I, Not-I and the Mysteries of Consciousness

Can a computer philosophize? What exactly is a neural correlate of consciousness? How do you obtain a picture directly from the brain? In carrying out a radical transformation of the human body, where should we begin? We put these and other questions to Viktor Argonov, Ph.D. physicist and mathematician, and a composer.

 
2045: Viktor, you recently visited the “Toward a Science of Consciousness” conference in Stockholm, which was devoted to issues related to consciousness. What were your impressions?

Viktor Argonov: Several hundred specialists from dozens of countries took part in the conference, and diverse fields of scientific knowledge were represented: philosophy, neurophysiology, medicine, physics, “normal” psychology and even religious scientists. The presentations and discussion topics were so diverse that representatives of one group often did not understand the research topics of other groups. However, despite the significant divergences in the specialists’ views, I did not see anyone seriously accuse anyone else of practicing pseudoscience or take the opposite view and accuse someone of “a lack of spirituality”.

The topic of consciousness is so complex that no generally accepted canonical theory has yet been formulated that could be used to refute all other points of view. People are searching for one. I admit that some people succumb to the temptation of proposing simple and barely scientific answers. But instead of judging them, it’s better to propose worthy alternatives yourself. For me, the most important thing is that the majority of participants understood that a wealth of unresolved, fundamental issues remain in the study of consciousness. That is very important, given that many people in the “standard” world of science tend to deny even the very idea that such problems exist.

 
2045: Could you give a short description of what exactly the problem of consciousness is and how it differs from the problem of intellect? Why do some scientists deny the problem of consciousness?

V.A.: The intellect is the objective capability of a system to perform some complex behavior. Consciousness is subjective impressions, emotions, etc. Some scientists believe that the subjective sphere should not be studied by science at all; that it is enough to study intellect alone. But of course that leaves behind a whole range of questions.

The first issue is that of the exhibition of consciousness. A person has both consciousness and intellect, but it does not follow that every intelligent system is conscious and that every non-intelligent system is unconscious. Does a rock feel? A robot? A hydra? No one knows. And many ethical questions are interwoven here. Should an embryo be considered a person? Should animals and robots of the future be treated humanely? At what stage does life begin to have ethical value, and does that value always remain constant?

The second problem is finding the neural correlate of consciousness, i.e. the parts of the brain that contain precise information about the subjective world of a human or animal. We consider humans to be conscious beings. But how, when studying the brain, can you tell which exact pictures a person is seeing at a given moment, or which smells and tastes he is sensing, or what thoughts he is having? If you discovered the neural correlate of consciousness, you would be able to record hallucinations, dreams, thoughts—that is, make the subjective world of a person fully (if materialism is true) or partly (if dualism or idealism are true) accessible to observational study.

The third, more frequently occurring problem is that regarding attributes. We all know what the difference is between the colors red and green, but we aren’t capable of explaining it in words. Many people assume that at their core, attributes cannot be reduced to verbal or digital information in the brain. They appear to us to be something fundamentally elemental, something basic. And if you locate the neural correlate of consciousness, then what will hold particular interest is the way in which information about “redness”, “greenness”, “sourness”, and so on, is encoded.

The fourth problem that I would identify is the integrity of consciousness, its indivisibility. This concept was developed back in the time of Descartes. A person is a large entity. The brain consists of an array of parts and perhaps in experiments is in fact divided. However, there is only one “I”, and any other conscious being is already “Not-I”—the division is fundamental. Sometimes Siamese twins with a joined brain have one integrated consciousness, sometimes two (two different personalities). What determines whether it will be one way or the other? In both cases the nervous system is combined—the difference is only in quantity. It can be supposed that it’s not the brain that is fundamentally indivisible but the neural correlate of consciousness. There must be some variable that sets a strict boundary of where my consciousness ends and where that of another being begins. Strictly speaking, that variable has not yet been identified.

 
2045: Has progress been made in the study of these issues?

V.A.: The most widely recognized achievements are in the area of the neural correlate of consciousness. For example, in recent experiments they’ve managed to obtain data on the content of mice dreams. There are also grounds to suggest that the correlate of subjective time is a gamma-rhythm frequency. However, we are still far from being able to obtain a picture directly from the brain, let alone one with information about subjective attributes. The problem of attributes has not in fact been resolved. For the most part, other problems have not been resolved either. There were many hypotheses proposed about this at the conference, but the majority of them make an array of assumptions that remain unproven. I simply don’t think we have enough experimental data at the moment to build an authentically scientific theory of consciousness.

2045: Tell us a little more about the topic of your speech at the conference.

V.A.: My presentation related to the first topic I named. I proposed a non-Turing approach to the exhibiting of consciousness in a computer. Alan Mathison Turing’s famous test is able to determine whether a robot’s behavior is intelligent. But, as I said before, it does not hold that all intelligent behavior is conscious. I propose an alternative test, in which a computer is tested for the ability not just to behave intelligently but to understand issues of consciousness. It must not have any knowledge of philosophy obtained from books and discussions either. If, with training, such a computer suddenly begins to ask philosophical questions about consciousness and correctly understands them all, I contend that this is grounds for considering it to be conscious.
 

2045: You are known not only as a scientist but also as someone who works in the field of music. And in your opera “Legend of the Unrealized Future”, there is a very similar scene, in which a computer that controls the government holds forth on consciousness.

V.A.: The difference is that the ASGM (Automatic System of Governmental Management) has access to human knowledge. It’s for that reason that even without having consciousness, it can imagine the general outlines of the problem. And it’s entirely possible that, given its intellect, it can do so even better than people can. But it doesn’t take my test correctly, because it obtained a great deal of knowledge from human civilization ahead of time—it didn’t reach the point it did on its own.

2045: What did your colleagues make of your presentation? Your ideas are simple and therefore probably conceal pitfalls of some kind. You are basically proposing a test that is even simpler than Turing’s: Want to find out whether a computer has consciousness? Talk to it about it.

V.A.: Yes, on the level of ideas, everything is rather simple. And certain philosophers, such as Avshalom Elitzur and David Chalmers, used to take a serious approach to them, revealing the paradox of phenomenal judgments: consciousness is subjective, but statements about it are possible in the objective world. Every book about consciousness is a material object. And it must be explained which physical processes led to its appearance. That is the thread that leads most directly to consciousness. For some strange reason, most people are not thinking in that direction. Such an approach was new to many people at the conference, though it was greeted with interest.

2045: Two prominent participants of the conference, Stuart Hameroff and Roger Penrose, are advancing the idea of quantum consciousness. How tenable do you and other participants consider that idea to be?

V.A.: The main problem that quantum theories of consciousness solve well is the problem of attributes. Adherents of these theories propose the hypothesis that “redness” and “greenness” are quantum states of some kind in the brain. Hameroff and Penrose, among others, presume that coherent quantum states in microtubules of nerve cells are responsible for consciousness. But at the base of their theories is their own exotic version of quantum mechanics, which has not yet been verified. Many view it with skepticism. Certain other researchers are advancing more cautious hypotheses that do not require long-lasting coherent states and that are based on completely traditional quantum mechanics. In any case, thorough research is needed in this area. Quantum mechanics lies at the core of a contemporary portrait of the world, and it’s entirely possible that it has some relation to consciousness.

2045: How do all these issues relate to the idea of radically transforming the human body?

V.A.: For me as a transhumanist, the study of consciousness is not only a philosophical problem but a practical one as well. After all, transhumanism presupposes a transformation not only of the body in the typical sense of the word but also of the psyche. But how can you do that without understanding its natural foundation? Which parts of the brain can you replace with artificial ones without risk of destroying personality? It’s impossible to answer that question without studying the problem of consciousness. I think that before initiating the radical cyborgization of the brain, the neural correlate of consciousness must be located. Does it have a physical or purely informational nature in the form of neurosignals? Is there a group of neurons that is directly responsible for consciousness? Or perhaps consciousness is produced by still smaller elements within neurons? How fragile is the neural correlate of consciousness? Can it be transferred separately into a robot, or does a significant piece of the brain have to be preserved whole and inviolate?

2045: There is an opinion that it’s sufficient to copy the information in our brain and transfer it into a robot.

V.A.: If you do that, there will be two people who are identical in terms of the information they have, two “copies”. And with that, paradoxes of indivisibility arise. Stanislaw Lem described these well in his book “Summa Technologiae”. After all, there can be only one “I”. I would prefer to see ahead of time—before I begin to be copied—what will stand “before my eyes” after the experiment: the same thing, or that which is in front of the robot? Which of the copies will be the authentic “me”? Many people don’t see a problem in this at all. They say that there will be two “me’s”. But I am convinced that that is simply a distortion of terms that masks a lack of understanding of the problems of consciousness. Personally, I am more struck by the perspective that it is not “pure” information that is responsible for the sensation of “I” but rather the actual material of certain neurons. But that idea also needs to be substantiated.

2045: Perhaps the immaterial soul is responsible for consciousness?

V.A.: Unlike obstinate materialists, I won’t deny such a possibility. The neural correlate of consciousness must be located, and then what exact role the material setup of the brain has in shaping consciousness will be visible. If it turns out that certain parameters of sensations are independent of it, then those parameters are not reducible to physical material.

2045: And lastly, a question regarding artistic creativity. Will your new opera, “The Little Mermaid” (http://argonov.ru/mermai.html), also include such philosophical issues?

V.A.: One issue that comes up is the idea of enacting a technological transformation of a living being with the goal of achieving immortality; questions of utilitarian ethics also play a part. The theory of consciousness will not be touched on in particular, since The Little Mermaid, in contrast to ASGM, clearly does not possess it. But for that very reason, she possesses not just abstract theories but also personal motivations, which engender significant ethical issues. Her way of solving these problems is a kind of challenge to standard human moral behavior. But, generally speaking, she is the kindest of the characters, because she loves people.