Tuesday, October 4, 2011

Brain Regions, Errors of Logic, and Willie Horton

Though the past few decades have yielded a wealth of information about how the brain works, there is much that remains mysterious. It therefore came as a surprise to many a neuroscientist when the hallowed New York Times published an editorial by "marketing expert" Martin Lindstrom claiming that we loved our iPhones in much the same way we love our families and our dogs.
I'd leave it to my fellow neuroscientists to explain why there isn't a shred of evidence to support that claim, except that it exemplifies a logical error pops up everywhere from neuroscience to politics. In the interest of helping you be a better observer of both of those things — and because I like writing about them — here's the claim, the problem, and the broader lesson.

The claim
Lindstrom used a method called functional magnetic resonance imaging, or fMRI, to study what parts of the brain respond to smartphones. He had eight men and eight women aged 18 to 25 watch videos of iPhones and listen to their characteristic sounds. (If you've even been around an iPhone, you know the ones I mean.) His claim: iPhones activate the insula, a brain region he writes is "associated with feelings of love and compassion." Therefore, he goes on, we love our iPhones.

The problem
Let's break this down. First, what does fMRI do, and what does it mean that the insula is "associated with" love and compassion? FMRI is a brain scanning technique that measures how blood flows in the brain, and it's most often used to measure how different brain regions activate in response to everything from emotionally-charged images to trading in simulated financial markets. Then, to say the insula is associated with love and compassion is to say that evoking feelings of love and compassion activates the insula.

What inferences can we draw? Despite the connection to love and compassion, just because someone's insula responds doesn't imply that person is in a loving mood. The insula responds to a host of other things, among them disgust and the violation of social norms, and love activates other parts of the brain, including the caudate nucleus and the anterior cingulate cortex. Not only is it wrong to say that insula activation implies feelings of love, it's also wrong to say those feelings are somehow centered in the insula.

In fact, the most current viewpoint is that networks of brain regions are responsible for emotional responses, not to mention your capacity to do arithmetic and follow baseball. Your response to smartphones is not somehow "in" your insula — it's in a complex pattern of responses that extends across your entire brain.

Logic and the broader lesson
Lindstrom's argument illustrates a common logical error: the belief that the converse of a true statement is true. 

Consider the statement, "if an animal is a person, it is also a mammal." Even though that's true, it's converse, "if an animal is a mammal, it is also a person," is obviously not. In Lindstrom's case, the underlying, reasonably well-supported claim is "something that evokes love activates the insula." Lindstrom infers that "something that activates the insula evokes love," but as the animal example illustrates, that doesn't necessarily follow.

If this kind of error seems harmless, it's not. Back in 1988, a political action committee that supported presidential candidate George H.W. Bush ran the now-infamous Willie Horton ad, focusing on a furloughed prisoner in Massachusetts who used his furlough weekend to escape and commit a series of crimes. Apart from the racist overtones and anti-Dukakis message of the ad, what the PAC wanted people to believe was that furloughed prisoners were going to rob your store and rape your women. See the flaw? One example, from which they hoped you'd generalize.

So when candidates start parading successful small business owners who were raised in a ditch or Canadians who've waited 28 years for a kidney transplant, ask yourself this: is this what always happens, or is this just one example?


  1. Nathan, look at the relevant letter in yesterday's Times –– or, better, this uncut version. It looks as though many people are with you.

  2. Thanks for link, John! I was not the first to comment on the neuroscience by any means; there was a splendidly-titled piece in Wired the day after Lindstrom's piece came out. One thing that I wanted to do — that others hadn't — was point out how this is an example of broader, quite common error of generalization.

  3. I wonder if you could comment on claims by researchers that experienced meditators show high activity in brain-regions 'associated with compassion', e.g. http://www.bbc.co.uk/news/health-16389183. Do they mean the insula, or are they looking at the more complete picture of networked areas being activated? Thank you