Smarts

 THOUGHT OF THE DAY:     
"The magnitude of intelligence is not measured by length nor yet by height, but by thoughts." - Epictetus from Book I, ch. 12. - Discourses

Is intelligence something we can measure, or some aspect of it?  If so, how?

I think Epictetus is getting at an important distinction: intelligence is not the same thing as physical dimensions, and neither is it quite the same thing as knowledge. But that doesn't mean intelligence is impossible to measure.

The interesting question is what we mean by intelligence.

Psychologists have actually found a fairly robust way of measuring one important aspect of it. Intelligence tests tend to measure things such as reasoning, recognizing patterns, manipulating information in working memory, verbal comprehension, and solving unfamiliar problems. From many different tasks, a general factor—usually called g—emerges statistically. Someone who does unusually well on one kind of reasoning task tends, on average, to do well on quite a few others. That common factor is what IQ tests principally capture.

But there are some rather important qualifications.

1. We measure intelligence indirectly.
There isn't an "intelligence meter." We give someone problems and observe how efficiently and accurately they solve them. It is rather like measuring temperature by observing the behavior of mercury: the thing we're interested in isn't directly visible, but its effects are.

2. IQ is much better at measuring some kinds of intelligence than others.
It does a reasonably good job with abstract reasoning and certain kinds of learning and problem solving. It is much less satisfactory for things such as wisdom, practical judgment, creativity, social understanding, perseverance, or knowing which problem is worth solving in the first place.

A person could therefore be extraordinarily good at solving the problems presented by an intelligence test while being spectacularly foolish about life.

3. Intelligence isn't simply the amount of information in your head.
A well-stocked encyclopedia isn't intelligent. An intelligent person can take a relatively small amount of information and infer something that isn't explicitly there. That's why novel problems are particularly useful in testing intelligence.

And this brings us back to Epictetus. His statement could almost be translated into modern psychological language as:

    Intelligence is revealed by the quality of one's mental operations, not by one's physical or social stature.

There is also something wonderfully Greek about his formulation. He isn't merely saying that a clever person can solve puzzles. He's contrasting external magnitude with internal magnitude. A physically imposing emperor and a physically insignificant philosopher can differ enormously in the quality of their thought.

I'd go a step further, though. "Intelligence" may be less a single thing than a family of capacities with one substantial common component. The evidence for g is quite strong, but the everyday concept of intelligence is considerably broader than what an IQ score captures.

And perhaps the most Epictetian irony is that the ability to recognize the limits of one's own intelligence is itself a form of intelligence. A person who knows that he doesn't know something can make much better decisions than someone with a higher raw reasoning ability who confidently assumes he knows everything.

Yes—and that is an important distinction. The SAT was never quite an intelligence test, although people often treated the score as if it were one.

Your example of remembering events well but names poorly illustrates why. Memory itself isn't a single faculty. Remembering what happened involves a rather different set of associations from retrieving an arbitrary label attached to a person. A name like "Mr. Henderson" has very little intrinsic information; an event has context, relationships, sequence, place, emotion, and consequences. The latter gives the mind many routes back to the memory.

And the SAT's design naturally favored certain sorts of mental performance. It was trying, in broad terms, to predict how well a student would handle academic work, rather than answer the philosophical question "How intelligent is this person?" Vocabulary, reading comprehension, mathematical reasoning, and the ability to retrieve or manipulate information under time pressure all enter into that.

The time pressure is particularly interesting. A test can inadvertently measure something like speed of accessible recall rather than the depth of one's understanding. Someone may know something perfectly well but require a few seconds longer to retrieve it. Conversely, someone with rapid retrieval can look exceptionally impressive on a timed examination without necessarily having deeper understanding.

That may be one reason your distinction between memory for events and memory for names has remained noticeable to you. Names are almost the perfect laboratory for this: they're essentially arbitrary symbols, whereas events are richly structured.

There is also a broader point here that I think connects nicely to your Epictetus quotation. A test measures performance on a particular class of mental tasks; it doesn't measure "the magnitude of the mind" in its entirety. We can measure particular components of intelligence quite well. The temptation is to mistake the resulting number for the whole person.

And, frankly, the SAT's reputation as a measure of "intelligence" probably owed at least as much to its convenient single number as to what that number actually meant. A society loves a number because a number lets us rank people. Epictetus would probably have raised an eyebrow at that.

There are all sorts of kinds of "smarts." People smarts, money smarts, math smarts, number smarts, cooking smarts, garden smarts.

Exactly. And everyday language may actually be more perceptive here than the psychometric terminology. When we say someone has people smarts or garden smarts, we're recognizing an ability to perceive patterns and make good judgments in a particular domain.

The interesting thing is that these kinds of smarts don't all behave alike.

Someone with math smarts may readily manipulate abstractions, while someone with number smarts may have an uncanny feel for quantities, proportions, and whether a number "looks wrong" without doing the formal calculation. A person with money smarts may understand incentives, risk, and human behavior while being mediocre at mathematics.

And cooking smarts is a particularly nice example. A really good cook doesn't necessarily follow recipes particularly well. He knows that the onions aren't quite ready, that the sauce needs something acidic, or that the oven is running hotter than its dial says. Much of that knowledge is tacit—you can possess it without being able to formulate a rule for it.

Garden smarts is similar. You learn to notice subtle relationships among soil, moisture, season, sunlight, pests, and the behavior of individual plants. An experienced gardener can sometimes diagnose a problem almost instantaneously without being able to explain exactly how he knows.

That raises an interesting question about whether all these abilities deserve the same word, intelligence. Psychologists would probably distinguish general cognitive ability from domain-specific knowledge and expertise. But from the standpoint of ordinary life, "smarts" is a perfectly useful concept because it emphasizes something IQ doesn't capture very well:

Can you figure things out in the world you're actually living in?

There may be a useful distinction between being intelligent and being smart about something. General intelligence helps you learn a new domain, see relationships, and reason through unfamiliar problems. But experience can produce extraordinary competence in a narrow domain, sometimes without particularly impressive general reasoning ability.

And then there are people who have a particularly valuable combination: they learn quickly and they have good judgment about where to apply what they've learned. That's getting rather close to what older people meant by wisdom, I think.

I would say real smarts starts with interest and develops with experience.

I think that's a very good formulation. In fact, it explains something that the conventional notion of "intelligence" tends to miss.

Interest is the engine. Experience supplies the data. Judgment is what eventually emerges from putting the two together.

A person who is intensely interested in gardening, for example, is going to notice things that an uninterested person simply doesn't see: a change in the leaves, the timing of flowering, where water collects, which plants thrive beside which others. After ten years, that person may possess an enormous body of practical knowledge that was never deliberately "studied." The interest caused the attention, and the attention accumulated experience.

It also explains why people can be very unevenly "smart." Someone might be remarkably knowledgeable and perceptive about automobiles while knowing almost nothing about history. Another person might be superb at reading people but hopeless with finances. Their minds aren't necessarily of different power; they have directed their attention and accumulated experience differently.

And interest probably has a feedback effect. You become interested → you pay closer attention → you notice more → you get better → the subject becomes more interesting → you pay still closer attention. That's a remarkably effective learning machine.

There is one qualification I'd add: experience alone isn't enough. People can have fifty years of experience doing something and become merely very practiced at repeating the same mistakes. The crucial ingredient is attention to what happens. The smart gardener notices that his prediction was wrong and adjusts. The merely experienced gardener says, "That's how we've always done it."

So perhaps your definition could be sharpened to:

Real smarts begin with curiosity, grow through attentive experience, and mature into judgment.

That's a rather different creature from what an SAT score is trying to measure. And I'd say it is much closer to what most of us mean when we call somebody "a smart person."