N.Wells
Posts: 1836
Joined: Oct. 2005
|
| Quote (NoName @ June 03 2015,12:39) | With all due respect, and that is considerable, I have to say that while I agree completely with the thrust of your arguments, there is an underlying issue that is not being addressed, and should be.
That question is whether or not 'intelligence' as commonly understood is univocal enough to permit a useful definition, operational or not.
It seems highly likely to me that 'intelligence' is one of those words that gets multiple usages, multiple definitions, and even in those cases where precision is most wanted, it gets tossed around as if "everyone knows" what it is.
Consider merely the difference in meaning between 'military intelligence', 'Central Intelligence Agency', 'Intelligent Design', 'Artificial Intelligence', and 'Intelligence Quotient'.
These are less easily distinguished at the definitional level because there are multiple abstractions collapsed into the term.
I am perfectly prepared to say that until there is a useful univocal definition of 'intelligence', no explanation for it can be sought. 'Intelligence' is not a phenomenon in the mold of 'velocity'.
This is, of course, one of the reasons that an operational definition is required -- operationalizing the term should render it univocal. Or expose the fact that it is inherently equivocal and that there are specific elements or aspects of intelligence that are necessary but not sufficient. And that there may be elements or aspects that are sufficient but not necessary.
But I am all but certain that we will find that any useful definition of 'intelligence' will point up the fact that 'intelligence' is a generalization that points to a set of phenomena such that the full set consists of necessary and sufficient conditions but useful subsets do not overlap in either or both necessary and sufficient conditions.
Gary's "bug navigation" pseudo-exmple certainly points in a direction where there may actually be actual-factual genuine intelligence at play, and yet utilizes only a subset of necessary and sufficient conditions strictly disjoint from those that apply in, say, the composition of a melody, or the recognition of a transposed melody.
We all, myself included, fall too easily into the trap of taking seriously claims to have an explanation for 'intelligence' and rejecting the explanation, rather than the misuse and abuse of the term.
Which takes us right back to the necessity of a clear, specific, operational definition for *any* researcher making claims about intelligence. |
I agree completely. I didn't mean to skip over that. Psychologists have trouble defining and measuring human intelligence, let alone making a quantifiable assessment of whether cats or dogs are more intelligent. That's part of why science is stuck at "we don't know" (also, intelligence is clearly complicated). (We'll probably get there about the time a computer tells us, "And by the way, all your ideas about artificial intelligence are wrong". :) ) However, for Gary to proceed as he wishes, among many many other problems, he needs good definitions, and good operational definitions so that he can quantify the differences in intelligence at very low levels, building up to questions like the difference in intelligence between an amoeba and a mushroom. We'd all benefit from better actual and operational definitions, but Gary has to have them, and instead he's got an utterly worthless pile of word salad that manages to be demonstrably far worse than standard bad definitions. He doesn't actually have include all common versions of "intelligence", but he does have to pick one that makes sense and stick to it, rather than jumping around between different versions in order to smuggle in his desired conclusions.
|
