Critique of Cartesian dualism II: Daniel Dennett’s multiple drafts theory

To call René Descartes a father of Western philosophy is no exaggeration. His contributions to philosophy and mathematics dominate our thought, ranging from his famous declaration ‘I think, therefore I am’ to his construction of the Cartesian coordinate plane. But what were once considered foundational paradigms are becoming increasingly becoming frustrated by advances in contemporary neuroscience. From Antonio Damasio’s Descartes’ Error, to Daniel Dennett’s disfigurement of the Cartesian theater in Consciousness Explained, to the common neurologist’s knowledge of the banality of the pineal gland, modern sciences have begun to push against Descartes’ philosophy. This essay series will discuss recent developments in neuroscience and psychoanalysis, and elaborate on a common thread that reveals widespread epistemological breaks with Descartes. What does this break mean for the future of Western philosophy in general, and cognitive sciences in particular?

Daniel Dennett’s multiple drafts theory

Dennett’s entire argumentative approach in his book Consciousness Explained is directed against the stubbornly-intuitive concept of the Cartesian theater. Recall that Descartes posited a kind of internal observer, a homunculus (Latin for ‘little man’) which represents the central integration site for all sensory experiences; hence the allusion to a theater. This is certainly corroborated by our everyday experience of reality. Thanks to what Kant called the transcendental unity of apperception, the gamut of our phenomenal experiences collect towards a singular point of observation; we have a single experience of a red, flaming hot, steaming lump of coal, not any kind of disjointed experience of redness, hotness, and seeing steam.

Dennett pushes back against this notion by invoking a series of clever thought experiments, complemented by empirical findings from contemporary neuroscience. The first among these is known as the phi color effect. Its parent effect, dubbed phi in the early 20th century, is a well-known visual illusion whereby two dots flashed in quick succession approximately 4 degrees away will appear to the viewer to be one single dot moving horizontally. You can try this out for yourself here: I had the best effects with 0 ms t preamble, 150 ms t disk, and 50 ms inter-disk, with the distance cursor set somewhere above the ‘s’ in reset (make sure to make both dots red at first). Our visual processing system seems to be filling in the gap between the two dots with an illusory experience of motion. Now, in the variation known as the phi color effect, we make the two dots different colors. Let’s say the first dot is red and the second dot is green. We already expect subjects to report seeing one single dot moving back and forth horizontally; how will the change in color influence this illusion? Under the same conditions as the above experiment, subjects report that the illusory, horizontally-moving dot changes color midway through its purported course.

Dennett notes that at first glance, this finding involves temporal contradiction. How could our brain give us the experience of a change in color midway through when the actual change in color doesn’t occur until after the second dot has appeared? First, let’s start with the basics: whatever the brain is doing to give us the green-midway illusion, it has to be doing it after the green dot occurs — there’s no information ‘coming from the future’ here. With this, Dennett rules out any magical precognition on the part of the brain. So then, some kind of revision is happening in our brain to give us the illusion of seeing a green-transforming dot before it ‘reaches’ the other side.

Here, Dennett distinguishes between two kinds of potential revisions. The first, dubbed an Orwellian revision, occurs post-experientially: there was a time T when you observed X, but milliseconds afterwards, the memory is replaced with an intrusive image of Y, making you report that you actually saw Y at time T. Like the news censors in George Orwell’s 1984, an event which occurred is later tampered with. The second revision, which he calls a Stalinesque revision, intrudes itself within the experience itself: your observation of X was contaminated with Y at time T, which is what makes you report that you actually saw Y at time T. This is similar to Stalin’s show trials, in which the proceedings were completely fabricated in real-time.

At a macro level this distinction between post-experiential and intra-experiential revision seems reasonable; Dennett will argue that when it comes to the brain, this is not so.

Let’s return to the phi color effect. The Orwellian answer to the paradox: how could the subjects report seeing green midway, before the green dot flashes on the other side? would be to say that after the red (A) and green (B) dots are flashed, someone in the edit room is unpleased with the experience and inserts the illusory images of the dot moving between and changing into green (C and D). You did have a genuine, conscious experience of A then B, but this editorial revision happens so fast that by the time you make your verbal report, it is influenced by the illusory inclusion of C and D. Alternatively, the Stalinesque answer says that some kind of ‘delay in consciousness’ occurs, a blackout which allows the edit room to create the sequence A C D B then proudly present it to your conscious experience. There was no uncontaminated experience of A then B: the presentation of the whole sequence was constructed by an unconscious editing process then presented to your conscious experience.

Essentially, we have the following distinction between the two accounts: the Orwellian revision involves a subjective (i.e., conscious) experience of A B, which is immediately edited into a memory of A C D B, while the Stalinesque revision contends that the editing happened ‘behind the scenes’ to bring you the subjective experience of A C D B in the first place. Of course, both accounts agree that the subject gives the illusory report “I saw A C D B”, but the Orwellian account says that a conscious experience of A B existed at some point in time, whereas the Stalinesque revision posits no such uncontaminated experience.

So, what does Dennett say about this distinction? That there is no possible way to distinguish between the two, and that therefore, both are relying on an erroneous conceptual notion, namely, that of a ‘Great Divide’. And here is the crux of his Multiple Drafts theory of consciousness: there is no meaningful definition we can give of the precise ‘moment of consciousness’ when it comes to a particular mental representation. This is also the final blow to the Cartesian theater, since that does posit the existence of a singular ‘observer’ of mental process who could say that there was a singular moment of being conscious of something.

Essentially, Dennett is saying (to a hypothetical defender of the Orwellian revision account): it doesn’t make any sense to say that you were conscious of A B at time T, since something happened immediately afterwards that made the memory of your consciousness that of A C D B. There is no such thing as a conscious moment which is nonetheless inaccessible to every later subjective experience: to say that would be a bizarre position which Dennett dubs the ‘objectively subjective’: “the way things actually, objectively seem to you even if they don’t seem to seem that way to you!”. A defender of the Orwellian revision would insist that there was a time when he was conscious of A B, but of course has no evidence for that claim. Dennett says that we cannot simply posit some consciousness of X and then say that the effects of have vanished: consciousness itself *is* the influences that a state has on subsequent actions and reports.

This makes the Multiple Drafts Theory an operationalist theory: Dennett himself calls it ‘first-person operationalism’. For Dennett, the ‘reality’ of consciousness is nothing apart from its later effects on the subject. It’s not as if we have a subjective experience of being conscious of something, and this experience can (or perhaps, if it’s revised in post, can not) influence other states. Consciousness itself is the thing which influences the later states. He puts it this way: “there are no fixed facts about the stream of consciousness independent of particular probes.” By probes, Dennett means any objective measure which tries to actualize the subjective state — asking a subject for a verbal report is a common example.

My recent experience at a concert will make a nice example here. The Voidz were playing songs from their second album, which I hadn’t listened to yet, and at one point lead singer Julian Casablancas responded to a fan swinging a neon glowstick around by saying, “this next one’s called neon propeller”. At the time, I didn’t know it was a joke, and I figured the fan was making that motion as a symbol for the song. Since the song was one of my favorites, I made sure to listen to it afterwards; when I wasn’t able to find it, I realized that it wasn’t actually the name of the song. I’ve since listened to every song on the album, but I still don’t know exactly which song was the one I misattributed the name ‘neon propeller’ to.

So the question is: what happened to that conscious experience of listening to song X while attributing to it the title ‘neon propeller’? The standard interpretation would say that although that subjective experience exists, it is no longer ‘accessible’ to me. But the Multiple Drafts theory would encourage us to think of the experience not as some positive kernel which is hidden in my mind, but rather, the series of influences it has since had on me. This is the essence of operationalism as a theory of consciousness: we cannot speak of the existence of consciousness in and of itself: it is defined as the measurements and influences it has on later states.

Dennett gives another well-known example in psychology to illustrate his Multiple Drafts theory. In a phenomenon called metacontrast, subjects who are flashed a disk for 30ms which is immediately followed by an outer ring report seeing only the ring: the memory of the disk seems to have disappeared.


Let’s rehearse our Orwellian/Stalinesque paradigm here: the Orwellian revision theorist would say that the disk was in your conscious experience, but was later deleted, while the Stalinesque revision theorist would say that the outer ring somehow delayed consciousness of the disk. By contrast, the Multiple Drafts theory would describe the momentary sensory impression of the disk as a functional position of the later, updated ‘draft’ of the outer ring. By this conception, it doesn’t make sense to ask when the subject was ‘conscious’ of the disk, as an illusory ‘moment of consciousness’ does not exist. Rather, the disk existed in some point in time in a purely information-processing sense, insofar as it contributed to the later ‘draft’ which presented the mental representation of the ring.

Dennett is fond of thinking of the mind in this operationalist manner: the multiple drafts theory is first and foremost an information processing theory which views consciousness as a series of information-organizing processes which are later measurable by verbal report, memories, and so on. For this reason, he stands on the A-side of Ned Block’s distinction between A- and P-consciousness, which comprises the functional-analytical aspect of consciousness. In fact, he believes that P-consciousness — what we would call the ‘raw feeling’ of subjective experience — is nothing more than a particularly stubborn illusion of immensely complicated A-consciousness. Block’s distinction between these two, as well as Dennett’s conception of qualia, are interesting topics for the study of consciousness and will be the subject of future posts.

Although I have my disagreements with Dennett’s scientific reduction of subjective experience, his arguments in Consciousness Explained are some of the most powerful and intriguing ones against the notion of the Cartesian theater. Anyone who wishes to defend the idea of a singular observer in the mind must at the very least contend with the examples that Dennett puts forth in this book. In my opinion, Dennett’s arguments are part of a growing body of evidence which suggest that the assumptions of Western thought with regards to the human subject are deeply flawed and need both scientific — and philosophical — revisions.

In part III, we’ll examine how psychoanalysis is also frustrating the idea of the Cartesian homunculus from a completely different perspective: an interrogation of the speaking subject.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s