Essentialism can be a difficult topic to write about. More and more I think it is at the heart of my beliefs about genetics and behavior, yet it can seem esoteric, as though I am more concerned with the way we talk about things than about scientific reality. But in fact, I think this is a rare instance where how we talk about things has a real effect on how we construct scientific beliefs. I usually talk about these questions in terms of the heritability of intelligence, but obesity is a more concrete example.
The Times had an article yesterday about obesity.
The original headline said something like, “We need to define what it means to have obesity,” but that has now been changed. It is now titled, “Are We Thinking About Obesity All Wrong?“. But an early sentence still refers to “expert groups around the world rush to define what it means to have obesity.“ Talk about “having obesity” is essentialist, as opposed to the descriptive way to put it, “being obese”. Obese is not necessarily a pejorative adjective; it is just a way of stating that someone is heavy. Outside of a scientific context, it doesn’t need a numerical definition any more than any other adjective.
Talking about “having” obesity suggests that it is some kind of internal syndrome with its own reality, and some specific set of causes, causes that can presumably be “treated” with effective interventions. But I don’t think weight is that kind of thing. I would put it this way. People vary in weight, for all sorts of reasons. Some people, inevitably for anything that varies, become very heavy. Some very heavy people become unhappy with their weight, either on their own or because a clinician of some kind has told them that they should be unhappy about it. People who are unhappy about their weight then have a choice about what to do about it. They can go on a diet, increase their exercise, take Ozempic, or learn to be happy with their body the way it is. Scientists or clinicians can offer data relevant to these decisions— what the health risks of obesity are, what the chances of success are for an intervention— but they can’t provide the scientifically correct answer because there is no scientifically correct answer.
These are all difficult decisions and I don’t mean to minimize them. My point is that because obesity isn’t a thing, an essence, there is no single medically correct way to think about it. Trying to force it into some medical category— defining what it means to have obesity— muddies things more than it clarifies them. Thinking about obesity has more in common with thinking about bankruptcy than it does with heart disease. In my upcoming book, I describe such things as an “outcome” as opposed to an essence: an end point that can be defined on its own terms, but which does not have any representation at a lower level of analysis (eg in the genes) or in some earlier developmental state. I have an earlier post making this argument in terms of depression here:
What does all this have to do with IQ? People commonly talk about IQ in the same way, especially when they are talking about groups. In an upcoming article in the Journal of Controversial Ideas, the authors argue that immigrants “have” low IQs. (Paige Harden and I have a comment that will be published along with the original article.) What immigrants “have,” is lower scores on standardized IQ tests, for fairly obvious reasons. I don’t mean that in the dismissive way, I am not saying that IQ is “just what IQ tests measure.” Or maybe I am saying that, but I am also saying is that what IQ tests measure— the ability to get correct answers to abstract questions, in the here and now— is an important characteristic for functioning in modern society. But IQ isn’t something that people “have”. Over the century since the concept was invented, we have gotten so used to the idea of IQ as an essential inner quality that we rarely question it, but intelligence is no different than any other behavior we have figured out how to measure and quantify. Like obesity, people differ in their ability answer questions correctly. There are many reasons why some people might not do well on such a test, and many things that might be done about it. But there is no inner IQ that can be addressed in a unified way.
Finally, one of the reasons essentialist thinking about obesity or IQ is so tempting is that they are both “heritable.” In humans, saying something is heritable means that people who are more genetically similar are also more phenotypically similar. This sounds as though it is implying that there is some kind of inner genetic essence, an “innate IQ” that might get passed down from immigrants to their children, but it doesn’t mean that. See my last post. Even the most complex and anti-determinist human traits are compatible with a modest degree of statistical heritability. That is the basis of the first law of behavior genetics. Heritability does not imply essence.
Hi Eric,
This is such an important topic that is rarely understood, let alone commented on, in scientific circles. As Alfred Korzybski stated, "the map is not the territory" and "the word is not the thing." I think the lack of recognition of this distinction is a huge problem in the social sciences, where conventional labels are often treated—at least behaviorally, based on people's actions—as natural kinds.
There is a profound difference between the labels used in physics and molecular biology, which are typically precise and discrete, versus those in the social sciences (which often carry unquestioned assumptions that are not empirical facts). You bring up obesity as a good example. I think this is a much bigger problem than is acknowledged, though.
How about 'disorder' in clinical psychology and psychiatry? The concept of a 'disorder' is a philosophical frame, NOT an empirical fact. The same with 'mental illness' (including psychological suffering in an enlarged category which was, prototypically, just used for phenomena that were caused biologically, could be transferred biologically, and that didn't suddenly vanish through a conversation, for example). And yet, any therapist with any level of real skill knows that a client can enter with a session with a diagnosed 'mental illness' and walk out at the end of the same session no longer meeting the (artificial) criteria for it (Schizophrenia being an obvious counterexample, showing that members of the same category can respond far more differently than the label might suggest on the surface). Often it takes several sessions, and often the same result happens through a change in life / social circumstances, but regardless it's still qualitatively a different experience than that of cancer, diabetes, etc. I think it's deeply counterproductive when we confuse the map and the territory. As start, we could start specifify what's implicit--what X calls Y. E.g. you meet the criteria for what, currently, is labeled as 'obseity' by [the national society of XYZ]. At least that makes clear there is a difference beweetn map-territory. Any disagreements?
P.S. To be fully accurate, there obviously are some Conventional Labels in molecular biology. E.g. 'gene'. The term, 'gene', is still far more empirically based, precise, and reflecting 'reality' than psychological labels though.