but I think when people use these kinds of purely statistical examples they tend to miss context.
The smoke isn't a white teacher scoring black students lower. The smoke is a white teacher scoring black students lower in culture with a history of racial discrimination and inequity. The framing of the situation changes the logical solutions.
Now, I get the point people try to make with these, in that correlation doesn't equal causation. I'm fully behind that.
Now, if you hand the teacher equal tests with white-sounding names or black-sounding names, and the black-sounding names score lower, you know you've probably got a problem.
And we've done that for women in STEM. There IS a bias, on a lot of different levels, for a lot of reasons, against women in certain fields. And it's not just 'sexism' in the classical sense, because women are just as likely to do it as men.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
Look, we've got three or more different analogies on the table, all of which cash out differently upon evaluation. Each analogy is freighted with a bunch of additional factors, specific to the particular analogy, which cause us to evaluate them differently:
Analogy #1: 'lump/cancer'. Most people, myself included, would say getting the lump checked is probably a good idea.
Analogy #2: 'smoke/fire'. If we think about this one carefully, I think we find it to be somewhat different than lump/tumor. I myself have been involved in 200-some-odd events of fire detection systems going off in my building. Of those 200, only two represented actionable events, and even in those cases anyone with even a passing familiarity with the layout of the building would not have been in any real danger. It's gotten to the point where I don't even get out of my chair when the smoke alarm goes off unless building security comes by and makes me do so. Most other people in my building feel the same way.
This is quite a bit different than lump/tumor. I'd get every lump checked. I wouldn't panic at every smoke alarm.
Analogy #3: 'dead body/Stairc killed someone'. Most of us would agree that it's simply unreasonable to interrogate Stairc every time a dead body is found.
The question is, which of these buckets if any does 'unequal outcomes/X-ism' fall into? Is it like 'lump/cancer'? 'smoke/fire'? 'dead body/Stairc did it'?
You and Stairc are both asking me to treat it as if it were more like 'lump/cancer' than 'dead body/Stairc did it'. The problem is, I don't see any reason to believe that. That reasoning is the very thing I've been asking for since the beginning.
It's not enough merely to point out that the consequences of workplace sexism are bad. The consequences of letting Stairc keep killing people are bad too. What you need to get a handle on is the Type I error rate. How often does this so-called warning sign signal X-ism and how often is it detecting other things? We all agree that letting Stairc continue a killing spree is bad, if he's on one. The problem is that our test to figure out whether he's on one has such terrible flaws that we can't rationally act on it.
Firstly, in Bayesian reasoning, "prior" doesn't mean "historical." It means something closer akin to "hypothetical" -- it's what you're assuming to be true in advance of examining the evidence. That's why in statistical hypothesis testing, one purposefully chooses the prior to be hostile to the effect one is testing for. In other words, if we were looking for sexism, we would always choose a prior with no sexism.
Secondly, that something was historically highly weighted in a prior does not necessarily means it merits equal weighting when designing a prior today. Despite history, today I'm not going to put very much of my prior probability mass on, say, the statement "fire is made of phlogiston."
Thirdly, choosing a prior favorable to your preferred conclusion is exactly what privileging the hypothesis is!
And indeed, the reason I would be suspicious of someone choosing a prior straight out of the 19th century is because in the intervening years there has been an unending stream of evidence and experience of less systemic discrimination throughout Western societies. Starting from an historical prior and updating on recent history should cause everyone's posteriors to shift radically away from X-ism.
My position is not affirmative. Rather, I allege a gap in the argument of those who say that observing non-random-seeming demographic outcomes in a field is an adequate sign of systemic discrimination. The gap is that nobody has explained why the outcomes should appear random. As far as I can tell, nobody here has closed that gap.
The closest we've gotten is to make some analogies where in some cases you'd act even if that gap weren't closed. My argument there is that there are other cases where you wouldn't. It is then the burden of the person on the affirmative position to show that we are not in one of the cases where you wouldn't. Further, all of these analogies are different from the matter under study in significant ways, so one should be suspicious of reasoning from those analogies by default.
Which if thou dost not use for clearing away the clouds from thy mind
It will go and thou wilt go, never to return.
Look, overt discrimination has decreased because it's actively illegal, but as I just posted there is still an inherent bias against female candidates, even if it isn't as overt as 'oh she's a woman so she's not as good or she'll get pregnant in a year'. The effects of hundreds or thousands of years on our cultural aren't going to disappear overnight.
And, as I mentioned in a previous post, they shouldn't be completely 'random' (I assume by random you mean a true random distribution which would, in theory, match the population at large with a sufficiently sized sample). You're taking it to an extreme here that no rational person would argue. No field will ever have a perfect distribution, for a lot of reasons that go beyond one characteristic.
You're also implying that the only thing people have to go on is the statistic itself (hence the gap you mention), which isn't true. We know there are problems with women in STEM (and other fields) because it's constantly being reported on. That doesn't mean it's overt sexism (or racism, or whatever x-ism) or that it's the sole (or even a major) factor, or that anyone in particular is to blame for the skewed demographics - just that it exists.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
How did you get that out of what I said? If anything, I am advocating not ignoring history -- specifically, the past 100 or so years. Freezing our Bayesian priors as we would have set them in 1920 is what I would call ignoring history, and taking history into account by updating on the intervening 100 years should cause our posteriors to be significantly less weighted towards the various -isms.
I must point out that I never said anything about American Indians. This is a new topic you are introducing, and all the words you've said here you've literally just put into my mouth. You didn't even bother to ask my position before presuming whatever it is you're presuming here. I don't know what explains the poverty of American Indians, I take no affirmative position on the matter, and my response to anyone taking the affirmative position that the poverty of American Indians today is caused by historical policies would be "prove it."
Yes, to be clear, by "random" I mean "appears as if it had been randomly sampled from the population at large." And I'm not the one taking it to the extreme! If someone does a study of a subpopulation, it results in a distribution that doesn't look like a random sample, and that person concludes that this is an indicator of discrimination, it's that person who's "taking it to an extreme." It's they who are relying on the hypothesis I'm questioning here.
The question isn't whether the effect being measured exists, though. If you show me a study that says black students are on average doing worse than white students in school, I'll believe that. The question is whether the existence of that particular effect is strong enough evidence of something we'd recognize as discrimination.
And all the studies purporting to go further than just measuring the gross effect, including the ones you posted, are highly contentious. Particularly those about gender gaps. For every study alleging a gender gap caused by discrimination there is another study purporting to explain it in terms of something that isn't discrimination. I'm not saying either side has the right of it, necessarily, but the waters are muddy enough where the issue isn't going to be resolved by the studies.
Again, we're looking for the cause of these patterns. Detecting the patterns is one thing; tracing them back to whatever cause you're alleging is another. It's the tracing-back part where all of my objections lie. The fact that overt discrimination has been greatly reduced is relevant to the 'tracing back' part, because if we start with a high-discrimination prior and update on overt discrimination being reduced, we end up with a lower-discrimination posterior.
And that was the point I was making to jlpitt -- when we actually update the 1920s prior to the 2020s prior, we're looking at a situation where it's much harder to start from an hypothesis of discrimination.
Which if thou dost not use for clearing away the clouds from thy mind
It will go and thou wilt go, never to return.
I think this is where we are misunderstanding each other here. It doesn't have to be discrimination. It can just be a cultural bias, with no one intentionally discriminating against anyone.
Well, yeah. The pay gap isn't because of active discrimination, it's because women aren't as assertive in pay negotiations as men are. Because they aren't socialized to be assertive. It's also part of the reason the 'glass escalator' exists with men in women dominated fields. And there are a bunch of other differences, revolving around child care and maternity leave, etc. I would think very rarely would you find that there is active discrimination.
Although I'd be interested to read a study that places the gender gap's causal factors on something other than a cultural norm.
Just in case we're still crossing wires here after what I just said, define 'discrimination' for me, in your words.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
I'll come back and address your other arguments when I have time a little later, but I want to zero in on this statement for one second.
I don't think it's anywhere close to proven or accepted that this statement explains the pay gap. There are many other reasons proposed for the pay gap, even among academic feminist sociologists.
But let's pretend for a second that the pay gap is 100% attributable to women being less assertive in pay negotiations. Why would you assume that this is entirely cultural? The effect of testosterone levels on things like aggressiveness, assertiveness, and the like is extremely well-studied. And anecdotally, you can read accounts of trans men who started taking testosterone and experienced big changes in their personalities, including increases in aggressive and assertive behavior.
Now, why would I assume that it's cultural? Maybe it isn't. It could just as easily be explained by women's role as caregivers in the family, and on average spend an extra hour a day caring for their family than their male counterparts, and that cuts into their actual working time over the course of the year. But... wouldn't that be a cultural issue? Why are women spending more time taking care of their families than their husbands or SOs?
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
1. Are testosterone levels cultural? If you make the claim that the 6.6% number derives from culture, you need to show that it could not be explained by "non-cultural" sources.
2. I have no idea whether womens' role as primary caregiver in the family is cultural or not. I will say I'm aware of no human culture, ever, in which women have not been primary caregivers. I will also note that females are primary caregivers in most (all?) non-human primates. This at least suggests the difference might not be cultural.
Plus, some studies have suggested that testosterone isn't necessarily the 'aggressive' horomone we've previously believed it to be.
It's most, not all.
And, to be fair, it doesn't matter all that much when I've already posted evidence that women get offered less money when equally qualified, even before negotiations begin. Since both men and women are guilty of it, it implies something beyond biology.
TerribleBad at Magic since 1998.A Vorthos Guide to Magic Story | Twitter | Tumblr
[Primer] Krenko | Azor | Kess | Zacama | Kumena | Sram | The Ur-Dragon | Edgar Markov | Daretti | Marath
You, Jay13x, are taking the affirmative position that the gender pay gap is explained by "culture" (which you seem to define, roughly, as anything that is not an innate biological attribute of humans).
I, bitterroot, am not taking any affirmative position. I am completely agnostic as to whether the pay gap even exists, and if it does exist, whether it is caused by biology, culture, or a combination of the two.
To convince me that I should believe you rather than someone else, you have to prove your case. That means you have to demonstrate that biology is not the explanation for the pay gap.
I offered testotsterone as an example of something that might conceivably explain the gap. You need to prove that testosterone cannot explain the gap, not merely offer reasons why testosterone might not explain the gap. (I certainly acknowledge that it might not explain the gap, this is why I'm agnostic here. But you have not eliminated the possibility that it might explain the gap).
Moreover, there are other possible biological bases for these differences that are not testosterone. You need to explain why we can dismiss these as well.
Let's assume for a moment it is true that women are offered less "even before negotiations begin."
Womens' (typical) role as primary caregiver might conceivably be the reason that women are offered less - employers (correctly?) expect women to work fewer hours. You have not shown that the tendency of women to act in a caregiving role is determined by culture as opposed to biology. Therefore you have not demonstrated that the tendency of employers to offer women less money is determined by culture as opposed to biology.
Why does lump/cancer get the pass if not because of the potential severity of consequences? This was the reasoning you offered in comment #101. What I’m trying to find out is, at what point can something be considered a warning sign without satisfying (2)? Is there any sort of general guidelines we can use?
Right, but hypotheticals are influenced by the past. I think we agree here, since you cite the last 100 years as important for how much mass to assign to sexism in our prior.
OK, let me first of all say that all I know about Bayesian estimation and credibility I learned studying for exams a little while back, and I haven’t formally applied it since. However, I don’t remember this being the case.
As an example, let’s look at fire alarms in your building. If you were to construct a prior right now about the likelihood of an alarm being false, would it look roughly like a binomial distribution with P(false alarm) = .99? Then, after you find out whether the alarm was false or not, you take the outcome and assign some amount of credibility to it, and end up with a posterior. If it turns out there really was a fire, you update your p to something like .90, and then next time there’s a fire this new distribution becomes your prior. Is that not the gist of it?
The point being, I understand the prior to be the best model of actual behavior we can come up with, independent of any hypothesis testing that may occur later. I'm not 'looking for sexism', I'm looking for an accurate model and then seeing whether sexism carries meaningful weight in that model.
This isn’t what I’m doing.
Re-reading my post, it does indeed look like I was advocating taking a prior straight out of the 19th century, but that’s not what I meant to do. What I mean to say is that sexism was so prevalent and entrenched in the workplace up through the 19th century that the last 100 years of progress have not made this information irrelevant. We have made remarkable progress, but we have not cured sexism.
So when you say this:
I completely agree that any prior we construct today is much less weighted toward sexism than it would have been 100, or even 60 years ago. However, there is still mass associated with sexism that is weightier than most of the other trillions of possible explanations for the disparity of sexes we observe today. This is all I’m really trying to say.
If this is what you’re arguing, then I think we might agree. But you seem to be saying more than this. The way you responded to Stairc, you seem to be arguing against viewing skewed demographics as a possible sign of sexism, not just an adequate one. It’s this stance, and not your quoted stance, that I am arguing against. I’m really trying to be careful about putting words in your mouth here, but I don’t know how else to interpret this:
Please let me know if I really am misinterpreting you.