I don't get it. How is IQ not malleable? To me, it obviously is. Let's look at the WAIS. One of the subtests measures the size of one's vocabulary. Let's say, you have a child who never used to read. She considered it boring. She was still smart though. Suppose she was good at logical reasoning. Hence, you she got through school and ended up doing well on tests and what not. Suppose she took the WISC as a kid and bombed the vocab section.
Suppose, as an adult she takes on a reading intensive major and starts reading a lot more. If you made her take the WAIS now, she'd score higher on the vocab subtest. Let's say her other scores stay constant. Hasn't her IQ increased? If this isn't convincing just imagine that she also reads up on current affairs a lot more as an adult. So, her information subtest score also increases. Again, hasn't her IQ increased?
Maybe potential is fixed but IQ isn't potential. IQ is something that imperfectly correlates to potential.
FWIW, I think IQ score is a rather imperfect measure of ability. I do think ability, or at least expression of ability (er, is that just performance?), can change, especially to the extent that a child's processing improves (over time or, say, with therapies). However, in the particular example you suggest, it seems to me that the adult version of the person demonstrates better performance rather than increased inherent ability. On the other hand, if ability is simply the potential for performance, perhaps you are correct. Maybe the question for this example is the extent to which vocabulary is truly indicative of ability as opposed to performance as a proxy for ability. I'm getting tongue-tied here...
