Did You Just Say What I Thought You Did?
I know you're probably thinking "What does Marcus care about Body-Mass Index?" And, to tell you the truth, I don't care much about it, except for how did it become so important to so many people and where did it come from? I've always been fascinated by scientific thinking, and its evil opposite twin - pseudo-science - and it occurs to me that telling you the tale of BMI might amuse you, while also giving me a chance to talk a little bit about science, intellectual honesty, and how failure to think critically about problems can result in bizzare social distortions.
Leucippus and Anaxagoras
In the early 5th century, BC, several presocratic Greek philosophers - in particular Leucippus and Anaxagoras - attempted to address the question of dividing things into smaller pieces. They observed that if you took a piece of something, let's say bread, and cut it in pieces, eventually you'd get crumbs. And what if you cut a crumb into pieces? You get increasingly fine dust. What if you cut the dust? Eventually, do we reach a point at which things cannot be cut any smaller, or not? Leucippus and Anaxagoras said that eventually one reached a size beyond which one could no longer subdivide. They gave these smallest particles the name 'atom' - a name that scientists use to this day. But, the greeks of the 400BC did have anything like the capability to repeatedly subdivide pieces of dust; they were, basically, just making it all up as they went along. It was how philosophers did things in those days: they'd announce a theory and sometimes they were right but more often (unless it was something obvious) they were wrong. Anaxagoras, for example, also claimed that The Sun was a burning chunk of iron about the size of all Greece. As you can see - he was off by a bit.
Science hadn't been invented, yet. The scientific method was invented later, and is probably the most important thinking innovation in human history. The scientific method is an repeated process of forming a hypothesis, testing it, weighing the results, and drawing conclusions. Anaxagoras got as far as floating a hypothesis, but didn't have enough information or even any useable way of further supporting or contradicting his hypothesis. Thousands of years later, philosophers of science like Karl Popper would add to the scientific method the idea that you can't even meaningfully hypothesize about something if you cannot concieve of a test that can support or contradict your hypothesis. In contemporary terms, Anaxagoras was "blowing smoke." It wasn't possible to prove that atoms existed until the 1700s when chemists like Lavoisier and Dalton began figuring out conservation of mass despite chemical reactions. Thompson and later Rutherford determined (and demonstrated) the existence of the electron and atomic nucleus and - by 1900 I suppose we could say that Anaxagoras was proved right. Why doesn't he get a lot of credit for having come up with the atomic model? Because there's no possible way that he actually knew what he was talking about. In science, the credit for discovery goes to the person who first understood and could demonstrate something, not the guy who made a lucky guess.
What does all this have to do with BMI? Well, it's important because there are a lot of things that are 'theories' that are just a bunch of stuff someone thought up on a fine afternoon and announced as if they were facts - and there are other theories that are broadly supported by exhaustive evidence. How do we tell them apart? Actually, "telling them apart" is almost the wrong question; there are so many completely ungrounded ideas out there that some of us have adopted a more extreme position of assuming that most of what we are told is probably wrong unless we take the time to do some research and convince ourselves that there's supporting evidence. This position is called "skepticism" and, in extreme forms such as that proffered by the 2nd century Greek philosopher Sextus Empiricus, resulted in the philosopher's withholding judgement about the truth of anything.
Modern "skeptics" generally challenge new ideas when they encounter them, to a greater or lesser degree. I consider myself to be inclined toward skepticism, which encourages me to fill my head will all kinds of trivia, whenever I encounter a new idea and feel obligated to research it before adopting it as part of my world-view. I'm pretty sure that what cemented me into this attitude was studying psychology - a field that is largely dominated by thinkers who appear to have just made stuff up and asserted it as truth. Ever since I read a bit of Freud, I realized that you're just a gullible fool if you believe even a fraction of what you read - especially in the "social sciences" - unless you've confirmed it from a couple of different angles. It's laborious, but if you like to read and do research, it can be fun.
Adolphe Quetelet was a pretty busy fellow, indeed. His greatest contribution was the combination of statistics with the "social sciences." In 1835 he began publishing papers on the topic of "social physics" - a term originally coined by August Comte. Comte disagreed with Quetelet's use of statistics and coined a new term: 'Sociology'. Quetelet's influence was considerable and inspired later social scientists like Alfred Binet, the inventor of the "IQ Test". Binet was part of the gigantic blossoming of the "social sciences" of the mid-late 1800s - a period in which quacks like Freud and Jung reigned supreme, making stuff up as fast as they could publish it.
(Adolphe Quetelet - he looks well-proportioned)
Now we can contextualize my earlier digression about Anaxagoras. Someone like Freud or Jung (or, as we shall see, Quetelet) was not performing science if all they did was pull theories out of their subconscious and print them as statements of fact. They were doing the same thing as Anaxagoras - blowing smoke. (although Freud was more inclined to "snort coke" than "blow smoke" - he was an early fan of the 1980s' favorite reality distorter) Quetelet's legacy to modern life, however, comes to us via a rather bizzare route.
Quetelet's "social physics" included attempts to determine by statistics who was most likely to be a criminal based on educational, cultural, and other factors. At one point Quetelet published what he called the "Quetelet Index"(1) - a table of ideal height/weight ranges. Quetelet's table of ideal height/weight ranges was derived using remarkably poor methodology: he asked his friends whom he considered 'well-proportioned' how tall they were, and how much they weighed. Then, he averaged the results in each height range and set the numbers to a curve. Why did he do that? Because he didn't bother to ask more than a couple dozen people, at most, and interpolating onto a curve was much easier than trying to find a well-proportioned friend who was, say, 5'11 tall. In other words, he didn't have enough data, so he fudged his numbers. Of course, to have produced something with any validity at all, he'd have had to try to ground the notion of 'well-proportioned' in something other than just one researcher's idea of what looked about right. For all intents and purposes Quetelet pulled the numbers out of his backside; that's how "social science" was done in those days.
In preparing for this journal entry, I thought about going through my collection of photos of models, and estimating the bra size of my models, and publishing a "Ranum Index" of ideal bra sizes. But then I realized that would probably take several hours to do and the answer would probably be 34B. So I'll just do like Adolphe Quetelet and say that 34B is the ideal, because, uh - well, because that's about right. How do you know it's the ideal? Because that's what the chart says! Would it be rooted in any kind of medical knowledge? Hell no. Would it be dependent on any kind of understanding of public health? Nope. It would be complete pseudo-science.
In the 1950s, Americans began to get fat. The Metropolitan Life Insurance Company decided to publish an actuarial table about "ideal height/weight" ranges and - guess what they did? They took Quetelet's table because, after all, it was very sciency. I remember, as a kid, they had a height/weight chart in the wall in our elementary school cafeteria - it was Quetelet's chart. So American school-kids circa 1970 were being compared to Quetelet's curve-fitted version of his 'well-proportioned' friends.
(The Quetelet Index, modern update)( Large )
One of my classmates in elementary school was pretty tall and slender, and was sent home with a note that he was "dangerously underweight." At the time I didn't have the word 'bullshit' in my vocabulary, or I might have recognized what was going on. Quetelet's chart was built based on his adult friends and completely failed to recognize that kids undergoing growth spurts sometimes grow out in different ways. The chart probably also suffered from weirdness resulting from the curve-fitting at the top and bottom of the chart. Quetelet probably didn't know anyone who was 6' 4" tall - or 4' 6" either. What's important about this - and why the Quetelet Index is pseudo-science - is there is no actual theory behind it; it's just a bunch of numbers. Basically, Quetelet's Index has only the predictive power to tell you whether or not you're likely to be what Quetelet would consider 'well-proportioned.'
The Quetelet Index is not sciency enough. So, in the late 1970s, some clever person back-fitted Quetelet's Index to a formula.
(The Body Mass Index Formula)
Basically, BMI is the curve that Quetelet used to extrapolate his Index, so that instead of a chart, you now can assess whether or not you're 'well-proportioned' (to Adolphe Quetelet) based on how far you are off the curve. So, if you're currently on the BMI - congratulations - you'd be considered 'well-proportioned' for a Belgian from the 1850s. In the opinion of one particular Belgian from the 1850s. Is there any indication that BMI says something about your health? Nope. Is there any indication that BMI says something about whether you are normal? Nope. The 6' 4" tall ex-special forces soldier that I ate lunch with today is "morbidly obese" according to the BMI - but as far as I could see he looked pretty 'well-proportioned'; I certainly wasn't going to call him "fatty." But the real reason I wouldn't call anyone "fat" or "thin" is because I don't base my notions of 'well-proportioned' on the opinion of a Belgian Quack who lived 150 years ago.
Backing Into Numbers
Anaxagoras may have been correct about atoms, but he didn't know about atoms. Adolphe Quetelet may turn out to be correct about his index of 'well-proportioned' but I doubt it. The important point is that we do not know. Yet, many people are being held up to this nonsensical wild-ass pseudo-science, as if it's valid. A skeptic, however, would respond, "this is unconvincing. Go back and try science and come back with something supported by evidence." Unfortunately, people tend to prefer to accept pseudo-science that seems to be in more or less the right ballpark, like Quetelet's wild-ass guesses do (unless you're short, tall, strong, or lanky).
(Bruce Lee: the old height/weight chart says 'underweight' Well, he is dead...)
In 2006, Spain banned models from performing on a public catwalk unless they have a specific BMI. Thinking as a skeptic and a scientist, we could reasonably ask whether the purpose of this law was to ensure that models were healthy, or whether they were 'well-proportioned' according to an 1850s Belgian pseudo-scientist. If the idea is to require models to be healthy, then the thing to do would be to use some kind of test that can be objectively shown to have some predictive power regarding health. It might, for example, make more sense to say that a model needs to be able to run 3 miles in 17 minutes, do 15 push-ups, and 40 sit-ups, or they are disqualified. That would probably dramatically cut down on the number of cigarettes the models smoke, if nothing else.
Some of you may be thinking "but - BMI appears to be pretty good, in general." You need to re-think that a little bit. If the premise is that BMI somehow has something to do with an individual's healthiness, consider that an 80-year-old who has had a quadruple bipass may have a "good" BMI whereas a 20-year-old paratroop commando may not. Any theory that claims to indicate something about health, which treats an 80-year-old cardiac patient as even being on the same page as a 20-year-old athlete is obviously, clearly, totally wrong. It appears to be pretty good, in general, if you subconsciously limit who you are applying the index to. If there were some kind of body fat measuring system that had any kind of medical/scientific validity it would almost certainly have to treat infants differently from pre-pubescents and pre-pubescents differently from geriatrics. If you see a metric like BMI and it's omitting factors that clearly affect its results, you need look no further: you're dealing with pseudo-science.
This journal entry is not intended to embroil me in a debate about anorexia or eating disorders; that's a completely separate topic. This journal entry is about pseudo-science, how it can become embedded in popular culture, and how embedded pseudo-science can eventually become part of public policy. That's the real tragedy, to me: if society needs to get involved in this issue, they should make sure that the technique they use to determine whether or not the problem is real or the solution works is empirically tied to the problem. Because otherwise, you wind up setting up a public policy that is nonsensical, since it is grounded on nonsense. How many children are going to suffer emotional distress because they were told that they were too thin/fat because they were too far from Quetelet's opinion of what is 'well-proportioned'? If we're going to use something like BMI, we should use something that is grounded in empirical results, not guesses.
Why do skeptics like me want to attack pseudo-science like this? Because people tend to accept misinformation and (as you can see has happened with Quetelet's Index) sometimes the pseudo-science gets so entrenched that it blocks anyone from ever developing any real science in that area. If you think for a few minutes about the factors that might have to make up a useful metric (assuming we actually need one!) of ideal weight, you can see right away that it would be complicated. Why leap to an 'answer' that is easy and wrong? Because it's easy. One of the most important things about being a scientist is to know when to say "I don't know." It's OK to leave a void in our knowledge, rather than filling it with nonsense.
(1)Isn't it amazing what you can find on the Internet?! : Quetelet A. Recherches sur le poids de l’homme aux different âges. In: Nouveaux Memoire de l’Academie Royale des Sciences et Belles-Lettres de Bruxelles. (1832) t. VII.