The New Phrenology: The Limits of Localizing Cognitive Processes in the Brain

by William R. Uttal

William Uttal is concerned that in an effort to prove itself a hard science, psychology may have thrown away one of its most important methodological tools—a critical analysis of the fundamental assumptions that underlie day-to-day empirical research. In this book Uttal addresses the question of whether psychological processes can be defined and isolated in a way that permits them to be associated with particular brain regions. New, noninvasive imaging technologies allow us to observe the brain while it is actively engaged in mental activities. Uttal cautions, however, that the excitement of these new research tools can lead to a neuroreductionist wild goose chase. With more and more cognitive neuroscientific data forthcoming, it becomes critical to question their limitations as well as their potential. Uttal reviews the history of localization theory, presents the difficulties of defining cognitive processes, and examines the conceptual and technical difficulties that should make us cautious about falling victim to what may be a "neo-phrenological" fad.

Genres: Psychology, Science, Philosophy, Nonfiction

276 pages, Paperback

First published January 1, 2001

Reviews

[Rob · Rating 3 out of 5]

In a nutshell? Uttal is arguing that the modern imaging technologies (e.g., fMRI) are toys used by cognitive neuroscientists that are following their theories and using these subtractive methods to come up with supporting data for what is otherwise intractable. In other words, much of modern cognitive neuroscience is on a fool's errand because we really don't even have a working definition for what "thought" and "mind" are and so how could we possibly hope to match up its specific component parts and processes with specific brain regions?

This is a highly technical text, to be sure. (Example: Uttal will throw out a term like "physiological psychology" in contrast to "cognitive neuroscience" without defining the two for differentiation. You are expected to know.)

The underlying thesis here is not that neural imaging is in any way bad or wrong, it is that many researchers are using these techniques in such a fashion that they have not stopped to adequately define the terms they are using or the questions they are asking. Uttal states repeatedly that there is no hard scientific evidence that the brain can be componentized or modularized; he suggests that these localization attempts are futile. (A striking example he gives is how subtractive fMRI was used to provide "evidence" of a "face recognition center" in the brain but how that same brain area showed the same kinds of activation when "recognizing" cars or birds or pictures of places.) Uttal's arguments can be difficult to parse because of their highly nuanced nature; that there is structural specialization within the brain is well-established for many things -- but those "things" are sensory or motor in nature and have no reflection on "cognition". He asks repeatedly: is "cognition" even something that you can define? And if you can define it, is it something that you can measure experimentally? Is cognition directly observable? Or are we limited to observations of cognition's artifacts? Its descendant behaviors?

The book is rigorous and technical; ultimately rewarding but certainly not something to approach casually. (But then again, it is targeted at scientists.)

[Trevor · Rating 5 out of 5]

I’ve finished this tonight and just have to write a review of it. I want to start with a disclaimer – I believe that capitalism has a fundamental bias towards eugenics. It has this because it presents itself as a system based on merit, and yet it is grossly unequal. To explain this gross inequality it relies on there being inequalities in people. That is, the system is equitable, but it is working with people who are deeply unequal – and so, the fairness of the system seems masked by the unequal individuals that make up the system. That inequality is intrinsic to people, they are born with it – and so eugenics is an inevitable consequence of capitalist justification, since capitalism inevitably results in very few having a lot, and the many having very little. If this division is just, then the difference between me and Elon Musk must be very great indeed. A trivial example of this is listening to Donald Trump talk about how fabulous his genes are – but inevitably, the successful are likely to believe one or other version of this modern fairy tale.

Now, that doesn’t mean genetic explanations for psychological differences are rubbish by default – but it does mean that I’m very, very likely to start off cynical and sceptical of such explanations – and the hurdle I will expect them to clear is going to be higher than some new theory of how stars form.

This book doesn’t muck around. Calling cognitive neuroscience ‘the new phrenology’ even made me gasp. And he clearly knows his stuff. I found this book very readable, much more readable than I generally find books on the brain. The reason for this, I think, is because other books that are convinced of a localised, modular brain are forced to talk at length about the different regions and their functions. This guy is essentially arguing that these modules have been overstated and so there is no need for him to do the whole gyro posterior hypo whatsit thing. I am almost prepared to accept what he has to say on the basis of this alone.

He gives a thumbnail history of the ideas of brain localisation and a wonderful explanation of things I probably should have been curious enough to look up myself, like how fMRI works. He sees this as a revolution in medicine that lesser mortals, such as myself, barely even realise has been a revolution. This book isn’t really about the lack of power of the tools available to us to help us peer into the workings of the brain – rather it is about the epistemological problems we have in interpreting what we see when we do look.

In the last chapter he provides 29 hypotheses that underlie the localisation view of the brain, and then 36 hypotheses for the counter view that such localisation is not likely to be verifiable. Really, a good review of this book would just quote those.

Okay, so what is localisation. Basically, it is the idea that our brains construct our psychological life using various modules. So, if there is a mental trait, say, attention, love or boredom, then there is likely to be a brain region linked to this trait. And we have all heard the classics of this – where some poor bastard gets stabbed up the nose with a sword, cutting a vital (but not too vital) part of their brain and end up with the paradoxical disability of being able to do all of the moves of the Hokey Pokey, while not remembering any of the words of the song. Would life be worth living in such a case? Almost certainly not.

He points out nearly endless problems with this localisation idea. The first is that the traits we are looking for are anything but obvious. I’ve given three – attention, love and boredom – but it wouldn’t be too hard to come up with a taxonomy of psychological traits that didn’t include one or perhaps all of these traits. He provides multiple taxonomies, no two of which are exactly similar. It also isn’t all that clear that psychological traits, even if the categories are well defined, might massively overlap. And if they do overlap, that doesn’t necessarily mean there is a simpler way of categorising that avoids these overlaps.

Another problem is in working out which bits of the brain are ‘on’ and which are ‘off’ at any given time. The first problem here is with setting the sensitivity of your detector. You know, if you sent your detector to notice anything that registers more than a three, you might just get lots and lots of noise. If you set your detector to notice when things are more than seven, you might miss all of the signal. And how do you know that a four in this part of the brain is the same as a four in this other part over here? Maybe in this other part, a four is nothing, and a seven is really a four.

The other bit to this is that most of this research works on control experiments. So, I put a rose in front of you. You have your eyes closed. I measure your brain waves. I ask you to open your eyes. You see the rose, your brain fires. I measure the difference between brain waves one and brain waves two. That’s what a rose looks like in brain waves and these are the brain regions that ‘see’ the rose.

Okay, that looks like it makes lots of sense. The problem is that we don’t really know if ‘more blood’ going through a bit of the brain tell us that bit of the brain is actively processing rose-ness. It could be doing lots of other things unrelated to roses, but strongly correlated to opening our eyes, maybe like suppressing a sneeze. Also, if we don’t know what the psychological categories our brains actually operate on are, we are unlikely to determine them from peak blood flow readings. Is there really likely to be a module in the brain that is linked to rosiness? It doesn’t seem particularly likely. But that might be more generally true. Just because we have had a deep sense of having been in love, is that necessarily a feeling we obtain from a distinct set of neural firings? Is the feeling of love a physiological construction in the brain, or might it also be influence by social and cultural factors? People like Pinker say the former – again, as good a reason as any to think it is probably wrong.

He is certainly not saying that thought is not connected to the brain, just that localisation and modularisation are metaphors that need to be justified rather than just accepted. The whole, this bit does this and that bit does this other thing is not supported by nearly as much evidence as people who talk in corpus occipital broca lobe thingo make out.

This guy wants to bring back behaviourism, well, sort of – a new version of behaviourism. His point is much the same as the reason we had behaviourism in the first place. You need to remember that behaviourism was initially a response to eugenics – a series of unsupported and unsupportable ‘truths’ that ‘explained’ why men were superior to women, white to black, rich to poor. Now neuroscience is presented as positive science that proves certain people have better brains and therefore deserve more social resources to be spent on them, that on the rest of us botched and bungled (as Nietzsche liked to refer to us as). As TS Eliot liked to say, although, admittedly, not of brain localisation and its relation to a new behaviourism, ‘I should be glad of another death’.

[Amro Fahmy · Rating 4 out of 5]

3.5/5

[Cheryl Faupel · Rating 5 out of 5]

very informative