Wednesday, May 29, 2013

1% Skepticism

I find myself, right now, 99% confident that I am who I think I am, living in a broad world of the kind I think I live in. The remaining 1% of my credence I reserve for all radically skeptical scenarios combined.

Most of us, I think, don't reserve even 1% of our credence for radically skeptical scenarios. Maybe if you're philosophically inclined and not entirely hostile to skepticism, you'd be willing to say, in certain reflective moments, that there is some chance, maybe about 1%, that some radically skeptical scenario obtains. But such acknowledgements are typically not truly felt and lived -- not in any durable way. Skeptical doubts stay in the classroom, in the office, in the books. They don't come home with you.

What if skeptical doubt did come home with you? What would it be like really to live with 1% of your credence distributed among radically skeptical scenarios?

It depends in part on what the scenarios are. Let mention three that I have trouble entirely dispelling.

Random origins skepticism. Most physicists think that there is a finite though tiny chance that a brain or brain and body could randomly congeal from disorganized matter. In a sufficiently large and diverse universe, we should expect this chance sometimes to be realized. A question that then arises is: Are randomly congealed beings relatively more or less common than beings arising from what we think of as the ordinary process of billions of years of biological evolution? A difficult cosmological question! I see room for some doubt about the matter, so it seems I ought to reserve some subjective credence for the possibility that such freak-chance beings are common enough that I might be one of them, and thus that I lack the sort of past, and probably future, that I think I have. (This is the Boltzmann brain hypothesis.)

Simulation skepticism. If it is possible to create consciousness artificially inside computers, then likely it is also possible to create conscious beings with radically false autobiographical memories and radically false impressions about the sort of world they live in. I don't feel I can entirely exclude the possibility that I am such a radically mistaken artificial being -- for example a being limited to a few hours' existence in a 22nd-century child's computer game. (This is a skeptical version of the "simulation" possibility, discussed non-skeptically by Nick Bostrom here and by David Chalmers here.)

Dream skepticism. Zhuangzi dreamed he was a butterfly. After he woke he wasn't sure if he was Zhuangzi who had dreamed he was a butterfly or a butterfly dreaming he was Zhuangzi. I'm inclined to think the phenomenology of dreaming is very different from the phenomenology of waking, and thus that my current experiences are excellent grounds for thinking I am indeed awake. But I am not absolutely sure of that. And if I might be dreaming, then I might not really have the family and career and past life I think I do.

I don't believe that any such skeptical scenario is true, or even very likely. On reflection, I am inclined to grant all such skeptical scenarios combined about 1% of my credence. In a way, that's not much. But in another way, that's quite a bit.

Suppose someone came to me with two ten-sided dice and said this: I will roll these two dice, and if they both come up "1", you will die. Would I rest in gentle confidence that a 99% chance of living is quite an excellent chance? Or suppose I were to learn that seven unnamed students from my daughter's largish elementary school had just been abducted into irrecoverable slavery. Would I feel sorry for those students but feel no real concern about my daughter, since the odds are so good she is not among them?

I sit on by back patio. If some radically skeptical scenario is true, then my daughter does not exist, or I will be dead within a minute as the disorganized soup from which I've congealed consumes me again, or my life will end in an hour when the child grows bored with the game in which I am instantiated. Should I be unconcerned about these possibilities because I judge it to be 99% likely that nothing of this sort is so?

I hear a voice from inside the house. It's my wife. How would it change things if instead of taking it for granted that she exists, I held it to be 99% likely that she exists? -- or 99.5% likely, if I allow a 50% chance of her real existence given a skeptical scenario?

Often, I visually imagine negative events that have a small chance of occurring. This morning, the newspaper told me of five teenagers who died in a street race; I looked at my own teenage son, who was readying for school, imagining his death. Or I'm passenger in an airplane descending through heavy fog and turbulence, and I imagine a crash. I hear of a streak of identity thefts in Riverside, see that someone has messed with our mailbox, and imagine our bank accounts drained. If I am a 1% skeptic, shall I then also imagine the world's suddenly splitting into void, Godzilla's rising over the horizon for the child's entertainment, my suddenly floating off into the air, opening my front door to find no ordinary suburban street but rather Wonderland or darkness?

I have, in fact, started to imagine these things more often. I don't believe them -- no more than I believe the plane will crash. Given non-skepticism, the plane is much less than 1% likely to crash. If I am a 1% skeptic, then I should probably think it more likely that am I Boltzmann brain or a short-lived artificial consciousness or a much-deceived dreamer, than that the plane will crash. Would it be more rational, then, for me to dwell on those possibilities than on the possible plane crash?

I am undecided about doing some chore. I could weed the yard. Or I could sip tea, enjoy the shade, and read Borges. I teeter right on the cusp about which is the wiser choice. But skepticism has not yet crossed my mind. Once it does, the scale is tipped. The 1% chance that the weeds are an illusion or a mere temporary thing -- the small but non-trivial possibility that this moment here is all I have before I die or the world collapses around me or I wake to something entirely new -- favors Borges.

If my credence in the durable reality of this patio and this book were to fall much below 99% -- if my credence were to fall, say, to only 80% or 50%, I would be laid low. My death might well be upon me within minutes. I would seek my family, my seeming-family; at the same time, my doubts would isolate me from them. I could not, I think, feel full proper intimacy while I regard my partner in intimacy as fairly likely to be mere froth or illusion. Even if my wife and children are not froth and illusion, such large doubt is disaster, for it would derange my choices and emotions, and few people would understand my doubts, even if my own careful reflection revealed those doubts to be well-grounded. Everyone around would see me only as insane with foolish philosophy.

1% skepticism does not have this effect though. I can still enjoy my Borges, even flatter myself somewhat with my excellent excuse for avoiding the weeds, which I am of course in fact almost certain do exist.

[For related reflections see Waterfall Skepticism.]


Howard Berman said...

I went through a skeptical phase when young, and though your post has some plausibility, I am skeptical of skeptical paradigms for the mere reason that they may be chance artifacts of particular cultures such as our own. Of course in a computer simulated universe the simulated people come up with the idea of a simulated universe. It is too glib and my guess is that skeptical ideas should be more challenging and not become something of a sport. So, first research question: do non modern, non western cultures have their skepticism?

Eric Schwitzgebel said...

Yes, Howard, there are skeptics in non-modern, non-Western cultures too -- like Zhuangzi mentioned above.

I agree that it's too easy for a skeptic to say "I doubt X, prove me wrong!" without any grounds for doubt. However, in the three cases above, I think there are at least some slender grounds for doubt. It's not as glib as just asking "why" until your interlocutor is exhausted or as asking for a proof of the reasonableness of deduction that doesn't assume the reasonableness of deduction.

But the post isn't really intended to support those three types of skeptical doubt -- they are too briefly presented for that! Rather, my aim is to explore the consequences for someone willing to allow a modicum of doubt for roughly those sorts of reasons.

howard berman said...

Thanks. I'll think about your post. I have a book somewhere on the history of skepticism that divides the field into Academic and Phyrronic. It sounds like modern skepticism has branched out somewhat- I mean in addition to Descartes

clasqm said...

It is possible that all this is illusory. The REAL reality is the universe in which the Boltzmann brain temporarily floats, or the 22nd century child playing the game, or the universe of the dreamer.

But the mere fact that we have no way of choosing between those three alternatives shows that we have no access to those realities. We can speculate, but we cannot enter into them. There are a few movies where a character supposedly steps off the screen and enters the real world. Unfortunately, WE are stuck in the movie.

So let's suppose for a moment it's the simulation scenario. You don't exist in the REAL reality, only in this secondary one. Your supposed past consists of lines of code and a data file. So do you. Well, then, what makes you think you have free will to decide to read Borges or not? It is predetermined by the coder and adjusted by the player. The best you can hope for is that your subroutine was hooked up to a random number generator somewhere along the line. Even your doubt itself is predetermined. The Boltzmann brain is in a similar position. If the particles had congealed slightly differently, your existence, including your doubt, would have been different, perhaps nonexistent.

The dream scenario is more interesting. But unless one of the religions is correct in its prescriptions of how to wake up (Buddhism would make this claim), we remain stuck in the dream.

But wait! "Dream" uses an analogy from *this* existence. So there are at least three levels of reality now: REAL reality, the dream-state that we cal normal wakefulness, and the double-dream that we normally call a dream. Even if you did "wake up" to the REAL, how would you know that there wasn't a fourth, REALLY REAL level above that? Skepticism just went to 1.1%

I need a coffee. Thank you for reading, my imaginary friends.

Anonymous said...

The more sceptical scenarios you give real probability to, the smaller the probability you should give to each one, assuming you continue to have 1% credence in 'I am in some sceptical scenario'. If you're only seriously considering a Boltzmann brain scenario where you have existed for a few seconds, that gets all your 1% sceptical credence, but when you start to also seriously consider a particular simulation scenario, since you have no reason to prefer it to the Boltzmann one, or vice versa, you should only have credence 0.5% in each. There are very many conceivable incompatible scenarios - it may well be that you shouldn't end up giving any of them a credence large enough to calculate or comprehend, even if you allot more than 1% of your credence to scepticism (even if you go up to 99% credence in scepticism in general, all you have to do is multiply the number of sceptical scenarios you know of hundredfold, which is surely possible if you devote your time to it, and you're back where you started).

So, the more sceptical you get, in terms of the number of scenarios you are willing to give any credence to, the less you should worry about the results of any particular scenario (that you will be dead in 5 secs, that things you remember and care about never existed, etc.). Since 'I am in some sceptical scenario' isn't very worrying on its own, if you are a committed sceptic you shouldn't have anything more to worry about than the rest of us.

clasqm said...

Anonymous: Fascinating idea: an Achilles-and-the-hare approach to doubt? But Eric's original piece states that he envisages his doubt growing to 50 or 80 % so I don't think your assumption holds. The poor man is at only 97% certainty already, 1% down for each of his scenarios!

I sense a story in there: present someone with another skeptical scenario each day, until the sheer volume of doubt makes them stop wanting to live.

No, Eric! Not you, You're real, Eric. Snap out of it!

Too late. He's gone. A true martyr to philosophy.

Please take note of that, class. If you're going to put a skepticism function in your AI, you must set the limit conditions at 1% and make sure the fail-safes kick in before your program shuts itself down. OK, that's it, don't forget the due dates for your term papers. See you Monday.

Alexander Kruel said...

You wrote: Skeptical doubts stay in the classroom, in the office, in the books. They don't come home with you.

Check out Roko's basilisk for an example that some people took home.

Eric Schwitzgebel said...

@ Alexander: Thanks for the link. Weird stuff!

@ Michel: Still here! Or, maybe, here for the first time with illusory memories of having been here before.

@ Anon: I agree, but since I'm only claiming 1% total for all scenarios combined, I don't think there's any inconsistency there. Obviously, it would be a problem to claim 1% for each of 101 mutually inconsistent skeptical scenarios....

Scott Bakker said...

There's a problem here with the term, 'radical,' I think, as well as a tension with your work on metaphysical crazyism. Metaphysical crazyism in the philosophy of mind actually *entails* radical skepticism, to the extent that 'radical' is taken to mean levels of incredulity of commonsense belief that the vast majority would consider 'crazy.'

This is one reason I've never been able to muster much interest in the three brands radical skepticism you mention: they strike me as small potatoes when you consider some of the ways science is directly cutting against the grain of intuitivity. There's physics, of course, but for me the big elephant in the room is meaning skepticism. Has anyone noticed how many 'silent eliminativists' seem to haunt neuroscience circles nowadays?

On a personal level this is the form of skepticism that I've been trying to escape for years. It often keeps me awake at night. Twinges of dizziness, nausea, are common... even prolonged periods of anxiety, something I imagine to be akin to derealization and/or depersonalization.

Ever since I began reading cognitive psychology fifteen years ago, and slowly came to the conviction of human theoretical incompetence outside the sciences, I have been stricken with the conviction that, all things being equal, our prescientific notion of ourselves will turn out to be every bit as blinkered as our prescientific notions of the cosmos - with the added wrinkle that we're talking about modes of sense-making that constitute the mandatory frame of our ability to cognize ourselves at all.

All in all, I probably spend less than 1% of my *time* thinking about these problems, but it's virulent enough to infect the other 99% of the time with the pervasive and corrisive odour of fraudulence.

Like apples, it only takes one, particularly bad 1% to spoil the whole bushel!

Eric Schwitzgebel said...

Thanks for the thoughtful comment, Scott! I agree that there are reasons to worry that I might be vastly deluded in my psychological self-conception too, though I find that issue a little more difficult to finger than the three types of skepticism I mention above. In my post on "Waterfall Skepticism" I discuss the possibility that I'm a madman. I am also working on an essay on moral self-skepticism tentatively titled "How Do I Know If I'm a Jerk?" Radical eliminativism about consciousness and meaning/reference based concerns don't compel me as much, for some reason which I probably ought to figure out how to articulate!

Pactura Observa said...

Hey Eric - tell me something. Wouldn't each moment of consciousness equal a lesser or - equal to the the parallel - skepticism vs. certainty that would be harder to offset a more built construct of conscious? That is to say, how can you be so sure you're unsure?

Eric Schwitzgebel said...

Possible: Only if the trials are sufficiently independent and I can trust my memory!

Callan S. said...

Surety - if it's stupid, but you don't die, it's not stupid.

If you do die, you don't know you were wrong anyway.

$9,000,000,000 Write Off said...

The part that caught me is how the skepticism affects decision making-- if you succumb, it tips you into "why bother with duty (weeds), indulge in pleasure (Borges)." Maybe your mind is working backward? "Weeding is a drag, I rather read Borges...and it probably doesn't matter anyway."
If so, you don't need esoteric alternative scenarios; the mundane 1% (or, more likely, 0.001%) possibility that your teenage son will weed tomorrow suffices to tilt you into indulgence. (Although, the esoteric scenarios are far more interesting and, thereby, make a better excuse.)

I guess I'm skeptical that your skeptical bent changes your behavior in comparison to a more credulous you.

blackboxdisease said...

Perhaps when you lock yourself into vivid dreams those percentages will change.

When a brain evolves to the point that it can manufacture and conjure thoughts from the subconscious into the conscience it could be more plausible to discern which reality is our true reality or maybe just complicate things more. Perhaps we as individuals are not living a single reality at all but only part of a bigger reality. That we are all connected as one consciousness at a level we are not yet aware of. There are many primitive species of insects and even micro organisms that alone, cannot accomplish many tasks but as a collective are able to solve complex solutions.

This may explain phenonama where groups of complete strangers find themselves at the same spot for the same reason, or how many people have the same ideas, dreams, inventions, cravings and more. I work at a resturant and come to notice that groups of people unknown to each other will order the same plates of the same entrees and when we finish that seating groups of other people will repeat the task only with different entrees.

I am heading on the path that we as a species, are connected at a quantum level, or a level yet discovered, separated only by measures put in place to have us appear as a single, personal entity. it would answer a lot of questions and undoubtedly create more

Unknown said...

That was very nice, thank you.

But I am again confronted with the search for THE solution. Perhaps it is instead a SEQUENCE solution. It's not necessary that Boltzmann's Brain construct us "as final product". It can be merely the kick-starter and Evolution finishes the job.

It's like saying I destroyed a two story house with a baseball. No, I threw the baseball into the snow at the peak of the mountain, that ball started rolling and picked up snow, such that the weight and velocity of the Earth's largest snowball . . . took out the house below. Though my baseball throw was required, it alone wasn't enough. The secondary contributing part of the rolling/adhering/growing snowball . . . completed the job.

Not all solutions are "single", some are "sequences". And neither part is right nor wrong nor complete. But parts of a whole.

The same "argument" occurs in Creationism vs Evolution. Why must it be "vs"??? Why can't the answer be Creationism + Evolution? I respectfully submit, it can.

Similar flaws occur in searching for "Energy Independence" (from oil). We can't use Solar Power, it only produced 15% what we need. We can't use Wind Power, it only produced 15% what we need. We can't use Wave Motion Power, it only produced 15% of what we need. But by my count, if we used all three, we're up to 45%. Why then is the 45% dismissed, because it's a multi-part (or sequential) solution, and solely because its not 'all in one'?

More possibilities might open up for your 1%, if you accept sequences as well as 'all in one' possibilities.