If I had been writing this blog a few years ago, I almost certainly would’ve reviewed Moral Minds by Marc Hauser. In his excellent book, Hauser makes a persuasive case for understanding the human capacity for morality in similar terms to our capacity for language. Using Noam Chomsky’s ideas about a universal generative grammar, Hauser attempts to identify the basic patterns of how we acquire a framework of moral judgment from our culture. Just like with language, he contends that the unique and distinct aspects of different moral systems are products of historical contingency and innumerable cultural variables. To simplify his point, if we look past these superficial differences, Hauser says that we will find a small number of variables that, like so many knobs, are set to a specific value by the culture of our upbringing. Ultimately, the existence of these ‘knobs’ is partly explained by conditions in our evolutionary past.
An interesting, if controversial thesis, to say the least. Accordingly, it made a big splash when it came out in 2006. Although Hauser was already quite prominent and recognized in his own right, Moral Minds came with some impressive endorsements from the likes of philosopher Peter Singer and Nobel Prize winner Daniel Kahneman.
Unfortunately for Marc Hauser, his appearances in the news in the last few weeks have not been so flattering. From what I can piece together from different reports, a former student has come forward and accused Hauser of inventing fraudulent experimental data and bullying other researchers working on the experiment into covering it up. After an internal investigation, it turns out the claims were founded and Hauser had committed academic fraud. He’s accepted responsibility for the “error” and has taken a year-long leave from his position at Harvard.
Pretty shitty. In context, the experiment in question was a limited, albeit prominent, study of primates’ response to musical tones. Its outcome has very low relevance to his central arguments in Moral Minds. And yet it all smells a bit fishy. Fudging data to stay a leader in the field seems pretty low.
And apparently it’s way more common than I imagined. By coincidence I was reading a book this week called Wrong by David H. Freedman. It’s an enchanting tale about the many ways that the media experts, pundits, public academics and various gurus (management, investing, etc…) manage to fuck up so spectacularly, so often. When I picked up the book I was pretty dubious that he could convince me that ‘most’ experiments and expert advice were wrong. Of course, I was mostly wrong myself. While I remain a little soft on what fraction of experts are actually dripping with fail, Freedman makes a strong case.
The institutional design of scientific investigation in the Western world lends itself to a great deal of bad science, fraud and hype. Human psychology has a lot to do with the demand for the particular varieties of flashy science that get churned out. And, of course, the human brain is full of plenty of biases that prevent us, even after extensive training and talent, from shitting the bed, so to speak. There are plenty of forces tugging us from the truth and a rare few humans can reliably evade them.
Freedman argues that we put too much faith in studies that pretend to show more than they do. Experiments on animals extrapolated to humans, peer-reviewed journals – even the gold standard of research: random, double-blind controlled studies – they all have significant flaws that can and do lead us down the wrong path. These flaws are known to us, and yet remain remarkably common.
Of course, the same can be said of almost every other domain of human experience. We misperceive and misinterpret a great deal in all that we do. That’s part of being human being – a humbling limitation of our abilities is never far away. But it’s not just the limitations of human knowledge manifest in ‘experts’ that we have to blame. Much of the cheating and pandering done by scientists is bolstered by the manipulative press and the gullible and rapacious public. From one point of view, there is a perfect storm of forces that lead scientists and experts to say ridiculous things, only to earn fame, prestige and wealth.
Long story short, the mix of human cognitive biases, our terrible (really, really terrible) intuitive grasp of statistics, our increasingly sensational media, our highly competitive academic institutions (and limited funds for research) and the widespread desire for quick fixes to complex, counter-intuitive problems adds up to one big recipe for mistakes and deceptions.
Freedman adds that not all experts are wrong, and certainly not all are wrong all the time. But more often or not, there’s plenty of reason to be deeply skeptical. Wrong is well worth its price if all it does is re-affirm the need to take all advice with a grain of salt. Fortunately it is also a shockingly entertaining read, given that it could be accurately described as a book about research methodologies (everyone’s favourite course in university…)
As for Hauser, I’m not prepared to come down too hard on him.* As I hope to discuss in a later post, one-off assessments of character don’t get us very far. He may have cheated this time (and earned himself a healthy dose of disgrace as a result) but I don’t think this invalidates his entire body of work. The weird thing about science is that getting caught cheating can have the paradoxical effect of strengthening the legitimacy of Hauser’s other work, providing that it stands up to the inevitable scrutiny it will now face.
* With one exception. This whole unpleasant episode adds yet further proof to my unassailable philosophical theory: people with goatees are not to be trusted. I think Freedman might have a goatee-lite, which I will accept on the condition that he does not nurture it to mature size.