Wednesday, November 27, 2013

Feynman's cognitive style

Some interesting finds in this 1966 AIP oral history interview with Feynman.

I have always felt that Feynman was cognitively a bit "lopsided" -- much stronger mathematically than verbally. This might be partially responsible for his way of learning -- it was often easier for him to invent his own solution than to read through someone else's lengthy paper. (Personality factors such as his independent streak, and his strong creativity, also play a role.) But this sometimes left him with gaping holes in knowledge. In contrast, Schwinger had at age 17 an encyclopedic understanding of what was known about quantum electrodynamics -- he had read and mastered all of the literature as a high school kid!

This excerpt reveals that Feynman did not understand the conventional formulation of QED even after Dyson's paper proving the equivalence of the Feynman and Schwinger methods. (When someone explained the action of a creation operator on the vacuum, Feynman reportedly objected "How can you create an electon? It disagrees with conservation of charge!" :-)
... I was struggling gradually to learn. I mean, I had to learn something to prove the connection between my thing and the same thing. Dyson had done a great deal in that direction. That didn’t satisfy me because I couldn’t follow that. Dyson told me, when he wrote his paper, “Don’t bother to read it, there’s nothing in it that you don’t know, except that it proves it’s the same as what everybody else knows, but it doesn’t say anything different or do anything different than is in your paper. Nothing more in it,” he told me.

... Yeah, because I remember him telling me not to worry about the paper. It hadn’t anything in it, you see. ... But then I thought I had to understand the connection, for publication purposes and others. And I had a good opportunity, because Case sent me his theorem — the manuscript of a big paper that he was going to publish in the Physical Review, which had all the steps of the theorem. Now, I argued in the meantime with myself, in my usual physical way of arguing, and concluded for several physical reasons, by some examples and other things — simpler examples that weren’t so elaborate as the calculations I made — that it couldn’t be true that the two methods would give the same result. ... I prepared a letter in which I wrote the physical arguments. Then I decided, that isn’t going to convince him. Nobody pays any attention to physical arguments, no matter how good they are. I’ve got to find a mistake in the proof. But the proof has creation and annihilation operators and all kinds of stuff. So I went to some students, in particular Mr. Scalator who was only fair, but he understood. He had learned in a pedestrian way what it all meant, and he explained to me what the symbols meant. So I learned like a little child what all this was about, so I understood what the symbols that he was using in the paper meant, and I tried to follow the proof, and I learned enough to be able to do that kind of mathematics, see — for the first time. So I followed the whole thing through, and I found a mistake, a very simple algebraic error, in the proof. He commuted some things that didn’t commute and so on.
Feynman never carefully read either Schwinger or Tomonaga's work:
Weiner: How about Tomonaga’s work? When did you first hear of it?
Feynman: I don’t know when I first heard of it. The work itself, I never knew exactly what it was, and I don’t yet know precisely what it was.
Weiner: You read his paper?
Feynman: No.
Weiner: I mean, there’s one paper that is often cited —
Feynman: No. No. I don’t think I read the paper. But this must be understood — I don’t mean anything disparaging. If Schwinger hadn’t been in the front yard at Pocono, or next to me, I wouldn’t have known what he did either. I got the same as everybody else. If you can do it yourself, why learn how somebody else does it? So I don’t know precisely what the relation of Tomonaga’s and Schwinger’s work is or the relation of his and mine. I think the relation of Tomonaga’s work to my work is very small. I mean, I think he’s gone around much closer the direction that Schwinger went.
Weiner: I think it’s the general impression.
Feynman: But I don’t know the precise relationship of their work. But I believe, if I’m not mistaken, although you’ll have to ask Schwinger, that everything that Schwinger did he did without knowledge of what Tomonaga did. I hear, but I don’t know, that Tomonaga did a very great deal, and did essentially what Schwinger did, except perhaps for working on certain practical problems. I don’t know. That’s what I hear. But I don’t know. I’m sorry, that sounds stupid, but I have never looked into it, and I never read Schwinger’s paper in a comprehensible way. I don’t know what’s in that paper of Schwinger’s.
Weiner: Haven’t tried to read it?
Feynman: Never. Tried in the sense that I looked at it and I flipped the pages, because it’s too hard. I read it at a time when I didn’t even know what a creation-annihilation operator was. I read it — you probably can prove that by the fact that I refer to it in various places, and get certain formulas out of it — I read it in the same way that I talk to him. When something looks like something, I know that’s it, you know? But I didn’t follow all the steps. I never followed all the steps.
Weiner: But you did know, when you talked to him at Pocono, and then —
Feynman: I know Schwinger — that’s what I say, I must have read it in pieces and bits. I know what Schwinger did; I know more or less how he did it. ...

Feynman: Yes, because we talked together, we had the physical idea of what starts it, but there’s a difference from that and checking all the equations, ... I don’t know whether he really read mine in detail or not. But he knows what’s in it, and I know what’s in his, but I can’t tell you. Perhaps if I look at his paper carefully, I can see that I really did read it, you know? I mean, I’d have to have it and look at it and see if I did read it. That’s a good way to look. I doubt that I read it in detail. I doubt that I looked at all of the various complicated sub-things that he had to worry about, like what to do with the longitudinal waves — because I don’t think there’s any problem with the longitudinal waves. I couldn’t pay attention to such a thing, see? So I doubt that I’ve ever read the paper in any careful way like a student would try to learn it. I don’t believe I’ve ever done that.
Finally, an interesting conversation between Feynman and Oppenheimer concerning the covariant propagator and positrons as electrons moving backwards in time:
So I went to the Physics Society and gave this paper, and I wanted Professor Oppenheimer to hear it, and other people like that. I particularly wanted Oppenheimer to hear it because he often said that there wasn’t anything to it. He understood Schwinger’s and he didn’t understand mine. And I thought he would be at the meeting. I’d kind of half thought about him when I prepared it. When I went to the meeting, he wasn’t there, but I gave the paper, and then Weisskopf got up and said, “This paper is so important and unusual” and so on “that we ought to give the man more time to express his ideas.” ... Then I stepped down, and just at that moment, Oppenheimer came in and sat down in the chair just ahead of me. And he turned around and said, “What did you talk about?” I said, “The idea of electrons going backwards,” meaning positrons. He said, “Oh, I heard all that. Oh, yes,” he said, “I heard that stuff, right? That stuff I heard.” I said, “Yeah, you’ve heard it, but you’ve never understood it.” Now, the response to that was an invitation I found in the mail when I got back to Cornell, to come to Princeton to the Institute and explain all my ideas, in as many lectures as I wished, two a week, as long a time as I wanted, expenses to be paid by the Institute, and so on. He’s a very great man, I know. I mean, I understand him. We’re good friends. You know. I mean, it’s not enemies. I said that because I was trying to get something across to him, that he didn’t understand it. That was honest. He knew that if I were driven to say that that was true — you know what I mean — and it was worth learning. So I said that, and his response was very generous — any length of time I want, any conditions. So I went to the Institute of Advanced Study.
In his eulogy, Schwinger described Feynman as "... the outstanding intuitionist of our age ..." :-)

Note added: I recalled another anecdote related to this post. At his Pocono talk Feynman was repeatedly asked by Dirac "Is it unitary?" (referring to Feynman's diagram method deduced from the path integral). Unfortunately, Feynman did not seem sure what "unitary" meant and responded "perhaps it will become clear as we proceed..." (a trick he learned from an earlier Schwinger talk). Feynman also did not seem to know what an S-matrix was!

But is it unitary?  :-)

See follow up post: Feynman and the secret of magic.


steve hsu said...

Note, this discussion is necessarily subjective and I don't ascribe a high degree of confidence to my comments. But it's fun to discuss. The Feynman v. Schwinger (and to some extent Dyson) observation is something I have thought of for many years. I have had the chance to interact with F and D a bit and have heard talks from S. Certainly S and D would score higher on a V test than F, although F is still very expressive and a good communicator (one of the best!).

I know people (I guess I am one) who can quickly scan papers and get the gist of things. (Actually the average scientist would be shocked at just how fast this can be done! A big part of it is simply preprocessing -- having a better mental map of the key issues across many areas. But part of it is a V-dependent ability to parse and decode what is written on the page.) If the paper has math in it then this is necessarily a combination of verbal *and* mathematical fluency. I also know people who are very good problem solvers yet are not fast readers and cannot do what I just described. They are much more comfortable solving a problem themselves as opposed to understanding someone else's solution, especially when that solution is encoded in the formal, academic language of a scientific paper. The first group tends to have a global understanding of the literature and the second group much less so. Reading is still the fastest bandwidth path into the brain! However, there are some otherwise smart people who don't learn primarily by reading. I don't know if the V factor is the best one to characterize this, but it seems reasonable to me. The second group also tends not to have as large a vocabulary as the first, etc.

Incidentally, a lot of programmers and engineers and experimental physicists are in the second group and their V scores are certainly lower on average than among theorists. (See Roe study and SMPY...)

If you read the whole AIP interview there is another revealing part where F discusses the kind of grilling typical at IAS. Note the questions often ask the speaker to relate his work to some larger context or very esoteric earlier work. Those kinds of questions, which F felt very uncomfortable with, because of the holes in his knowledge, were typical high V theorist questions -- i.e., from erudite know-it-alls who have been absorbing and structuring/compressing knowledge at a high rate throughout their lives. F felt more comfortable with "practical" self-contained questions, e.g., about the method of calculation or some numerical factor.

Richard Seiter said...

Thanks, Steve! That was illuminating. It might help put my questions/comments in perspective if I note I was friends with three of your contemporaries at Berkeley in physics (I think). Despite having fairly similar test scores to the one I knew best I was in awe of him (as was one of the other two people ;-) (he is one of the people I was thinking of in my comments on the Bezos article). He was clearly superior to me in math/physics (as a freshman he intimidated a fair number of the upperclass physics majors IIRC) and I have always been curious how that failed to show clearly in the standardized tests. Much of it probably stemmed from his great love of the field (working hard problems for fun almost obsessively), but I am convinced there is some aspect of innate intellect in play that is not captured by the tests (and is a major differentiator of top tier abstract thinkers). If you have any interest in who/how I knew them please send me an email (we corresponded about Anne Roe's papers).

I'm a programmer/engineer and I think my relative V/M balance (still unbalanced towards M just not so much as most in those fields) makes me a bit of an oddity there (i.e. your assessment rings true to me).

The mental map you describe is key. I think the ability to form/manage that is part of the reason behind the apparent multiplicative effect intelligence can have (highly tied to cumulative effort which I think is the major cause). I'm not as good at it as what you describe, but still far better than most and it is a huge aid to rapid assimilation/learning.

Thanks for the IAS pointer. I don't know the field/people well enough to get good value for time spent reading the entire transcript. The Bethe quote was great: ‘Gentlemen, if you knew what I was going to say, why did you invite me to speak? Now, I want to make an uninterrupted speech, unless you have a specific, detailed, and sensible question.’

steve hsu said...

> there is some aspect of innate intellect in play that is not captured by the tests <

This is undoubtedly true!

Richard Seiter said...

I agree that extraction is undoubtedly true (the funny thing is it reduces to the same thing the IQ test haters say, just without the dismissal of the utility of IQ). How do you feel about following up that idea further? Any thoughts on relative importance of the non-test factors vs. the test factors? Are they orthogonal or is high IQ a necessary precondition? How many non-test factors does it take to explain the people we are talking about (e.g. do S and D excel in different non-test factors than F, or does IQ explain most of the ability of some subset of them, or...)? Where do idiot savants fit into this discussion? Or people who just seem to be from another planet like Ramanujam? How much does spatial intelligence describe the non V/M aspect? etc. etc.

I find this topic fascinating. It's a big part of the reason I enjoy your blog and hope we get to see results from the BGI research at some point. If you'd like to say/discuss more I'd enjoy that and hopefully some of the other blog commenters would join in.

As a side note, another interest of mine is the short term variability of intellect. My thinking clarity varies dramatically (I would say by at least 1 SD of IQ) over fairly short time intervals (hours, days, weeks). I think this is largely in response to environment (perhaps the body reallocating resources to deal with toxins or ?, or perhaps just a simple interference with finely tuned biochemistry, or...). I have spent years thinking about this and would appreciate any insights anyone has to offer. I have been unable to find any research literature that discusses this directly.

steve hsu said...

re: utility of g, if you are admitting thousands of kids year after year to a freshman class, then statistical validity is all you need ask for from the test. Even a fraction of an SD difference between two groups (e.g., Berkeley vs UCLA admits) could have a big impact downstream: e.g., on Nobel prizes won per alumnus.

If you are only selecting a *single* programmer or grad student for your team, I'd still say an SD difference between two candidates is worth thinking hard about (other things being equal). So, I'm certainly not sympathetic to the claims of the anti-g crowd.

I think the biggest non-testable factor is motivation, drive, work ethic. We try to measure that using proxies but it's challenging, not least because the factor is itself time and context dependent.

Richard Seiter said...

I agree with you about lack of sympathy for the anti-g crowd (I think you have to have a great willingness to ignore data to be strong anti-g rather than just skeptical of it as a be all). I just find it intriguing that we all share the positive assertion you made. I'm a big fan of trying to find points of agreement with those who disagree with me (I'm not placing you in this category, it seems like we are largely in agreement on these ideas).

I also agree about the importance of motivation, drive, work ethic. It's only approximate, but I think the marshmallow test is the best simple proxy I know. I suspect you could go a long way in predicting peoples outcomes given results from both an IQ and marshmallow test (better than any other combination of two metrics?). FWIW I think there is a good chance self control can be improved by environmental/nutritional measures in significant subpopulations (e.g. inmates, there has been some interesting nutritional research in this area).

Any thought on how to evaluate the other non-testable factors relating to intellect? (ignoring the important second order effects of the factors already mentioned) The only ways I can think of are solving hard problems (testing concrete manifestation of the abstract factor, but generally requiring too much specific knowledge to be useful for talent identification rather than talent selection) and references from people who have a substantial quantity of the factor. Any other ideas? It really seems there should be a way to measure the differences in thinking between e.g. S/D and F. Perhaps fMRI? Or something even simpler like some variant of inspection time?

To be clear, given prior discussions of IQ and achievement it seems clear that the answer isn't just a harder IQ test. What needs to be different?

P.S. Responded. Thanks.

Christopher Chang said...

I think there's an additional distinction worth making between reading formal papers and directly reading others' source code/technical drawings/etc.: programmers/engineers/experimental physicists can be very effective at learning from others through the latter pathway even while having mediocre paper-reading ability.

steve hsu said...

> High V individuals who seem to have low M are just mathematically ignorant. <

I know many counterexamples, even from childhood, to this statement.

Richard Seiter said...

Can you give an idea of how you evaluate this and what sort of deltas you are observing? Are you seeing V>M differences of the same magnitude as the M>V differences seen in many engineers etc.? I'm particularly curious about how you distinguish mathematically ignorant (uneducated) but potentially high (or moderately so) M from low M.

ben_g said...

Kind of hard to believe that a guy who could write and tell stories as well as him would have low V

Yan Shen said...

Old info, but apart from the controversy over his supposedly low aggregate IQ score, the two main hard data points we have for Feynman regarding cognitive profile are

"He attained a perfect score on the graduate school entrance exams to Princeton University in mathematics and physics—an unprecedented feat—but did rather poorly on the history and English portions.[18]"

"Language delay is a risk factor for other types of developmental delay, including social, emotional, and cognitive delay. Some children may grow out of these deficits, even coming to excel where they once lagged, while others do not. One particularly common result of language delay is delayed or inadequate acquisition of reading skills. Reading depends upon an ability to code and decode script (i.e., match speech sounds with symbols, and vice versa). If a child is still struggling to master language and speech, it is very difficult to
learn another level of complexity (writing). Thus, it is crucial that children have facility with language to be successful readers.

Neuroscientist Steven Pinker postulates that a certain form of language delay may be associated with exceptional and innate analytical prowess in some individuals, such as Albert Einstein, Richard Feynman and Edward Teller.[1]"

Cornelius said...

Can you please elaborate?

I have never met a high V (3+ sd) and apparently low M (2- sd) person who truly had a low M IQ. Most of these individuals pursued high V interests from a young age and were scared away from math for one reason or another. American culture devalues math so it could be that in the US the number of people with low math literacy and great math potential outnumber those with high g but truly low M.

It should be relatively easy to gauge the M IQ of the math illiterate. Even humanities professors with a fear of math have learned basic arithmetic. A test of their ability to do mental math normalized against a diverse high g group should reveal whether these individuals have truly low M. Give everyone a week to prepare for those who are rusty. I would bet money that the M IQ of the high V folks turns out to be much higher than they suspect.

steve hsu said...

Cornelius: V and M are positively correlated -- this is the basic observation leading to g, etc.

I realize you and I were not describing the situation. I meant that there are many individuals who are, say, +3V, +1M, so "low M" is in relative terms, not absolute. I don't think I've met anyone who is +3V, -1M, although I suspect they exist (just very rare).


Richard Seiter said...

Steve, what do you think about Cornelius's edit: "The reverse was not apparent: high mathematical ability did not seem to indicate concomitantly high verbal ability." This seems surprising given that the reverse correlation did appear. The SMPY population is far enough out on the tail that I think we have to be careful about assuming things which are true in general are also true in this population (though that is certainly the way to bet without additional data).

+3/+1 seems like a good split for a talking point. I agree +3/-1 would be rare [and probably due to a specific problem (or possibly beneficial mutation?) or injury]. +3/+1 maps fairly well to what I see in engineering and I'll take your word for it you see the reverse split with similar frequency (I have too little close exposure to possible high V/low M and too little confidence in my ability to discern +1M from mathematically ignorant +3M from farther away, e.g. I would never guess my sister's M from a distance). My high school (all 6 grades) had ~1000 people (from a less selected population than the one you would have seen in a college town) so +3 was uncommon (and my college/career was/is biased high M). I wonder if your childhood could have had some bias towards high V/low M due to subpopulations of the professors.

Do you know of a scatter plot for the SMPY individuals? The main conclusion I draw from the majors/occupations plots is that Science (with a relatively small count) is the only one in the upper right (which I think was your point ;-). It would be interesting to see the relative density of points in each quadrant for the SMPY population. It would also be interesting to see the same plot with the axes normed against the general population (I suspect that would make some difference since I think the V SD is larger than the M SD in the SMPY)

It is worth noting that the SMPY would presumably exclude +3V/+1M people. Is there a SVPY? ;-) The "high V/low M" there would seem more likely to be something like +4V/+3M.

Cornelius, thanks for that link. There were some interesting tidbits. Any additional suggestions like that?

Norkuat said...

What about physicists with higher V than M? (Delta \approx 2 s.d.) Are they all into stuff like complex systems and such?

Diogenes said...

g falls apart at high ability somewhat. the plot of subtest vs g is heteroskedastic with lower conditional variance at lower g. the stupid are stupid in every way. the smart are often very smart in one way and not so very smart in others. one would expect greater maximum subtest differences with higher and with lower scores if the g vs subtest plots were homskedastic. inmho this supports the conclusions of this author

it's been my observation that high sat m and mediocre v is much more common than the reverse.

but m is not a factor of any iq test. the third factor which hasn't been mentioned but which does explain a lot of why iq is so poorly correlated with achievement is the attention factor. one who's high v and high s but average a is going to be limited. and limited more than someone who's high v high a and average s.

Diogenes said...

the three factors of the wechslers are verbal, visuo-spatial, and ATTENTION not math.

a biographer of sam walton remarked that walton's great talent was focus. vos savant has remarked that without superior attention other talents are useless. unfortunately, although amphetamine for add is the single most reliable treatment for any psychiatric disorder, its benefit for those with already superior attention is negligible, or so i've read.

Diogenes said...

cornelius, mathematicians distinguish algebraic and geometric. the ability to do mental arithmetic and to do higher maths i expect is no more correlated than any two cognitive abilities. talent vs realized ability is just a word game, because it's impossible to assess them independently for anything. but one of my grandfathers was an invigilated prodigy at mental arithmetic with an "innate intuition for physics" even though he was a high school drop out and spent a year in prison for bookmaking in the 20s. his invigilator was another relative who's in who's who for fluid dynamics.

Diogenes said...

but he'd then not be an example of ashkenazi over-achievement as the ashkenazim are average in spatial ability.

Diogenes said...

"I think the biggest non-testable factor is motivation, drive, work
ethic. We try to measure that using proxies but it's challenging, not
least because the factor is itself time and context dependent."

indeed? the distinction between ability and the desire for ever more challenging intellectual environments may be without a difference.

and now that life expectancy at birth in the solomon islands is 74!!! i must agree with diogenes of sinope and with

"Since analysis is hard, here is an anecdote:

I know a guy who has barely done a day’s work, pushing fifty, he spends
his time cultivating his 2600 (or something, I don’t know, chess appals
me) world ranking. I understand father was a genial CEO type who never
had the luxury to be a board-game bum. The IM naturally sees his father
as a Philistine-with-trust-fund.

I hate to sound like a Commie – my moustache is a lot narrower than
Stalin’s – but hyper smartness is a bourgeois phenomenon; (and, yes,
there were Soviet geniuses, but they ran on residual bourgeois fuel.)"

and that whatever their iqs, david sinclair, lenny guarente, david murdoch, aubrey degrey, etc. make all living physicists look stupid.

Diogenes said...

"What needs to be different?"

you're laboring under a delusion that human beings and their societies are simple. father's income is a bettr predictor of son's income than iq in the us, but not in denmark. rot on!

Diogenes said...

maybe it can't be trusted, so trust this:

or just trust my test scores. but then again the long-term reliability of iq tests has had scant investigation. are there "late bloomers"? are there "early bloomers"? does the current "configuration" advantage the early too much? are there people who keep developing up until their final hour? are there people who give up at 21 and decline thereafter?

the narrative is that at 17 one's iq is fixed until death. is there actually any evidence for this?

5371 said...

Some of those people (I had to look them up) seem to have a problem with irreproducible results.

Cornelius said...

Unfortunately, that SMPY study is the best I've seen to support the argument that high V leads to high M.

My view developed in grad school. I was very social in grad school and so got to know students from many different departments. The students in the typical high V fields were much better at math than what my bias led me to expect. On the other hand, the engineers really were as low V as I expected.

I was also a TA for the lab portions of the intro physics courses for engineers. Those kids were HORRIBLE writers and this was a top school.

Based on those samples I formed my current bias. The gap between V and M for high V types is much less than the gap between V and M for high S types. I assume that M is mostly constructed from components of V and S. S seems to be more useful for M, geometry being an extreme example, BUT some areas of M such as mathematical logic lend themselves more to V.

I suspect that the thing we call g is mostly just fitness. It might even make sense to call it cerebral fitness. g is just how well the brain developed. Undoubtedly there are genes that control how "well constructed" a person's brain is. That is why g seems to underie V, S and M. Even if separate genes determine V and S, the effect of those genes is dominated (in most individuals) by the genetic and environmental factors that control overall brain development.
Very high g individuals and some regular individuals from ethnic groups with a higher frequency of high V or S alleles (e.g. Jews and Native Americans) show a more lopsided intelligence profile. If intelligence were just g, there would be no ethnic groups with lopsided intelligence profiles.

Of course much of this is speculation, but this is the hypothesis I would work with if I were at BGI.

5371 said...

Since that's nothing like what biologists mean by fitness, I think you need a different word.

Richard Seiter said...

Heh. I don't think I have that delusion. What I do have is a belief that a small number of simple but powerful ideas can go a long way towards helping to understand people/societies. Just have to be careful not to confuse the model with the reality. If anything I think my cognitive style tends more towards overcomplicating things rather than oversimplifying.

Richard Seiter said...

Interesting. Thanks.

I would divide your description of g into anatomic and metabolic components. The anatomic side (relatively fixed) provides a ceiling (and other constraints/ characteristics like robustness to environmental changes) while the metabolic side involves shorter term fluctuations (e.g. think pre/post coffee or smart drugs). For me the metabolic side seems to mostly affect ability to focus (I think similar to what @Diogenes means by attention, I agree on its importance but think it correlates strongly with self control, e.g. marshmallow test, so tend to talk about that instead), but I think it also affects g (e.g. ability to reason abstractly and/or working capacity).

While we are on this topic, I encounter people (who seem reasonably but not exceptionally smart otherwise) who appear utterly unable to think abstractly (saw this a lot in high school, some in engineers, but more in business types). Is this all M (how I usually interpret it) or does it also load on V and S?

P.S. I'm not sure fitness is such a bad word here. This usage is nothing like what biologists mean by the word, but I think it corresponds fairly well to the idea of physical fitness which is a good analogy for communicating with large groups of people (the question being how valid is the analogy. I tend to think it has a fair amount of validity). I think this notion of fitness also implies certain things about the correlation of g and physical attributes which are likely to be true.

Richard Seiter said...

I don't think talent vs. realized ability is just a word game even though I agree it is impossible to assess them *completely* independently. I think this is important because I believe there is value in identifying people who have talent (like your grandfather) and giving them an opportunity to develop it into realized ability.

Richard Seiter said...

I think Eysenck's theory about stimulation and extra/introversion may be relevant here. Experimenters also used amphetamines (and depressants) to change people's levels of stimulation. Any thoughts on whether attention/focus is correlated with extra/introversion?
If this observation has merit one might expect amphetamine to have a negative effect on attention for those with normally superior levels.

I'm not seeing the same breakdown you do for Wechslers:

Richard Seiter said...

Thinking some more about the mental map idea. It seems to me the influence of people with a physics background on other fields might be partially caused by physics (along with the requisite applied math) providing more effective mental maps (power and general applicability) than most fields. Does this seem plausible to you? (I can't decide if the previous is too obvious a question to even ask, so I'll err on the side of asking) If so, any thoughts on which parts of physics/math offer the most "bang for the buck" in this area?

P.S. I think one of @Cornelius comments in was a pretty good cut at an answer to a more general version of my last question.

Cornelius said...

This is all anecdotal:

Language delay runs in my wife's family. They are all relatively high g but don't seem to be lopsided towards V, S or M. I would characterize them as having high fitness. They are all tall, healthy, attractive, smart and have better than 20/20 vision. A particularly interesting feature of their language delay is that all the individuals with language delay jabber incoherently before they start speaking coherently and once they do start speaking, they speak in complete sentences.

Yan Shen said...

Some further comments related to autism and math ability...

"The ability to perform better on intelligence tests that measure spatial rather than verbal ability is inherited in people with autism, according to a paper published online in October in Human Genetics.

Traditional intelligence quotient (IQ) tests measure a combination of spatial intelligence — such as the ability to create a pattern with blocks of different colors — and verbal intelligence, which includes
language measures such as vocabulary. These are separately referred to as performance IQ (PIQ) and verbal IQ (VIQ). People with autism have been shown to perform better on PIQ than on VIQ measures."

"The participants in the study were 36 children between the ages of 7 and 12. Half had been diagnosed with autism. Each group had 14 boys and four girls, as autism disproportionately affects boys, the researchers noted.

All of the children had IQs in the normal range and showed normal verbal and reading skills on standardized tests administered as part of the recruitment process for the study. However, researchers found that on the standardized math tests, the children with autism outperformed the other children.

After the math test, researchers interviewed the children to assess which types of problem-solving strategies each had used: Simply remembering an answer they already knew; counting on their fingers or intheir heads; or breaking the problem down into components, a comparatively sophisticated method called decomposition.

The children with autism displayed greater use of decomposition strategies, suggesting that more analytic strategies, rather than rote memory, were the source of their enhanced abilities, the researchers said.

The children then worked on math problems while their brain activity was measured in an MRI scanner, in which they had to lie down and remain still. The brain scans of the autistic children revealed an unusual pattern of activity in the ventral temporal occipital cortex, an area specialized for processing visual objects, including faces, according to the researchers."

BTW, if anyone here is interested, there was a documentary made about the 2006 British IMO team called Beautiful Young Minds. It's all up on YouTube.

Part 1 of 6 above. One of the most interesting things discussed was the high prevalence of autistic children that year amongst those featured in the video.

"Beautiful Young Minds was a documentary first shown at the BRITDOC Festival on 26 July 2007 [1][2] and first broadcast on BBC 2 on 14 October 2007.[3] The programme follows the selection process and training for the British team to compete in the 2006 International Mathematical Olympiad (IMO), as well as the actual event. Many of the young mathematicians featured in the film had a form of autism,which the movie links to mathematical ability. The team goes on to win numerous medals at the olympiad, including four silver and one bronze. It was directed and shot by Morgan Matthews, edited by Joby Gee, and featured music by Sam Hooper. It was also screened at the Bath Film Festival in October 2007.

stevesailer said...

My impression is that most of the famous physicists of the 20th Century also had highbrow tastes in the arts and humanities. Feynman stood out for having regular guy tastes.

stevesailer said...

Here's an analogy: it's not uncommon for professional musicians to be unable to read music.

In contrast, I have zero music ability, but after a year of piano lessons at age 8, I could read music better than many real musicians. But I still had poor rhythm, no ability to harmonize, and only the most theoretical knowledge of keys.

In this analogy, Feynman would be like a master jazz soloist for whom reading music was mostly a waste of time.

Diogenes said...

but lopsidedness is according to the white gentile mean? so looked at another way white gentiles are lopsided with high s low v.

Diogenes said...

my experience is that the ability to write well and verbal iq are not well correlated. of course what makes writing good or bad is subjective, and in general the simpler the ideas one is trying to communicate the "better" the writing.

i've known many verbally gifted people who found writing the most difficult thing. talk show host charlie rose has said it's the hardest thing for him.

is there really anything such thing as good writing? see waugh on joyce. ulysses was rated best novel of the 20th c by some large body.

Diogenes said...

the number of factors is as many as you like so long as there is any residual which isn't pure noise, and the wechsler has been revised recently. digit span and arithmetic would load high on a.

if one can't sit still in a quiet room and read his verbal iq is going to be less all else being equal than one who can.

Diogenes said...

iirc, one great who couldn't read music was dave brubeck.

Diogenes said...

but the bigger point is that when 90% of people in the developed world die from age related diseases and the correlation between income and life expectancy is so small, why does any really smart person not pursue the very simple question, put metaphorically, "why does a man live 7 x longer than a dog?"

one reason is that there is no funding for it. less than 1% of the nih budget is dedicated to research in the biology of aging.

if you think this question has been answered you're wrong. even today no one has any idea what the answer to that question is.

for this reason i've known since age 12 that my society is insane.

perhaps steve is looking forward to up-loading himself.

stevesailer said...

Feynman was one of the great storytellers in American culture.

Yan Shen said...

Of course, Feynman like Einstein, etc had uh other sources of inspiration as well...

"While a professor at Caltech, and after winning the Nobel Prize, Feynman would frequent a Pasadena strip club to work on his calculations, stopping to look at the girls when he was having difficulty. It is obvious from his recounting of the story that Feynman derived a great deal of self-satisfaction from his part in bridging the great divide between theoretical physics and female nudity. Perhaps it was Feynman’s mother who registered the most amusing and poignant reaction to this aspect of Feynman’s character; when Omni, a widely read popular science magazine at the time, named Feynman the “Smartest Man in the World”, her response was brief: “If that’s the world’s smartest man, God help us.”

Matthew Stern said...

High V low M?

Luboš Motl said...

An interesting topic but I disagree with your views, Steve, see:

Cornelius said...

"The M threshold was set at a fairly anemic 500 when looking at the high V types..."

The averages for SAT V and M among college-bound students were 430 and 500 respectively. A score of 500 for a 12-year-old is quite high. Only one student scored below 500 and we don't know if that individual scored 490 or 400 so let's neglect that outlier for now. If we take 500 as a cutoff, then the average high V student scored somewhat higher than 500. If we use a naive ratio IQ, that means the cutoff was around a ratio IQ of 150 (approx. 140 deviation IQ), and this was the 1980s before the higher education bubble, so the 18-year-olds we're comparing to probably had an average deviation IQ of 110. Our high V kids probably had a cutoff M deviation IQ of about 150.

I don't know about the connection between language delay and high S. I see a relationship between language delay and high g. This is definitely an area for further study. Did Einstein have S >> V? He had language delay but he also spoke in complete sentences once he started speaking and he read Kant at 13. That sounds like a high V profile to me.

I have noticed the relationship between S >> V and autistic qualities. It's an interesting relationship. Do people with low motivation to interact with others just end up concentrating their effort into other areas, which don't develop verbal and emotional intelligence - although the potential exists? Or does the low V and/or EQ come first? Feynman is an interesting case because he was a great communicator and had high EQ.

Cornelius said...

Williams syndrome sufferers have low IQ across the board, but they have large vocabularies and well developed communication skills. Their extremely high motivation to interact with people may have more to do with their communication skills than their V IQ does.

I think of domestic dogs as having the equivalent of Williams syndrome in canines. They are dumber than normal canines, but they have very high EQ and an intense desire to connect with others. Dogs have very little social fear; they want to be everyone's friend. They can pick up on human emotions even better than many humans can. And they suffer from many health problems due to their genetic abnormalities.

Richard Seiter said...

@Luboš Motl you make many excellent points in your blog post, but I think criticizing Steve for ad hominems against Feynman is a bit rich given the number of them against Steve that appear in your post. I think everyone in this thread would agree Feynman was a genius (whatever that word means precisely). However, I think it is valid to discuss how his genius differed from that of many of his peers.

Richard Seiter said...

Interesting. Thanks for that link. You don't happen to have a link to free full text (I could not find any)?
I agree with you about the importance of attention/focus. I also think that, if it is possible, the best way to increase g in children/adults is to address this dimension.

I had noticed attention in the subtest details of my link, but thought it was too uncommon there to be a principal component of g. Your link indicates that view was incorrect.

BTW, the paper literally said: "Verbal Comprehension, Perceptual Organization, and Freedom from Distractibility". Is Perceptual Organization equal to or similar to spatial ability?

efalken said...

Feynman was a genius, but he was human too, meaning, his genius was limited to various domains that he applied his efforts and attention. Supposedly his papers had heavy editing on their English. Einstein's English compositions were banal, and they were probably highly polished by other, and his understanding of non-physics was pretty poor. Like athletes, experts in everything are experts at nothing, and while a great soccer player is probably better than average at American football, he's also probably rather inferior relative to a typical college football player. Specialization is necessary, and this implies great inconsistency of excellence. Feynman's grasp of common sense was far superior to Gell-Man, who was definitely more of a stereotypical intellectual. Feynman's "Cargo Cult" lecture, his Challenger findings, his fixation on fun, highlights he understood the difference between, as he would say, the man and the uniform.

Hal Swyers said...

A lot of our views on intelligence are arguable capricious and arbitrary...

Diogenes said...

the factors are given whatever name the person doing the analysis likes. they don't mean anything by themselves, but looking at the subtests with the highest loadings for a given factor one can decide for himself what aspects of cognition these factors correspond to if any.

for the sb, passage memory is another attention loaded subtest. and attention is a rather "large" and vague term. believe it or not, i've known otherwise intelligent people who can't read for any length of time without becoming sleepy. and some tire sooner than others. at the extreme there may be some who can sit down and read an entire tractate of the talmud without a break.

steve may have a digit span of 15 and so doesn't notice it, but even if one understands all the concepts putting them together in long complex relationships takes a third factor beyond v and m or v and s, imho.

provigil is another medication which may enhance attention even in those without a problem.

5371 said...

Did you ever read Aldous Huxley, "After many a summer"?

Sagredo said...

I think that he just presented himself as a plain guy to emphasize his genius. He disguised his work and efforts. Feynman was an artiste. "He was more of a clown than he was a scientist sometimes" (Gell-Mann)

peterlund said...

Kierkegaard was extremely funny (and not that hard to read when you already speak the language).

ohwilleke said...

"I have always felt that Feynman was cognitively a bit "lopsided" -- much stronger mathematically than verbally."

Feynman is perhaps the most verbally articulate of all major physicists bar none. While Einstein may have a catchy quote or two to his name, Feynman's fame is as much as product of his voluminous non-professional writings as it is from his impressive professional contributions to physics. His explanations of QED are a lucid as the popularizations of science by Carl Sagan who is renowned for his verbal abilities. How many other modern physicists are often quoted at weddings and funerals?

If anything, his weakness was in making sense of the math and physics jargon in the professional literature which is more of a mathematical thing than a verbal thing. Sure, he was a physicist and as such far better than the average bear at math while only modestly talented verbally - but he came closer to a balance between the two than just about any other scientist of his era.

Diogenes said...

yes. lord something continues to "develop". as man is like
an ape which never fully develops, he turns into an ape eventually. i even listened to it on tape when i was very little.

the answer may be that aging is so complex that the only way man can live
longer is to become a bowhead whale, so to speak. BUT the evidence so
far is that very simple interventions have dramatic effect on lower
animals. the knockout of ONE gene combined with cr multiplies the
lifespan of c. elegans by 8! similar results in single knockout mice.
and rapamycin has been shown to retard aging in a large sample of mouse

an even more interesting question is why some
animals DO NOT age. this is true of the nmr and of all vertebrates which
never reach a fixed size. crocodiles, turtles, rockfish, etc. DO NOT
age. the nmr is the only mammal which doesn't age.

aging = increase in mortality rate with chronological age.

but any one who claims to be smart wouldn't have ANY interest in black hole entropy or whatever.

Diogenes said...

when i was 12 i was concerned i was like winston smith, "a minority of 1" (except for the biochemist quoted in the omni article which made me a minority). but then my japanese exchange student at 13 told me this was HER interest. a guy in hs told me it was HIS. he went on to become a dentist, couldn't get into med school, i guess. she couldn't pass the entrance exams in japan.

here's a story. the founder and leader of what was once, may still be, the most valuable company in the world died from cancer in his 50s. his company was a TOY MANUFACTURER and its market cap wasn't from a bubble, it was justified by real earnings evidenced by it's buy back "yield". what killed him? cancer. and the incidence of cancer is 100 x greater in the 6th decade than at ages 11 and 12 when human mortality from all causes is at its lowest. so the chances are 99/100 that the ultimate cause of his death was aging. his billions couldn't save him.

5371 said...

Aren't you worried that if avoiding death became easier than it now is, people would take cowardice, cruelty and egotism to heights they can never reach even in our world?

5371 said...


Diogenes said...

(n)aked (m)ole (r)at

all such concerns are premature. it may be that a human living to 150 would be like a human flying or living underwater. that is, for these the human would have to be a bird or a fish. humans already live 3x longer than would be expected from their body size and basal metabolic rate.

BUT i think this IS an open question.

Hacienda said...

People who obsess about death are stupid. Death is the greatest mutational change. As always, nature is more creative than humans will ever be. The only choice left for Westerners is to depopulate, justify their history by expanding into space or some other equally preposterous adventure, and the return the Earth to those who can live on it in peace (minimal mechanization). Speak not of savages, for you do not know them.

Diogenes said...

and you do know savages, like schneebaum?

as conway said, "the romans reached as far as hot baths without the fatal knowledge of machinery. the only thing the east should take from the west is plumbing."

"People who obsess about death are stupid."

true. and about health too. orthorexia is apparently a real disease.



Nat Philosopher said...

One thing I found interesting about Feynman's cognitive style was his discussion in Surely you must be joking... of learning techniques to pickup women. An expert in pickup gives him some pointers like: never buy a girl a drink before she has agreed to sleep with you and means it. Rather than simply applying the advice, which is also how modern primers on pick-up would teach it, Feynman approaches it by deep method acting, putting himself into an emotional state where he strongly resents women for trying to get a drink from him on false pretenses. His method acting is successful, but because he is uncomfortable with the emotional state he abandons the method after one or two tests. its interesting that such a cerebral guy would have such a method acting rather than detached approach to the pickup method taught, and also interesting that he was able to implement it.

Blog Archive