Just went to Howar Gardner's talk at the OU. He's the one mostly famous for the theory of multiple intelligences. Don't think he's the only one claiming such things, but reading through the wikipedia article I kind of gathered the existence of some experimental evidence behind these ideas. It says:

An intelligence, he feels, is substantiated when it is demonstrated by evidence of * Potential isolation by brain damage. * The existence of idiots savants, prodigies and other exceptional individuals. * An identifiable core operation or set of operations. * A distinctive development history, along with a definable set of 'end-state' performances. * An evolutionary history and evolutionary plausibility. * Support from experimental psychological tasks. * Support from psychometric findings. * Susceptibility to encoding in a symbol system. (Gardner, 1983, p. 62-69)

He's recently also come out with a new intelligence, the naturalistic one:

"The core of the naturalist intelligence is the human ability to recognize plants, animals, and other parts of the natural environment, like clouds or rocks. All of us can do this; some kids (experts on dinosaurs) and many adults (hunters, botanists, anatomists) excel at this pursuit." From an interview with Howard Gardner by Ronnie Durie in Mindshift Connection, a publication of Zephyr Press.

But as a couple of cognitive psychologists pointed out after the talk (which was centred on his new book, the "Five Minds"), how are the above criteria used? Is there anything that grounds these assumptions? In other words, why five and not six or seven or eight? Gardner simply could not provide an answer: the workaround is his claim to use the five minds idea as a "policy of an educator" in setting up a learning design. Are we satisfied with that?

I think I agree with what the wikipedia says, this time:

...one of the major criticisms of the theory is that it is ad hoc. The criticism is that Gardner is not expanding the definition of the word "intelligence"; rather, he denies the existence of intelligence, as is traditionally understood, and instead uses the word intelligence whenever other people have traditionally used words like "ability". In this view, it is intellectually dishonest to relabel all of a person's talents as "intelligences". This tactic has been criticised by Robert J. Sternberg (1983, 1991), Eysenck, 1994, and Scarr, 1985.

And ultimately, it not new this not-so-scientific stance he's taking up:

Ultimately, it would certainly be desirable to have an algorithm for the selection of an intelligence, such that any trained researcher could determine whether a candidate's intelligence met the appropriate criteria. At present, however, it must be admitted that the selection (or rejection) of a candidate's intelligence is reminiscent more of an artistic judgement than of a scientific assessment. (Gardner, Frames of Mind: The Theory of Multiple Intelligences, 1983)