“Neurobabble” vs Real Science

8 Comments

• Olivier Morin 21 December 2010 (13:58)

  I share your perplexity. After reading Burge's piece, many [i]New York Times[/i] readers will probably retain that psychology is divided into two antagonist fields: whacky neuroscientists and serious vision researchers. The two fields have nothing in common: brain scientists, for example, have nothing interesting to say about vision, because "Brains-do-not-see ; people-do". Of course, you and I (and Tyler Burge) know that this is wrong, wrong, wrong. We know that vision scientists and brain scientists have been making reciprocal contributions to each other's work for decades. So, why is Burge spreading these misconceptions in the [i]New York Times[/i]? Well, again, no mystery here. He objects to the success of fMRI research. And here, he has a point. There are many excellent scientists who use fMRI as a tool to answer clever psychological questions. But some practitioners of fMRI use it for the sake of it, because they know it makes their research sexy. And it works well with funding agencies. Back to your question: Should you be happy with Burge's article? Well, weirdly enough, I think you should. Some day, maybe soon, you may be facing a funding board with a clever but hard to sell project, for which you want to collect behavioral data before forming anatomical hypotheses. Imagine. That day, you are competing with a team who wants to use fMRI to "See which areas of the Brain Light Up when we read James Joyce". Wouldn't it be nice to know that someone in the jury read Tyler Burge's piece?

• Davie Yoon 27 December 2010 (18:03)

  Thanks Olivier for your comment. I do agree that Burge seems to have a specific beef with (certain instances of) fMRI research. However, I wonder if his message would be most effective directed at scientific funding agents or science journalists. Many (hopefully most!) neuroscientists would agree with Burge that society does not benefit from research that consists of a poorly specified question plus brain pictures. As these same neuroscientists are part of grant peer-review panels (e.g., for NIH agencies), it is unlikely that such topics as the MRI of love or James Joyce would be judged valuable contributions worth funding. Presumably, Burge would be pleased with neuroscientists who would take such a stance -- such neuroscientists would stand hand in hand with Burge and with their colleagues in behavioral psychology. So where should Burge most profitably direct his ire? Perhaps the right target is not funding agencies (for the reasons above) or the many thoughtful cognitive neuroscientists that go beyond the caricatured "MRI of X" experiments. Maybe the people he wants to reach are science journalists -- and he would certainly not be the first (e.g.,Neuroskeptic,Neurocritic,Bad Science. Here, I'm with Burge, 100%. I'm just not sure the other directions he moves in the article are the right ones.

• Olivier Morin 28 December 2010 (10:23)

  So this is merely about bad journalism? There is [i]no funding from agencies[/i] for "poorly specified questions plus brain pictures"? Let me pick the very first example that springs to my mind. Check this study that was funded by the National Institute for Medical Health (note the complete absence of psychiatric relevance) (the paper is herehere). Bad journalism is not to blame, for once, since the people at Washington University wrote the press release themselves. My memory is vague. Maybe the study was not as useless as the authors made it appear to be. Maybe there really was a point in using fMRI this time. But, considering the cost of fMRI, and the wealth of methods in the field of mental imagery, cheaper methods were probably available to get to the same (flimsy) result. Only thing is, they wouldn't have attracted so much attention from the general public and (yes!) money from the NIMH. (and thanks again for a great post! Belated XMas wishes)

• Davie Yoon 22 January 2011 (00:36)

  Hm. I think I will take a more vigorous stand for the proposal that neurobabble is propogated more by poor popscience journalism or web discussion and less by funding agencies and scientific practitioners. Not that funding agencies and practitioners are blameless here -- I think the problems highlighted by Burge and others show the need for better scientific communication. But still. Olivier: you linked to Speer et al, 2009, Psychological Science: "Reading Stories Activates Neural Representations of Visual and Motor Experiences." Without defending the authors' specific methodology or soundness of interpretation, let me defend the underlying research question here, because I don't think you would actually feel AS infuriated about the misuse of federal funding on this project if the theoretical context were clearer. Maybe you would still be slightly aggravated and if you were on a panel, you would vote against funding such projects -- fair enough. But I don't think your ire would rise to Burge-ian levels. This project can be said to address theories of embodiment/simulation. I think many philosophers find this topic worthy of study and debate, as evidenced by the debate over mirror neurons and simulationist/embodied accounts of action understanding (e.g., the Interdisciplines debate moderated by Sperber and others). Here the theory is taken to a more abstract level -- instead of asking how one would represent/comprehend an action on observes visually, the question is how one represents/comprehends a written account of an action? The mental health implications of this research include social communication disorders where action understanding may be impaired and disorders of semantic understanding where interpretations of communication (written or otherwise) become degraded -- so things like autism and Alzheimer's. Now I'm not sure that the authors would agree with my characterization of their work, and possibly if I read their papers more thoroughly, I would have a different take on the situation. However, I wanted to take a stab at the specific example you so gamely threw my way. So here we have an example of a project you and possibly Burge and others may disdain as "neurobabble", likely from casual descriptions of the project in a press release or blog post. Yet a more careful reading and richer theoretical context could give a different picture. Therefore, I think we have found an example of propogation of neurobabble by journalists/bloggers rather than funding agencies/researchers (although you can still complain that they didn't explain the research clearly enough). Reaction? By the way, we can add Rober Scruton to our list of philosophers who take issue with neuroimaging work. Although he prefers the term "neurotrash" to "neurobabble". I think it's all very well and good for Burge, Scruton and others to remind us not to overinterpret neuroimaging findings. But the level of ire and apparent targets of this ire seem misguided.

• Davie Yoon 22 January 2011 (17:03)

  I was web-stalking Stephen Kosslyn, who is joining Stanford this year, and came across the following passage which I think says more elegantly how neuroimaging approaches can contribute to theories of mental imagery. Kosslyn writes about his book, Wet Mind: [quote]It took 9 years to write another book. The problem was that the pure cognitive science approach--at least for me--had run its course. It just wasn't clear how to answer fundamental questions about mechanisms that underlie cognition with the tools at hand (behavioral data, logical analysis, computer models, etc.). Somewhere around 1985, I came across a now-classic paper by Roger Tootell and his colleagues, illustrating the topographic structure of primary visual cortex in the monkey brain. I was amazed to see a picture of what the animal had been looking at literally projected on its cortex. It had never occurred to me that literally depictive representations in the brain might exist. (Ignorant me: This had been known at least since the early 20th Century.) Of course the mere fact that certain brain areas are physically organized topographically does not necessarily imply that the representations in those areas depict information, but it would be a surprise if they didn't (and they do in fact turn out to be functionally depictive). The idea of mapping cognitive activity into the brain afforded all sorts of new ways to inform theory. I called this the "Wet Mind" approach--taking advantage of the idea that "the mind is what the brain does" to make use of what is known about the brain to understand the mind. [/quote]

• Davie Yoon 23 January 2011 (16:33)

  Just this last tidbit, because it hits right on the nose, Seeing a thing makes you think about it, say experts: [quote]SEEING a picture of someone doing something makes you think about the thing they are doing, according to new research. Scientists at the Institute for Studies have finally established that when human eyes see a thing the brain will often generate a thought that is in some way related to the thing that has just been seen. Professor Henry Brubaker said: "We applied the seeing-thinking forumula to smoking and found that it followed exactly the same pattern. "We got a bunch of smokers together and showed them a picture of a cigarette. We asked them if this made them think about cigarettes and they all said 'yes'."[/quote]

• Steven Gross 23 January 2011 (17:53)

  The following might go without saying, but just in case it's useful: As Morin notes, Burge of course does not think that neuroscience plays no role in explaining psychological phenomena. He does think, however, that neuroscience on its own cannot [i]exhaustively[/i] explain, for example, perception. On his view, there are aspects of explanation that are proprietary to psychology -- for instance, those that involve ineliminable appeal to representations and their functional role. You'll find lots and lots of references to neuroscientific work in his book Origins of Objectivity. (Indeed, his son Johannes is a vision scientist who's published in neuroscience journals.)

• Olivier Morin 23 January 2011 (22:39)

  First, let me restate what we have been and will be agreeing upon: the criticisms of Scruton and Burge are weird caricatures that don't resemble the science we both know. Deep down, Tyler Burge knows that there is a lot of excellent neuroscientific work, like the neuroscience of 25 years ago cited by Stephen Kosslyn, that have rendered many services. That is not their target. Their target is the current craze for overpriced, over-hyped, pointless neuroimaging studies. They are not always cautious enough to distinguish the two, or to signal the existence of some excellent neuroimagers. Since they are well-informed and clever philosophers (as Steven Gross signals in the case of Burge), there is only one conclusion left: their identity has been usurped, as so often happens on the web, by a couple of trolls who used their name to publicize outrageous claims. It happens, and to the best of us. Now, you defend Speer et al.'s paper by noting that it relates to three topics that seem to hold many promises for other fields and for society at large: Embodiment, mirror neurons, and a cure for action comprehension disorders like autism. Davie, I have a confession to make. For a short while, I did some work in neuroscience on precisely these topics. And I came back thoroughly convinced that talk of embodiment-mirror-neurons-and-a-cure-for-autism is the purest expression of Neurobabble/Neurotrash ever. This paper by Gregory Hickok is, I think, a definite demonstration of the bankruptcy of the most popular theories surrounding mirror neurons (see also here). I cite it because he takes two pages to criticize a paper that I wrote, and every criticism he makes is perfectly accurate. This is just to say that I feel competent enough to judge, and also that I take my share of the blame. Yes, my friend! Heaven forgive me, I am a sinner like everyone else. But you don't need to be told this. You know it all. You know Vicky Southgate's fantastic paper debunking mirror-neuron theories of autism. We both know it: we have never met a neuroscientist who could say sincerely that their work on mirror neurons will help cure autism, unless they're severely deluded. Davie, would it be possible that your identity was stolen by a troll who wants to give ICCI readers a very bad image of neuroscience, by damning it with faint praise? I suspect so. Because the Davie Yoon I know would not think for a second that Speer et al.'s selling points are anything but a lot of hot air. Agree? By the way the mirror neuron craze is a case in point to answer our question. Was the craze generated by neuroscientists, or by bad journalists? I would say: both, but scientists sparked it. Need proof? - This is Prof. Ramachandran talking about mirror neurons - This is Prof. Iacoboni, one of the main defender of mirror neurons, producing the worst kind of science journalism (according to [i]Nature[/i]) without any journalist. And there's plenty more where those came from. People with a sincere hope of finding a cure for autism are pouring their precious money into research on mirror neurons. Do you know any neuroscientist in that field who actively discourages them? No? Why? Can journalists take all the blame?