The cogito and time-delay

I’ve been thinking about how well Descartes’ cogito argument works given the following plausisble thesis:

1. Every perception, including introspection, has a time delay.

Consider:

2. I am in pain.

3. If I am in pain, then I exist.

4. So, I exist.

Supposedly, (2) is clear and distinct. But wait (!). By (1), I only introspect premise (2) with a time delay. In other words, by the time I introspect premise (2), the pain is over. It is one thing to be in pain—obviously, when I am in pain, I am in pain—but it is another to be aware that I am in pain.

In other words, at the present moment, if I am to stick to the indubitable, all I get to say is:

5. I was in pain.

6. If I was in pain, then I existed.

7. So, I existed.

Now, if eternalism or growing block is true, I still get to conclude that I exist simpliciter, but not indubitably so (since I need to rely on the arguments for eternalism or growing block).

But there is an even more serious problem. Once we accept the time delay thesis (1), we no longer have indubitability in our introspection of pain. For suppose the time delay from being in pain to being aware that one is in pain is a microsecond. But now consider the half-microsecond hypothesis that the universe came into existence, fully formed, half a microsecond ago. If so, I would still have the introspective awareness of being in pain—without having had a pain! The half-microsecond hypothesis is crazy, but no crazier than the evil demon hypothesis that Descartes cares so much about. So now we don’t have indubitability about (2) or (5).

And what goes for pain goes for any other conscious state, i.e., for anything that Descartes calls “thought”.

We might now want to deny the time-delay thesis (1), and say that:

8. Whenever I have a conscious state Q, I am immediately thereby aware of having state Q.

But a bit of introspection shows that (8) is false. For being aware is itself a conscious state, and so if (8) were true, then whenever I have a conscious state, I have an infinite sequence of conscious states of meta-awareness. And I clearly do not.

Indeed, introspectively reflecting on the states of meta-awareness shows that sometimes the time-delay thesis is true. Let’s say that I am aware that I am in pain. It takes reflection, and hence time, to become aware that I am aware that I am in pain. So the time-delay thesis is at least sometimes true.

Now it might be that we are lucky and the time-delay thesis is false for introspection of first-order conscious states, like being in pain. I am a little sceptical of that, because I suspect a lot of non-human animals are in pain but don’t even have the first meta-step to perceiving that they are in pain.

So let’s grant that the time-delay thesis is false for introspection of first-order conscious states. Now it is no longer true that, as Descartes thought, his cogito could be run from any conscious states. It can only be run from the ones for which the time-delay thesis is false. But it’s worse than that. Even if the time-delay thesis is false for some introspective perceptions, it is not indubitable that it is false for them. The claim that these introspections lack time-delay is far from indubitable.

Yet all that said, isn’t it true that even in the half-microsecond world, I exist? Even if I didn’t have the pain that I think I had, surely to think that I had it requires that I am! Yes, but I only become aware that I think I had a pain with a time-delay from my thinking that I had a pain, because the time-delay thesis is empirically true at all the meta-levels.

This is all very strange. Maybe one can save something by supposing that awareness of a conscious state Q is always partly constituted by Q, and even with a time-delay we have indubitability. Maybe in the half-microsecond world, I couldn’t be aware of having had a pain when I didn’t have the pain, because the second-order awareness is partly constituted by the occurrence of the first-order awareness, be that occurrence past or present. Maybe, but the partial constitution thesis seems dubitable. And once we get to some meta-levels it seems implausible. Couldn’t I be mistaken in thinking that I aware that I am aware that I am aware that I am aware of Q, while in reality I only had two meta-levels?

I am feeling disoriented and confused now.

The unity of consciousness

I am now simultaneously aware of the motion of my fingers and of the text on the screen. Call this co-awareness. Co-awareness is not the same thing as awareness by the same subject. For if I type with my eyes closed and then stop typing and open my eyes, the tactile and visual experiences still have the same subject, but there is no co-awareness. Perhaps co-awareness is awareness by the same subject at the same time. But experiments on split-brain patients suggest that it is possible to have one subject with two simultaneous awarenesses that are not co-awarenesses.

Consider this very simple theory of co-awareness: it is not possible to have co-awareness between two distinct awarenesses. The case I started this post with was poorly described. Strictly speaking I had a single awareness of the conjunctive state of affairs of my fingers moving and there being text on the screen. I did not have an awareness of my fingers moving, nor did I have an awareness of text on the screen, but only of the conjunction.

On this view, rather than my co-hosting a quale of moving fingers and a quale of black markings on a white background, I am hosting a conjunctive quale of moving-fingers-and-black-markings.

All this, however, seems implausible. It certainly doesn’t fit with how we talk: everyone would say that I was aware of my fingers moving.

Similarly, I note, if Alice were to tell me that Bob was lazy and stupid, I would be correct to report that Alice told me that Bob was lazy, even though Alice did not in fact express the proposition that Bob was lazy, but only the conjunctive proposition that he is lazy and stupid. It is good use of ordinary language to attribute the statement of a conjunct to someone who stated a conjunction containing that conjunct. The same is true of awareness: we can attribute the awareness of a conjunct to someone who is aware of a conjunction. Maybe the right way to talk about this is to distinguish non-derivative and derivative, or focal and non-focal, senses of assertion and awareness. Alice non-derivatively asserts that Bob was lazy and stupid, and derivatively that Bob was lazy. I am non-derivatively aware of the conjunctive state of affairs of motion of my fingers, the text on the screen and a variety of other things, and derivatively of each conjunct.

With this distinction, we can build on the simple theory of co-awareness:

1. It is not possible to have co-awareness between two distinct non-derivative awarenesses.

2. Co-awareness occurs between two derivative awarenesses A and B provided that there is a non-derivative awareness C such that I count as having A and B in virtue of C being an awareness of a conjunction that includes the object of A as well as the object of B as a conjunct.

In a way, this simply shifts the difficulty of figuring what makes it be the case that an awareness is an awareness of a conjunctive state to the difficulty of figuring out what makes a non-derivative awareness of a conjunction be an awareness of a conjunction. That is, indeed, a tough problem. But it is a problem that is just a special case of a general problem that we would need to solve even if we had solved our original co-awareness problem in some other way: the problem of the logical structure of the objects of perception. If I see a shape in the distance that looks like a dog or fox, what is it that makes me have an awareness of a disjunction between a dog or a fox? If I see something that looks like it’s not a dog, what is it that makes me have a negative awareness of a dog?

It may seem puzzling how there can be a logical structure to qualia. I don’t see why not. But then I am strongly inclined to a representationalism that holds that the differences in the qualitative properties between conscious states are determined by the differences between the states’ representative properties. And representative properties have a logical structure.

Unfelt pains

Here is something everybody should agree on: there are no unfelt pains.

The obviousness and clarity of this strongly suggests:

1. Pain is the very same concept as awareness of pain.

But if (1) is true, then we should be able to put “awareness of pain” wherever we have “pain”. Thus:

2. Awareness of pain is the very same concept as awareness of awareness of pain.

And we can repeat the substitution:

3. Awareness of awareness of pain is the very same concept as awareness of awareness of awareness of pain.

This leads to an endless regress. I won’t worry about that. Instead, I will worry about the fact that from 1–3, the following follows:

4. Anyone who is in pain is aware of awareness of awareness of awareness of pain.

But 4 is empirically false. It is especially false in the case of intense pains that are so overwhelming as to make the multiple levels of awareness in 4 impossible.

So, we should reject 1. How, then, do we explain why there are no unfelt pains?

I think the answer is to say that “x feels a pain” or “x is aware of a pain” can be understood in two ways:

5. x is aware of their state of paining

6. x is paining.

I think that in ordinary usage of “feels a pain”, 6 is the right understanding even if 5 is a more literalistic translation. Given that to be aware of a pain just is to pain, it’s trivial that there are no unfelt pains, since anyone who is in pain is paining just as anybody who is engaged in a dance is dancing.

(If instead we opted for the unordinary sense of 5, then it would be false that everyone who is in pain feels a pain, since one might have the first-order pain without the second-order awareness of that pain.)

So far this sounds like the familiar adverbial theory of perception. But I don’t like the adverbial theory of perception. After all, to feel is to be aware, and to be aware is to be aware of something. What is one aware of when one is feeling pain? The natural answer is that one is aware of pain. But that gets us back to 1–4.

So if it’s not pain we are aware of, and yet we don’t want pure adverbialism for pain, what are we aware of? Thomas Reid noticed that we have a word for the hardness of a physical object, namely “hardness”, but not one for the corresponding phenomenal state. In the case of pain, it seems to me we have the opposite predicament: we have a word for the mental act of sensing, namely “pain”, but no word for the property that the act of sensing represents. (Reid's account here is that pain is a mere sensation, without anything represented, but I don't like that.)

But we have a word that comes pretty close. Anyone who feels pain feels unwell. And to feel unwell is to sense (one’s) unwellness (in a non-factive sense of “to sense”). So to feel pain is to sense a particular kind of unwellness (there are other kinds of unwellness, like the ones sensed in nausea or itching). We don’t have a word for that particular kind of unwellness, though we can describe it as the kind of unwellness that is properly sensed in pain. (By the way, the word for the genus of sensations of unwellness seems to be “discomfort”. Every pain is a discomfort, but nausea and itching are discomforts that aren’t pains.)