The Baby and the Bathwater: Concepts That Are Flawed But Still Useful

I think there's an important distinction between "this concept is being used in a flawed and damaging way" and "this concept is inherently useless and damaging". But too often, I see people confusing the two, and assuming that if there are valid criticisms of a concept, it's an entirely useless concept that can only be harmful.

Here are some concepts I think still have some use, despite their flaws.

Mental Age

Mental age is the idea that if you compare an atypically developing child's scores on a cognitive test to the results of NT children of varying ages, very often their pattern of functioning will be similar to that of most NT children at a particular age. This age at which responses are similar to the atypical child's responses is termed that atypical person's "mental age".

This has traditionally been taken to mean that atypical children should be like NT children with the same mental age in all sorts of ways, including ways not actually assessed by the test used. For example, a 15 year old with a cognitive mental age of 5 years is assumed to also have the emotional maturity of a 5 year old, the interests of a 5 year old, the academic abilities of a 5 year old and the level of independence of a 5 year old. This can result in unfair restrictions on this person, missed opportunities to do things a 5 year old wouldn't be ready to do, and generally treating this person like a small child when he or she would rather be treated like a teenager.

For an added wrinkle, some atypical people have skills that are not standard for any age of NT. For example, an autistic child who communicates through complex echolalic but mostly situationally appropriate sentences has a language profile that is not typical of NT children at any age, so assigning a mental age to that child's language abilities is fairly arbitrary.

And lastly, not all NT kids are actually taught to their full potential. For example, there is evidence that under the right circumstances, NT 1 year olds can actually be toilet-trained. In addition, no real effort is usually made to teach reading to children under 6 years old, but some children have learned to read much younger with direct instruction. This means that even if most NT kids at the same 'mental age' aren't being taught a skill yet, direct teaching might enable a delayed child to master that skill.

But despite all this, I think mental age is still a useful concept, provided you a) confirm that they are actually showing a normal pattern of development at a different rate, and b) remember that the mental age given by a test only applies to the skills assessed by test.

For example, on most cognitive tests, adult non-human great apes score similarly to human preschoolers. Social cognition is similar to a typical human 3 year old, language is more like a 2 year old (in language-trained apes), while folk physics understanding is more at around a 5-6 year level. (Working memory, meanwhile, is better than most human adults.)

This data pretty reliably predicts what cognitive toolkit a typical non-human great ape has for solving problems.  For example, when it comes to social problems (eg "how do I keep that guy who's bigger and stronger than me from taking away this snack I just found?" or "how do I do this thing I want to do and not get in trouble for it?") non-human apes solve these problems about as skillfully as 3 year old humans do. For example Koko, a sign-language trained gorilla, when asked who ripped the sink off the wall, blamed her cat All Ball - about as plausible a lie as the ones many 3 year old humans make.

And yet, if you left a group of 2-6 year old children in a rainforest by themselves - even if they'd grown up in that forest from birth - I doubt they'd survive very well. This is a challenge that suitably prepared non-human apes can easily manage. Indeed, some non-human apes grow to be very wise and experienced leaders in their societies, guiding the younger generations.

If you take mental age as 'the age they really are inside', you're doing it wrong. But if you take mental age as 'they will probably be able to do X task in skill area Y because typical kids with the same mental age in skill Y can', then it's a pretty useful concept.

IQ Score

This is a closely related concept to mental age. In fact, originally IQ scores were derived by dividing 100 by the person's actual age and then multiplying by their mental age. Now, they're usually determined more by the percentage of same-age peers who score higher or lower, but the link still applies, and IQ can be misused in the same way that mental age is.

However, there are also some misuses that are more directly associated with IQ. Since IQ is always determined by standardized testing (while mental age can be determined in lots of different ways), many of the criticisms of IQ relate to problems with standardized testing.

The biggest problem is underestimating performance. If a person fails to answer a question on an IQ test correctly, did they really not understand the concept? Or did something else interfere? If they didn't understand the instructions, weren't physically capable of responding correctly, interpreted the question in an unusual way or just plain decided not to answer correctly, their IQ score will be an underestimate of their true cognitive ability. This is the main reason why ABA can increase IQ in autistic children - it doesn't make them smarter, but it does make them more compliant, better at understanding commands, and teaches them what kind of behavior is expected in a testing situation. (Basically discrete trial training is like non-stop testing with the correct answers being given to the child until the child learns them.) In other words, they become more 'testable', and so their IQ goes up.

Another criticism of IQ comes from the myth that IQ should predict adult performance. Many people express skepticism about the idea that a person with a genius-level IQ could be working at McDonalds instead of revolutionizing some field of study. But IQ was never intended as a fortune telling device, and there's a lot that determines success apart from IQ - socioeconomic status, disability, motivation, resilience, educational opportunities, and pure luck and timing. Einstein wasn't just smart. He had supportive parents, he was fascinated by physics and mathematics, he had enough mental resilience to withstand a bad school experience, and he entered the field at a time when the conditions were ripe for the discovery of relativity.

There are people who are every bit as smart as Einstein, but these other factors haven't come together as much. Torey Hayden's books One Child and Tiger's Child feature a girl with an IQ of 182, which is higher than the common estimate of 160 for Einstein. But at the age of 6, she was in a special education class as a last resort before going into a state institution. Why? Because she was from a poor, neglectful and severely abusive home (she was hospitalized once because her uncle stuck a knife in her vagina), and had severe emotional problems (she poked out the eyes of the classroom fish). Given her struggles, the fact that, as an adult, she was able to live independently, stay out of trouble and be employed was remarkable. Could she have revolutionized a scientific field? Maybe, if she'd had the opportunity. But it's not really surprising that she didn't, given how much the tables were stacked against her.