What does a brain scan discover

Brain scanner or behavior?

Why do psychology and other social sciences even need brain research? My thesis: The silver bullet for human science is still behavioral studies

To my article on misophonia or selective noise intolerance (misophonia: the next psychological disorder?), Which sparked lively discussions in the Heise forum, I also received some reactions by e-mail. Among them was Gerheid Scheerer-Neumann, a psychologist and professor emeritus for elementary school education from the University of Potsdam. She wrote:

In one of my advanced training events, a participant reported enthusiastically about a study that was able to prove the success of an intervention in the case of reading and writing disorders in the brain as well. But: EVERY learning changes the brain, whether we can measure it or not!

Gerheid Scheerer-Neumann

Everyday trivialities

In my article I have the discovery of British neuroscientists that unpleasant noises in the cortex in the brain of noise-sensitive people lead to higher activity than trivial, even tautological criticism. After all, reactions in connection with unpleasant emotions are very often found in this area of ​​the brain. Above all, the researchers' claim that they have discovered nothing less than "the brain base of misophonia" is dubious.

I have been criticized here and there for the tone of my article. However, these critics do not seem to have a problem with the fact that some scientists acquire the morale of a used car dealer. Be that as it may, Ms. Scheerer-Neumann's email gave me the idea to think a few steps further.

Reasons for Neuro Popularity

Under what conditions do psychology and other social sciences need brain research to confirm their findings? I think the popularity of neuroscientific procedures like functional magnetic resonance imaging (fMRI) in these disciplines is due to two reasons:

First, some behavioral scientists suffer from a lack of self-esteem because they (supposedly) do not do "hard science". Second, researchers must increasingly raise funds in the competitive science market; And would you guess into which area such funds have been increasingly flowing since the 1990s, the decades of the brain in Europe and overseas?

Example language knowledge

To make my answer more specific, I came up with an example. Two groups learned language skills: Group 1 used index cards for this. Group 2 flashcards while classical music was playing. The test subjects learned the same length of time and had to take a language test at the end. Based on known literature, we expected music to promote learning. The result is as follows:

Money for the brain scanner

Thanks to this result, we feel encouraged to submit a research proposal for EUR 100,000. Half is for a doctoral candidate (which is available on a special offer in Germany) for the rough, the other half we need for a few preliminary examinations and then the study in the fMRI with 2 x 20 test subjects. Now let's consider what this fMRI study can bring.

In Scenario 1 despite serious efforts, we find a behavioral effect again (N = 40, p <0.01, Cohen's d = 0.81), but no difference in brain activation between the groups when taking the language test. We can rule out methodological errors. Is there therefore no important difference between the learning processes?

Nonsensical neuro-realism

This is actually how a so-called neuro-realist would argue. Neuro-realism is a pattern of reasoning that Eric Racine and colleagues have found repeatedly when examining neuroscientific science communication. In her words: "Observed patterns of brain activation are accordingly presented as the ultimate proof that a phenomenon is real, objective and effective ..." (Soc. Sci. Med., 2010, 71, p. 728; my translation)

It is just like the press office of the University of Nijmegen once advertised an in-house neuroscientific study on burn-out: Now that the researchers have found differences in the brain between people with and without burn-out for the first time finally proven that the disorder really does exist.

You probably already guessed it: Neuro-realism is nonsense. Psychology and other social sciences have more than a century of methodological research behind them. The effects that can be found with this and with the help of clean experimentation do not need to hide behind any apparatus, no matter how expensive - and also not behind so-called "hard" sciences (which can only describe the world with the help of statistics).

So let's come on Scenario 2: After preprocessing the fMRI data, our doctoral student presses the evaluation button and after a few minutes we see a yellow-orange cloud of statistical values ​​superimposed on anatomical images of a brain.

The activation cloud

We don't need a lot of imagination where such a cloud could be: We have linked listening to music with a language learning task. The upper turn of the temporal lobe (lat. Superior temporal gyrus) or the Sylvian furrow (lat. Lateral fissure) will probably be there. Hooray, we have brain activity!

Before anyone gets me wrong, brain studies can of course be interesting. The starting question, however, was what this adds to psychology and other social sciences.

Differentiate systems

One could imagine that there are two different systems for learning in psychology, system A and system B. These systems are associated with different locations X and Y in the brain. If you now find stronger brain activation in place X, then could that would be a reference to system A.

But let's be honest enough that the minimal differences in the oxygen content of the blood that we make visible with fMRI tend to encourage speculation. And how was that again with the comparisons? If the activation at X is greater than at Y, then that at Y is less than at X. Perhaps the latter is much more interesting.

Find the mechanisms

Markus Knauff, professor of psychology at the University of Giessen, recently described a good example of how psychology and brain research can interlock in the exploration of rational thinking in Brain & Mind. And he got by completely without neuro-exaggeration. An exception as pleasant as it is rare.

Or, of course, you can always say: It is a legitimate research question what the mechanism of a psychological process on the underlying level. But who says that this level is neuroscientific and that the complex psychological process of learning is not broken down into its individual cognitive-emotional parts with further behavioral studies? They are also available almost free of charge.

Puzzles of the brain

It is the case that the (mostly speculative) fMRI findings raise new questions rather than provide answers. Last but not least, this has to do with the open basic questions about the measured BOLD signal. As an exception, I quote myself (will brain researchers soon be able to decipher dreams?):

You have to be aware that the researchers' approach, for example using the ideas with the receptive fields, is quite clever, but we are still dealing with a very coarse-grained signal, the BOLD- measured with the fMRI. Reaction (see my post brain work).

The researchers here have already used a better resolution than the average (10 instead of 27 cubic millimeters), because after all they only needed to measure part of the brain (especially the occipital cortex with the primary visual areas) and were able to work with fewer and thinner layers .

Ten cubic millimeters of cortical tissue still contain 200,000 to 1,000,000 neurons, of which an average value is then recorded every second. You shouldn't expect too much accuracy from the model and I think the researchers have already extracted the maximum amount of information here.

It is doubtful whether merely increasing the spatial / temporal resolution will solve the problem. From a neurobiological point of view, the BOLD signal is just a correlation with cell activation, which could be electrically measured more precisely by several dimensions.

Stephan Schleim

Neural secrets

True to the motto "attack is the best defense", some neuroscientists try to divert attention from such uncertainties with linguistic eyewash. Talking about the "neural correlates" or even the "brain base" of a psychological process, as our misophonia researchers have just done, is often showing off.

And a much more general problem: the neurons don't tell us what they're doing by themselves. Conversely, neuroscience often needs psychology to understand how the brain works.

Prima psychologia - psychology first!

It is time to conclude. Of course there are burning questions for brain research. But then you have to formulate it on the neuroscientific level, which is not so trivial at all. The common practice of just looking to see if you can find a difference somewhere in your brain, around which you then knit a story, is a waste of time and money.

Psychology and the other social sciences are - from the point of view of knowledge - very seldom dependent on the neurosciences, rather the other way round. Since the majority of human perception, feeling and thinking is expressed in behavior, this is still the focus of observation.

One cannot, but does not have to ask about the mechanisms of this behavior; even if you do that, it doesn't necessarily lead you into brain research. Instead, you should think of another junction, namely that of the application! In contrast to the ixte brain activation study, it can be of great benefit to transfer a stable (for example: learning) effect into practice.

Ceterum censeo I would like to close with a quote from Joseph Dumits, a neuro-anthropologist from the very beginning. Freely translated he wrote:

If we know that a meditation technique helps students learn, for example, then we should consider how to teach them to students instead of wasting money on research into the neural processes involved.

Joseph Dumits

P.S. To avoid misunderstandings: The example with classical music was made up, if it was not already apparent from the text. But there really is research on this.

This article also appears on the author's blog "Menschen-Bilder".

Read comments (43 posts) https://heise.de/-3629658Report an errorPrint