This is part two of my response to Larry (see previous card)...

Larry asks the following questions: "Do you think intelligence can be measured by something as simple as reaction time? Is it possible that this is a culture-fair measure? Do central processors really exist?"

The idea is that IQ measures the "efficiency" of a central decision maker, where efficiency means speed (fast on the straigtaways, slow on the curves).

I have alot of problems with this idea. First of all, as I explained in my previous card, I believe that the brain consists of many processors, not just one. Measuring reaction times while reading, like most standard IQ tests, focuses on the tip of the processing iceberg. IQ tests, including Larry's reaction test, are essentially LANGUAGE tests.

For this reason I fear that the reaction time test as Larry describes it addresses only one facet of intelligence and is also NOT a culture-fair measure. Wherever there is language there is cultural bias. When measuring reading you are measuring things like vocabulary which is in turn a product of education which is in turn a product of income, social standing, etc. Language is a social act and measuring language is measuring society.

(I am oversimplifying here and I know Larry has heard these arguments before. I will gladly expand and defend if asked.)

This problem would be ameliorated somewhat if a broader range of reaction tests were employed. But I still have a problem with the notion of efficiency. There is more to intelligence than coming up with efficient solutions. The hallmark of a great mind, it seems to me, is the ability to come up with STRANGE solutions, which are almost by definition NOT efficient.

You may respond that you are measuring not just speed, but speed at the appropriate moments. And yet who defines what decisions are appropriate, or what parts of the text deserve the most attention. Cultural biases may sneak in at this point.

Or perhaps you will argue that strange solutions are the result of some kind of low-level efficiency. Maybe. But this notion still seems to ignore many key complications. For example, an efficient algorithm may be slower in the short run but faster or more productive in the long run. And again, who defines "productive"?

All of this raises two deeper and more troubling questions. First, what in the hell IS intelligence, anyway? Intelligence is one of the most nebulous concepts in all of language. An IQ test is necessarily based on some kind of arbitrary and inhumanly narrow definition of intelligence, and each definition is oozing with cultural biases. So the problem of cultural bias is at the root of any test right from the beginning.

And second, are IQ tests really a good idea? Professionals know that IQ tests measure the ability to take IQ tests, but many people take them far too seriously. Often the people most eager to reduce the million dimensions of a human mind to a single number are also people in positions of authority. A dangerous situation indeed!

Still, I must admit that the experiments in this area are intriguing and may shed some light on what is still a very murky understanding of the human mind. Tell us more, Larry!