An unnecessary defense of neuroimaging (comment on Paul Bloom)

Paul Bloom has an article in Seed today lamenting the sway brain imaging research holds over the public and media compared compared to other less-feted areas of psychology. The response in the blogosphere has been passively favorable, so I thought I’d try to provide a spirited defense of the opposite view. I should mention that I conduct most of my research with fMRI, so I make no pretense to be impartial (but I do pretend to have reasons for studying what I study!) I should also mention that I have an enormous amount of respect for Bloom (for both his popular and scientific output), so none of this should be construed as personal criticism of any sort.

The crux of Bloom’s argument isn’t really that there’s anything wrong with imaging itself (though he does seem rather unimpressed with much of what it’s produced), just that the excessive focus on it detracts from other areas of cognitive science and psychology:

… [Imaging] is more than just phrenology. But it is not so dazzling that it should usurp other areas of research.

I’m not sure Bloom’s advocating for anything like an actual reallocation of the resources imaging receives, but for the sake of argument, let’s suppose he is. His characterization of the current state of affairs is, I think, completely accurate: it’s true that the media disproportionately focuses on fMRI studies, that fMRI research gets more funding than cognitive psychology research, and that people ooh and aah over pictures of brains lighting up in ways they don’t over eloquent verbal descriptions of Methods and Results sections. But I don’t think this is much of a problem–not for the media, not for the public, and not even for cognitive psychology. Here’s why:

• It’s far from clear that the exaggerated media focus on fMRI research takes anything away from the rest of psychology. While it’s certainly true that the number of stories covering imaging research has gone up in recent years, I don’t see any reason to suppose relative coverage of other areas of psychology has gone down (can anyone furnish empirical data?). It’s not as though in the absence of fMRI, reporters would be climbing over each other to report on new tweaks to the ACT-R model, or publishing expositions on brave new approaches to transformations of RT distributions. And there certainly are areas of (mostly applied) psychological research that continually pique the media’s interest. Cell phones and driving performance are one example; parental influence on childrens’ behavior are another perennial favorite (and some of the reporting on parent-child relationships makes fMRI coverage look positively Pulitzer prize winning!)
• A large number of neuroimaging researchers (likely the largest contingent) are trained as psychologists, and have at least a cursory familiarity with other branches of the field (most commonly cognitive). As such, a good deal of imaging work aims to address issues first posed by cognitive psychologists, or is informed by cognitive psychological models. My guess is that if anything, public awareness of major issues in psychological research has increased as a result of the focus on neuroimaging. Naturally, the sophistication of the cognitive theories presented in imaging papers doesn’t rival that of full-length cogpsych articles. But it shouldn’t be expected to. Criticizing cognitive neuroscientist for not getting the psychology perfectly right is on a par with the frequent charge leveled by cognitive neuroscientists that many of the explanatory models psychologists propose aren’t neurally plausible and seem at odds with current knowledge of biology. It’s true enough, but isn’t particularly bothersome, because biological implementation usually isn’t a major concern among cognitive psychologists. Conversely, the level of abstraction cognitive psychologists strive for isn’t necessarily an objective cognitive neuroscientists share.
• Neuroimaging is a rapidly growing field. While it may reach saturation in the next few years (it’s hard to say right now), it’s worth remembering that PET and fMRI have only been around 20-30 years, and fMRI has only been widely deployed for about 10 of those. Setting aside the question of whether imaging offers better information than traditional cognitive psychology, it pretty clearly offers different information. Given that we’re only scratching the surface so far in terms of what we can expect to accomplish with fMRI, it’s not unreasonable to increase imaging funding until, say, every major university in the U.S. has a research scanner (probably still a decade or so away). That’s a level of support that’s pretty much on par with other branches of expensive science. Moreover, neuroscience is a much larger field than academic psychology (SFN draws annual crowds of 30,000; APA around a third of that), and it’s not unreasonable to expect cognitive neuroscience to eventually end up somewhere intermediate to the two disciplines in terms of size and resource allocation.
• There are several other areas of science that attract as much or more media coverage as imaging research. Cosmology and population genetics come to mind right away. And once again, developmental psychology (a field Bloom himself has made important contributions to) isn’t that far behind. So it’s probably unfair to focus attention solely on neuroimaging. Sure, people are wowed by any mention of the brain. And as Bloom observes (and has been instrumental in demonstrating), biological explanations have a stronger allure for most people than purely psychological ones do. But genetic explanations are pretty pervasive and compelling too these days, and that’s no reason to stop doing genetic research. What’s important is whether the results advance our understanding of the world, not whether lay people find them interesting or not.
• Bloom argues that the study of reaction times in psychology has told us much more about the mind than the collective field of neuroimaging. Well, maybe and maybe not; there’s really no good way to operationalize that statement. I certainly know several people who would lean in the opposite direction, but simply asserting as much doesn’t make it that way. More to the point though, it would hardly be surprising if the study of RT has produced more knowledge; after all, people have been studying RTs for a hundred years and taking pictures of the brain for only a couple of decades. If anything, the fact that the comparison between the two is even worth making could be taken as evidence that cognitive neuroscience has come a long way very quickly.
• Aesthetics, aesthetics, aesthetics. Having partaken in an embarrassingly large number of “my discipline’s better than yours!” arguments over the years (I don’t mean to suggest this is what Bloom’s saying. I don’t think he is), I’ve come to the conclusion that much of what drives people to choose once scientific discipline over another is just an aesthetic preference for some levels of explanation over others. There’s rarely any objective way to arbitrate between the utility of different scientific disciplines (of course, one can always appeal to practical application–but in that case virtually all of psychology fares poorly compared to most of the natural and biomedical sciences). Instead, what one generally finds is a two-way street where people working at a higher level have predictable criticisms of those working at a lower level, and the converse. Philosophers of mind think psychologists are obsessed with details and overlook the really important problems; psychologists think much the same of neuroscientists, with the added concern that neuroscientists worry too much about physical implementation and lack sufficient regard for theoretical models; cognitive neuroscientists think all of the above of molecular or systems neuroscientists, and so on. When pushed in the opposite direction, the criticism tends to focus on how nebulous the constructs are at the higher level and how they lack any realistic implementation in biological structures. Well, who’s right? Everyone and no one, as far as I can tell. But I’d love to hear a good argument in favor of disciplinary superiority, if anyone knows one. (Note that I don’t extend this relativistic stance outside of science; I think there are plenty of good reasons for privileging the scientific approach over, say, literary theory).

Having said all this, I’m well aware of the irony implicit in writing a long piece in defense of a discipline that’s already hogging the limelight. And I’ve certainly experienced my fair share of frustration at the poor quality of media coverage of neuroimaging studies. Still, I think the focus on neuroimaging is justified, for the most part, and I don’t see how it detracts substantively from the rest of psychology (though it may bruise a few egos).