I’ve been having an interesting discussion with Michael Egnor, a neurosurgeon and member of the creationist organization the Discovery Institute, over the question of whether the mind can be explained as arising purely from physical processes of the brain. Egnor does not believe it can be, and claims that neuroscientific research supports his position. My position has been that he fails to understand this research and there remains no evidence that some functions of the mind can only be explained by “immaterial” processes or entities. The most recent of my articles on this subject, as well as links to the first two, can be found here. Egnor has now responded to my articles via the Discovery Institute’s “Mind Matters” blog, so let’s have a look at what he has to say.
Most of the article is devoted to a discussion of the definition of “abstract thought”, which need not detain us long. Nothing he says there is particularly difficult or controversial. However, he then goes on to say I do not understand this concept and, in support of this claim, refers to a figure I had cited from the following paper:
The global landscape of cognition: hierarchical aggregation as an organizational principle of human cortical networks and functions.
Now, as I had said in my earlier post, I do not claim to fully understand this very complicated paper. I mainly mentioned it as an example of the complexity of the research that Egnor has to dismiss or ignore in order to assert his claim. And that he dismisses it is clear from his response. He obviously made no attempt to understand the paper, if he even read it at all. Instead, he merely claims, “It is noteworthy that the very research Ali cites is research on concrete sensory processing by the brain, which we all agree is material. The research is not on abstract thought.” That Egnor misunderstood the paper is made evident by reading this plain English summary. Understanding the neural correlates of abstract thought is the whole purpose of the paper! What confused Egnor was the reference in the graph (which I suspect is all he read of the paper) to cortical areas of the brain designated by sensory modalities. This emphasizes how woefully out of his depth Egnor is here. His familiarity with the literature seems to be largely limited to research that was radical and cutting edge – in the 1940’s. There have been a few advances made since then, he might be interested to know.
At this point you might be wondering why you should care about this disagreement between two non-neuroscientists regarding the subject of neuroscience, and I wouldn’t blame you. What would be most helpful would be to get the views of some actual, you know, neuroscientists. Well, as luck would have it, one of this month’s issues of Scientific American (Vol. 321, Issue 1) has a cover story on the very topic we are discussing, entitled “How Matter Becomes Mind”, by Max Bertolero and Danielle S. Bassett. It is likely still on the newsstands as I write this, or can be purchased thru SA’s website. So my advice is to stop reading now and, instead, get your hands on this article. OTOH, I realize not everyone is going to do this, so I’ll do my best to summarize my own understanding of this field of research.
Max Bertolero 
Danielle S. Bassett
Still here? OK, then. The difficulties involved in neuroscientific research have not just arisen from the complexity of the brain. They also arise from the very nature of how the brain operates. It has no moving parts. Its workings cannot be directly observed in the manner that we can observe the processes by which muscles and tendons exert forces on bones and joints to cause our limbs to move, or by which the heart contracts to force blood through the circulatory system. In this sense, the brain is something of a “black box”. Or, at least, it has been through most of human history. The relationship between anatomical structures of the brain and sensory, motor and cognitive functioning has historically been investigated through observing the effects of specific injuries or lesions to the brain. Research on non-human animals has also been very important in this regard, but is limited by the fact that many neural structures are unique to our species.
Another line of investigation, upon which Egnor relies heavily, is the neural stimulation techniques pioneered by Wilder Penfield and others in the 1940’s and 50’s, in which discrete areas of the cerebral cortex were stimulated in the brains of conscious patients, who could then report what they experienced.
In recent decades, however, neuroscientific research has been revolutionized by the advent of functional neuroimaging techniques, such as PET, SPECT and, most recently, fMRI. These techniques allow the visualization of regions of activity within the brain, which can then be correlated with specific stimuli that are presented to the subject, or tasks that the subject is asked to perform. This has allowed a more detailed mapping of the cortex in terms of the areas that are involved with particular sensory, motor and cognitive modalities, as illustrated below:
Most of those modules’ names are self explanatory, though should not necessarily be taken too literally (e.g. when a blind person is reading Braille, interestingly, it is processed by the visual modules). Each of these modules tend to function independently of one another. However, more complex tasks and functions require cooperation between two or more of these modules, and understanding the nature of these interactions has led to the discipline of network neuroscience. This entails applying graph theory, a branch of mathematics that deals with the relationships within networks of objects, to data from functional neuroimaging. In this way, a map of sorts can be drawn of the physical and functional connections between various points of the brain that may not be physically contiguous. This is illustrated below:
Each circle is termed a “node” and represents a neuron or set of connected neurons. The lines connecting them are (somewhat confusingly) termed “edges” and, in this context, represent a connection between nodes that is demonstrated by synchronization of their firings either at rest or when the subject is performing a task. As you can see, there are certain nodes thru which different modules are connected to each other, and these are called “hubs”. These are of particular importance in more complicated cognitive tasks. The simplified schematic above is illustrated more accurately by the stunning Diffusion Tensor Image below, which shows the white matter tracts in a human brain
Isn’t that pretty? That image comes from the website of the Human Connectome Project, a collaboration that has been ongoing since 2009 and endeavours to map out the functional and anatomical connections that exist in the human brain. It’s a big project, as the synaptic connections in the human brain number 300 trillion.
So returning to to Michael Egnor’s questions: How is it that, in Penfield’s experiments, it was never possible to produce a complex abstract thought? And why do “intellectual seizures” not exist? Neural connectivity theory provides an answer: Such complicated cognitive tasks require precise and specific interactions between multiple areas of the brain, coordinated through the hubs that connect the various functional modules of the cortex. When a surgeon stimulates part of the brain with an electrical probe, or when this occurs spontaneously through an epileptic seizure, usually only a single node or set of nodes in a single module of the brain is stimulated. This will provoke specific physical or emotional sensations, evoke a specific memory, cause movement of part of the body, etc. However, such stimulation is too crude and localized to produce a more complex response such as an abstract thought. In their article, Bertolero and Bassett employ an extended metaphor in which the brain is compared to a symphony orchestra. A single musician can play a single line on his instrument, and the other musicians in his section may do the same. However, to perform a symphony requires precisely controlled and coordinated interactions between the various sections of the orchestra.
For Egnor to insist that we have already learned all we need to know from the experiments of Penfield and their like is the equivalent of telling physicists that they should limit themselves to spinning buckets of water and throwing rocks off of towers, and forget about using particle accelerators. It is the equivalent of telling geneticists to stick to learning about Mendel’s experiments with pea plants, and just pretend Watson and Crick never discovered the structure of the DNA molecule and the Human Genome Project never happened.
Egnor tries to dismiss this research by invoking what he calls the “pleonastic fallacy.” I have to admit, I had never heard of that one. On looking it up I find that it appears to be a fallacy concocted by a single writer to deal with a particular argument made by philosopher Daniel Denett and, moreover, may not even be a fallacy at all. In any event, this supposed fallacy does not apply to the line of argument I am making here. Egnor reports an observation (“Stimulation of discrete areas of the cortex does not cause complex abstract thoughts”) and arrives at a hypothesis (“The brain does not produce complex abstract thoughts”). However, there is another hypothesis: Complex abstract thoughts requires the coordinated actions of multiple discrete areas of the brain, in a specific manner that cannot be replicated by randomly stimulating parts of the brain with an electrode. This latter hypothesis is now supported by an increasing body of empirical evidence. From this, the only reasonable conclusion is that Michael Egnor’s claim that the book is now closed on this question is, at best, highly premature.
Another way of looking at the question: If one accepts, as Egnor does, that the correlation between stimulation of the brain and a particular mental function suggests that the brain is responsible for that mental function, then this should also be true when it is demonstrated, by fMRI, that activity in regions of the brain is correlated with a particular mental function.
As I complete writing this, I notice that Egnor has just posted yet another response. His latest argument is easily dismissed:
Dr. Ali contradicts himself. First, he argues that it’s no surprise that cutting the brain in half and separating the cerebral hemispheres, as Sperry did, has no effect on abstract thought. Then he argues, in connection with Penfield’s seizure research, that abstract thought is the wholly material result of delicate complex connections within the brain: (“what is needed [for abstract thought from brain connections] is a [metaphorical] trained concert pianist”).
In short, he argues that it’s no surprise that cutting the brain in half has no effect on abstract thought, and then he argues that delicate complexity and interconnection is essential for abstract thought. Dr. Ali is debating Dr. Ali.
No, no contradiction. That abstract thoughts require complex interactions between various areas of the brain does not mean it requires complex interactions between every part of the brain. If abstract thought remains after the corpus callosum is cut, this may merely demonstrate that abstract thought does not require an intact corpus callosum.
This is not very complicated.
Interesting series. Thanks.