Descartes and Knowing One is Awake

Descartes maintained that our experience never permits us to judge with certainty that we are awake: “[T]here are no certain indications by which we may clearly distinguish wakefulness from sleep.” So I would be unjustified in judging that I am now, in fact, awake with a computer screen in front of me. 

(At the end of the day, of course, Descartes did think I could be so justified, but before that day’s end I would have to traverse through his cogito, taxonomy of mental contents, clear and distinct ideas, ontological argument, and his argument that God is trustworthy. Not being able to negotiate that entire journey, I want to see where we end up on dreams if we decline to go beyond Descartes’s First Meditation.)

I don’t know about you, but when I ask myself the question whether I am awake or asleep, I almost always conclude that I am awake, and I am always wrong. That is because I never ask myself if I am awake when I am awake. It is a question that arises only when I am dreaming, usually because there is something that seems faintly suspicious, for example, about the way the bear in the woods turned out to be a high school friend in football uniform reading Plato. 

It is central to the issue, I think, that the dream state is an impaired state for making judgments about anything. That I make a certain sort of mistake in an impaired state is not, in general, good evidence that I would make the mistake in an unimpaired state. A’s intoxicated judgment that A is not intoxicated and that B has two heads is not as good as B’s sober judgment of single headedness and A’s intoxication. 

“Wait,” I imagine I hear, “You are begging the question. Minimal intellectual honesty requires you to treat equally the possibility that you are asleep, and so impaired, and the possibility that you are awake and unimpaired. Descartes’s point is that if you are dreaming you will very likely impaired-conclude that you are unimpaired-awake. It is not something you would conclude on the evidence provided by the dream were you awake, but you do so conclude because you are asleep.”

Fair enough. I cannot simply assume unimpaired faculties when evaluating the impairment of my faculties, but then, neither should I assume away the possibility that the evidence will support the conclusion of non-impairment so strongly as to satisfy the requirements of methodological neutrality.

That I never ask if I am awake when I am awake is already enough to show that there must be some important differences between dreaming and waking life –

differences bearing directly on Descartes’s skeptical claim. Are the differences of such magnitude and character that I can be confident that I am awake – keeping always in mind that my ability to make such judgments will be impaired if I am, in fact, asleep? 

I have memories of what dreams are like, although often very imperfect memories. I also have a stock of memories of waking experience, a much larger stock, quite a number of which are in reasonably high resolution. A natural way to decide whether I am now awake or dreaming is to compare my present experience with both classes of memories. 

Is my present experience more similar to my memories from the “when awake” batch or the “dreaming” batch?  The first batch, conveniently, includes some memories from yesterday and the day before, which are still pretty fresh and clear, and do seem very much like what I am currently experiencing. My memories of five minutes ago are even clearer and more detailed, and very like what I am experiencing now, but probably comparing them would be cheating as they might well be part of the same dream that I might be in. Perhaps I can safely count my pre-breakfast computer session among the awake memories, but that is unnecessary. The catalogue of awake memories and their likeness to what I am now experiencing is enough without committing to any potentially controversial line as to when my dream began if I am now dreaming.

You may be concerned that fair neutrality as to whether I am dreaming would mean distrusting my division between dreaming and waking memories. (You, of course, know that I was not dreaming because dreams do not have the right sort of causality for you to be reading a post I merely dreamed that I wrote. You should not give up your concern about appropriate neutrality, however. You can transpose everything I say here into your own first person thought experiment. After all, in the end, what you want to know is whether you are well justified in thinking that you are now awake. So it is your own hypothetically impaired division of memories between the awake and the dreaming that worries you. This same transposition you may make throughout this post.) 

Descartes’s argument itself depends upon there being a difference between waking and dreaming and our having distinguishable memories of each. He draws upon the specifics of his own memories of dreams to argue to his skeptical conclusion. So my line of inquiry is dialectically permissible, and I am going to take your question of the last paragraph as coming down to the magnitude of the risk that dream-impaired judgment would get the division wrong, and for that reason, perhaps make it false that my current state is more like the awake memories. 

Just what is it that makes dreams and waking experience different? (If you are currently dreaming, you will, I think, still grant, that you are not so impaired as to fail to have some grasp of some differences.)

Weird content.  

Bears turning into people are far from the weirdest of dream contents. Sometimes dreams are so jumbled as completely to evade description. It is easy enough to check that my current experience is not anything like this – a finding that seems immune to concerns about impaired judgment. Unfortunately, however, this observation does not get me very far because not all dreams are weird. Some are fairly realistic. Their script is straight out of everyday life. So absence of weirdness in what is now happening cannot really count towards justifying the judgment that I am awake. So no progress yet.

Beyond-my-capability content. 

As it happens (really) a Brahms sonata is playing in my library now. I take what I am hearing to be very strong evidence that I am awake. I am not nearly a good enough musician to be making it up in a dream and I don’t have nearly a good enough memory to recall it from the last time I heard it, I think many months ago. If I am dreaming, then the best explanation of what is going on is that I am subconsciously conjuring up some piano-ish sounds and impaired-judging them to be extraordinary music of high complexity, and subtlety. It is impossible to believe this. More telling, the reasoning that is required even to lay out how the impaired judgment might work seems to be well beyond what my memories of dreams reveal that my dreaming-reasoning has ever been able to muster. This does, then, count against the hypothesis that I am dreaming.

Life Story Connectedness.  

In making the judgment that I am not suffering from impaired judgment due to intoxication, leading me to make an erroneous judgment that I am not intoxicated, I can review my recent history. I take note that I have had no alcohol since dinner two nights ago when I had about three ounces of wine. I have not been injected and have taken by mouth nothing with the slightest probability of having been spiked or fermented. All this would contribute significantly to the conclusion that I am not now intoxicated, should the question arise.

My judgment that I am now awake has similar support. I have good memories of my last several goings to sleep and awakenings. They connect consistently with my memories of the larger story of my life, as does everything in the room in which I write this. Dreams, in my experience, often have no connection to my larger life story. I am simply there in a free floating episodes, it not being any part of their content what I was doing immediately previously. 

Even the most realistic of dreams have only a thin connection to my history. Perhaps I remember an incident from my past, but never is there a systematic fitting in with a whole life story. In my dream, of course, I am not in the least troubled by that. In striking contrast, I can check detail after detail of my present experience and recent memories, and find they fit very nicely indeed with my larger biography as evidenced by my memory and my surroundings. This difference in the ability to connect up to my life story between my current state and any dream I can remember is strong evidence that I am not now dreaming. Again the evidence is of such nature as to resist the hypothesis that it only looks strong because I am evaluating it under impairment. 

Dream phenomenology. 

What is probably the most common reaction to the challenge one cannot be confident one is not dreaming runs something like this: “I know what it feels like to dream, and I know what it feels like to be awake, and this is being awake.”  Here, of course, there does seem to be some vulnerability to Descartes's observation that when dreaming he made exactly this judgment of his phenomenological circumstance, and got it wrong. So it may seem that my assessment of my current phenomenology can contribute nothing to the justification of my belief that I am awake. I think, however, that it does contribute. 

Consider, again, intoxication. A friend says, “I am not drunk, I only have a little buzz on.” This might be sincere, but wrong. Suppose now, on another occasion, we think that some punch may have been spiked. Our friend had quickly downed a considerable quantity of same 15 minutes ago, and says “I know what it feels like to be drunk, and I am definitely not drunk. I don’t feel the slightest effects.”  Absent additional extraordinary circumstances, we would accept our friend’s statement as very strong evidence that the punch was not spiked. It is one thing to mistake, in intoxication, very for modestly. That is within the range of effects of typical intoxication-impairment. To mistake very for not at all is something that an experienced and honest subject is not going to do. 

I run through in my mind one after another phenomenological aspects of my present experience: the character and quality of images, of sounds, of touch, of smell, of body position, of little aches and pains, of the way what I intend to do becomes what I do. I am awake. 

Although I have reached this same conclusion when I was in fact dreaming, my memory of what my experience was then like, and of the cursory survey of my dreaming phenomenology that led me to that conclusion, reveals a very marked difference to what I am now experiencing. If I am making a faulty comparison, I must be giving too much credit to my current experience or too little to the “I’m awake” experience as I remember it from my dreams. The story would have to be that I am impaired in a way that causes me to mistake two things that are in fact the same, my reality markers analysis now, and my markers analysis as remembered in dreams, even though I am now able to catalogue a significant list of differences – including that the now process involves, as the dream process does not, the analysis of that significant list. 

Although it is not logically impossible that I am asleep, I conclude, as I sit here typing on my keyboard, that I am certainly awake. I so conclude for the pedestrian reason that this is what being awake feels like. This conclusion is reinforced by the way my current experience connects with my life story and that fact that it includes components that could not possibly be a product of my own imagination. No dream-impaired-judgment hypothesis can be rigged to caste any but an extremely implausible doubt of my now being awake.

Perhaps an extremely implausible doubt is all Descartes needed? He did say, “no certain indications by which we may clearly distinguish wakefulness from sleep.” If Descartes intended only that I cannot now have absolute certainty, that a rational confidence after the decimal point of a million 9s isn’t good, enough, then he should have skipped dreaming and gone immediately to the Evil Deceiver. The reliability of judgments of wakefulness is no longer of interest if absolute certainty is demanded.

(For my own part, I don’t think that our subjective probability even for “3 + 2 = 5” should reach quite to 1. See

https://lawrencecrocker.blogspot.com/2015/09/nothing-is-certain-cromwells-rule.html. 

.)