The conversation of intelligence comes up a lot in my circles, largely because therapy asks us to meet out clients where they are functioning, and so being able to gauge their intelligence plays a large role in helping them to properly heal. The problem is, defining intelligence is really one of the more difficult metacognitive tasks. Intelligence must define itself, a difficult task as it observes from within. As a result, in many ways, intelligence is like porn: we can’t define it, but we know it when we see it, the only difference is we want people to know we have intelligence, and want to be known publicly for having a lot of it.
Now granted, we do have what are supposed to be “intelligence tests,” but these are based off of testing for specific tasks we feel are parts of what intelligence could be, such as working memory, comprehension, etc. and they really don’t account for a lot in fields where success depends on intelligence. Your ability to succeed may require you to have the ability to perform at a certain advanced level on these tests, but merely being able to perform does not in any way guarantee success. In practice, these tests are more often used not to predict success or discover highly intelligent people (we allow actual success to find the intelligent for us), but rather we use them to diagnose problems, where discrepancies in between performance on different tasks allows us to diagnose learning or cognitive problems.
Now this doesn’t mean the tests should be discarded, or say nothing about intelligence, as they have a utility noted above, and are tested to show their reliability before use. Each subtest that measures performance on a specific task is measured against what we feel is “general intelligence,” or “g.” The tests may give us a very fuzzy picture of g, but they give us a picture none, the less, and, unfortunately, give us the best picture of any other tool. In these validity studies, there is something that has given us an interesting result that may give us a clue to understand intelligence better.
As it turns out, the subtest that is found to be most highly correlated with g is that of vocabulary. The more intelligent you are, the better you perform on this test. This is an interesting result because the specific task that vocabulary tests it memory recall, and has nothing to do with reasoning, abstraction, or other higher order cognitive functions. And so, the test that doesn’t directly test these higher order functions is actual the best measure of these functions.
So why is this so? How can a test that does not measure higher functions predict the presence of those functions? If you look at humans and language, the answer may be found in the process of acquiring language itself. People do not learn language through dictionaries and word drills that merely focus on storing and recall of words. Rather, learning language is a complex process that requires people to understand context, reason out rules, infer definitions, learn vicariously from the language of others, etc. These methods of language acquisition require the exercise of higher order cognitive skills. As such, it may be the reason that vocabulary correlates so highly with g, is that correlation does not reflect only the act of recall, but the complex act of acquisition that information being recalled.
This concept can be taken from the realm of vocabulary and into the thinking and intelligence as a whole. A lot of people like to emphasize living according to rules. We are regularly surrounded with rule filled self-help books, political and religious ideologies, codes, and other forms of rule based thinking. The irony is: rule based thinking is not actually thinking. In its most basic form, rule based thinking is about looking for a situation in which a rule is applied, and then applying the rule to give your the result. Like vocabulary, in execution it depends largely on recall, not higher cognitive processes. Intelligence, then, is about acquiring rules, and not just applying them. Intelligence is the continual investigation and testing of rules, not the adherence to them.
This idea that intelligence as being an acquisition and not adherence to rules is born out as we look at the most restrictive examples of rule adherence. The most extreme form, that of dichotomous black/white, good/bad, all-or-nothing thinking is regarded as indications of developmental and cognitive restriction within psychology, and largely seen as pathological. But even when you leave that extreme of borderline thinking behind, the more rigid a person is, the more they rely on simple and strict rule based thinking, the less they are able to actually understand and function successfully in the world.
So to intelligence is more about how you search than what you have, its how you play with and acquire the information, not what you have. With the rules we live by, its shows more as you investigate, acquire and increase the complexity of the rules than just have the rules themselves. Which makes a lot of sense, because rigid rule adherence requires you to follow prescription and an other’s thinking, rather than actually thinking for yourself.