A Controversial Approach to Brain Imaging Studies

My first introduction into neuroscience came one day in my graduate course in bilingual education. There, I learned a brief lesson on functional MRIs, or fMRIs, and their ability to detect brain activity. Of course, this brief lesson was tied to the benefits of a bilingual and didn’t speak much to any other brain activity the fMRI could detect. Still, I was intrigued.

From brain autopsies in the late 1800s to specialized MRIs today, we know where to locate different functions in the brain such as the five senses, speech, and motor skills. Yes, it was incredibly exciting to me, but it was not surprising that neuroscientists were taking the next step to mapping all 10 billion neurons in the brain. What did surprise me was another researcher who believes that technological advances, specifically brain imaging studies, will never give us a complete understanding of how our brains work and their use is hurting the field of neuroscientific research.

William R Uttal argues that brain imaging is not a sufficient study tool, and we must bring theories of behaviorism into the study of neuroscience and cognition. Right now, as I have not read his Mind and Brain: A Critical Appraisal of Cognitive Neuroscience, published last year, I can only rely on a review on Metapsychology Online Reviews. Uttal does not dismiss the anatomical discoveries of the brain, useful in medicine, but he does dispute the application to all cognitive processes.

In an article published in 2004, Uttal acknowledges that functions such as senses, motor, and emotion can be localized, that is, they have specific locations within the brain. He argues, though, that these are basic functions, whereas “high-level cognitive processes” are not. Such cognitive processes can not be localized as they take place all over the brain. Brain imaging is the go-to method for neuroscientific research, but Uttal believes that the method is in fact very unscientific: when synthesizing multiple research studies, Uttal demonstrates that replication is not there. On average, either researchers have not tried to validate previous studies, or they were unable to.

Uttal also explains that the brain is too complicated to be analyzed; rather our current technology is unable to study neuronal interaction, so researchers are left to hypothesize and request more money for research with limited research methods. One final argument Uttal gives is the need for neuroscientific research to be purely scientific. Most research studies are considered quasi-scientific, meaning that there are variables or outside influences that can't be controlled. In a recent interview, Uttal talks about the errors of neuroscientific research not being purely scientific:

In my new book, in which I deal more with meta-analyses, there was a fascinating thing that happened. I started listing and tabulating the number of potential biases and errors that could get into analysis of that kind. And before I was done, I had four pages of potential errors, potential confounds, potential mistakes. Some of them were interpretive, some of them were statistical, some of them were mechanical, some of them were design errors in the experiments.

I found Uttal's arguments interesting. I'm especially interested in reading his analysis on brain imaging and his newest book, due to be published next month, Reliability in Cognitive Neuroscience.

What I'd really like to know are his thoughts on the newest brain imaging equipment being used by the Human Connectome Project. Would he argue that it too is not specialized enough to study high-level brain function? Would he criticize the study has being too advanced for our current scientific limitations? I searched for any discussions Uttal might have had on the subject, but I haven't found any yet.

I would also be interested in his response to the study from Switzerland exploring the specificity of neuron synapses. Is their technology also not sufficient, and, therefore, their study, faulty?

I do agree that our brains are highly complex. That fact is what most excites me about the field of cognition. But I'm not sure I can agree with Uttal when he says that fMRIs are not useful to neuroscientific research and that most results of that research are no more than a hypothesis proving another hypothesis.

I'll be requesting Uttal's current book through inter library loan, so hopefully, I'll be able to do my own review soon. His new book isn't due out till next month, so that one will have to wait!