My comments on “Consciousness and Neuroscience”

Crick has done neurobiology a service by taking on the challenge of trying to find specific types of neuronal activity that are correlated with consciousness. This project has dramatic philosophical implications because so many philosophers have thought that issues like understanding consciousness are beyond the power of materialistic reductionistic science. “Consciousness and Neuroscience” is not much different after 8 years from “Towards a Neurobiological Theory of Consciousness” (1). This in itself could be taken as good evidence that most neurobiologists are correct when they think that it is premature for them to study consciousness. This sense of prematurity is amplified by the fact that there are only a few rare sources of data that Crick and Koch have so far found useful for analysis of the “neuronal correlate of consciousness”: 1) cases of human brain damage that alter consciousness, or results form lesioning studies in animals, 2) Non-invasive brain scanning methods like functional MRI and electrophysiological recordings from animals which are used in attempts to identify specific types of neurons with activity patterns that match the perceptual states of the animals. Some times these methods can be combined with method #1, and 3) electrical stimulation of specific brain regions that seem to be able to shift an animal’s conscious perception of a stimulus.
In the first part of this article, Crick and Koch summarize some recent results from these kinds of experiments.

When Crick took up the challenge of trying to identify a neuronal basis for consciousness, he must have thought that there was a real possibility that something relatively simple like 40 Hz feedback loops among thalamic and cortical brain regions would solve the problem. This was not a silly idea, as some philosophers have suggested, only overly optimistic. The most interesting parts of “Consciousness and Neuroscience” are the indications that Crick is now admitting that he will have to widen his search strategy and accept the fact that consciousness is a tougher problem to solve than the structure of DNA.

A main reason for the broadening of Crick’s analysis of consciousness is that simple tricks like the 40 Hz feedback signals have not solved the problem. Crick now accepts the idea that additional brain regions (such as prefrontal cortex) beyond the early visual sensory processing regions are important for even the most basic forms of visual sensory awareness, while the 40 Hz signals seem largely restricted to “lower” sensorymotor brain regions.

In a section called “Future Experiments” Crick and Koch issue a call for additional experimental techniques that could aid in the search for neural correlates of consciousness. It has recently become possible to reversibly inactivate specific populations of neurons using genetic engineering tricks. It should be possible to extend these methods to the primates used for consciousness research. Crick and Koch also point out that recent understanding about the molecular controls of the patterns of connectivity between neuron populations should lead to methods for specifically eliminating types of brain connections. This is a good idea, but it is interesting that “virtual elimination” of connection pathways is easy in computer models of the brain, and is one of the powerful tools available to people who work with computer models of the brain. Crick has never shown much interest in computer models of the brain, probably because he has hoped that solving problems like consciousness will be simple enough that we will not need to use computers as tools for aiding our understanding of how brains work. I suspect that he is overly optimistic about how easy it will be to understand things like consciousness in neurobiological terms.

Crick and Koch have provided a few short sections where they mention philosophy of mind. A section called, “Why are we conscious” deals with the idea of “zombies”. In a section called “Philosophical matters” they deal with qualia and “The problem of meaning”. Their general attitude towards philosophy of mind is reflected in the statement, “In recent years the amount of discussion about consciousness [by philosophers] has reached absurd proportions compared to the amount of relevant experimentation.” However, it is interesting that they thank Chalmers and Searle for helpful comments concerning this article. As a biologist, I am fairly comfortable with what Crick and Koch have to say about zombies and qualia, but I doubt that it is very satisfying for many philosophers.

In the section called, “The on-line system” Crick and Koch say, “It is obviously important to discover the difference between the on-line system [of vision, as can be shown to operate in people with blind-sight], which is unconscious, from the seeing system, which is conscious.” Contrast this with what they say near the end in discussing what will happen once neurobiologists have found a neuronal explanation for consciousness, “It is likely that scientists will then stop using the term consciousness except in a very loose way.” They compare this to the disinterest of biologists in debating the question of whether viruses are alive or dead. This is a very weak analogy by which the fact that viruses are in the fuzzy gray region between alive and dead is taken to imply that the word “alive” is not really of much use to biologists and that the word “consciousness” is similar in this regard. This is very misleading and misses to point entirely. The word “alive” is not usually found within the literature of biology because most of the time the distinction between what is alive and what is not alive is clear to everyone. We would no more expect to see an economist pausing frequently to point out to his audience what elements of an economic system are money, he and his readers all know this. What elements of ecosystems are alive, what elements of economies are money, and which aspects of brain activity are the neural correlates of consciousness are all issues of fundamental importance.

What I find most interesting in this paper is that Crick and Koch show signs of moving in the direction that Gerald Edelman proposed for consciousness research 10 years ago. One key aspect of Edelman’s theory of consciousness was that he proposed that consciousness is a global brain function, meaning that many widely distributed parts of the brain are important for consciousness. As mentioned above, 10 years ago Crick and Koch were pursuing the idea that a mechanism for a simple form of consciousness, visual awareness, might be found with just a few brain regions, the closely linked thalamic and visual cortex regions. Thus, it is remarkable to find Figure 1 in this article showing whole series of sensory and motor regions and even the environment. This is the kind of diagram which is found throughout Gerald Edelman’s work. The point that Crick makes with Figure 1 is that there are many possible control systems in a brain for taking sensory input and producing adaptive behavioral responses based on the sensory input. These systems seem to exist in a hierarchy from very quick and unconscious to conscious systems that, while slower, can be more refined. What is the basis of this “refinement” in behavior that is made possible by consciousness? Crick and Koch are reluctant to say the words, but the key is learning and memory. The brain contains several memory systems which allow animals to combine current sensory inputs with past experience to produce adaptive behavior that makes sense for the animal in terms of the environment that the animal exists in.

Why are Crick and Koch so reluctant to admit that the problems of learning and memory are fundamentally involved in consciousness? Basically, because learning and memory are the real “hard problem” of mind. Crick has hoped that there is some simple trick that brains use to produce consciousness that can be found without us first having to understand learning and memory. However, I think Crick is wrong. I favor Gerald Edelman’s approach to consciousness which places memory squarely at the center of the problem of consciousness.

This fundamental issue (the relevance of memory mechanisms and learning in our attempt to understand consciousness) is most directly (and it is not very direct at all) confronted in the section called “The problem of meaning.” I agree with the idea that the problem of meaning has two aspects: 1) how is meaning expressed in neural terms?, and 2) how does the expression of meaning arise? These are the two issues at the heart of Gerald Edelman’s theory of consciousness. The answers to these questions are simply 1) memory and 2) learning. Importantly, Edelman’s approach is to tackle the two together, since memories are the result of learning. I think Crick is admitting (very quietly) that the unconscious-consciousness hierarchy (see Figure 1 of "Consciousness and Neuroscience") is defined by the memory mechanisms that are involved in each level of the hierarchy. Crick is admitting that understanding the neuronal correlate of consciousness would be a sterile result in itself because what is really important to consciousness is meaning. A person can be conscious of sensory inputs, but if those experienced inputs have no meaning for a person, then no sensible behavior will result, you would have a zombie that has low-level awareness, but nothing more. It is hard to see the distinction between such a meaningless conscious existence and unconsciousness. The distinction between consciousness and unconsciousness depends on the meaningfulness of conscious experiences and memory is the source of meaning. Edelman saw this all clearly, and I think, in the end, Crick will have to admit openly that Edelman was correct. This is hard for Crick to do because 10 years ago he issued a scathing attack on Edelman’s theory of mind and Crick went off in a different direction. It is at least promising to see Crick finally starting to open the door towards his acceptance of the importance of memory and learning in our attempt to understand consciousness. I see the same glimmer of hope in Rey’s CRTT, in which he admits that there must be a mechanism for getting meaning (semantic content) into his LOT. Memory and learning are tough problems for neurobiologists, but progress is being made. It is not too much to hope that the neurobiology of learning and memory will soon begin to allow us to start to understand consciousness. Edelman was the first to show how this can work in theory, only the details of the explanation remain to be filled in.   Back to first Crick page.

1. Francis Crick and Christof Kock "Towards a Neurobiological Theory of Consciousness" (1990) Seminars in the Neurosciences, 2: 263-275.

Go to John's Book Page.

Go to John's Home Page.