But with that said, any definition that says that Feynman wasn't a genius is obviously wrong. In every environment that he was in, he was given the genius label because he obviously was one compared to (usually quite intelligent) peers. If a test says that he is not a genius, then the test isn't capturing what common usage says that genius means.
Conversely the flip side of the point stands as well. The fact that someone is labeled a genius by a test does NOT mean that they have whatever intellectual abilities would be required to better society. There is a correlation - they are more likely to have those abilities. But not a guarantee. And the fact that we see high IQ people failing to use their brains life is not always evidence that those people's gifts are being used. The true gifts may simply not be what was recorded on the test.
Perhaps Feynman was a poor example. I was going by his claim that his own IQ was 125 (or something to that effect). I suspect that there is more to that number and that we shouldn't take it at face value. I see no reason why he wouldn't score much higher.
You are so wedded to your theory that you are unable to accept facts at face value. Presented with evidence that Feynman's IQ was not exceptional, you conclude that he was not a genius. Presented with evidence that he was a genius, you conclude that there must be an external reason that he did not do better on the IQ test, and the test result can be thrown out.
The conclusion that you seem to think impossible is that someone can clearly be a genius and yet that genius would not be reflected on IQ tests.
Let me offer you two alternate hypotheses.
The first fits the criticism in the above post. Many questions on IQ tests are ambiguous - there are multiple logically possible answers, and to find the "right" one you have to figure out what line of reasoning the test maker was most likely to have followed. This is a question about conventions, and Feynman clearly was extremely unconventional. He therefore is an example of someone who was literally too creative to do well on the test.
A second possibility is that Feynman's gifts were overwhelmingly mathematical, not verbal. His graduate entrance exams stand as evidence of this, as do comments from people who have examined his work. See http://www.psychologytoday.com/blog/finding-the-next-einstei... for an example. Since an IQ test has a large verbal component, on which Feynman would have only been somewhat above average, he did not score exceptionally well.
I like the first theory better, but there is significantly more evidence for the second. But either way there is no evidence from anyone who knew him, including Feynman, that the IQ score he got is not what he should have been expected to get based on what an IQ test measures.
The fact is that there is overwhelming evidence that IQ tests highly correlate with other measures of intelligence (SATs, GREs, job advancement, etc). I'm sorry but one outlier does not invalidate the whole body of research regarding IQ tests. The reasonable conclusion regarding Feynman is to suspect other factors at play. I'm simply changing my opinion in the face of new evidence.
>A second possibility is that Feynman's gifts were overwhelmingly mathematical, not verbal.
And of course this would be the "other factors at play". There are plenty of people who are savant-like in specific areas but aren't a "genius" in the generic sense. In a mathematical/scientific setting, Feynman would appear as a genius. In other settings, he would not. There's nothing surprising about any of this that would warrant a redefinition of genius.
Where to begin? I'll just make random points in no particular order.
- When a man widely recognized as one of the top geniuses of his generation is determined by a test to not be a genius, it is not Feynman's status as a genius that should be questioned.
- "Correlated with" is very different from "the same as." See the second half of http://bentilly.blogspot.com/2010/02/what-is-intelligence.ht... for some sample calculations on exactly how meaningful the correlations measured between an IQ test and other measures of intelligence likely are. (Short summary. If there is a 0.7 correlation between IQ and "true intelligence", then people with an outstanding IQ should be expected, on average, to merely be of somewhat above average intelligence. And vice versa.) Thus the measured correlations are actually evidence that IQ tests are at best moderately effective at identifying true genius.
- The body of data that we have on IQ and intelligence is COMPLETELY CONSISTENT with the prediction that we should expect a top mathematical genius to have only a somewhat above average IQ. Feynman's relatively modest IQ is therefore not a surprise. The fact that it is not a surprise is again evidence that IQ tests are a flawed method of identifying genius.
- We actually have no real evidence that IQ, g, or other related measures are measuring anything directly meaningful at all. The arguments about this are complex, please see http://masi.cscs.lsa.umich.edu/~crshalizi/weblog/523.html for an explanation. As a side note, to the extent that by "genius" we mean something other than the weighted average of random abilities actually measured by IQ tests, we should expect that IQ tests are not that good at identifying geniuses.
Yes, I know that I just gave you a lot of complex stuff to digest. But please give me the courtesy of assuming that I have at least a passing familiarity with the subject of IQ tests, have put some thought into it, and my opinions should not be immediately disregarded without investigation.
At the very least I hope I gave you enough to cause you to reconsider whether "top 2% in IQ" effectively captures what the word "genius" means in common usage.
In Feynman's case it seems quite plausible that the 125 is not an accurate report of a rigorous testing environment.
He is certainly the type of fellow who would intentionally misreport his score for personal/philosophical/fun reasons, or not care enough to remember the number correctly decades later.
Why do people keep trying to argue that an IQ of 125 is not reasonable for someone like Feynman?
The 125 figure might or might not be exactly correct, but it is reasonable given other data we have about him. For instance, according to James Gleick's biography, Feynman was the only person to ever get a perfect score on the math and physics graduate entrance exams to Princeton. At the same time he scored below average on the verbal and history tests. (Note that the "average" at Princeton would be well above average in the general population.)
Feynman was an amazing genius. I have the deepest respect for him. That fact notwithstanding, his genius apparently did not evenly extend to all types of intellectual abilities. If so, it is unsurprising that he would merely score very highly on an IQ test.
But with that said, any definition that says that Feynman wasn't a genius is obviously wrong. In every environment that he was in, he was given the genius label because he obviously was one compared to (usually quite intelligent) peers. If a test says that he is not a genius, then the test isn't capturing what common usage says that genius means.
Conversely the flip side of the point stands as well. The fact that someone is labeled a genius by a test does NOT mean that they have whatever intellectual abilities would be required to better society. There is a correlation - they are more likely to have those abilities. But not a guarantee. And the fact that we see high IQ people failing to use their brains life is not always evidence that those people's gifts are being used. The true gifts may simply not be what was recorded on the test.