Into the thicket of variation: More on the political orientations of parents of sons and daughters, and a return to the tradeoff between internal and external validity in design and interpretation of research studies

We recently considered a pair of studies that came out awhile ago involving children and political orientation: Andrew Oswald and Nattavudh Powdthavee found that, in Great Britain, parents of girls were more likely to support left-wing parties, compared to parents of boys. And, in the other direction, Dalton Conley and Emily Rauscher found with survey data from the United States that parents of girls were more likely to support the Republican party, compared to parents of boys.

As I discussed the other day, the latest version of the Conley and Raucher study came with some incoherent evolutionary theorizing. There was also some discussion regarding the differences between the two studies.

Oswald sent me some relevant comments:

It will be hard in cross-sections like the GSS to solve the problem of endogeneity bias. Conservative families may want to have boys, and may use that as a stopping rule. If so, they will end up with disproportionately large number of girls.

Without longitudinal data, it will not be easy to get rid of this reverse-causality.

But Conley and Rauscher only looked at the sex of the first child, so this should not be a problem, right?

Not quite, as Oswald explains:

Yes, Powdthavee and I argued this too about the eldest child. So does Washington in her cross-section data study. The problem is that in principle one has to control for the covariate of family size, and that is endogenous, and there is no obvious instrumental variable that satisfies an exclusion restriction.

So, just to clarify, if you don’t control for family size, and you only look at sex of first child, then you should be ok. But one then should realize that the effect of a daughter includes many different things, including, possibly, the total number of children. Suppose, for example, a person has a first child and now you consider two branches:
(a) First child is a daughter. If conservative, the parent continues to have more children. If liberal, the parent might stop having kids. [I don’t know if the data show such a pattern, but the point is that such correlations could arise.]
(b) First child is a son. Reverse patterns as above.
In that case, there’s an interaction between party identification and the sex of the first child, and one of the outcomes is #kids. So, for example, if having lots of kids (of either sex) has an effect on attitudes, this could show up as an effect of the sex of the first child.
Nonetheless, this would still be an effect, even if not a direct effect.

Oswald is arguing that Conley and Rauscher’s study is a sort of reduced-form analysis, and it that it is being interpreted as a direct analysis.

I sent the above to Oswald, who added:

There is one way to make his result and ours consistent (though not Washington’s I am afraid). We actually, because we are doing a panel analysis on the birth of a child, are picking up the effect of a baby girl or baby boy. It might be that very young daughters turn me Left, and then much older daughters turn me Right.

However, this is one of the few cases I have seen where it could be that different countries really do give different social-science results. You implied that.

The Republicans are actually not greatly like the Tory party in my country. Obama is actually not very different from the Tory party. So that is a complication in cross-country comparisons.

This is related to the famous tradeoffs between internal and external validity. Comparing sex of first child has bulletproof internal validity but it affects the interpretation, in a way similar to “intent to treat” analyses.

It’s also interesting how confusing are the issues of interpretation, even in this relatively clean example. To pick up on one of our recent themes, it is necessary to retreat from the usual implicit assumption of large, persistent, and invariable effects (“having a boy makes you more liberal” or whatever) and into the thicket of variation.

P.S. I sent the above to Dalton Conley, who wrote:

I totally agree that what we are doing is a reduced form analysis, essentially.

First born girl may affect marital status (which we know it does from literature, hence our control) and other things we cannot control for.

Doesn’t mean that the total effect isn’t what it is, though . . .

Oswald has bigger problems in his sample though since it’s not limited to bio children. I think that should be a bigger concern (than reduced form versus direct effects): non random selection into sex of offspring. Panel data, of course, does nothing to solve that problem if it’s correlated with unobservables that are themselves correlated with partisanship….

The part of our paper that I [Conley] regret is our crazy biological interpretation. I don’t know what I was thinking or why reviewers didn’t spank me on that…

Good discussion, all consistent with the idea that in social science we have to attack problems from many directions. When it comes to causal identification and validity, there is no magic bullet.

64 thoughts on “Into the thicket of variation: More on the political orientations of parents of sons and daughters, and a return to the tradeoff between internal and external validity in design and interpretation of research studies

  1. Andrew:

    I’ve only skimmed this quickly but I think communication would be much more effective with a causal diagram. I quickly sketched one and posted it here: (PDF) https://dl.dropboxusercontent.com/u/15842577/DAG1.pdf

    I am not sure it captures what you have in mind but that is the point. Had you drawn one yourself I would have not doubt.

    The DAG shows how computing effects within strata defined by number of kids opens a back-door path between gender of first kid and party ID at time t=0, while blocking a mediator of the effect of gender of first kid.

    • Fernando, as an off topic question, what software did you use to draw this? The graphs look great! I realize DAGs aren’t too crazy to create, but I still haven’t found a program that makes drawing these really easy. Any suggestions?

      • Mark:

        Thanks! For quick sketches like this one I use libre office draw. Resize page and export as PDF.

        In latex I use graph library for TikZ. In R igraph and others.

        I don’t think any of them is perfect.

        • PS What I would like is to generate the code by drawing the graph and not the other way around. Then revise using code. Surprisingly we seem to be far from that.

        • Hi Fernando,
          You might want to use TikzEdit, which allows you to draw a graph, which is then translated into Tikz / LaTex code on the fly. You can then paste that code into LaTex and edit accordingly. I used it in the past to draw DAGs, but nowadays I just end up re-using my old LaTex code.

        • Rahul:

          Looks interesting! R igraph also has autolayout options. Great for large network graphs.

          For a while I’d thought of doing a Twitter feed with a “DAG of the week”, the idea being to post a dag illustrating some identification controversy. In Twitter bc that is how much economy can be achieved with graphics.

          Maybe one day.

    • I’m thoroughly confused now reading the post & your comment and other comments. Have we just shown various plausible ways how everything might indirectly influence everything else? Is there a way out of this morass of confounders?

      Is it ever possible to know by more studies whether the parents of girls support any one party or not at all? How? All this is making me feel very nihilistic. Help! :)

      • Rahul:

        This is not my topic but I think what it shows is how contingent on assumptions all observational causal inference is.

        For example even the gender of first born need not be exogenous once you allow for abortion.

        PS the book by Morgan & Winship has a very gentle intro to DAGs.

  2. Can somebody explain me this

    Conservative families may want to have boys, and may use that as a stopping rule. If so, they will end up with disproportionately large number of girls.

    . I always thought that you can’t change the sex ratio by manipulating the number of children. Basically, whatever cultural pattern you have in mind if you decide to have another child it still be a boy or a girl with the correct probability ratio. The same holds for the first child only. No matter what you have in mind after having the first child, it cannot change the sex ratio for the first born. so what am I missing?

    • I think Andrew may have gotten the math wrong here.

      Assume fertility is unlimited and you stop at the first boy child. This gives probabilities that are roughly

      p(B) = 1/2
      p(GB) = 1/4
      p(GGB) = 1/8
      p(GGGB) = 1/16
      p(GGGGB) = 1/32
      ...
      

      The expected number of boys is (1 * 1/2 + 1 * 1/4 + 1 * 1/8 + …) = 1

      The expected number of girls is (1 * 1/4 + 2 * 1/8 + 3 * 1/16 + …) = 1

      You’re going to have to be willing to have a lot of kids either way!

      (Now if you assume the probability of a boy is closer to the real figure of roughly 0.51, then the expected number of boys is still 1, but the expected number of girls is closer to 92%.)

      • Bob:

        Right, under that scenario, expected #kids is 2. And, under the alternative scenario (stopping after kid #1), the expected #kids is 1. So “party id” and “sex of 1st kid” interact in predicting #kids.

        The confusion may be in my use of statistics jargon. I wrote, “there’s an interaction between party identification and the sex of the first child.” That sounds like I’m saying there’s a correlation between party identification and the sex of the first child. But I wasn’t trying to say that. An interaction between x1 and x2 is not a correlation between x1 and x2; rather, in statistics jargon, it’s a nonlinearity of x1 and x2 when predicting some third variable, y.

    • “In probability theory, the optional stopping theorem (or Doob’s optional sampling theorem) says that, under certain conditions, the expected value of a martingale at a stopping time is equal to the expected value of its initial value. Since martingales can be used to model the wealth of a gambler participating in a fair game, the optional stopping theorem says that on the average nothing can be gained by stopping to play the game based on the information obtainable so far (i.e., by not looking into the future).”

      http://en.wikipedia.org/wiki/Optional_stopping_theorem

  3. I have a really hard time believing that there would ever be a group-level preference for girls. Liberal or not. If the assumption is that liberals act as the opposite side to conservatives, the opposite to preference for boys is not preference for girls; it’s the elimination of (culturally based) gender bias. Hence, good liberals would stop regardless whether the first child was a boy or girl. Unless their preference was two or more children, which should be randomly assigned. In that case, only conservatives would have a correlation between large families and therefore many daughters(?).
    I’m not saying this is true, but this explanation should be as equally plausible as liberals having a preference for girls, especially since I have never heard of large-scale preference for girls, ever.

    (Ok, I did hear about the matriarchal society Mosuo i China. But even they lacked preference for girls.)

  4. I had a question related to D.O.’s — I’m also having trouble thinking through Oswald’s first claim connecting the stopping rule to being the parent of daughters.

    It seems to me that if we look at the extreme case where all conservatives adopt the rule ‘continue having children until you have a boy’, there’ll be a minority (25%) of conservative families that will have at least 2 girls before they have a boy. However, there’ll be a large number (50%) that have a single male child.

    Thinking more about what this means for the theory being tested, it seems to me that if ‘being the parent of daughters’ is measured by having at least one daughter, then the stopping rule Oswald suggested actually has feedback in the opposite direction: if we assume this stopping rule for conservatives, only 50% will have at least one daughter, while any decision rule for liberals that involves a positive probability of having multiple children, and no gender-specific stopping rule, will result in a probability of more than 50% of having at least one daughter. This means (I think), that the conditional probability of being conservative ought to actually decrease if you condition on having at least one daughter.

    Maybe, though, Oswald instead has in mind that ‘being the parent of daughters’ ought to be measured as having at least two daughters. Even here, though, If we assume liberals choose uniformly over having 1, 2, or 3 children, 25% of conservatives (under the stopping rule) and 25% of liberals will have at least two daughters.

    I think Oswald is right that under this stopping rule Conservatives as a group could have more daughters, but it’s harder for me to see how this impacts the theory if ‘being a parent of daughters’ is measured by the presence of a daughter.

  5. On further thought, the constancy of sex ratio can be put to a good use. Just consider definition of “a parent of boys” and “a parent of girls” as proportional to number of boys and girls. For example, the father of 3 boys and 2 girls comes down as contributing 3 units to the “parent of boys” rubric and 2 units to the “parent of girls” rubric. This will make the weighted sample skewed toward people with a lot of kids (presumably, mostly conservative), but if our goal is to look at difference, it should not matter much. Of course, it won’t work if there is strong nonlinearity in political attitudes based on family structure, but at least it gets rid of all the funny problems with the sample selection.

  6. There’s another potential confounder: one study found that, as best as I can recall, husbands with one or more sons were 9% less likely to get divorced than husbands with only daughters. (I can’t recall if they adjusted for family size, etc.) I’d like to see that replicated, but it’s not prima facie implausible.

  7. By the way, the celebrated Gender Gap in the 2012 election was small compared to the obscure Marriage Gap. In the Reuters-Ipsos post election panel survey of over 40,000 voters, the Gender Gap was only 3.8 points, but the Marriage Gap was 21.4.

    http://www.vdare.com/articles/the-gop-s-other-problem-marriage-gap-huge-in-2012-but-marriage-declining

    The GOP is essentially the party of married white people, and the Democrats are the party of everybody else.

  8. I went on to compare the Marriage Gap to various other demographic factors in the 2012 election, and Marriage dominated most of them, except for race. Having children isn’t all that important relative to marriage. Unfortunately, the Reuters-Ipsos panel didn’t collect whether the children were male or female, but I doubt if the electoral effect was much compared to being married or not.

  9. Based on Figure 1 of the Conley paper it seems there are big things going on in 2 kid families. If the two kids are males, the party affiliation is more than 2:1 Democrat:Republican. If the two kids are females, it’s basically 2:1 the other way. That’s a whopping effect size for that subtype of family, and at 20% of the sample (half of 400) it contributes more than its fair share to the final effect size.

    I’m not up on how often these results have been replicated. The 2013 paper is not especially highly powered for what it is trying to show. Simulating some differences in proportions suggests that these differences are pretty close to the .05 level, and in the published data they are driven by a heavy subcomponent of the data. If this is a stable effect that can be reproduced in multiple surveys and with sharp resolution, then its more interesting for sure. But if not, then having a couple of studies find “effects” in opposite directions might have a simpler explanation.

  10. By the way, here’s a hilarious new NYT column by Charles Blow about how having female loved ones turns men into evil Republicans:

    “The problem with having your message powered by machismo is that it reveals what undergirds such a stance: misogyny and chauvinism. The masculinity for which they yearn draws its meaning and its value from juxtaposition with a lesser, vulnerable, narrowly drawn femininity.

    “We have seen recent research suggesting that men with daughters are more likely to be Republican and a study finding that men with sisters are more likely to be Republican.

    “The study of men with sisters was conducted by researchers at Stanford Graduate School of Business and Loyola Marymount University. A report from Stanford about the study concluded, “Watching their sisters do the chores ‘teaches’ boys that housework is simply women’s work, and that leads to a traditional view of gender roles — a position linked to a predilection for Republican politics.””

    http://www.nytimes.com/2014/02/01/opinion/the-masculine-mistake.html?hp&rref=opinion

    • The claim that catches my eye is the assertion that it’s watching sisters doing chores that’s doing the work to turn boys into machismo cads. How does that possibly get justified?

      Here’s what I read in one account of the study:

      “The NLSY survey, conducted by the federal Bureau of Labor Statistics, included interviews with children as young as 10. They were asked if they regularly helped with straightening out their room, keeping the rest of the house clean, doing the dishes, and cooking. Girls were more likely than boys to perform each of these tasks than boys, but the differences were substantially larger for one variable than for the others: doing the dishes. Considering everyone in the sample who had a younger sibling, 60 percent of boys responded that they helped with the dishes, compared with 82 percent of girls. Healy and Malhotra estimate that boys with sisters were 6 percent to 7 percent less likely to do the dishes than boys with brothers.

      These effects represent a potential mechanism underlying the main results. Men with sisters don’t seem to grow out of the idea that housework isn’t their problem. In 1997, adult men with all sisters were 17 percentage points more likely to say that their spouse did more housework compared with men with all brothers, according to the PSP data.”

      http://www.businesswire.com/news/home/20130723005454/en/Young-Men-Sisters-Republican#.Uu0Ql7SdASk

      So, if I understand this right, it’s all about the dishes? And only a 6 percent differential? I mean, seriously?

      The infinite power of Patriarchy moves in mysterious ways.

      • “Household chores” always seem to be defined as washing dishes, doing laundry, making beds, dusting, etc. They never seem to be defined to include mowing the lawn, washing the car, running to the store for milk and bread, taking the dog out for a walk, cleaning out the garage, painting the picket fence, etc. Thus, wives and daughters usually are seen as doing the bulk of the “household chores,” simply because the many chores that husbands and sons tend to do are discounted altogether.

        • Sorry, but this is just not true. The American Time Use Survey, which is the source of much of the data on housework in the US, identifies many different kinds of housework, including those that are stereotypically masculine (mowing lawns, paying bills, etc.).

          Mansplain much?

        • “The NLSY survey, conducted by the federal Bureau of Labor Statistics, included interviews with children as young as 10. They were asked if they regularly helped with straightening out their room, keeping the rest of the house clean, doing the dishes, and cooking.”

          Which of those cited “household chores” is “stereotypically masculine?”

  11. One mechanism that should be kept in mind is that the more children you have, the less plausible sound liberal theories of 100% Nurture over Nature.

    For example, I’m an only child and my parents were convinced that everything I did, good or bad, was the result of their nurturing. I sort of felt that way too until we had our second child and it quickly became obvious that they were very different and that our nurture had distinct limits imposed by our sons’ genetic diversity.

    Having children of both sexes makes it particularly obvious that more dogmatic sort of feminist orthodoxy about social conditioning etc. is silly.

    • Strawman. “100% Nurture over Nature” is almost idiotically silly. What makes you think that’s the mainstream liberal position? Ignorance or malice?

      Is the analogous Steve Sailer theory “100% Nature over Nurture”? Like if you’d been raised by a pack of wolves you’d still grow up to be this Steve Sailer we all love?

      • Now now Rahul, there are plenty of people who adhere to radical “100% Nurture over Nature” worldview. They usually strongly believe any other viewpoint is not only wrong but immoral, and they’re not shy about lording their righteousness over everyone else. Go to just about any university in the US and you can’t swing a giraffe without hitting someone like that.

        The Sailorites seems to believe in a ~50% Nature, ~50% nurture mix for no other reason than because a bunch of research seems to show that.

        Steve’s point sounds plausible enough to me. I often hear parents and new grade school teachers express surprise at how strongly and stubbornly little boys act boyish and little girls act girlish.

        • I think for every trait the Nature to Nurture contribution varies. Something like height or skin or eye color is obviously heavily Nature. OTOH, say, “becoming-a-good-statistician” undoubtedly has a huge Nurture component.

          Look, there’s all kinds of idiots (even or perhaps especially at Universities) but if indeed one must truthfully choose a median liberal position I sincerely doubt it can be labelled “100% Nurture over Nature”. Just as I won’t label the median conservative position as “100% Nature over nurture.”

          PS. Are you saying that these “parents and new grade school teachers” were ascertained to be also specifically liberal by you? Otherwise, all your observation shows that a large fraction of our population is stupid and / or naive. I’m hardly surprised.

        • if indeed one must truthfully choose a median liberal position I sincerely doubt it can be labelled “100% Nurture over Nature”.

          No, it certainly is that. The standard, median liberal position is an incoherent mishmash of normative and descriptive considerations, as I already noted. That standard position is that it is unethical to ever consider nature, at all, unless nurture has been definitively discounted, which amounts, in practice, to a 100 percent nurture over nature position.

        • I think it’s pretty fair to say that the classic liberal position on the Nature vs. Nurture question, when it comes to any kind of socially important trait, is to say that the role of Nature is essentially trivial and can be ignored.

          Basically, they understand that it would be completely dogmatic and transparently anti-scientific to declare that Nature has no role whatever, so they assert instead that, sure, Nature’s there alright, but it’s unutterably insignificant.

          Of course, on any rational ground, the claim that each and every socially important trait is only trivially affected by Nature is effectively as dogmatic as saying that it just doesn’t play any role at all. But that doesn’t change their tune.

        • No. No, no, no, no, no. You are underestimating the dogmatic moralism of the liberal position, which is that any nature should be ignored unless definitively demonstrated at 100 percent. If variation in some trait is even 1 percent nurture, whatever that means, there is a moral imperative to ignore the other 99 percent that is nature.

        • Recall the broad popularity of Malcolm Gladwell’s 10,000 Hour Rule for being a star at anything. (E.g., Macklemore rapped about it on his grammy winning record.)

          But, perhaps the winds of change are blowing. Sports Illustrated reporter David Epstein just published a book called “The Sports Gene” about the importance of what Epstein calls “human biodiversity” in sports achievement, with many attacks on Gladwellism. President Obama was seen buying Epstein’s book on his Christmas shopping trip.

        • Don’t like Gladwell myself, but if at all, it makes sense to think of it as “It needs ~10,000 hours to be a star at anything, even provided you have the aptitude, motivation & other prerequisites” not the sillier “Give any fool 10,000 hrs & he can become a star at anything he chooses”

        • Rahul:

          Amazingly enough, there are people who week to believe the strong version of this statement. For example, here’s Steven Levitt:

          Last spring, I jokingly (okay, maybe half-jokingly) wrote about my quest to make the Champions Tour, the professional golf tour for people over the age of 50. In that post, I made reference to the ideas of Anders Ericsson, an old friend whom Dubner and I wrote about in our New York Times column back in 2006, and whose ideas later became the centerpiece of a number of popular books. Anders is the one who thinks that talent is unimportant. Oversimplifying a bit, he argues that with 10,000 hours of the right kind of deliberate practice, more or less anyone can become more or less world-class at anything. I’ve spent 5,000 hours practicing golf, so if I could just find the time for 5,000 more, I should be able to compete with the pros. Or at least that is what the theory says. My scorecards seem to be telling a different story!

          I’m a bit confused by this. On one hand he’s saying that Ericsson’s ideas are good enough to have been endorsed in Levitt’s NYT column; on the other hand, the last bit about the scorecards indicates that Levitt doesn’t really believe it. So I don’t know.

        • Well, reminds me of the John B. Watson quote:

          Give me a dozen healthy infants, well-formed, and my own specified world to bring them up in and I’ll guarantee to take any one at random and train him to become any type of specialist I might select – doctor, lawyer, artist, merchant-chief and, yes, even beggar-man and thief, regardless of his talents, penchants, tendencies, abilities, vocations, and race of his ancestors. I am going beyond my facts and I admit it, but so have the advocates of the contrary and they have been doing it for many thousands of years.

          Ericsson doesn’t seem a hack or a quack. Perhaps, there’s a nuanced position, a middle ground: You may not be able to train any random bloke to be a Wimbledon star but perhaps the point that Ericsson, Watson et al are making is that we commonly overestimate the genetics bit. With 10,000 hours of training a large swathe of the population might already posses the genetics necessary to attain fairly high levels (though not the top). Much higher levels than what we expect or are commonly told that we are capable of.

        • A man named Dan McLaughlin read about the 10,000 Hour Rule in Gladwell’s “Outliers,” so he decided to take up golf at age 30, engage in 10,000 hours of intense Direct Practice, and make the PGA Tour:

          http://thedanplan.com/a-summary-of-the-dan-plan/

          After 3.75 years of near full time practicing, he’s down to an excellent 4.1 handicap. Unfortunately, PGA tour players are around -6, so he has about ten strokes to go. And he’s only improved by about two strokes over the last year and a half, so it looks like diminishing marginal returns are going to prevent him from succeeding in this interesting experiment.

        • Just to reinforce my point in my previous post.

          Try to find a liberal and a socially important trait for which the liberal asserts that that trait is, quite likely, or even quite possibly, substantially based on genes.

          It’s not going to be an easy exercise. And it’s going to be a close to impossible exercise if one takes it a step further and requires that the distribution of the trait varies across population groups, and does so substantially on the basis of genes.

          This demonstrates pretty clearly the tiny orbit in which liberal thinking about evolution and biology must move.

        • Even were you to fine one the liberal response would just be something like the following: well, there’s nothing we can do about that so we have a moral obligation to focus on the inputs to the trait that are nurture-based. Until a liberal is forced to admit that a trait is 100 percent nature they will simply retreat to the moralistic position that nature should be ignored.

        • Look, unless one locate a liberal who will grant a substantial role for genetics in socially important traits (and again especially when it comes to distributions across population) it’s moot how they might react to that possibility.

          The more crucial point here is that simply can’t allow the possibility to be seriously entertained, so damaging is it to their movement in its current form (before WWII, though, progressives certainly did accept the possibility, so it’s not really inherent to more basic liberal values). Identity politics reigns supreme in the liberal movement of today, and that just can’t countenance differences between groups based substantially on population groups — or gender, for that matter.

        • Not sure if you’re familiar with JayMan at JayMan’s Blog but he is an outright (mixed race) liberal who is every bit the adherent to HBD as Sailer. He’s a heavy-hitter and does original analysis.

        • One further point.

          It is pretty amusing to see how the debate over Nature vs Nurture always proceeds when liberals must reckon with it.

          Rahul’s response is classic: How dare you say that liberals all believe that Nurture is 100%! I defy you to find a liberal who makes that reckless claim! That, sir, is a Straw Man!

          But, again, what does every single liberal assert when it comes to socially important traits, and especially across populations or genders? That the role of genes is, at most, utterly trivial and can be ignored.

          You know, as if that position represents a stage of scientific and rational enlightenment far removed from the “Straw Man” of claiming a 100% role for environment.

        • Which is why it’s clear that rejecting nature is strictly, for liberals, a moral consideration; facts have absolutely no relevance because of that.

        • Eye color is “heavily nature,” Rahul? Pray tell, what nurturing trait or trick by the Taylors gave their daughter Elizabeth her so-called violet eyes? As for “becoming-a-good-statistician [sic],” I seriously do doubt that nurture plays a “huge” role in such an outcome, since it seems overwhelmingly likely to me that one either is born with the necessary type of brain for it or not. Nurturing might destroy that native capacity, but I do not believe that it can create it, anymore than I believe that “proper” or “better” or, even, “optimal” nurturing would have allowed me to sing like Pavarotti.

      • Meaning ain’t in the head – WVO Quine

        What liberals “think”, Rahul, is simply what they do and what policies they advocate. The standard liberal position is that it is a moral imperative to craft every public policy as if all behavioral variation were 100 percent nurture over nature unless definitively demonstrated otherwise. In practice, this ends up being identical to advocating policies that are predicated in variations in behavior being 100 percent “nurture over nature”.

        What goes on in your head is entirely irrelevant. The reality is that you, Rahul, advocate policies that are identical to those that would be advocated by someone who believes in 100 percent nurture over nature. Ergo, “100 percent nurture over nature” is what you “think”.

        Again, Meaning ain’t in the head – WVO Quine

      • Is the analogous Steve Sailer theory “100% Nature over Nurture”?

        Nature creates nurture. If you think otherwise, then, I invite you to offer a competing explanation for where nurture originates. God, maybe?

  12. Pingback: Weekend reads: One researcher resents “cyberbullying” while another wishes peer reviewers would spank him | Retraction Watch

  13. @ Rahul regarding the Watson quote

    Why stop at a doctor or lawyer? Why not teach them to be Albert Einstein or John Watson? Dead serious. See, what explains Watson’s comment is arrogance, extreme arrogance. Someone like Watson views the average doctor or lawyer in the same way that the average doctor or lawyer views someone flipping hamburgers at McDonalds. Watson is not offering a sober analysis but preening “look at me, I’m John Watson”. Further, giving someone with the mind of John Watson a controlled environment to concentrate on one person is irreplicable on a mass scale, therefore, it’s irrelevant to social policy.

    Look, dude, the blank slate is completely busted. Time to get a new ideology.

    • Asher:

      John Watson has been dead for over 50 years. Give the guy a break. Also, this (from Wikipedia): “That Watson did not hold a radical environmentalist position may be seen in his earlier writing in which his ‘starting point’ for a science of behavior was ‘…the observable fact that organisms, man and animal alike, do adjust themselves to their environment by means of hereditary and habit equipments.'” If you think he “is not offering a sober analysis but preening,” that’s your problem, not ours.

      • I just took the quote at face value and responded to it. I have encountered such blatant dishonesty and dissimulation from blank-slaters that every time I encounter one it’s like a shark with blood in the water.

        • So your experiences have conditioned you into your natural response? Are you sure you’re playing no part in muddying the waters?

      • Andrew,

        I don’t see how much your quote from Watson proves much about whether he was a “radical environmentalist”.

        All Watson’s invocation of “hereditary” equipments would seem to entail is that, say, a human being is different from an animal in his or her capacities to respond to learning. So, yes, Watson would certainly accept the unremarkable fact that a gerbil can’t be trained to be Gauss. I wouldn’t describe the contrary position so much as “radical environmentalism” so much as I’d describe it as psychotic.

        So, Yay, Watson, you’re not psychotic!

        But the real issue has always been what kind of flexibility human beings have to become any desired kind of human being in the right environment. And I don’t know how else one can read the original quote from Watson without concluding he was as radical an environmentalist in answer to that question as possible.

        • Highly:

          This is just getting ridiculous. If you’re interested in this topic, you might want to read a bit about psychology research in the early 1900s. The Watson wikipedia entry would not be a bad place to start, then you can follow the links from there.

Comments are closed.