# Bizarre Illusions

Illu­sions are cool. They make me think some­thing is hap­pen­ing when it isn’t. When offered the clas­sic illu­sion pic­tured to the right, I won­der at the color of A and B. How weird, bizarre, and in­cred­ible.

To­day I looked at the above illu­sion and thought, “Why do I keep think­ing A and B are differ­ent col­ors? Ob­vi­ously, some­thing is wrong with how I am think­ing about col­ors.” I am be­ing stupid when my I look at this illu­sion and I in­ter­pret the data in such a way to de­ter­mine dis­tinct col­ors. My ex­pec­ta­tions of re­al­ity and the in­for­ma­tion be­ing trans­mit­ted and re­ceived are not lin­ing up. If they were, the illu­sion wouldn’t be an illu­sion.

The num­ber 2 is prime; the num­ber 6 is not. What about the num­ber 1? Prime is defined as a nat­u­ral num­ber with ex­actly two di­vi­sors. 1 is an illu­sion­ary prime if you use a poor defi­ni­tion such as, “Prime is a num­ber that is only di­visi­ble by it­self and 1.” Build­ing on these bad as­sump­tions could re­sult in all sorts of weird re­sults much like di­vid­ing by 0 can make it look like 2 = 1. What a tricky illu­sion!

An op­ti­cal illu­sion is only bizarre if you are mak­ing a bad as­sump­tion about how your vi­sual sys­tem is sup­posed to be work­ing. It is a flaw in the Map, not the Ter­ri­tory. I should stop think­ing that the vi­sual sys­tem is re­port­ing RGB style col­ors. It isn’t. And, now that I know this, I am sud­denly cu­ri­ous about what it is re­port­ing. I have dropped a bad be­lief and am look­ing for a re­place­ment. In this case, my vi­sual sys­tem is dis­t­in­guish­ing be­tween some­thing else en­tirely. Now that I have the right an­swer, this op­ti­cal illu­sion should be­come as un­in­ter­est­ing as ques­tion­ing whether 1 is prime. It should stop be­ing weird, bizarre, and in­cred­ible. It merely high­lights an ob­vi­ous re­al­ity.

Ad­den­dum: This post was ed­ited to fix a few prob­lems and er­rors. If you are at all in­ter­ested in more de­tails be­hind the illu­sion pre­sented here, there are a hand­ful of ex­cel­lent com­ments be­low.

• Why do I keep think­ing A and B are differ­ent col­ors?

Mean­while I am think­ing ‘Wow! My brain can au­to­mat­i­cally re­con­struct a 3D image from limited 2D in­put and even com­pen­sate for shad­ows and light­ing. That is or­ders of mag­ni­tude more com­plex than the re­verse, gen­er­at­ing such images from a model such as those we add 3D cards to com­put­ers for’.

I don’t par­tic­u­larly con­sider this an ‘illu­sion’, es­pe­cially when it is not si­mul­ta­neously ac­knowl­edged that it is an ‘illu­sion’ that A and B are squares on a 3D ‘square X a bit’ thing that also has a cylin­der on top of it.

• Wow, good point! I never thought about it like that. It raises the ques­tion: Why are peo­ple amazed when you say, “Tiles A and B are ac­tu­ally the same color—check for your­self!” but they roll their eyes when you say, “There are no squares in this image—check for your­self!”? In both cases, you can re­spond with, “Well, yeah—if you don’t in­ter­pret it like the scene it’s try­ing to rep­re­sent!”

I’m not a very good artist, so learn­ing about how to cre­ate these illu­sions sounds like a good rea­son to take an art class, and help me ap­pre­ci­ate what artists are do­ing. (Why didn’t the first ma­jor break­through in cog­ni­tive sci­ence come from painters and sketch­ers?)

Of course, it prob­a­bly wouldn’t do much to help me un­der­stand why they can count ran­dom smears on a can­vas as “art”...

• Why are peo­ple amazed when you say, “Tiles A and B are ac­tu­ally the same color—check for your­self!” but they roll their eyes heir eyes when you say, “There are no squares in this image—check for your­self!”?

It’s about ex­pec­ta­tions. Peo­ple ex­pect to be able to take (phys­i­cal) ob­jects that ap­pear to be differ­ent col­ors, ex­am­ine them un­der a va­ri­ety of con­texts, and always per­ceive them as differ­ent. Peo­ple in­cor­rectly ex­trap­o­late that ex­pec­ta­tion to images, and thus find the fact that re­mov­ing the con­text re­veals these images to be the same col­orRGB sur­pris­ing. They also ex­pect to be pre­sented with rep­re­sen­ta­tions, so point­ing out the fact that they’re look­ing at a rep­re­sen­ta­tion seems silly to them.

• Of course, it prob­a­bly wouldn’t do much to help me un­der­stand why they can count ran­dom smears on a can­vas as “art”...

By way of re­vers­ing the ADBOC con­cept, I dis­agree de­no­ta­tion­ally but con­firm your con­no­ta­tion. As you ex­plain in the cousin sev­eral times re­moved post, many kinds of art are bul­lshit. Cul­tural prefer­ences that would not be par­tic­u­larly likely to be re­dis­cov­ered if all trace was re­moved. This differs from other forms of art which are more speci­fi­cally di­rected at aes­thetic prefer­ences in­trin­sic to hu­mans.

Of course, im­mers­ing your­self in a cul­ture and ex­pe­rienc­ing the flow of sta­tus first hand is the per­haps the best way to get an in­tu­itive an­thro­polog­i­cal un­der­stand­ing. I found, for ex­am­ple, that hav­ing done a re­search de­gree in a sub­field of AI helps me un­der­stand how peer af­fili­a­tion by per­sist­ing with re­search­ing silly ideas can be counted as ‘sci­ence’.

• That last line is com­ing from a de­cid­edly un­ra­tional state of mind!

• How so? I wasn’t spout­ing the usual greedy/​fake re­duc­tion­ist cliches; I was talk­ing about the paint­ings that look like a 3-year-old made a mess, yet get clas­sified as art, and not­ing that an art class prob­a­bly wouldn’t con­vince me this is ap­pro­pri­ate.

What spe­cific crit­i­cism of that claim do you have?

• I was talk­ing about the paint­ings that look like a 3-year-old made a mess, yet get clas­sified as art, and not­ing that an art class prob­a­bly wouldn’t con­vince me this is ap­pro­pri­ate.

What spe­cific crit­i­cism of that claim do you have?

Short ver­sion:

High art is about a lot of things, not least of which is im­pact on the viewer.

In the case of Pol­lock, for in­stance, a lot of the in­ter­est­ing thing that was go­ing on there, was that he de­picted a pro­cess—not by paint­ing a rep­re­sen­ta­tion of him­self do­ing it, but by ac­tu­ally do­ing it. You can look at a Pol­lock and see how he con­structed it with­out be­ing dis­tracted by ex­actly what he was try­ing to con­struct. And be­ing able to see that as­pect of art and be aware of it, will in turn give you a greater ap­pre­ci­a­tion of me­dieval cathe­drals and Greek sculp­ture.

Pos­si­bly re­lated: no one knows what sci­ence doesn’t know

• To add to this with a similar ex­am­ple, con­sider that some peo­ple pre­fer listen­ing to for­eign lan­guage vo­cal­ists be­cause it al­lows one to ap­pre­ci­ate the sound of the vo­cal in­stru­ment with­out fo­cus­ing on the words.

• To me, most mu­sic sounds like a for­eign lan­guage (though one that sounds ex­actly like English), un­less I’m fa­mil­iar with the lyrics be­fore­hand, in which case I can “hear” them just fine.

• Like this?

• Yes, most peo­ple don’t care much about the ac­tual lyrics. Which ex­plains the phe­nomenon thomblake was try­ing to use for ten­u­ous sup­port of an­other hy­poth­e­sis, yet re­mains mod­ded to 3 for some rea­son (7 if you in­clude his par­ent com­ment).

• Right, be­cause the words (i.e. the lyri­cal se­man­tics, as differ­en­ti­ated from the qual­ities of the sounds the words make) are a small, per­haps neg­ligible com­po­nent of what peo­ple like about many of these songs.

If you were try­ing to draw some other in­fer­ence from this fact, you’re go­ing to have to be more spe­cific about why that in­fer­ence fol­lows.

• Definitely re­lated: Truly Part of You

If I erased your knowl­edge (and ev­ery­one else’s) of what the kewl kids had clas­sified as “good art”, would it grow back? Would you even­tu­ally re-rec­og­nize the same works as be­ing good, with the same rel­a­tive merit, for the same rea­sons?

If your an­swer is no, that’s a big red flag that you’re deal­ing in bul­lshit.

(The cor­rect re­ac­tion to the parable of The Em­porer’s New Clothes is not, “Well of course a kid isn’t go­ing to see the clothes! What’s your point?”)

• If I erased your knowl­edge (and ev­ery­one else’s) of what the kewl kids had clas­sified as “good art”, would it grow back? Would you even­tu­ally re-rec­og­nize the same works as be­ing good, with the same rel­a­tive merit, for the same rea­sons?

I imag­ine that the an­swer to this is yes for a great deal of art. I don’t know much about it my­self, but when I think about art that I like I can find rea­sons aside from cul­tural sig­nifi­cance or peer pres­sure.

This as­sumes the ques­tion is ig­nor­ing the lens cre­ated by my limited ex­pose to art. I highly doubt that any of the artists I like would have been ex­pe­rienced by me if oth­ers hadn’t con­sid­ered them worth­while.

Mu­sic is an eas­ier anal­ogy for me to make. I can more ac­cu­rately de­scribe what I like in mu­sic be­cause I know a few more terms. But also, when I listen to a song, I find that my opinions are more dis­tinct. I as­sume this is be­cause my tastes are be­com­ing re­fined; I am open to other in­ter­pre­ta­tions.

• That’s not re­lated.

If you took away ev­ery­one’s knowl­edge of English, and some­one laid King Lear at your feet, what would you do with that? The fact that art is rooted in cul­ture and con­text, some of which is the re­sult of stochas­tic pro­cesses, does not mean you’re deal­ing in bul­lshit.

• I would be very sur­prised to dis­cover that a King Lear in an un­fa­mil­iar lan­guage had been pro­duced by an ape. I am not sur­prised by hoaxes like this. I think that is in­dica­tive of a mean­ingful differ­ence.

• From the link:

After Peter had cre­ated a num­ber of paint­ings, Ax­els­son chose what he con­sid­ered to be the four best and ar­ranged to have them ex­hibited in an art show at the Christina Gallery.

Em­pha­sis added to in­di­cate flaw in ex­per­i­men­tal pro­to­col.

Edit: This point is much weaker than it ap­pears at first glance. See re­sponses.

• I would still be sur­prised if the mon­key King Lear was cho­sen as the very best of the mon­key’s liter­ary oeu­vre.

• Yeah, you’re right—odds of a mon­key pro­duc­ing a King Lear to choose are quite low.

• Yeah, I no­ticed that too. I felt that it was still a valid test of crit­ics’ abil­ity to in­ter­pret art con­sid­er­ing that most artists will do the same thing with their col­lec­tion be­fore en­ter­ing an ex­hi­bi­tion.

• And, on re­flec­tion, se­lec­tion is a very weak form of op­ti­miza­tion.

• That’s a far cry away from “even­tu­ally re-rec­og­niz[ing] the same works as be­ing good, with the same rel­a­tive merit, for the same rea­sons.”

• Eras­ing ev­ery­one’s knowl­edge of English is a far cry from eras­ing their knowl­edge of “what the kewl kids had clas­sified as ‘good art’”.

• ? Was this sup­posed to be a sep­a­rate re­ply to my ear­lier com­ment? I think it brings up a valid point, but looks a bit like a non-se­quitur where it’s at now.

• It’s a refer­ence to that ear­lier com­ment, which is its great-grand­par­ent, but also a di­rect re­ply to its par­ent. I think it makes sense if you read the full thread.

• But it does mean that the writ­ing of King Lear is less of an epistemic achieve­ment than, say, the laws of physics, which are not de­pen­dent on a par­tic­u­lar species’ form of com­mu­ni­ca­tion.

If King Lear is (claimed to be) a good work, given a cer­tain lan­guage (hu­man­ity? evolu­tion­ary his­tory? poli­ti­cal his­tory?), does the recog­ni­tion of its sup­posed great­ness sur­vive dele­tion of the knowl­edge about what the kewl kids think is great?

If peo­ple con­tinued to speak English, but King Lear fell out of fash­ion and later was found, but dis­con­nected from any­one’s recom­men­da­tion, would peo­ple still de­cide it was bet­ter than most other works? Would they de­cide it for the same rea­son?

Do chil­dren spon­ta­neously flock to King Lear at a cer­tain age, even when it’s not recom­mended to them by a True Liter­ary Author­ity?

• If King Lear is (claimed to be) a good work, given a cer­tain lan­guage (hu­man­ity? evolu­tion­ary his­tory? poli­ti­cal his­tory?), does the recog­ni­tion of its sup­posed great­ness sur­vive dele­tion of the knowl­edge about what the kewl kids think is great?

Of course not. It doesn’t even come with 3D spe­cial effects!

• If your an­swer is no, that’s a big red flag that you’re deal­ing in bul­lshit.

I se­ri­ously doubt that the cor­rect an­swer is “no”.

Ob­vi­ously, there would be a lit­tle bit of wob­ble—I might not care who Pol­lock is, but I ex­pect there would be some­thing else I’d find that would illu­mi­nate the same as­pects of the aes­thetic ex­pe­rience. But I think be­ing the first to do it that way counts for some­thing.

Thanks for the link—I read that ar­ti­cle a while ago, but I hadn’t re­al­ized Drew had been refer­enced here.

• Sorry, all I got out of that was a name-drop and (what seemed like) a dodge.

Could you an­swer again, and this time maybe ex­plain it a lit­tle differ­ently? Speci­fi­cally:

-Are you claiming that Pol­lock dis­cov­ered a way of satis­fy­ing the aes­thetic senses that al­lowed gen­er­al­iza­tion of the method in other forms?

-Let’s say I knew a wacko who be­lieved that “By his­tor­i­cal ac­ci­dent, Pol­lock be­came a fo­cal point for peo­ple of high-sta­tus to iden­tify each other, de­spite there be­ing noth­ing spe­cial about his work.” What ev­i­dence would you point me to that has a low Bayes fac­tor against that hy­poth­e­sis?

• I just re­al­ized we seem to be ar­gu­ing over wine again. I fold.

• Yes, we are.

If you spend ten years as­so­ci­at­ing wine with a good time, and are ex­pected to have a re­fined palette for wine to be part of the kewl kids club, then guess what—you can make your­self like wine! The fact that you like wine in such a sce­nario does not, to me, count as a gen­uine lik­ing, in the sense in which I judge bev­er­ages. Any sub­stance, even bat urine, will find con­nosieurs un­der those con­di­tions!

What I want to do is, find out what’s good about some­thing, that isn’t sim­ply an ar­ti­fact of prac­tices that can make any­thing look good.

That’s why I’m not im­pressed by “en­joy this be­cause peo­ple are tel­ling you to en­joy it”, which the sup­port for much high art and al­co­hol amounts to.

In­stead of re­fus­ing to en­gage the is­sue, maybe you should start to think about the re­cur­sivity of your crite­ria for qual­ity?

• What I want to do is, find out what’s good about some­thing, that isn’t sim­ply an ar­ti­fact of prac­tices that can make any­thing look good.

This seems ex­ceed­ingly ar­bi­trary. The ex­act evolu­tion­ary pro­cesses that made ice cream taste good gave rise to the con­noisseur phe­nomenon. Our abil­ity to “pre­dict” evolu­tion and make some­thing “taste like vic­tory” doesn’t make en­joy­ment of those things less real, let alone less “good.”

Be­sides, we could keep fol­low­ing this rab­bit hole to the end of time. What makes it en­joy­able for you to not suc­cumb to the trends of sta­tus? Why is the good-feel­ing re­ward that gives you bet­ter than the good-feel­ing re­ward that some other ac­tivity gives some­one else?

• So wait—are you con­ced­ing that art is just about sig­nal­ing that you like what­ever-high-sta­tus-peo­ple-like? Or that “you will get higher sta­tus for say­ing you like this” is a valid rea­son to judge a work as be­ing art?

This seems ex­ceed­ingly ar­bi­trary. The ex­act evolu­tion­ary pro­cesses that made ice cream taste good gave rise to the con­noisseur phe­nomenon. Our abil­ity to “pre­dict” evolu­tion and make some­thing “taste like vic­tory” doesn’t make en­joy­ment of those things less real, let alone less “good.”

It mat­ters for the same rea­son the placebo effect mat­ters. Pills can make you bet­ter, merely by virtue of be­liev­ing they’ll make you bet­ter. But, for a rigor­ous sci­ence of cur­ing peo­ple, we want to know what makes peo­ple get bet­ter even more re­li­ably than just be­liev­ing they will.

Like­wise, there are prac­tices that can make peo­ple like some­thing. But there’s no point to say­ing, “Hey, af­ter this prac­tice, peo­ple like it!” That con­veys no in­for­ma­tion—it’s true for ev­ery­thing. Like with placebo cures, I want to know what is good above and be­yond that that re­sults from stan­dard “make some­thing seem good” tricks.

And there is a differ­ence: I liked choco­late be­fore I knew what any­one thought about it. In con­trast, very few peo­ple liked al­co­hol be­fore they found so­cial or non-taste rea­sons to drink it.

If all you care about is the fi­nal level of lik­ing, why not spend all this effort mak­ing your­self like healthy foods? And why the broad re­luc­tance for peo­ple to ad­mit, “okay, wine is re­ally just about show­ing off sta­tus”? Why do I have to pry teeth to get any­one to talk about this?

• The rea­son peo­ple are re­luc­tant to ad­mit it is be­cause you are sim­ply wrong. I like beer bet­ter than wine, even though wine has higher so­cial sta­tus and greater psy­cholog­i­cal effects. I would drink beer in pri­vate if it had the same taste but no al­co­hol, and I would definitely pre­fer it to a milk­shake, on taste alone.

What makes you so re­luc­tant to ad­mit that some peo­ple might have differ­ent tastes from you?

• What makes you so re­luc­tant to ad­mit that some peo­ple might have differ­ent tastes from you?

The in­ves­ti­ga­tion doc­u­mented here led me to re­ject that ini­tial, more ob­vi­ous and prob­a­ble the­ory.

• My ques­tion was some­what rhetor­i­cal, in re­sponse to your “Why the broad re­luc­tance...”

In fact I read through the thread that you link to and found it quite un­per­sua­sive.

It’s true it’s some­what sur­pris­ing that so many peo­ple said they preferred the taste of milk­shake. But in re­al­ity that’s partly a ques­tion of con­text. If you’re com­par­ing the taste of sweet things with the taste of non-sweet things, it can de­pend on what you feel like at the mo­ment. Some­times you have a de­sire for sugar, some­times you don’t.

• Did you read it all? It wasn’t just the milk­shake com­par­i­son. It was the fact that, if you ig­nore the ques­tion “do you like al­co­hol?” and sim­ply ask about the sup­posed im­pli­ca­tions of lik­ing al­co­hol, my an­swers match up with ev­ery­one who claimed to like al­co­hol. Yet I char­ac­ter­ize my state as “not lik­ing al­co­hol”, while oth­ers char­ac­ter­ize it as the re­verse.

See the check­list.

Again, the point is to sub­tract away the in­fluence of fac­tors that can make you like any­thing. If ap­ple­juice made me happy and kil­led my usual in­hi­bi­tions, I’d “like it”. I might even get over the taste. I might even show off my pick­i­ness about which ap­ples must be used be­fore I will con­sider to drink it.

But this is a HUGELY differ­ent sense of lik­ing than ex­ists for a milk­shake. Or milk. Or smooth­ies. Or mocha pep­per­mint frap­pu­ci­nos. Or any of the other things that I didn’t have to con­sume many, many times to fi­nally de­cide I like the taste of.

• The check­list doesn’t seem very strong ev­i­dence to me:

“-Think milk­shakes are bet­ter tast­ing than the best al­co­holic drink.” I don’t think this. And even for peo­ple who do, many peo­ple like the taste of some things more than oth­ers, with­out dis­lik­ing the taste of the lat­ter.

“-En­joy the taste of al­co­holic drinks when it is drowned out with some other fla­vor.” Sure, if it’s a good fla­vor. But I also en­joy the taste of the al­co­holic drinks when it isn’t drowned out at all.

“-Believe it changes our men­tal states in a good way.” Pos­si­bly, but this doesn’t show that it wouldn’t taste good with­out this effect.

“-Could not com­fortably chug down a al­co­holic drink the way we might a milk­shake.” I think this hap­pens with strong drinks be­cause the al­co­hol causes a cough­ing re­flex, not be­cause of the taste. But I can definitely drink a beer com­fortably just as fast as a milk­shake, and I can do the same with wine if a lit­tle wa­ter is added (and it still tastes like wine, in­di­cat­ing that it isn’t a ques­tion of taste.)

• “-Could not com­fortably chug down a al­co­holic drink the way we might a milk­shake.” I think this hap­pens with strong drinks be­cause the al­co­hol causes a cough­ing re­flex, not be­cause of the taste.

Okay, I hope state­ments like this show what I’m deal­ing with on this topic. We have sub­stances that pro­voke the chok­ing re­flex in peo­ple, as your body protests against this sub­stance en­ter­ing you, just as it would for toxic smoke, clean­ing fluid, and en­g­ine oil, and yet peo­ple ca­su­ally ig­nore that and say with a straight face, “oh, what a plea­sure it is for me to drink this deli­cious bev­er­age! Why would not oth­ers so en­joy it?”

• ...then why do I put hot sauce on my bur­rito?

• Do you drink the hot sauce di­rectly? Do you put so much on that it pro­vokes a chok­ing or winc­ing re­ac­tion? Then I don’t think it’s com­pa­rable.

ETA: Oh, one more rhetor­i­cal que­si­ton: Do you act sur­prised that there are peo­ple who aren’t will­ing to pay in­sane prices to in­jest bur­ri­tos with so much hot­sauce that they have to suffer through eat­ing it?

Be­cause that’s what it would take for me to have the same per­plex­ion as I do about al­co­hol.

• I don’t—I’d choke or wince, and I don’t want that. But I still like hot sauce on my bur­rito.

What I am ar­gu­ing—and I be­lieve this was Un­knowns’ ar­gu­ment—is that the effect of in­creas­ing rate of in­take is not in­dica­tive of whether a sub­stance is en­joy­able at the lower rate of in­take. I wouldn’t eat a tray of lemon squares, but I’d eat one piece.

• What I am ar­gu­ing—and I be­lieve this was Un­knowns’ ar­gu­ment—is that the effect of in­creas­ing rate of in­take is not in­dica­tive of whether a sub­stance is en­joy­able at the lower rate of in­take. I wouldn’t eat a tray of lemon squares, but I’d eat one piece.

Okay, give me a lit­tle credit here. I “get” that much—I mean, even a milk­shake will give you a brain­freeze.

The point is (and I ad­mit I’ve had a hard time ex­press­ing it with ex­am­ples be­cause of the con­found­ing fac­tors), peo­ple strangely start to use a defi­ni­tion of “en­joy drink­ing X” that ex­pands to cover as­pects that they ad­mit are very dis­plea­surable. Hard liquors will in­duce the cough­ing re­flex (the be­gin­ning of it), for ex­am­ple, even at very low rates of con­sump­tion.

This would seem to dom­i­nate the ex­pe­rience, but then, even in the midst of what is quite clearly painful, they en­joy it—and are some­how able to dis­cern “good” hard liquor from “bad” hard liquor.

Tak­ing the whole ex­pe­rience into ac­count, I can ac­cept that there’s a lot to like—just not the act of drink­ing.

• Might dill pick­les be a use­ful ex­am­ple? I had to be co­erced into try­ing them sev­eral times be­fore I came to find them ed­ible, but I en­joy them now, and there’s not much if any sta­tus in­volved there.

• Dill pick­les don’t have nearly the same per­plex­ity fac­tors that al­co­hol does, so I don’t think they’re a use­ful ex­am­ple.

• Peo­ple aren’t ul­tra-par­tic­u­lar about which dill pick­les they like, be­yond them not look­ing gross.

• Peo­ple don’t claim to be able to dis­cern all the differ­ences.

• The taste of dill pick­les doesn’t serve as a con­ve­nient ex­cuse for get­ting high.

• Dill pick­lers aren’t reg­u­larly used to get high, and aren’t in dan­ger of be­ing banned or over­reg­u­lated.

You get the point.

• I’m not sure that those fac­tors can be fully, or even partly, sep­a­rated from sta­tus sig­nal­ing. For ex­am­ple, I ex­pect that I could tell the differ­ence be­tween differ­ent kinds of pick­les, and de­velop a fa­vorite among the brands that ex­ist. I have no par­tic­u­lar rea­son to do so, and if I did, I wouldn’t talk about it, but if pick­les be­came trendy, and the pickle com­pa­nies started mak­ing sub­tly-differ­ent types to satisfy the de­mand for sig­nal­ing tools, I prob­a­bly would at least try the va­ri­eties and pick a fa­vorite. (I have a fa­vorite brand of may­on­naise, af­ter all, and am that picky about which brand of Ma­cadamia nuts I’ll eat.)

• You as­sume that the bad effects will dom­i­nate, but I’m not sure that would be the case. If you like the taste it­self enough, that might bal­ance the bad effect. And “good” hard liquor might (and does, in my limited ex­pe­rience) re­duce the bad effects.

Fur­ther, com­par­a­tively few peo­ple like (neat) hard liquor, and the di­rect un­pleas­ant effects are sig­nifi­cantly re­duced in beer, wine, and mixed drinks.

I agree with you that sta­tus con­sid­er­a­tions will of­ten make peo­ple in­clined to get an Ir­ish Cream when they’d pre­fer a milk­shake or fool them­selves into think­ing that ex­pen­sive wine tastes bet­ter. But you’re ap­par­ently mak­ing a strong claim (no­body re­ally likes the taste of al­co­holic bev­er­ages), on weak ev­i­dence.

• The point is (and I ad­mit I’ve had a hard time ex­press­ing it with ex­am­ples be­cause of the con­found­ing fac­tors), peo­ple strangely start to use a defi­ni­tion of “en­joy drink­ing X” that ex­pands to cover as­pects that they ad­mit are very dis­plea­surable. Hard liquors will in­duce the cough­ing re­flex (the be­gin­ning of it), for ex­am­ple, even at very low rates of con­sump­tion.

And hot sauce will in­duce a burn­ing sen­sa­tion at even very low con­cen­tra­tions of cap­saicin. Like BDSM, some­times peo­ple ac­tu­ally do like that.

• Right—some­times. Not “the over­whelming ma­jor­ity of the adult pop­u­la­tion, which also hap­pens to get high while do­ing so.” It’s the ubiquity, not just the strangeness, that con­fuses me.

• You’re sure of that ubiquity part? I just think you should put off en­dors­ing com­pli­cated be­liefs un­til you are sure they are based on good data. In this case, I be­lieve that means a proper so­ciolog­i­cal study.

Edit: Such a study may also make it eas­ier to con­firm the ex­tent of var­i­ous pro­posed mo­ti­va­tions.

• I don’t think we’re go­ing to need a so­ciolog­i­cal study to ver­ify that the vast ma­jor­ity of adults drink, and claim to like it when they do. That’s all I meant by ubiquity. I should have said “com­mon­ness” or some­thing equally awk­wardish.

• No, you seem to be claiming that they’re do­ing it for the taste. They do it in spite of the taste, or in­differ­ently to the taste, and add sweet fla­vors to mask it. How­ever, peo­ple can grow to like the taste over time, and some of the fla­vors in the taste are good by them­selves, though not as good as, say, a choco­late milk­shake.

You also seem to be as­sum­ing that a taste has to be re­pul­sive or deli­cious, in­stead of just neu­tral, or all right in cer­tain con­texts, or oc­ca­sion­ally de­sir­able.

• No, you seem to be claiming that they’re do­ing it for the taste.

Er, no, that’s pretty much the op­po­site of what I’m claiming. I’m claiming that they say they do it for the taste, but mainly (or solely) want the psy­choac­tive effects.

• I’m claiming that they say they do it for the taste

Sorry for get­ting it wrong. Any­way, I don’t think they say that they pri­mar­ily do it for the taste, which is an em­piri­cal ques­tion.

I think they say they “like” it, and they mean they like the over­all ex­pe­rience, and you’re in­ter­pret­ing that to mean that they like the taste. Or they say they like the taste be­cause they grew to like it over time, or be­cause they mix it with other fla­vors, and you’re in­ter­pret­ing that to mean they pri­mar­ily drink it for the taste.

• Do you act sur­prised that there are peo­ple who aren’t will­ing to pay in­sane prices to in­jest bur­ri­tos with so much hot­sauce that they have to suffer through eat­ing it?

I am not sur­prised when some­one does pay to do such a thing to their bur­rito.

• Even if it were very com­mon, and a prac­tice con­cen­trated in the top 10% wealthiest peo­ple?

• I think it is quite com­mon for peo­ple to eat food that is hot enough to cause dis­com­fort or even pain, at least in some cul­tures. The un­com­fortably-hot curry is a Bri­tish tra­di­tion that of­ten goes hand in hand with the con­sump­tion of beer. In my non-sci­en­tific per­sonal ex­pe­rience will­ing­ness to eat (and en­joy) food with lev­els of heat that cause dis­com­fort cor­re­lates some­what with wealth/​sta­tus—it can be seen as a marker of open­ness to ex­pe­rience and em­brac­ing cul­tural di­ver­sity.

• Peo­ple get de­sen­si­tized to hot sauce af­ter a while; it takes more to cause dis­com­fort in some­one who rou­tinely eats hot sauce than in some­one who doesn’t.

I’m fond of spicy food. (My father, who I sus­pect is a su­per­taster, isn’t.)

• Peo­ple do get de­sen­si­tized to hot/​spicy food over time but I think peo­ple who en­joy the sen­sa­tion tend to in­crease the dosage to com­pen­sate. Speak­ing from per­sonal ex­pe­rience, I still like hot food to burn slightly, it just takes more chilli than it used to to achieve that. The burn­ing/​dis­com­fort isn’t an un­for­tu­nate side effect of the pleas­ant taste of chillis for me, it’s an es­sen­tial com­po­nent of the en­joy­ment of eat­ing hot food. I’ve heard that the rea­son peo­ple en­joy spicy food is that chilli stium­lates pain re­cep­tors and causes the re­lease of en­dor­phins and it is the en­dor­phin re­lease that peo­ple crave but I don’t know if that is true.

• Peo­ple do get de­sen­si­tized to hot/​spicy food over time but I think peo­ple who en­joy the sen­sa­tion tend to in­crease the dosage to com­pen­sate.

Yes, that’s what I meant.

Another thing I’ve no­ticed is that some hot pep­pers have good fla­vors in ad­di­tion to the burn­ing sen­sa­tion (jalapenos, for ex­am­ple), but oth­ers seem to be prac­ti­cally taste­less apart from it.

• Mmm… I think I missed some­thing. How I would I stop be­ing not sur­prised if it were a com­mon prac­tice?

Uh, I mean, why would I start be­ing sur­prised if it were a com­mon prac­tice [to pay in­sane prices to in­ject bur­ri­tos...]?

• I grant the re­flex is a way of your body protest­ing. I just don’t think it has to do with TASTE. And I gave ev­i­dence for that from the the fact that if it is diluted, it has the same taste, but not the same protest.

Also, this re­flex is differ­ent from nau­sea, which I would ad­mit would be a protest to the taste, and if I dis­like the taste of some­thing suffi­ciently, it causes nau­sea in me. Noth­ing like this hap­pens with al­co­holic drinks.

• Dilu­tion doesn’t change the taste of a drink, and al­co­hol doesn’t cause nau­sea …

And I’m the one that’s ra­tio­nal­iz­ing a re­futed po­si­tion?

• 1)dilu­tion weak­ens the taste (and the other effects like the chok­ing protest), but it doesn’t change it to an­other taste; 2) I’m not talk­ing about get­ting drunk, I’m talk­ing about the effect at the mo­ment of drink­ing it.

• Would some­one like to make a falsifi­able claim about how a per­son is likely to re­act to al­co­hol over their first few in­stances of drink­ing it? If so, I’d be will­ing to be a guinea pig.

The only times I’ve had al­co­hol were over a decade ago, and in­volved ei­ther hav­ing com­mu­nion at church or my father in­sist­ing that I take a sip of his beer. I’ve never ex­pe­rienced an al­co­hol buzz. I dis­like be­ing in the kinds of situ­a­tions in which one would drink so­cially, but am cu­ri­ous about how al­co­hol might af­fect me sep­a­rately from that. I do find the smell of wine and beer aver­sive (but not nau­se­at­ing), which I un­der­stand might af­fect the out­come, but I’d be will­ing to try them any­way. (I’d been con­sid­er­ing try­ing wine cool­ers, but, hey, it’s for sci­ence.)

• Given that you don’t like the smell of beer and wine, you likely won’t like the taste at first ei­ther. But some peo­ple do like the taste even the first time, so this isn’t strong ev­i­dence for SilasBarta’s po­si­tion.

You might like wine cool­ers which tend to be a lit­tle sweeter. Ac­tu­ally, one thing cor­rect in Silas’s po­si­tion is that many peo­ple like sweet drinks be­cause they are sweet, not be­cause of the al­co­hol, and they are some­times un­will­ing to ad­mit this for so­cial rea­sons.

• Sure, I’d love to make such a pre­dic­tion, but those who dis­agree with me know all too well what the re­sult will be and will try to ra­tio­nal­ize away the pre­dictable re­sults of you try­ing al­co­hol … oops, too late.

• I don’t think any­one dis­putes that peo­ple usu­ally don’t like al­co­hol the first time they try it. I’m dis­put­ing that, af­ter lik­ing it, there’s a differ­ence be­tween this lik­ing of al­co­hol and the lik­ing of your apt ex­am­ple, milk­shakes. The two likes are the same.

• No, they’re not the same, be­cause you have to go through a pro­cess to like al­co­hol, which would just the same cause you to like bat urine. You don’t have to do that for milk­shakes.

• After that pro­cess has hap­pened, they’re the same.

• Yes, once you use a pro­cess that will cause peo­ple to like ANYTHING, in­clud­ing bat urine, it will cause them to like al­co­hol.

If you look closely, that state­ment has no in­for­ma­tion con­tent, and it’s equiv­a­lent to yours.

• “Any­thing” is too broad. If you made drink­ing bat urine suffi­ciently high-sta­tus, or gave it some re­ward other than taste, then yes; oth­er­wise, no.

EDIT: Taste in this case = “plea­sure from first taste”

• Good thing al­co­hol doesn’t have other re­wards (like pleas­ant men­tal states) or im­pacts or your sta­tus.

Other­wise, the situ­a­tions might be par­allel!

• You are bring­ing your own as­sump­tions into it, as well, like that “any­thing” isn’t cyanide.

• So make pre­dic­tions about what will hap­pen af­ter I’ve had N drinks. Or what would hap­pen af­ter I had N drinks of near-beer, if it’s the psy­cholog­i­cal effect you’re con­cerned with and not just the so­cial one.

• my an­swers match up with ev­ery­one who claimed to like al­co­hol.

I com­pletely agree with you about wine tast­ing speci­fi­cally. But there are those of us who ac­tu­ally like the taste of some al­co­holic drinks, even with­out the psy­cholog­i­cal effects, sig­nal­ing, or need to ac­quire the taste. It doesn’t look like your an­swers match up with that.

• I made my main point in the other com­ment, and I don’t want to in­clude these two com­ments to­gether be­cause I don’t want the other to be ig­nored, but health is an ob­jec­tive mea­sure, whereas plea­sure is not.

First of all, I think you’re ig­nor­ing that there are some prac­tices that, de­spite mak­ing some peo­ple like an ac­tivity, will not make other peo­ple like the ac­tivity—i.e., that placebo will work on some peo­ple, but not other peo­ple, so to that ex­tent, there is some­thing marginally “real” (un­der your defi­ni­tion) there.

I un­der­stand what you mean very much; I’ve spent a ridicu­lous amount of time think­ing about it over the past decade. Cog­ni­tive dis­so­nance seems like a weak trait when you no­tice it in some­one else, ei­ther to change your val­ues to dis­like the in­ac­cessible or the re­verse, to change your val­ues to like the ac­cessible.

But why? I tend to like things I’m bet­ter at than most of the peo­ple I know, like math and ar­gu­ing and point­ing out other peo­ple’s cog­ni­tive dis­so­nance. Why should I ex­pect other peo­ple to be any differ­ent?

In the end, the “lik­ing” part is re­ally, like you pointed out, lik­ing the taste of sta­tus more than the taste of al­co­hol. But I en­joy spicy food, de­spite not lik­ing it origi­nally, ei­ther. I didn’t like hip-hop, but I figured there must be some­thing there that at­tracts so many peo­ple; now I like some. I didn’t like a bunch of pop­u­lar TV shows, but I didn’t want to as­sume that all the ways I’m differ­ent from peo­ple who did like those TV shows were ways I was bet­ter; what if they were ways I was worse? So I watched a bunch of them. Most of them still suck, but I found I like House and Big Love, de­spite think­ing be­fore­hand only idiots could like those shows.

I agree with you to some ex­tent—if I have to have some­one tel­ling me I’m cool for me to en­joy it, I don’t want to par­take. But that’s not be­cause it’s less “pure,” it’s be­cause I’ve done ac­tivi­ties like that be­fore and it’s not fun not be­ing in con­trol of when I can en­joy my­self.

I have been tel­ling peo­ple for years that the differ­ence be­tween peo­ple who like al­co­hol and peo­ple who don’t is that the peo­ple who don’t didn’t have a peer group to pres­sure them to drink their first 5 beers. That’s true with me and with al­most ev­ery­one I know (al­though there are some who claim they liked beer right away, and I even be­lieve a few of them).

But now even when I’m alone, do I en­joy hav­ing a beer and re­lax­ing? Yes, very much. Would I like beer if it weren’t for the al­co­hol’s effects of re­lax­ing me? Prob­a­bly not, no. But that doesn’t change that the al­co­hol has changed how much I en­joy the taste of beer, be­cause now it ac­tu­ally tastes good. That doesn’t seem dis­in­gen­u­ous to me. I don’t ex­pe­rience the en­joy­ment any less, so it’s hard for me to dis­credit it due to the fact that the way I got to that point was through try­ing to not look stupid to my friends when they gave me a beer for the first time.

• I have been tel­ling peo­ple for years that the differ­ence be­tween peo­ple who like al­co­hol and peo­ple who don’t is that the peo­ple who don’t didn’t have a peer group to pres­sure them to drink their first 5 beers. That’s true with me and with al­most ev­ery­one I know (al­though there are some who claim they liked beer right away, and I even be­lieve a few of them).

Yes, I think that’s what my point comes down to: so you like beer af­ter be­ing pres­sured by friends to drink it for five years. Then, by sim­ple force of habit, you come to like it—your tastes change.

But I say: so what? What does that tell me about beer? Like I keep say­ing, if you go through this (five years of drink­ing it with friends who pres­sure you to drink it) with any drink, you will end up “lik­ing” it. So there’s noth­ing about the hops or the spe­cial micro­brew­ery or the yeast or this or that. It’s com­pletely ar­bi­trary.

I would much pre­fer to drink some­thing that ac­tu­ally tastes good. If I want to fur­ther en­hance this with a group of friends, great. But stop tel­ling me beer tastes good. Keep­ing up with habits you’ve de­vel­oped in pleas­ant situ­a­tions is what “tastes” good. The psy­choac­tive effects of a so­cially-ac­cept­able product “taste” good. Beer, how­ever, does not taste good.

ETA: Similarly, Homer’s The Odyssey isn’t good. Rather, a bunch of peo­ple have a tra­di­tion of read­ing it that they pass on and get the next gen­er­a­tion to per­pet­u­ate. But what the heck am I sup­posed to learn about good writ­ing from that, other than:

1) Here are some refer­ences you can make that you can ex­pect peo­ple to “get”

2) If you want to start a cult, here are some things you can do that will trick peo­ple into lik­ing your holy texts.

• I would much pre­fer to drink some­thing that ac­tu­ally tastes good. If I want to fur­ther en­hance this with a group of friends, great. But stop tel­ling me beer tastes good. Keep­ing up with habits you’ve de­vel­oped in pleas­ant situ­a­tions is what “tastes” good. The psy­choac­tive effects of a so­cially-ac­cept­able product “taste” good. Beer, how­ever, does not taste good.

It looks to me like you’re try­ing to curry the 2-place pred­i­cate “tastes good to X” into a 1-place pred­i­cate “tastes good”, with­out re­ally spec­i­fy­ing the X that you’re sup­ply­ing as an ar­gu­ment. Surely X isn’t “ev­ery­one”. And it can’t just be “many/​most peo­ple”, since you’ve at­tached other con­di­tions (like “psy­choac­tive taste changes don’t count”).

In my ex­pe­rience, most things taste differ­ent the sec­ond or third time around. The stom­ach and in­testines are con­nected to the ner­vous sys­tem, ya know—you get di­rect neu­ral feed­back on the things you put in your body. If that feed­back is nega­tive, you might find that sub­stance A isn’t quite so tasty the sec­ond time around. Does any mod­ifi­ca­tion of a taste count as a “re­fine­ment” in your eyes, rather than an ar­bi­trary change?

Anec­dote: I con­cluded long ago that coffee tasted like dirt, and I wanted no part of it. Once ev­ery few years, some­one would say “oh, try this, it’s good coffee!”, and I’d try it, and it tasted like dirt. Then a cou­ple years ago I was eat­ing break­fast at a diner and de­cided to just choke down some free caf­feine. It tasted like dirt, but it did the job. A few days later, I forced down an­other cup of dirt­wa­ter. The next day I felt a sud­den, sharp crav­ing for coffee. That third cup of coffee tasted vastly bet­ter than the first two. I think the mechanism is ob­vi­ous enough.

But the drug-in­duced shift in my per­cep­tion of coffee wasn’t as sim­ple as “it used to taste bad, now it tastes good”. Pre­vi­ously, all coffee tasted the same to me. As soon as coffee stopped tast­ing like dirt, the differ­ences be­tween types of coffee be­came much more no­tice­able. I’m no con­noisseur, but you’ll have a hard time con­vinc­ing me that I pre­fer a French-pressed Su­ma­tran blend to the drip coffee available at my job be­cause of peer pres­sure.

There was a TED talk (I don’t re­call which one) where a neu­ro­scien­tist spoke about some fMRI re­search done on mon­keys perform­ing a task re­quiring dex­ter­ity in the hands and fingers. They found that the re­gion of the mon­key’s brain mapped to its rele­vant digits ba­si­cally grew with prac­tice. This “adap­tive re­s­olu­tion” as­pect of per­cep­tion means if you re­strict X to “tastes good the first time” in an at­tempt to filter out noise from sta­tus games, you’ll also throw away ev­ery­thing that can’t be eas­ily per­ceived with the ini­tial chunk of al­lo­cated brainspace.

To sum­ma­rize, your ver­sion of “tastes good” ap­pears to be an over­sim­plifi­ca­tion. Our taste sense seems to be quite in­her­ently adap­tive. So I dis­agree that “there’s noth­ing about the hops or the spe­cial micro­brew­ery or the yeast or this or that”. I think your per­cep­tion of the sub­tler qual­ities of beer is prob­a­bly just too low-res, be­cause your ini­tial re­ac­tion to the taste of al­co­hol is pre­vent­ing your brain from al­lo­cat­ing ad­di­tional re­sources to fla­vor de­cod­ing.

In­ci­den­tally, choco­late is one of the foulest-tast­ing things I’ve had the mis­for­tune of plac­ing against my tongue. Dark, milk, cheap, fancy—it all tastes the same.

• In­ci­den­tally, choco­late is one of the foulest-tast­ing things I’ve had the mis­for­tune of plac­ing against my tongue. Dark, milk, cheap, fancy—it all tastes the same.

Heresy!

• In­ci­den­tally, choco­late is one of the foulest-tast­ing things I’ve had the mis­for­tune of plac­ing against my tongue. Dark, milk, cheap, fancy—it all tastes the same.

I hate a lot of choco­late, too! I can’t stand Her­shey’s Kisses, choco­late cake, choco­late milk, M&Ms, or choco­late ice cream. I do like hot choco­late, choco­late chip cook­ies, Three Mus­ke­teers bars, and Nes­tle Crunch bars.

• Okay, thanks for ex­plain­ing all of that. It re­ally sheds light on the dy­nam­ics at play here. My thoughts:

1) Even if what you’re say­ing is true, about the brain al­lo­cat­ing more mass to a given ac­tivity the more you do it, giv­ing a plau­si­ble mechanism for greater abil­ity to dis­t­in­guish coffees, that still doesn’t differ­en­ti­ate it from bat urine. We can ex­pect the same thing would go on there. Once you’re ac­cus­tomed to bat urine, you’ll be able to tell all the differ­ent kinds apart, you’ll have a newfound ap­pre­ci­a­tion for its “taste”, etc., all be­cause of your neu­ral plas­tic­ity.

So it still comes back to my origi­nal ques­tion: given this strange path to a per­son’s judg­ment that they like wine/​coffee/​bat urine, what is the ap­pro­pri­ate way to de­scribe this kind of lik­ing? Are we use­fully carv­ing con­ceptspace by putting this kind of lik­ing with milk­shakes, which most peo­ple like the first time, and all sub­se­quent times? Is the lik­ing-bat-urine a differ­ent phe­nom­e­nal ex­pe­rience than lik­ing-milk­shakes?

2) Even though your ac­count of the chang­ing taste for coffee may be right, are you sure about the sen­si­tivity to nu­ances? Have you given your­self blind taste tests for ran­dom beans? Keep in mind, that when sci­en­tific con­trols are in place, wine “ex­perts” in­evitably fail mis­er­ably to make the dis­tinc­tions they claim are im­por­tant.

It’s ac­tu­ally not that un­ex­pected to dis­like the office’s coffee in fa­vor of your own. I’m still at the stage of not lik­ing coffee un­less it’s ul­tra-sweet­ened (frap­pu­ci­nos ftw), and even I can tell what’s bad coffee. Not nec­es­sar­ily the taste, as the fact that bad coffee, um, dou­bles as a lax­a­tive.

3) My taste in beer hasn’t changed de­spite drink­ing it for ten years. The best I can say about any beer is that it “doesn’t hurt that much go­ing down”. (Guiness wins in this re­gard.) The best ex­pla­na­tion seems to be that my “su­per­tast­ing” abil­ity makes me very sen­si­tive to the al­co­hol, blur­ring out any other taste, and keep­ing me from adapt­ing to the nu­ances. The prob­lem, though, is that su­per­tasters are es­ti­mated at 25% of the pop­u­la­tion. So why aren’t 25% of peo­ple voic­ing my opinion on al­co­hol? Why would they stay silent about hat­ing it, while drink­ing it for so­cial and psy­choac­tive benefits?

• 1) Even if what you’re say­ing is true, about the brain al­lo­cat­ing more mass to a given ac­tivity the more you do it, giv­ing a plau­si­ble mechanism for greater abil­ity to dis­t­in­guish coffees, that still doesn’t differ­en­ti­ate it from bat urine. We can ex­pect the same thing would go on there. Once you’re ac­cus­tomed to bat urine, you’ll be able to tell all the differ­ent kinds apart, you’ll have a newfound ap­pre­ci­a­tion for its “taste”, etc., all be­cause of your neu­ral plas­tic­ity.

I don’t think lik­ing is in­her­ently tied to differ­en­ti­a­tion. It seems more like shift­ing your fo­cus—you’re per­ceiv­ing fun­da­men­tally new taste data, some of which you may find pleas­ant. I doubt that be­com­ing an ex­pert bat urine taster would im­part much love for the urine rel­a­tive to do­ing the same for coffee or beer. If Seth Roberts is right, we en­joy the com­plex fla­vors of fer­mented stuff like beer be­cause they’re mark­ers for valuable biotic di­ver­sity. The same is prob­a­bly not true of bat urine.

2) Even though your ac­count of the chang­ing taste for coffee may be right, are you sure about the sen­si­tivity to nu­ances? Have you given your­self blind taste tests for ran­dom beans? Keep in mind, that when sci­en­tific con­trols are in place, wine “ex­perts” in­evitably fail mis­er­ably to make the dis­tinc­tions they claim are im­por­tant.

I wouldn’t call the things I’m sen­si­tive to “nu­ances”—the differ­ence in fla­vor be­tween “Su­ma­tran” and “other” beans I’ve tried seems pretty ma­jor. There are prob­a­bly other similar beans that would be in­dis­t­in­guish­able, my prefer­ences on the sub­ject are qu­a­sitran­si­tive at best. I haven’t tested it, nor have I tested my abil­ity to dis­t­in­guish ale from lager.

The prob­lem, though, is that su­per­tasters are es­ti­mated at 25% of the pop­u­la­tion. So why aren’t 25% of peo­ple voic­ing my opinion on al­co­hol? Why would they stay silent about hat­ing it, while drink­ing it for so­cial and psy­choac­tive benefits?

In­ter­est­ing, I’d never heard of su­per­tasters be­fore. I don’t see what the prob­lem is here, though. You ap­pear to be seek­ing a con­cept of “gen­uine” fla­vor, and you’ve ruled out psy­cholog­i­cally adapted tastes. But that’s a bit tan­gent to the situ­a­tion where some­one starts out dis­lik­ing beer and ac­quires a taste for it from psy­choac­tive re­in­force­ment. They still prob­a­bly end up with higher-res per­cep­tion of it than some­one who hates the taste of it and drinks it any­way. Note the differ­ence be­tween “drink­ing for psy­choac­tive effect” and “drink­ing be­cause pre­vi­ous psy­choac­tive effects led to a mod­ified per­cep­tion of fla­vor”. Peo­ple in the lat­ter cat­e­gory have no cause for com­plaint (ex­cept for al­co­holics, of course).

If they’re re­ally drink­ing it for so­cial benefits, the mo­ti­va­tion to stay silent is prob­a­bly also so­cial benefits.

• I don’t think lik­ing is in­her­ently tied to differ­en­ti­a­tion.

I didn’t say it was. I ac­cept that you also started lik­ing the taste it­self; I just claim that this would hap­pen for any­thing, in­clud­ing bat urine, so I don’t put it in the same class of stuff that tastes good be­fore sig­nifi­cantly mold­ing your mind to make it so.

If Seth Roberts is right, we en­joy the com­plex fla­vors of fer­mented stuff like beer be­cause they’re mark­ers for valuable biotic di­ver­sity. The same is prob­a­bly not true of bat urine.

Sounds like a de­spised “just-so” story to me. You can just as well find mark­ers of biotic di­ver­sity in bat urine (at the very least, di­a­betic bat urine) that de­rives from the va­ri­ety in their diet, and the differ­ent kinds of bats, etc.

I wouldn’t call the things I’m sen­si­tive to “nu­ances”

I know. I was refer­ring to your newfound “abil­ity to dis­t­in­guish differ­ences to a higher de­gree of pre­ci­sion” and didn’t know a shorter term. Please don’t crit­i­cize some­one’s ter­minol­ogy un­less you offer an al­ter­nate, su­pe­rior term that you would not ob­ject to. Like what I did in a differ­ent dis­cus­sion over here in point 2.

I haven’t tested it,

Okay. Scien­tists have, though, and usu­ally you can get away with swap­ping out “high qual­ity” stuff for low qual­ity stuff and peo­ple won’t no­tice. They will throw a sta­tus-driven hissy fit if they find out what you did, though.

If they’re re­ally drink­ing it for so­cial benefits, the mo­ti­va­tion to stay silent is prob­a­bly also so­cial benefits.

So would you agree that my the­sis is at least ac­cu­rate for a por­tion of the pop­u­la­tion? (The the­sis was, “Peo­ple don’t re­ally like the taste but use the sup­posed taste and other rea­sons as an ex­cuse for get­ting high in a so­cially ac­cept­able way and keep­ing it le­gal to do so.”)

• I don’t think lik­ing is in­her­ently tied to differ­en­ti­a­tion.

I didn’t say it was.

True, what you said was

Once you’re ac­cus­tomed to bat urine, you’ll be able to tell all the differ­ent kinds apart, you’ll have a newfound ap­pre­ci­a­tion for its “taste”, etc., all be­cause of your neu­ral plas­tic­ity.

which I read as im­ply­ing that differ­en­ti­a­tion causes “lik­ing” (“in­her­ently tied” was im­pre­cise ter­minol­ogy on my part). What did you ac­tu­ally mean?

Sounds like a de­spised “just-so” story to me.

Uh, taste as an evolu­tion­ar­ily-shaped nu­tri­tion-de­tec­tor isn’t ex­actly a novel just-so hy­poth­e­sis. If your real ob­jec­tion is with the as­ser­tion of com­plex fla­vor prefer­ences or the link be­tween such fla­vors and biotic di­ver­sity, I don’t know what call­ing it a “just-so story” even means. You were prob­a­bly look­ing for a slightly less gen­eral re­tal­i­ate but­ton.

You can just as well find mark­ers of biotic di­ver­sity in bat urine (at the very least, di­a­betic bat urine) that de­rives from the va­ri­ety in their diet, and the differ­ent kinds of bats, etc.

Valuable biotic di­ver­sity. The kind of stuff that gar­ners pos­i­tive feed­back from the tract.

I know. I was refer­ring to your newfound “abil­ity to dis­t­in­guish differ­ences to a higher de­gree of pre­ci­sion” and didn’t know a shorter term. Please don’t crit­i­cize some­one’s ter­minol­ogy un­less you offer an al­ter­nate, su­pe­rior term that you would not ob­ject to.

I wasn’t “crit­i­ciz­ing your ter­minol­ogy”, I was at­tempt­ing to cor­rect a per­ceived mi­s­un­der­stand­ing in progress. You used the word “nu­ance” and then went on to talk about dou­ble-blind taste tests, which taken to­gether led me to be­lieve that I hadn’t effec­tively com­mu­ni­cated the scale of dis­tinc­tion I had in mind. Hence the com­par­i­son to ale and lager. I’m well-aware of wine snobs and their em­bar­rass­ing track records.

As­sum­ing that my ter­minolog­i­cal cor­rec­tion is some in­effec­tual, off-topic crit­i­cism of your choice of words is as­sum­ing I’m ba­si­cally act­ing in bad faith. Not very pro­duc­tive.

So would you agree that my the­sis is at least ac­cu­rate for a por­tion of the pop­u­la­tion?

Yes.

• True, what you said was … which I read as im­ply­ing that differ­en­ti­a­tion causes “lik­ing” (“in­her­ently tied” was im­pre­cise ter­minol­ogy on my part). What did you ac­tu­ally mean?

I was list­ing the differ­en­ti­a­tion, and the lik­ing of taste, as two sep­a­rate phe­nom­ena, with any pos­si­ble causal re­la­tion­ship, not nec­es­sar­ily the differ­en­ti­a­tion caus­ing the en­joy­ment.

Uh, taste as an evolu­tion­ar­ily-shaped nu­tri­tion-de­tec­tor isn’t ex­actly a novel just-so hy­poth­e­sis.

Yes, we do have (what can be called) nu­tri­tion de­tec­tors, but none of them work any­thing like what would have to be pre­sent for the one you posited: 1) in the EEA, we didn’t nor­mally taste the in­gre­di­ents of beer, 2) 25% of the pop­u­la­tion is dis­tracted by the taste of al­co­hol and un­able to use the in­for­ma­tion, 3) the nu­tri­tion de­tec­tors we do have evoke pleas­ant re­sponses in al­most ev­ery­one, from a very young age (i.e. aren’t ac­quired tastes).

I call it a “just so story” be­cause it doesn’t pass many ob­vi­ous san­ity checks.

Valuable biotic di­ver­sity. The kind of stuff that gar­ners pos­i­tive feed­back from the tract.

None of the things in beer “gar­ner pos­i­tive feed­back from the tract”. And knowl­edge of what fruits and meats the bats in the area are able to eat would definitely sig­nal the di­ver­sity in the area. If you meant GI tract mi­coor­ganisms, beer came around way too late, and is way too dis­similar to other things we con­sume to have been adapted for as a gauge of use­ful di­ver­sity.

I wasn’t “crit­i­ciz­ing your ter­minol­ogy”, I was at­tempt­ing to cor­rect a per­ceived mi­s­un­der­stand­ing in progress.

What is the brief ap­pel­la­tion you be­lieve I should have used to de­scribe what I was refer­ring to? If you don’t have one, you should have ac­cepted the speci­fic­ity/​brevity trade­off I made in try­ing to sum­ma­rize what you just said, and re­sponded to the sub­stance of the point, say­ing what I got wrong there.

If you do have one, you just passed up your sec­ond op­por­tu­nity to be helpful by tel­ling it to me. What’s your goal here?

As­sum­ing that my ter­minolog­i­cal cor­rec­tion is some in­effec­tual, off-topic crit­i­cism of your choice of words is as­sum­ing I’m ba­si­cally act­ing in bad faith.

No, tel­ling me what I did wrong with­out tel­ling me what would have been right, is bad faith, be­cause it leaves me in the po­si­tion of hav­ing to get per­mis­sion from you ev­ery time I want to briefly re­fer back to some­thing you said.

So would you agree that my the­sis is at least ac­cu­rate for a por­tion of the pop­u­la­tion?

Yes.

Okay, thank you. I just wish I didn’t have to pull teeth to talk about these things.

• “None of the things in beer “gar­ner pos­i­tive feed­back from the tract”″.

Not true. One of the things I like about beer is that when I’m hun­gry, it tastes REALLY good. It tastes like I’m eat­ing a meal. This doesn’t hap­pen with wine, which is just a drink.

• LOL! What’s funny is, this isn’t the first time I’ve heard this line of “rea­son­ing”.

“Okay okay, high-carb sub­stance X might not taste good, but it tastes REALLY good when you’re en­ergy starved (in con­trast to all those high-carb food/​drinks that don’t taste re­ally good in such a cir­cum­stance).”

• I men­tion it be­cause it tastes bet­ter than other high carb food and drinks in those cir­cum­stances. That’s a fact, at least re­gard­ing my taste.

And there’s re­ally some­thing wrong with your man­ner of ar­gu­ment, since you could say some­thing similar about any rea­son why any­one would say any­thing ever tastes good. You might as well say you dis­like the taste of milk­shake, but just like the effects of fat and sugar on your body, or some­thing like that.

• [re­ply­ing sep­a­rately to this tor­tured meta sub-thread]

What is the brief ap­pel­la­tion you be­lieve I should have used to de­scribe what I was refer­ring to? If you don’t have one, you should have ac­cepted the speci­fic­ity/​brevity trade­off I made in try­ing to sum­ma­rize what you just said, and re­sponded to the sub­stance of the point, say­ing what I got wrong there.

Look, it wasn’t clear to me at all that you were mak­ing such a trade-off. I wouldn’t have men­tioned the word “nu­ance” at all if I thought you were you just ab­bre­vi­at­ing my in­tent. Mis­in­ter­pre­ta­tions are a dime a dozen in these sorts of con­ver­sa­tions, no need to take a re­trans­mit so per­son­ally.

What’s your goal here?

To have a clear ex­change of ideas. Do you sus­pect an­other?

No, tel­ling me what I did wrong with­out tel­ling me what would have been right, is bad faith, be­cause it leaves me in the po­si­tion of hav­ing to get per­mis­sion from you ev­ery time I want to briefly re­fer back to some­thing you said.

Em­pha­sis mine. You’re tak­ing it per­son­ally. It could just as eas­ily have been poor phras­ing on my part. I’m more in­ter­ested in en­sur­ing that the thing you read is the thing I’m try­ing to write than I am in figur­ing who’s to “blame” for some ter­minolog­i­cal “er­ror”.

• 1) in the EEA, we didn’t nor­mally taste the in­gre­di­ents of beer

Not sure what you mean here. In the EEA we could still prob­a­bly taste the rough sig­na­ture of a fer­men­ta­tion pro­cess.

2) 25% of the pop­u­la­tion is dis­tracted by the taste of al­co­hol and un­able to use the information

Again you im­ply that su­per­tasters are un­able to get past their ini­tial re­ac­tion to the taste of al­co­hol, de­spite the ut­ter plau­si­bil­ity of psy­choac­tive re­in­force­ment lead­ing to a mod­ified sense of taste.

3) the nu­tri­tion de­tec­tors we do have evoke pleas­ant re­sponses in al­most ev­ery­one, from a very young age (i.e. aren’t ac­quired tastes).

Source? My tastes changed slowly but con­tinu­ally as I aged. Is your as­ser­tion that none of the com­mon shifts from child­hood to adult food prefer­ences are linked to nu­tri­tional con­tent?

I call it a “just so story” be­cause it doesn’t pass many ob­vi­ous san­ity checks.

If the above or­dered list con­sti­tutes your “ob­vi­ous san­ity checks”, then I ques­tion their ad­e­quacy. If you’re refer­ring to some other san­ity checks, I’d be in­ter­ested in hear­ing them.

To clar­ify, I’m not ac­tu­ally ad­vo­cat­ing Roberts’ the­ory. I brought it up be­cause I think it’s plau­si­ble, which is all that’s re­quired to doubt the coun­ter­in­tu­itive as­ser­tion that de­vel­op­ing a taste for bat urine is akin to de­vel­op­ing a taste for beer.

If you meant GI tract mi­coor­ganisms, beer came around way too late, and is way too dis­similar to other things we con­sume to have been adapted for as a gauge of use­ful di­ver­sity.

Came around way too late? Die­tary adap­ta­tions can be pretty rapid (e.g., adult lac­tose tol­er­ance). But I doubt your as­ser­tion that beer is too dis­similar to other things we con­sume—get­ting a mes­sage like “this is fer­mented, caloric, and not ob­vi­ously toxic” from your tongue is prob­a­bly good enough.

• Not sure what you mean here. In the EEA we could still prob­a­bly taste the rough sig­na­ture of a fer­men­ta­tion pro­cess.

Then you’d have to show how it has se­lec­tive power. What in­for­ma­tion is gained from the fer­men­ta­tion stage, and why would it shift our makeup so quickly?

Again you im­ply that su­per­tasters are un­able to get past their ini­tial re­ac­tion to the taste of al­co­hol, de­spite the ut­ter plau­si­bil­ity of psy­choac­tive re­in­force­ment lead­ing to a mod­ified sense of taste.

Not one that just hap­pens to line up with a con­voluted mechanism that just hap­pens to jus­tify lik­ing beer.

Source? My tastes changed slowly but con­tinu­ally as I aged. Is your as­ser­tion that none of the com­mon shifts from child­hood to adult food prefer­ences are linked to nu­tri­tional con­tent?

We change what we like, but we keep the cat­e­gory of sweet (de­tec­tion of sug­ars). There is no sci­en­tific sub­stan­ti­a­tion for a “fer­ment­ed­ness” cat­e­gory de­tec­tor: just sweet, salty, sour, bit­ter, and the re­cent meaty one. That gives a se­ri­ous pre­sump­tion against this kind of mechanism.

To clar­ify, I’m not ac­tu­ally ad­vo­cat­ing Roberts’ the­ory. I brought it up be­cause I think it’s plau­si­ble, which is all that’s re­quired to doubt the coun­ter­in­tu­itive as­ser­tion that de­vel­op­ing a taste for bat urine is akin to de­vel­op­ing a taste for beer.

Then I just have to show equal plau­si­bil­ity of the use­ful­ness of bat urine, which I’ve done. Di­a­betic bat urine con­tains sugar, which in turn con­tains sweet­ness, which in turn con­tains in­for­ma­tion in­for­ma­tion about the plants in the area. This re­sult can be ex­tended to nor­mal bat urine, in which the fruit con­tent of the area will de­ter­mine bat urine bit­ter­ness, which we would then “en­joy” drink­ing, just as peo­ple learn to “en­joy” beer’s bit­ter­ness.

And, as a bonus, urine was con­sumed for a sliver of our evolu­tion­ary his­tory.

Sure, it’s con­voluted and im­plau­si­ble, but good enough to keep con­sump­tion of bat urine nice and le­gal, which is re­ally all it has to do.

• Then you’d have to show how it has se­lec­tive power. What in­for­ma­tion is gained from the fer­men­ta­tion stage, and why would it shift our makeup so quickly?

Why would it need to be a quick shift?

Not one that just hap­pens to line up with a con­voluted mechanism that just hap­pens to jus­tify lik­ing beer.

Huh? In ear­lier com­ments you seemed to have no prob­lem with the idea that peo­ple de­vel­oped a taste for things that got them high, but now the idea is sus­pect be­cause it sup­ports an ex­pla­na­tion for lik­ing beer?

We change what we like, but we keep the cat­e­gory of sweet (de­tec­tion of sug­ars). There is no sci­en­tific sub­stan­ti­a­tion for a “fer­ment­ed­ness” cat­e­gory de­tec­tor: just sweet, salty, sour, bit­ter, and the re­cent meaty one. That gives a se­ri­ous pre­sump­tion against this kind of mechanism.

And there’s no need for a “fer­ment­ed­ness” cat­e­gory de­tec­tor, any more than there’s a need for cones that se­lec­tively per­ceive yel­low.

Sure, it’s con­voluted and im­plau­si­ble, but good enough to keep con­sump­tion of bat urine nice and le­gal, which is re­ally all it has to do.

Ah yes, the “grand so­cial con­spir­acy to ward off pro­hi­bi­tion” hy­poth­e­sis emerges again. I’d be in­ter­ested in hear­ing more about how you think this is sup­posed to work.

• Ah yes, the “grand so­cial con­spir­acy to ward off pro­hi­bi­tion” hy­poth­e­sis emerges again. I’d be in­ter­ested in hear­ing more about how you think this is sup­posed to work.

A large frac­tion of the pop­u­la­tion is hugely en­thu­si­as­tic about some­thing, and acts to pre­serve it? It worked for choco­late—what makes you think al­co­hol in­spires less en­thu­si­asm?

• the re­cent meaty one

It’s called umami.

Then I just have to show equal plau­si­bil­ity of the use­ful­ness of bat urine, which I’ve done.

You haven’t shown equal plau­si­bil­ity for your “bat urine” hy­poth­e­sis as Roberts has for his “fer­mented food” hy­poth­e­sis. Go ahead and scan his blog un­der the cat­e­gories fer­mented food and umami hy­poth­e­sis. (I don’t agree with ev­ery­thing Roberts has writ­ten on the sub­ject.)

That said, I think it was an er­ror for loqi to bring up Roberts’s ideas at all—when he talks about fer­mented food, he means things like yo­gourt, soy sauce, natto, miso, fish paste, and kom­bucha, not the prod­ucts of al­co­holic fer­men­ta­tion. (ETA: No, ap­par­ently he in­cludes al­co­holic fer­men­ta­tion.)

• So would you agree that my the­sis is at least ac­cu­rate for a por­tion of the pop­u­la­tion? (The the­sis was, “Peo­ple don’t re­ally like the taste but use the sup­posed taste and other rea­sons as an ex­cuse for get­ting high in a so­cially ac­cept­able way and keep­ing it le­gal to do so.”)

I think peo­ple usu­ally ei­ther find a taste they like when they drink (some­times mix­ing in sweet drinks), or drink just for the al­co­hol and grow to like the taste over time. I doubt many peo­ple claim to drink solely for the taste: I’ve never heard any­one say this, though peo­ple who en­joy the buzz of al­co­hol also say they like the taste.

• I think peo­ple usu­ally ei­ther find a taste they like when they drink (some­times mix­ing in sweet drinks), or drink just for the al­co­hol and grow to like the taste over time.

Again, this is some­thing that could make any­thing taste good—it’s no ev­i­dence of lik­ing the al­co­holic drink. It’s one of the very rea­sons I rol­led my eyes at when peo­ple tried to con­vince me that I must ac­tu­ally like al­co­hol, be­cause I like a cer­tain drink that heav­ily dilutes the al­co­hol taste through sweet­ness.

I doubt many peo­ple claim to drink solely for the taste: I’ve never heard any­one say this, though peo­ple who en­joy the buzz of al­co­hol also say they like the taste.

I’ve cer­tainly seen peo­ple put on that pre­tense, and, in any case, they cer­tainly claim it’s a driv­ing fac­tor, if for no other rea­son than the vastly vary­ing prices for the same amount of al­co­hol.

• I do not par­tic­u­larly like the high of al­co­hol. How­ever, I re­ally like Bel­gian beer, and it has al­co­hol in it, some­times large amounts, and it’s a side effect I am will­ing to han­dle for the taste! Un­for­tu­nately, that side effect does mean I am forced to limit my­self to about 3 beers in one sit­ting.

I won­der if you have never drank suffi­ciently good beer. It doesn’t have to be that ex­pen­sive even, su­per-high end beers are much cheaper than su­per-high end wine. $5-7 for a nor­mal bot­tle,$30 for a bot­tle of the best beer in the world. http://​​www.rate­beer.com/​​beer/​​west­vleteren-abt-12/​​4934/​​

If you’re ever in Pitts­burgh, I’ll buy you a real beer at the Sharp Edge.

I also ad­mit that your point is prob­a­bly cor­rect and I am some­thing of an out­lier—and it’s re­ally just Bel­gian beer that I would drink de­spite the al­co­hol; most other beer and wine and liquor is noth­ing spe­cial.

• Thanks for the offer, and your in­put.

Pay at­ten­tion, ev­ery­one. This is what it looks like when you re­ally like drink­ing some­thing, rather than its effect on your mind:

I do not par­tic­u­larly like the high of al­co­hol. How­ever, I re­ally like Bel­gian beer, and it has al­co­hol in it, some­times large amounts, and it’s a side effect I am wiling to han­dle for the taste! Un­for­tu­nately, that side effect does mean I am forced to limit my­self to about 3 beers in one sit­ting.

When you start run­ning into hard limits about how much of the stuff you can con­sume be­fore dele­te­ri­ous effects on your body, and this is a down­side to you, that definitely sounds like a se­ri­ous en­joy­ment. (That’s where I am re­gard­ing ice cream and many other sweets.)

In con­trast, when there are very nar­row situ­a­tions in which you en­joy its “taste”, and drink “just enough” to ac­com­plish mild re­lax­ation when you want to, um, mildly re­lax, well, then I start to get skep­ti­cal.

• I’ve cer­tainly seen peo­ple put on that pre­tense, and, in any case, they cer­tainly claim it’s a driv­ing fac­tor, if for no other rea­son than the vastly vary­ing prices for the same amount of al­co­hol.

I think I un­der­stand. We’re talk­ing about two differ­ent things.

You’re say­ing, if I un­der­stand cor­rectly, that there’s a great deal of snob­bery in al­co­hol drink­ing: peo­ple claim that ex­pen­sive wines or liquors taste so much bet­ter, and this claim doesn’t stand up to scrutiny.

Out­side of this snob­bery, though, just in terms of friendly so­cial drink­ing, al­most ev­ery­one agrees that they drink be­cause they en­joy the feel­ing, and the taste is just some­thing they grew to like over time, or they mix it with some­thing sweet to make it taste bet­ter.

• Out­side of this snob­bery, though, just in terms of friendly so­cial drink­ing, al­most ev­ery­one agrees that they drink be­cause they en­joy the feel­ing, and the taste is just some­thing they grew to like over time, or they mix it with some­thing sweet to make it taste bet­ter.

Um, no, and that’s the prob­lem. I have never been able to get peo­ple to ad­mit that it’s just about the men­tal effects, and that they have to find ways to make them­selves tol­er­ate the awful taste. Not with­out a lot of teeth-pul­ling, and peo­ple tel­ling me about all the won­der­ful ar­gu­ments against this po­si­tion.

Again, it’s the in­sis­tence that they like “this par­tic­u­lar drink” be­cause it’s “so good” that both­ers me. No, it’s about get­ting high, and no one will talk about this.

• I’m sur­prised by this ex­per­i­men­tal re­sult. In my ex­pe­rience most peo­ple say that it’s about the men­tal effects as well as the taste. Just to be clear: over half the peo­ple you ask say that they don’t drink al­co­hol for the men­tal effects at all, and it’s solely about the taste?

I won­der if part of this is due to the way you’re ask­ing. You use lan­guage like “tol­er­ate the awful taste”, “suffer through”, and com­pare it to hot sauce and en­g­ine oil. Ob­vi­ously you strongly dis­like the taste of al­co­hol. Not ev­ery­one does though; while I drink pri­mar­ily for the men­tal effects, I also en­joy and have ac­quired a taste for some differ­ent types of al­co­hol, and I like some com­bi­na­tions of fla­vors when hav­ing a beer with food.

So maybe you’re get­ting strong re­ac­tions in con­trast to your ex­treme state­ments that al­co­hol tastes awful and no one could ever like the taste.

• It’s more like this:

me: I think I’m strange. I don’t like al­co­holic drinks. I mean, I like the effect on me, but not the taste, not the pro­cess of drink­ing it.

them: Yeah, that is strange. I mean, I like mar­gar­i­tas.

me: Oh re­ally? What do you like about them?

them: Well, I like them when I go out danc­ing...

me: No, I mean, like, about the taste.

them: Well, I like those re­ally frozen ones with lots of differ­ent fruit fla­vors.

me: So you like the taste of those mar­gar­i­tas? What is it about the taste?

them: Um, well, it helps me to re­lax. [Alter­nate: It’s kind of a so­cial thing/​so­cial lubri­cant.]

me: *falls out of chair* Okay, so about the taste. Do you like the taste more than that of a milk­shake?

them: Hm, that’s a good ques­tion, I’ve never even thought of that. No, I like the milk­shake much bet­ter.

me: *loses hope in hu­man­ity*

• What would hap­pen if you asked some­one this:

me: So you like the taste of those mar­gar­i­tas? What is it about the taste?

And they an­swered your ques­tion? Speci­fi­cally, would one good re­sponse cause you to re­think your the­ory on the sub­ject? How many re­sponses would you need to be con­vinced?

I am not say­ing I have those re­sponses. I am just cu­ri­ous.

• It sounds like the per­son here is say­ing he drinks for the men­tal effects (“it helps me to re­lax”), and that he doesn’t mind the taste be­cause it’s mixed with things he likes (“fruit fla­vors”). This seems like the an­swer I’d ex­pect.

Whereas it seems like you ab­solutely de­spise the taste, most peo­ple who drink don’t mind it, and some­times like it, es­pe­cially when mixed with fruit or sweet tastes.

• But they don’t like it, “es­pe­cially when mixed with fruit or sweet tastes (and taste-bud numb­ing ice, but what­ever)”. Rather, they like sweet, fruity, cold drinks, and still find them good, even if it is wors­ened with a lit­tle al­co­hol.

That, I think is the ap­pro­pri­ate way to char­ac­ter­ize it.. Again, re­mem­ber my in­ces­sant point about baseline com­par­i­sons: if some­one likes fruity sweet­ness, it’s go­ing to make pretty much any­thing (that doesn’t clash) taste good. But so what? That doesn’t mean they like the stuff its mixed with. It just means they like that fruity sweet­ness, and their en­joy­ment may per­sist even if the drink is de­graded with other, worse fla­vors.

What’s more, con­ver­sa­tions like these (alarm­ingly typ­i­cal) re­veal that peo­ple aren’t even think­ing about the dis­tinc­tion be­tween lik­ing a drink for its taste, and lik­ing it be­cause they like get­ting high—and nor are they in­ter­ested in learn­ing.

• I’ve been watch­ing this thread for a while, and as a fre­quent al­co­hol-drinker, I thought I would try to re­port my ex­pe­rience as hon­estly as I can man­age:

• Beer: In an ab­solute sense, I don’t like the taste. Since some beers taste less bad—or more in­ter­est­ing—than oth­ers, I will some­times com­ment that a par­tic­u­lar beer tastes “re­ally good”. What I mean though, is that it tastes “re­ally good” for a beer. I drink quite a lot of beer, be­cause I usu­ally pre­fer the slower, gen­tler, more con­trol­lable buzz to that of harder al­co­hol. I’ve heard plenty of women say they don’t like beer. In some cir­cles, it’s con­sid­ered un­manly for a man to say he doesn’t like beer, and I ex­pect that’s why I hear it much less from men. In some situ­a­tions, I take the praise of beer as short­hand for “I know we all don’t have much in com­mon, nor any real rea­son be­yond com­pany for hang­ing out, so lets go through the mo­tions of af­firm­ing our mu­tual love for some­thing that is safe to af­firm mu­tual love for.”

• Wine: This is definitely all about the taste, but it’s not at all the same cat­e­gory of taste as sugar or a milk­shake. This is all about the com­plex­ity of dozens of in­ter­act­ing fla­vors. It is a kalei­de­scope that you “see” with your tongue. It’s a taste ex­pe­rience by defi­ni­tion, but that doesn’t mean that it is any­thing like the tasti­ness of a milk­shake. The thrill is in the rich­ness of the pat­terns that ex­ist in the taste. Im­por­tantly, I find that only cer­tain wines at cer­tain ages pro­duce this effect to a worth­while de­gree. Lousy wine tastes lousy. A re­ally good caber­net franc, say, can be the kind of amaz­ing that makes me bolt up­right in my chair and go wide-eyed. Really. As far as the al­co­hol com­po­nent, it is such an in­trin­sic part of the taste-or­ches­tra that I, un­for­tu­nately, find it im­pos­si­ble to spec­u­late on whether I’d still drink wine with­out it. I think I would still drink it if it did not pro­duce a buzz, al­though there would be one less rea­son. I think I would still like swish­ing it around, even if I was only go­ing to spit it out. Need­less to say, I find it to be an ex­tremely pleas­ant way to get buzzed.

• Mixed drinks/​Hard al­co­hol: As far as I’m con­cerned, these have always ex­isted solely as a fast-track to get­ting buzzed or drunk. For me, they might as well be an IV drip. I, how­ever, ad­minister them in the nor­mal way, be­cause it is nor­mal and prac­ti­cal. Also, sip­ping lets me roughly cal­ibrate my dosage to match oth­ers, and the situ­a­tion.

• Bat urine: I don’t think this is a fair ar­gu­ment at all, be­cause you can­not sep­a­rate your dis­gust re­ac­tion from pure taste, in the ex­pe­rience of drink­ing some­thing. Sev­eral bod­ily fluids have lit­tle/​fairly neu­tral taste, but the ex­pe­rience of drink­ing it would still be dis­gust­ing.

I agree that most al­co­hol con­sump­tion is mainly about the buzz. I like differ­ent states of con­scious­ness. The one in­duced by al­co­hol is not my fa­vorite, but it’s one I can en­joy with­out hav­ing to sneak around or worry about prison, so I make do with the (some­what per­son­ally dis­ap­point­ing) poli­ti­cal free­doms I have. I do drink wine for the taste—it just isn’t the same kind of taste as any­thing else. It’s a sen­sory-over­load ex­pe­rience that hap­pens to be de­liv­ered by the tongue.

• Isn’t it pos­si­ble that a lit­tle bit of com­plex, as­tringent bit­ter­ness can ac­tu­ally make a sweet fruity drink more palat­able? I wouldn’t drink a vir­gin margher­ita; I hon­estly be­lieve the tequilla and triple sec make it taste bet­ter.

• Isn’t it pos­si­ble that a lit­tle bit of com­plex, as­tringent bit­ter­ness can ac­tu­ally make a sweet fruity drink more palat­able?

Hey, if that helps keep it le­gal and so­cially ac­cept­able to get high … sure, why not?

• It seems like this whole ar­gu­ment is mo­ti­vated out of a wish to make it so­cially ac­cept­able to say “I don’t like the taste of beer” by try­ing to paint ev­ery­one who dis­agrees as liars.

• No, I think he sim­ply hates the taste of al­co­hol so much that he can’t con­ceive that some­one could hon­estly like it.

• You need to read my his­tory again, for the first time. I ini­tially did be­lieve that I was just weird in not lik­ing al­co­hol, or that it would come with time. It’s the ob­vi­ous, fa­vored, sim­ple hy­poth­e­sis.

But I can only hold be­lief in it for so long un­til the shower of dis­con­fir­ma­tory ev­i­dence hits. When I look be­hind the veil and find out what it means for other peo­ple to like al­co­hol, and find that it matches up with what I con­sider not lik­ing al­co­hol, well … if any­thing, I held on to the be­lief too long.

• Did you no­tice that I said that I don’t match up with your crite­rion? Be­sides the fact that even that to­tal list didn’t seem to show that a per­son nec­es­sar­ily didn’t like the taste of some­thing.

You could at least mod­ify your be­lief to “some peo­ple don’t like the taste of al­co­hol but claim that they do for such and such rea­sons...” and then it would be­come more ac­cu­rate, since surely this is likely true of at least some peo­ple, while it is surely not true of all who claim to like it.

For ex­am­ple, an area where your po­si­tion has some truth is that there are guys who ba­si­cally dis­like any type of al­co­hol ex­cept sweet drinks, and these they like only be­cause of the sweet­ness, but they are un­will­ing to ad­mit it be­cause this is thought to be “girl­ish”. But at the same time, this is definitely un­true of many oth­ers.

• I ask that you take se­ri­ous note of the sym­pa­thy with which I’ve char­ac­ter­ized these liars. I com­pletely un­der­stand why they have to put on a show: any­thing that does to your mind what al­co­holic drinks do, but doesn’t have wide-scale so­cial sup­port from re­spectable peo­ple, is go­ing to get banned or oth­er­wise given se­vere re­stric­tions. Such a pre­tense doesn’t strike me as so wrong here.

What both­ers me is the wide­spread re­fusal to ac­knowl­edge this, even in pri­vate.

• I think you’re miss­ing a sig­nifi­cant fac­tor.

Many peo­ple don’t drink al­co­hol pri­mar­ily for the men­tal effects. Rather, there is a strong sta­tus penalty to drink­ing non-al­co­holic bev­er­ages. Most non-al­co­holic bev­er­ages are strongly as­so­ci­ated with chil­dren, at least in the af­ter­noon (juice and milk are OK at break­fast, not at din­ner). Adults can’t or­der them with­out send­ing an un­de­sir­able sig­nal about their ma­tu­rity.

Among the ac­cept­able drinks, you’re left with other “ac­quired tastes” (coffee and tea) or drinks that of­ten give other low sta­tus sig­nals (wa­ter alone is cheap, soft drinks are lower-class).

Once you’ve es­tab­lished that it’s a sta­tus is­sue, the re­fusal to ad­mit it is un­der­stand­able, since open con­cern for sta­tus is gen­er­ally a low-sta­tus trait. I don’t agree with all of Robin Han­son’s sta­tus ex­pla­na­tions, but it makes sense here.

The mind-al­ter­ing effects play into it as well. Even then, there are im­por­tant sig­nal­ing effects in play (Robin put up a post on that a bit ago). And ig­nor­ing taste to­tally is a mis­take. Even if I might pre­fer a milk­shake to an Ir­ish creme, I definitely pre­fer an Ir­ish creme to Ever­clear.

• Btw, I think your milk­shake com­par­i­son needs to be be­tween equal caloric por­tions.

I’d pre­fer 600 calories of milk­shake to 600 calories of beer. But I would rather have one beer than one milk­shake. For cer­tain val­ues of beer, beer is more deli­cious than milk­shake per calorie.

• Why could per-calorie be the rele­vant met­ric? And why would a met­ric re­quiring you to con­sume the full five beers be helpful?

• any­thing that does to your mind what al­co­holic drinks do, but doesn’t have wide-scale so­cial sup­port from re­spectable people

I’m con­fused. Are you say­ing that al­co­hol doesn’t have wide-scale so­cial sup­port from re­spectable peo­ple? What so­ciety are we talk­ing about?

I would guess that of the adult pop­u­la­tion in the US who drinks, at least 75% drink pri­mar­ily for the men­tal effects and would have no prob­lem say­ing so.

• Do you have trou­ble read­ing a full clause?

any­thing that does to your mind what al­co­holic drinks do, but doesn’t have wide-scale so­cial sup­port from re­spectable peo­ple, is go­ing to get banned or oth­er­wise given se­vere re­stric­tions.

Un­der­stand now? If al­co­hol didn’t have the so­cial sup­port it does, it would be Just Another Mind Al­ter­ing Sub­stance that would be banned, or that you’d need a pre­scrip­tion for.

Please, finish sen­tences be­fore re­spond­ing to them

• SilasBarta, I too am puz­zled at why Blue­berry mi­s­un­der­stood you, but your re­sponse was need­lessly rude.

• Would you say it was more or less rude than clip­ping a sen­tence in two and re­spond­ing to one that mis­rep­re­sented what I said?

Do any of you in­tend to crit­i­cize/​mod down Blue­berry for his/​her rude­ness, or do you just re­serve your re­bukes for the dili­gent?

• Do you have trou­ble read­ing a full clause?

I sub­mit that

a) you’re not re­ally cu­ri­ous, b) ex­pect any an­swer to come back nega­tive, and c) aren’t in­ter­ested in ar­gu­ing whether he can read a full clause any­way.

• If al­co­hol didn’t have the so­cial sup­port it does, it would be Just Another Mind Al­ter­ing Sub­stance that would be banned, or that you’d need a pre­scrip­tion for.

I find this difficult to swal­low. Al­co­hol pro­hi­bi­tion was a widely ac­knowl­edged dis­aster (or does this col­lec­tive mem­ory also count as “so­cial sup­port”?). The “Joe Six­packs” of the na­tion aren’t croon­ing over Miller’s exquisite blend of hops, but they’d be up in arms if you tried to take it away.

And drug policy (at least in the US) isn’t par­tic­u­larly con­sis­tent—if you don’t be­lieve me, feel free to con­duct your own ex­per­i­ment with some high-po­tency salvia ex­tract.

I doubt most peo­ple are wor­ried even sub­con­sciously about the rein­tro­duc­tion of pro­hi­bi­tion. Why pos­tu­late a co­or­di­nated so­cial re­sponse to such a non-threat?

• I find this difficult to swal­low. Al­co­hol pro­hi­bi­tion was a widely ac­knowl­edged dis­aster (or does this col­lec­tive mem­ory also count as “so­cial sup­port”?).

Yes it was a dis­aster—be­cause of al­co­hol’s wide­spread so­cial sup­port, that led to the black mar­ket, in­abil­ity to en­force, etc.

Hence my statement

If al­co­hol didn’t have the so­cial sup­port it does, it would … be banned, or … need a pre­scrip­tion.

You also said:

I doubt most peo­ple are wor­ried even sub­con­sciously about the rein­tro­duc­tion of pro­hi­bi­tion. Why pos­tu­late a co­or­di­nated so­cial re­sponse to such a non-threat?

There are many mea­sures short of pro­hi­bi­tion that re­strict al­co­hol. In try­ing to im­pose them, as so­ciety im­poses re­stric­tions on mind-al­ter­ing sub­stances, leg­is­la­tures butt up against the so­cial sup­port for al­co­hol. Re­tain­ing this sup­port is nec­es­sary for pre­vent­ing these (oth­er­wise rea­son­able) re­stric­tions on al­co­hol.

• Salvia is both new and lit­tle known in com­par­i­son to mar­ijuana, LSD, co­caine, heroin, meth, etc.

• This is get­ting per­ilously close to poli­tics, but the differ­ence with al­co­hol is the great his­tory of hu­man use. Al­co­hol was one of the first drugs reg­u­larly con­sumed by hu­mans. A lot of cul­ture has de­vel­oped around that. Pro­hi­bi­tion failed be­cause it tried to out­law the cul­ture. Cannabis and psychedelics were also used by pre-mod­ern hu­mans, but the gov­ern­ment could out­law the other drugs with­out a peo­ple’s re­volt be­cause the av­er­age per­son didn’t use cannabis and psychedelics. The av­er­age per­son did and does use al­co­hol.

• (Guiness wins in this re­gard.)

Of note, Guin­ness has a lower al­co­hol con­tent than most beer.

• The prob­lem, though, is that su­per­tasters are es­ti­mated at 25% of the pop­u­la­tion. So why aren’t 25% of peo­ple voic­ing my opinion on al­co­hol? Why would they stay silent about hat­ing it, while drink­ing it for so­cial and psy­choac­tive benefits?

More anec­do­tal ev­i­dence: Over 14 of the peo­ple I know do not drink al­co­hol in any form. The so­ciety I am from is prob­a­bly atyp­i­cal in this re­gard.

• I would think train­ing your­self to like choco­late would be a lot eas­ier than train­ing your­self to like coffee.

• Ma­nipu­lat­ing coffee into a good tast­ing form isn’t too hard; just add a lot of sugar and dilute it with enough milk, and it’ll prob­a­bly taste pretty good even if you think black coffee tastes like dirt. (And then, if you want, you can re­duce the amount of milk and sugar over time.)

• Yes, I think that’s what my point comes down to: so you like beer af­ter be­ing pres­sured by friends to drink it for five years. Then, by sim­ple force of habit, you come to like it—your tastes change.

[...] But stop tel­ling me beer tastes good.

I like the taste of a par­tic­u­lar beer and came to that con­clu­sion af­ter hav­ing about 4 beers be­fore it in my life­time. Not 4 serv­ings of that par­tic­u­lar beer, but 4 serv­ings to­tal. I un­der­stand ac­quired taste and claim that this does not qual­ify for that la­bel. I liked it. I though it tasted good.

I as­sume you are not say­ing, “beer always tastes bad for ev­ery­one un­til they get con­di­tioned by so­ciety.” That is what it sounds like to me, how­ever. Is beer just an ex­am­ple for the sake of con­ve­nience?

ETA: Similarly, Homer’s The Odyssey isn’t good. Rather, a bunch of peo­ple have a tra­di­tion of read­ing it that they pass on and get the next gen­er­a­tion to per­pet­u­ate. But what the heck am I sup­posed to learn about good writ­ing from that, other than:

Is it plau­si­ble that it is good? Or has that sce­nario been com­pletely re­jected from your wor­ld­view? I haven’t read it, per­son­ally, so I re­ally have no idea if it is or isn’t.

• There are adap­ta­tions of The Odyssey that are pretty fun to read. I agree that di­rect trans­la­tions tend to have is­sues, though; thou­sands of years of cul­tural change, the loss of lyri­cal el­e­ments through trans­la­tion, the change in medium from oral recita­tion to print, and many other fac­tors I can’t think of at the mo­ment all make the story much less im­pres­sive than it must have been back in an­cient times.

• Like­wise, there are prac­tices that can make peo­ple like some­thing. But there’s no point to say­ing, “Hey, af­ter this prac­tice, peo­ple like it!” That con­veys no in­for­ma­tion—it’s true for ev­ery­thing. Like with placebo cures, I want to know what is good above and be­yond that that re­sults from stan­dard “make some­thing seem good” tricks.

Is it fair to say that you are look­ing for a way to pre­dict “good” art be­fore it en­ters the cul­tural sta­tus stream?

• Not in the sense that I want to pre­dict the next big thing.

What I’m look­ing for is, what por­tion is due to ac­tual merit of the art­work, that peo­ple would ap­pre­ci­ate even in the ab­sence of oth­ers pres­sur­ing them to like it, or the sig­nal­ing effects of dis­play­ing it to oth­ers?

I have of­ten fo­cused on sce­nar­ios where you can get a judg­ment be­fore cul­tural effects in­terfere, but these aren’t strictly nec­es­sary. Like with the placebo ex­am­ple I keep giv­ing, there are ways to see what is due to some effect that will make any­thing look good, and what effect is due to the ac­tual merit. The hoaxes that oth­ers have refer­enced are good ex­am­ples of this.

Does that an­swer your ques­tion?

• Yes, this does an­swer my ques­tion.

The fol­lowup ques­tion: How use­ful is be­ing able to iden­tify “bad” art? Is it a step to­ward the same di­rec­tion of iden­ti­fy­ing merit?

(Good and bad as I am us­ing it means value from ac­tual merit and ig­nores all peer pres­sure or sig­nal­ing effects.)

• Very use­ful: in the fu­ture, we want to have ma­chines that can make the same (peer-pres­sure-free) art clas­sifi­ca­tions that hu­mans would, so they can pop out art them­selves. Bad art is nearly as use­ful as good art in helping to train such a ma­chine and iden­tify the al­gorithms hu­mans use to make these judg­ments.

But when the field of art has been cor­rupted to the point where it’s just a pure sta­tus game, there is no such clas­sifier that can be learned. The only ma­chine you’re go­ing to be mak­ing is one that looks hu­man, and hob­nobs its way up the so­cial lad­der so that it can learn what the elites think, and ren­der judg­ments in that way.

(I made the same cri­tique about some Ja­panese re­searchers’ quix­otic at­tempt to build a ma­chine that de­ter­mines how much hu­mans will like a given wine, based on chem­i­cal anal­y­sis. Hey guys—it ain’t the chem­i­cal com­po­si­tion of a wine that makes peo­ple like it!)

• Cool. Yeah, I pretty much agree with ev­ery­thing here and don’t have any­thing to add. I think this com­ment nails the sub­ject on the head.

• My point is that it doesn’t mat­ter if it’s about sig­nal­ing or not. Quests for sta­tus per­vade ev­ery as­pect of hu­man life and are in­escapable. Th­ese peo­ple be­lieve what they be­lieve and get up­set when you bring it up for the same rea­son that you will ob­ject if I said you’re only in­ter­ested in point­ing out their sta­tus-quest­ing for your own sta­tus-quest­ing. “I don’t care about sta­tus” is ev­ery­one’s con­ceit.

EDIT: Just to ex­pand on this a lit­tle bit—I’m say­ing that the de­sire to point out their cog­ni­tive dis­so­nance is mo­ti­vated by sta­tus, as well, and that fur­ther, nei­ther of these is worse than the other when rat­ing by sincer­ity or hon­esty.

• My point is that it doesn’t mat­ter if it’s about sig­nal­ing or not. Quests for sta­tus per­vade ev­ery as­pect of hu­man life and are in­escapable.

Yes, and the placebo effect in cures is in­escapable. But there’s still a part of the cure that is due to gen­uine bio­chem­i­cal effects from the medicine rather than the be­lief that it will work.

Like­wise, I want to know the por­tion of art—and al­co­hol—that is due to more than just those things that could rook any­one into lik­ing them. If, as it seems, in many cases, there is no such por­tion—if it’s all about be­ing con­di­tioned to like it in a way that could work for bat urine—then I don’t con­sider those things good, and I wish peo­ple would stop putting on the pre­tense that they are.

Science passes this test with fly­ing col­ors: no amount of phony, mean­ingless pa­pers by sta­tus jock­ey­ing sci­en­tists and en­g­ineers is go­ing to get an air­plane off the ground (with­out rip­ping apart) or an ex­tremely pow­er­ful bomb to go off. The buck stops some­where. Where does the art buck stop? Where does the drink qual­ity buck stop?

Th­ese peo­ple be­lieve what they be­lieve and get up­set when you bring it up for the same rea­son that you will ob­ject if I said you’re only in­ter­ested in point­ing out their sta­tus-quest­ing for your own sta­tus-quest­ing. “I don’t care about sta­tus” is ev­ery­one’s con­ceit.

Yes, that would ex­plain why some­one’s won’t say to my face the real rea­sons they drink. But in an on­line dis­cus­sion with 90% anony­mous han­dles: what’s hold­ing them back?

Just to ex­pand on this a lit­tle bit—I’m say­ing that the de­sire to point out their cog­ni­tive dis­so­nance is mo­ti­vated by sta­tus,

That may be a part of it. But read the link thomblake gave to my ear­lier thread: I was ex­pe­rienc­ing re­ally weird data. Peo­ple seemed to be ex­pe­rienc­ing the same in­ter­nal state as me, but us­ing differ­ent la­bels for it.

• The buck stops with you, be­cause art isn’t a com­pe­ti­tion. Maybe it is for the artists, but not from your end—it’s just what you en­joy.

I have a copy of a paint­ing hang­ing in my liv­ing room that I won’t name here, but it’s very pop­u­lar and fa­mous (and there­fore kind of stupid to have hang­ing in my liv­ing room, be­cause it doesn’t re­ally show off my taste as re­fined). But I get a lot out of it. I love look­ing at it.

If an art stu­dent came in and wanted to try to con­de­scend to me about my taste in art, what could I do? I’d look at him and say, “This paint­ing does for me what art is sup­posed to do for peo­ple. I don’t have the time or en­ergy to de­vote to re­fin­ing my taste. I ad­mit your taste in art is more re­fined and you might get more out of a Pi­casso than I do, be­cause I don’t get much.”

If he still wants to look down his nose at me, who gives a shit? Get out of my house, right? But I think the true art-lover will say, “I’m glad you ex­pe­rience some­thing that’s so mean­ingful to me, even if your taste is blunter and cruder than mine.”

I think this is analo­gous to if the art stu­dent came to me and said, “I never re­al­ized how cool the Pythagorean the­o­rem is be­fore. It’s amaz­ing.” Do I look at him and say, “Wow, you’re an idiot”? I would hope not; I would hope to think to my­self, “Well, it’s a start,” and say, “Right?!”

ETA: I’d be call­ing him an idiot be­cause he’s only get­ting it now, and not back when he learned it for the first time in high school and I re­al­ized how cool it was.

• The buck stops with you, be­cause art isn’t a com­pe­ti­tion. Maybe it is for the artists, but not from your end—it’s just what you en­joy. … If he still wants to look down his nose at me, who gives a shit? Get out of my house, right? But I think the true art-lover will say, “I’m glad you ex­pe­rience some­thing that’s so mean­ingful to me, even if your taste is blunter and cruder than mine.”

I’m sorry, but that’s a very naive view of “how it works”. The elite art cadre cer­tainly pro­motes the be­lief that there’s a lot more to art than what you or I per­son­ally like. They’re the ones that in­fluence, by their sta­tus, what stu­dents will be in­doc­tri­nated in, and what art­works they will be ex­pected to deem good, even as con­struc­tion work­ers mis­take the “good” stuff for trash. (This has hap­pened be­fore.) Even as the “art” in front of pub­lic build­ings, un­der the full en­dorse­ment of the art elite, is a blight on the land­scape.

If it were just a mat­ter of “en­joy what you like”, I’d have the same view as you do. But there is sig­nifi­cant money spent in­doc­tri­nat­ing stu­dents in one view of art—which un­like sci­ence, lacks a stop­ping-buck. There is the pre­tense that you have to en­joy Shake­speare, or the lat­est splotches on a can­vas, to “truly” ap­pre­ci­ate art. And as long as they pro­mote their priest­hood that de­cides which art is blessed, and gets the huge grants for mu­se­ums to “study” and pro­mote it, even as they cant sub­stan­ti­ate their opinions … well, then I have a prob­lem.

• But why do those things bother you, ex­cept in that you don’t like be­ing told you’re low sta­tus un­less you jump through cer­tain hoops?

• Do I re­ally need to ex­plain why it’s bad for peo­ple to be wealthy and high sta­tus dep­site never hav­ing pro­duced any­thing of value, and spend all their time per­pet­u­at­ing what is es­sen­tially an in­for­ma­tion cas­cade?

• Quite a judg­ment there, “noth­ing of value”! Be­cause peo­ple have to be trained to ap­pre­ci­ate it, it’s of no value?

• It’s not just the fact that peo­ple have to be trained. After all, peo­ple must be trained in or­der to read or use a com­puter.

The prob­lem is that there’s no clear stan­dard for what counts as suc­cess­ful train­ing. You can check for whether some­one can read (at a given level) us­ing tests that ev­ery­one will agree about for the re­sults. How do you know when some­one’s got­ten the right “art ap­pre­ci­a­tion train­ing”? “Oh, well, you see, you have to join our club, and hand around only our peo­ple for years and years, and then we still get fooled by mon­keys …”

• How do you know there’s no clear stan­dard? You’re not an artist.

• Falsifi­a­bil­ity, ba­si­cally. Or lack thereof.

• Well, my first hint was when the work of a mon­key was mis­taken for that of an award-win­ning artist...

• I’m will­ing to give them the benefit of the doubt on this one. Why shouldn’t I?

• Arts fund­ing with tax dol­lars is one par­tic­u­larly di­rect ex­am­ple.

• “But in an on­line dis­cus­sion with 90% anony­mous han­dles: what’s hold­ing them back?”

Once again, this is sim­ply very strong ev­i­dence that you are wrong. The rea­son peo­ple are in­sis­tent is be­cause they hap­pen to know what they like.

• Where does the art buck stop?

With, in the ab­sence of any­thing bet­ter, bucks.

Where does the drink qual­ity buck stop?

Well, there’s always the blind taste tests where peo­ple of var­i­ous de­grees of drink qual­ity naivety rank drinks. Which re­li­ably pro­duce a nega­tive cor­re­la­tion with the elite feed­back loop!

• I’m claiming that Pol­lock’s work demon­strated eas­ily-ne­glected and valuable parts of the aes­thetic ex­pe­rience.

As for ev­i­dence against that hy­poth­e­sis, I think that de­pends largely upon how se­ri­ously you take some of the rele­vant premises in your wacko’s model. Ac­cord­ing to some, there is vir­tu­ally noth­ing to all of cul­ture other than sta­tus games (though in this case the clause “de­spite there be­ing noth­ing spe­cial about his work” would make lit­tle sense). Ac­cord­ing to oth­ers, there re­ally is quite a bit to aes­thet­ics, and per­haps it’s worth listen­ing to the folks who’ve spent their lives study­ing it.

There are a lot of differ­ent kinds of things in the world, and many of them are valuable in un­ex­pected ways.

• So you re­ally can’t think of any­thing that is less likely to be ob­served if “it’s all bul­lshit” than if it’s not? There isn’t any kind of aes­thetic feel­ing you could feed to the wacko that he couldn’t help but burst out in ap­pre­ci­a­tion for?

As for ev­i­dence against that hy­poth­e­sis, I think that de­pends largely upon how se­ri­ously you take some of the rele­vant premises in your wacko’s model.

Not when I’m ask­ing for ev­i­dence with a low Bayes fac­tor, rather than a guaran­teed low pos­te­rior.

Maybe an ex­am­ple would be in or­der. Let’s say Bob is the wacko, but about quan­tum physics. Bob be­lieve that the claims of quan­tum physics are just a big sta­tus game, and so are the re­sults of the par­ti­cle ac­cel­er­a­tors and ev­ery­thing. I could point to ev­i­dence like the atom bomb. If they were just ar­gu­ing over mean­ingless crap the whole time, and as­sign­ing truth purely based on who has the most sta­tus, how did they ever get the un­der­stand­ing nec­es­sary to build an atom bomb, Bob?

There are a lot of differ­ent kinds of things in the world, and many of them are valuable in un­ex­pected ways.

Right, like Halo. Ex­cept that mil­lions of peo­ple like Halo even in the ab­sence of a well-funded in­doc­tri­na­tion cam­paign, and the fact that ex­press­ing ap­pre­ci­a­tion for Halo won’t en­dear them to the kewl kids of art.

It’s not very im­pres­sive if peo­ple start to en­joy some­thing af­ter they’ve

spent their lives study­ing it.

You have to ad­just for stuff like that.

• Well said!

• Meh.

Never trust a site with that many ban­ners.

I, for one, have never heard se­ri­ously dis­parag­ing things about 19th-cen­tury or Aca­demic art in gen­eral, and the es­say sounds a bit… ra­bid. It seems like the writer was op­er­at­ing un­der the com­mon heuris­tic “I don’t un­der­stand it, so it must be stupid.”

• And I never trust any­one that says, “No, no, this stuff is good, it re­ally is! Just pay for 10 years worth of ed­u­ca­tion in this spe­cific area, and then you’ll see the light!”

(Note the similar­ity to Scien­tol­ogy prac­tices...)

• But then what do you do when some­thing re­ally does take that long to ex­plain? Peo­ple say cat­e­gory the­ory is beau­tiful; is the non­math­e­mat­i­cian sup­posed to call them liars?

• Cat­e­gory the­ory doesn’t take 10 years to ex­plain. You should be able to ex­plain to a will­ing, in­tel­li­gent friend in two full days, and get them to a point where they see the beauty.

I’ve done similar things, like ex­plain­ing the el­e­gant beauty of air­craft com­po­nent struc­tural anal­y­sis—got a de­cent ap­pre­ci­a­tion across in 10 min­utes. (“You know how a chain is as strong as its weak­est link? A com­po­nent is as strong as its weak­est failure mode...”)

The point is, you can ex­plain it. That’s a lot more than you can do for (much of) art, it seems.

• The point is, you can ex­plain it. That’s a lot more than you can do for (much of) art, it seems.

Art can be ex­plained. There just aren’t that many peo­ple ca­pa­ble of ex­plain­ing it; ex­plain­ing things is a difficult skill.

• If you know that art can be ex­plained, then pre­sum­ably you’ve en­coun­tered an ex­pla­na­tion of it. Any chance you could point us in the di­rec­tion of it?

• How are you so sure that’s the prob­lem? What dis­t­in­guishes the state of art right now, from one where there is no jus­tifi­ca­tion for why one work is more artis­tic than other, and it’s just the blind lead­ing the blind, ev­ery­one even­tu­ally claiming they can see the Em­porer’s clothes?

And FWIW, there re­ally aren’t many things that are hard to ex­plain (given enough time) -- just peo­ple who don’t un­der­stand their own fields.

• How are you so sure that’s the prob­lem? What dis­t­in­guishes the state of art right now, from one where there is no jus­tifi­ca­tion for why one work is more artis­tic than other, and it’s just the blind lead­ing the blind, ev­ery­one even­tu­ally claiming they can see the Em­porer’s clothes?

The world is big and com­plex, and con­tains lots of differ­ent things. There is plenty of pseudo-art out there, no ques­tion—just like there is plenty of pseudo-sci­ence. But for rea­sons that aren’t im­me­di­ately ob­vi­ous (though some hy­pothe­ses do sug­gest them­selves af­ter a bit of thought), peo­ple re­gard pseudo-art as diminish­ing the cred­i­bil­ity of Art as a pur­suit, while not do­ing the same with pseu­do­science vis-a-vis Science.

I should ad­mit that I’m re­ally only ex­pertly-fa­mil­iar with one par­tic­u­lar art form (mu­sic), so the fol­low­ing is largely ex­trap­o­la­tion as it ap­plies to oth­ers—but I can’t imag­ine that the situ­a­tion isn’t similar.

There is a di­rect ana­log of in­fer­en­tial dis­tance in art. In fact, “ana­log” may not be the right word; it may just liter­ally be a form of in­fer­en­tial dis­tance. Ex­pe­rience and train­ing make a huge differ­ence with re­spect to how a work is per­ceived. This is an effect quite in­de­pen­dent of so­cial clique-for­ma­tion; it’s sim­ply the re­sult of one’s brain work­ing along a cer­tain path for a long time, af­ter which it be­comes difficult for oth­ers who haven’t trav­eled the same path to fol­low one’s thoughts. Un­for­tu­nately, this fact is not suffi­ciently ap­pre­ci­ated; peo­ple sim­ply ex­pect in­fer­en­tial dis­tances to be short.

Given this, it’s clearly pos­si­ble that one could slowly re­trace the path for the benefit of oth­ers, in many small steps, even­tu­ally bring­ing them along to where one is. But most peo­ple with ad­vanced artis­tic knowl­edge do not have this skill, and most peo­ple with­out ad­vanced artis­tic knowl­edge don’t ex­pect them to, be­cause they don’t ex­pect art to need to be ex­plained. So it shouldn’t be sur­pris­ing that there aren’t a lot of re­ally good art ex­pla­na­tions around.

• I would be care­ful with in­vo­ca­tion of in­fer­en­tial dis­tance. It’s not a get-out-of-ex­pla­na­tion-free card you get to use when­ever you have a hard time jus­tify­ing a be­lief (not that you were try­ing to use it this way, but some stan­dards have to be met—see be­low).

There are many rea­sons why you could have a hard time ex­plain­ing a con­cept to some­one. It could be that the con­cept is mush to be­gin with, and only kept afloat by a com­mon agree­ment by in­sid­ers not to call any­one out. It could be that you don’t ac­tu­ally un­der­stand it, in the ra­tio­nal­ist sense of hav­ing a mov­ing-parts model, where black boxes play a min­i­mal role, and which is con­nected deeply to the rest of your un­der­stand­ing of the world.

And fi­nally, it could be that there are many in­ter­me­di­ate con­cepts that you mis­tak­enly, but rea­son­ably, as­sumed oth­ers were fa­mil­iar with, and which take a lot of time to ex­plain. That is the prob­lem of in­fer­en­tial dis­tance. But you have to first rule out the first two pos­si­bil­ities. Then—and only then—do you get to cite in­fer­en­tial dis­tance.

You say that most peo­ple who claim to un­der­stand art just can’t bring oth­ers through. But why can’t you point me to one who has? Isn’t it kind of strange that there are sources that can take you through the in­fer­en­tial dis­tance for all of those top­ics that aren’t BS, but you don’t even know of the ex­is­tence of one that could close the gap for art? And that you don’t be­lieve you can close the gap?

And if it were truly clique-in­de­pen­dent, why would we see things like mon­key art hoax, where the art ex­pert—and it was some iden­ti­fied in ad­vance as an ex­pert—vi­o­lated pretty ba­sic con­ser­va­tion of ev­i­dence prin­ci­ples. On be­ing in­formed it was a mon­key rather than some­one the en­tire com­mu­nity has given awards to, her re­ac­tion was: “Well, I guess it looked kind of rushed.”

Why don’t these ex­perts have the un­der­stand­ing nec­es­sary to say, “you liar”?

You say you are an ex­pert in one area of art. I ac­cept that you have a broad knowl­edge of mu­sic. I ac­cept that you have passed the stan­dards typ­i­cally re­quired to count as an ex­pert in mu­sic. What I ques­tion is whether you pass the crite­ria here to count as an ex­pert, which are about whether you have made it truly part of your­self:

If you deleted your knowl­edge of art in this area, would it grow back? By some method other than some­one giv­ing you the blankly solid facts about which pieces are an aren’t art? Would you rate them the same way, for the same rea­sons? If you har­bor any doubts you need to go back to the first two hy­pothe­ses I men­tioned be­fore you can start ap­peal­ing to in­fer­en­tial dis­tance.

• At least where cat­e­gory the­ory is con­cerned, you don’t have to pay.

• Dis­lik­ing Pol­lock is ir­ra­tional. As is dis­lik­ing Cage. Or Joyce. Or PEZ.

• I love 4′33″. It helps me get to sleep.

• Peo­ple can get the hu­mor and still down­vote you.

I didn’t vote one way or the other.

• Edit: OMG! Is my hu­mor too sub­tle peo­ple? PEZ!

It was, yes—maybe we need to make emoti­cons more nor­mal here, since this is a re­cur­ring prob­lem. :P

(Down­vote re­moved)

• maybe we need to make emoti­cons more nor­mal here, since this is a re­cur­ring prob­lem.

But then how will we sig­nal our so­phis­ti­ca­tion and mu­tu­ally af­firm our sta­tus as an in­tel­lec­tual sub­cul­ture? ;)

• Neu­tral vote. I like the PEZ jux­ta­po­si­tion but ‘ara­tional’ would fit bet­ter. A sim­ply false as­ser­tion doesn’t fit well with the irony.

• As it was mock­ing bgrah’s as­ser­tion, and bgrah used “un­ra­tional”, and in my es­ti­ma­tion his mean­ing was closer to “ir­ra­tional” than “ara­tional”, I used the former. Per­haps us­ing “un­ra­tional” would have been bet­ter, though.

• Just con­sider it ev­i­dence of the level of cul­ture you’ll find here­abouts. Sav­ages.

• Wow! My brain can au­to­mat­i­cally re­con­struct a 3D image from limited 2D in­put and even com­pen­sate for shad­ows and light­ing.

I know! How bizarre! ;)

• I am be­ing stupid when my eye looks at this illu­sion and I in­ter­pret the data in such a way to de­ter­mine dis­tinct col­ors.

Not at all. In the con­text of the scene that this pic­ture rep­re­sents, A and B are ab­solutely differ­ent shades. On the con­trary, I think your per­cep­tual sys­tem would be poor in­deed if it did not re­con­struct con­text, and un­der-in­ter­preted the pic­ture as a mean­ingless 2D ar­ray of pix­els.

(BTW, as with the necker cube, I find that I can con­sciously ex­ert to ex­pe­rience the in­ter­pre­ta­tion that I choose, with­out too much difficulty.)

• Hmm. I can with the necker cube, but not at all with this one.

• I was never able to do it with this one be­fore, ei­ther. What I’m do­ing now is con­cen­trat­ing hard on the two tiles of in­ter­est, un­til the rest of the pic­ture fades into the back­ground. The two tiles then seem to be float­ing on a sep­a­rate top layer, and ap­pear to be the same shade.

• That worked! Cool!

• If you go to an Art or De­sign school. See­ing and pro­duc­ing illu­sions like this is one of the as­sign­ments that they usu­ally will give you in a 2D de­sign class.

As it has been de­scribed above, if you can con­cen­trate (if school, we learn how to look at them by squint­ing as we would when dis­cern­ing sim­ple shape or color—or, if you have ever learned how to look at one of those weird 3D images made out of what looks to be paint splat­ter) on the two squares, then you will be able to see that they are in­deed the same shade (not color, color is used to de­scribe some­thing else)

• Ah, that worked for me. For peo­ple won­der­ing how to do the tech­nique to see “Magic Eye ” images, you fo­cus your eyes so that the image dou­bles and and over­laps the image. That causes a stere­o­scopic illu­sion when done on any things that over­lap. You could prac­tice it here. Fo­cus your eyes so that the the first abc over­laps the sec­ond abc—now you have three abc’s in your vi­sion, the 1st and 3rd abc are be­ing seen out of one eye and the abc in the mid­dle ap­pears to be al­most 3d.

a...b...c......................a...b...c

In this case, I could see that A and B are the same color by tilt­ing my head and then fo­cus­ing so I saw a dou­ble image of A over­lap­ping B.

• Ex­actly… We spent a to­tal of 6 weeks in Art School de­sign class learn­ing how to do this spe­cific trick with a va­ri­ety of images. From color, to line length (you know those “which line is longer” tricks that make you think one line is longer when they are usu­ally the same length), to line thick­ness, to shad­ing and tint­ing ali­as­ing.

We spent those weeks con­sum­ing a lot of as­pirin and Tylenol.

• In­ter­est­ing, when I try this tech­nique the shades seem even more dis­tinct.

• It takes some prac­tice.

We were taught that if you put your nose right in the cen­ter of the image, and then let your fo­cus go, and pull back from the image, that at a cer­tain dis­tance from the image (as your fo­cus is still at ∞) var­i­ous struc­tures of the image will be­gin to re­solve. So con­trasts, similar­i­ties, and shades will all re­solve at differ­ent fo­cal lengths from the image.

It was rare that any one per­son would be able to pick up im­me­di­ately upon all the effects per­cep­ti­ble in an image. I was able to pick up on cer­tain shades of the color green that are used in con­trast to red, but it took me a long time to get the shad­ing of black-white (as in this op­ti­cal illu­sion—and it is but one of many).

When we were tested on this, we would not be told what was similar, or where op­ti­cal tricks were used, and we would have to pick them out of an image (and this was long be­fore the in­ter­net, so we couldn’t just go on­line to do a search for op­ti­cal illu­sions to find images to study that had their illu­sions spel­led out for us). So, it is a skill that can be learned. For me, even­tu­ally I had to learn how to fo­cus upon each square with a differ­ent eye, while squint­ing, and let­ting the fo­cus go back and forth be­tween my right and left eye. even­tu­ally, I get the images re­solved as a sin­gle shade as I go back and forth be­tween my eyes.

• I found two ways to do it my­self:

(A) Cover up ar­eas of the image to see what causes what to change color in your per­cep­tion. Slowly re­veal the full image again and some­times A and B look alike

(B) Let your fo­cus drift un­til the lines of the image get fuzzy. Look at the two squares with­out ac­tu­ally look­ing at them. I find that the col­ors look al­ike here. If I “snap” fo­cus back they still look al­ike but noth­ing is fuzzy any­more.

B works bet­ter.

• The point of the illu­sion is that they seem differ­ent in con­text. Ig­nor­ing con­text to make them ap­pear similar isn’t a proper re­s­olu­tion.

• I don’t un­der­stand this. Are you say­ing that A and B are not the same color?

• AndyWood gave a good ex­pla­na­tion, but let me elab­o­rate. If you saw the scene de­picted, but in real life—rather than on a flat pa­per or 2D screen—you would be cor­rect to in­fer that the ac­tual, in­var­i­ant col­ors of the tiles are differ­ent. But, since they are just pix­els on pa­per or a screen, their in­var­i­ant col­ors are the same, and yet your eyes tell you oth­er­wise.

So are the eyes “wrong” in any se­ri­ous sense? Well, let me put it this way: do you want

a) a vi­sual sys­tem that gives the right in­ter­pre­ta­tion of scenes that you are ac­tu­ally go­ing to en­counter of­ten, but is tripped up by care­fully de­signed op­ti­cal illu­sions?

or do you want:

b) a vi­sual sys­tem that gives the right in­ter­pre­ta­tion for care­fully de­signed op­ti­cal illu­sions, but fails to catch many at­tributes of com­mon scenes?

(Yes, there is a trade­off. Your vi­sual sys­tem en­coun­ters an “in­verse op­tics” prob­lem: given the retina images, what is the scene you’re look­ing at made of? This is ill-posed: many scenes can gen­er­ate the same reti­nal images. E.g. a given square could be far away and big, or close and small. To con­strain the solu­tion set, you need as­sump­tions, and any set of as­sump­tions will get some scenes wrong.)

Yes, you are wrong to think that the tiles have differ­ent col­ors. You are not wrong to pre­fer a vi­sual sys­tem that gets most scenes right at the cost of get­ting a few scenes (like this one) wrong.

(In­ci­den­tally, I re­ally like this op­ti­cal illu­sion, and have it by my desk at work. What’s so great about it is that once you see it, you can ac­tu­ally strip away ev­ery­thing that you think is caus­ing the illu­sion, and yet they still look differ­ent!)

• Your un­der­stand­ing of the word ‘colour’ does not match what prop­er­ties of the world your brain is try­ing to iden­tify and cat­e­go­rize when it in­ter­prets ‘colour’. The in­ter­est­ing con­stant prop­erty of ob­jects in the world that makes ‘colour’ use­ful to your vi­sual sys­tem for pur­poses of ob­ject iden­ti­fi­ca­tion and cat­e­go­riza­tion is re­ally the sur­face prop­er­ties that in­ter­act with in­ci­dent light­ing. Your brain at­tempts to ig­nore effects due to light­ing vari­a­tion and as­sign a ‘colour’ la­bel to ob­jects that is more or less an in­var­i­ant prop­erty of the sur­face un­der a va­ri­ety of differ­ent light­ing con­di­tions. This is in gen­eral not a solv­able prob­lem since the same in­ci­dent pho­tons can be pro­duced by a num­ber of differ­ent light­ing and ma­te­rial com­bi­na­tions. Op­ti­cal illu­sions like this merely re­veal the heuris­tics your vi­sual sys­tem uses to iden­tify the rele­vant con­stant as­pects of the scene and ig­nore the ir­rele­vant light­ing vari­a­tion. They gen­er­ally work quite well.

• When we cov­ered this phe­nomenon in my psy­chol­ogy de­gree it was referred to as colour con­stancy. I now work as a 3D graph­ics pro­gram­mer and so know a lot about the physics of light trans­port. The illu­sion does not sur­prise me any more, in fact it seems a lit­tle sur­pris­ing that I ever could have thought that the RGB colour value of an on­screen pixel was di­rectly re­lated to the prop­erty of ob­jects in the real world that we call ‘colour’.

• Well said (in­clud­ing your later com­ment about color con­stancy). Along the same lines, this is why cam­eras of­ten show ob­jects in shad­ows as blacked out—be­cause that’s the ac­tual image it’s get­ting, and the image your own reti­nas get! It’s just that your brain has clev­erly sub­tracted out the im­pact of the shadow be­fore pre­sent­ing it to you, so you can still see sig­nifi­cant con­trast and col­ors in the shad­owed ob­jects.

• Along the same lines, this is why cam­eras of­ten show ob­jects in shad­ows as blacked out—be­cause that’s the ac­tual image it’s get­ting, and the image your own reti­nas get! It’s just that your brain has clev­erly sub­tracted out the im­pact of the shadow be­fore pre­sent­ing it to you

That doesn’t ex­plain why faith­ful re­pro­duc­tions of images with shad­ows don’t prompt the same rein­ter­pre­ta­tion by your brain.

• Blacked out shad­ows are gen­er­ally an in­di­ca­tion of a failure to gen­er­ate a ‘faith­ful’ re­pro­duc­tion due to dy­namic range limi­ta­tions of the cam­era and/​or dis­play medium. There is a fair amount of re­search into how to work around these limi­ta­tions through tone map­ping. High Dy­namic Range cam­eras and dis­plays are also an area of ac­tive re­search. There’s not re­ally any­thing to ex­plain here be­yond the fact that we cur­rently lack the cap­ture or dis­play ca­pa­bil­ity to faith­fully re­pro­duce such scenes.

• Sure it does—Faith­ful re­pro­duc­tions give the shad­owed por­tion the ap­pro­pri­ate col­ors for match­ing how your brain would per­ceive a real-life shad­owed por­tion of a scene.

• Umm, that’s not what I meant by “faith­ful re­pro­duc­tions”, and I have a hard time un­der­stand­ing how you could have mi­s­un­der­stood me. Say you took a pho­to­graph us­ing the ex­act vi­sual in­put over some 70 square de­grees of your vi­sual field, and then com­pared the pho­to­graph to that same view, try­ing to con­trol for all the rele­vant vari­ables*. You seem to be say­ing that the pho­to­graph would show the shad­ows as darker, but I don’t see how that’s pos­si­ble. I am fa­mil­iar with the phe­nomenon, but I’m not sure where I go wrong in my thought ex­per­i­ment.

* photo cor­rectly lit, held so that it sub­tends 70 square de­grees of your vi­sual field, with your head in the same place as the cam­era was, etc.

• I thought you meant “faith­ful” in the sense of “see­ing this is like see­ing the real thing”, not “see­ing this is learn­ing what your reti­nas ac­tu­ally get”. If you show a pho­to­graph that shows ex­actly what hit the film (no filters or pro­cess­ing), then dark por­tions stay dark.

When you see the scene in real life, you sub­tract off the av­er­age col­or­ing that can be de­ceiv­ing. When you see the photo, you see it as a photo, and you use your cur­rent real-life-back­ground and light­ing to de­ter­mine the av­er­age color of your vi­sual field. The dark­ness on the photo de­vi­ates sig­nifi­cantly from this, while it does not so de­vi­ate when you’re im­mersed in the ac­tual scene, and have enough in­for­ma­tion about the shadow for your brain to sub­tract off the ex­ces­sive black­ness.

Been a long day, hope I’m mak­ing sense.

• As oth­ers have pointed out, the difficulty here is more in the se­man­tics of “color” than in the op­tics.

As a sim­plifi­ca­tion, we can con­sider the pro­jected color.P of a tile to be a product of its sur­face prop­er­ties (color.S) and the in­ten­sity of the in­ci­dent light. The illu­sion straight­for­wardly con­trives one of these terms—the light in­ten­sity—so that the color.P of tile A equals the color.P of tile B. But the brain, in­ter­pret­ing the image as a 3D scene with light and shadow, re­ports the color.S-es of the tiles, which are differ­ent un­der that very rea­son­able and use­ful in­ter­pre­ta­tion.

I’m sorry if this is a big dis­trac­tion from the point of your post. I’m still in­ter­ested in the point, so per­haps you can find an­other way of get­ting it across.

• As oth­ers have pointed out, the difficulty here is more in the se­man­tics of “color” than in the op­tics.

Yeah. I missed the se­man­tic shift. All it took was some­one point­ing out that there were two uses of Color drift­ing around and al­most all the com­ments snapped back into mak­ing sense.

I’m sorry if this is a big dis­trac­tion from the point of your post. I’m still in­ter­ested in the point, so per­haps you can find an­other way of get­ting it across.

The point is that an illu­sion gen­er­ally gives off a sense of bizarreness be­cause we are ex­pect­ing X but the illu­sion gives us Y. In the case of the color ex­am­ple, I once ex­pected boxes A and B to ap­pear to be the same color (per­ceived) if and only if they were the same color (RGB). The illu­sion shows this is not the case. Be­ing cu­ri­ous, I sought to un­der­stand the un­der­ly­ing prin­ci­ples be­hind why we per­ceive two differ­ent col­ors. Once this is un­der­stood, the illu­sion should no longer seem bizarre but a triv­ial ex­am­ple of the un­der­ly­ing prin­ci­ples.

In try­ing to find where I went wrong with the post, I come up with this:

• “Color” is an ex­tremely am­bigu­ous term. I should have seen this one com­ing.

• I think some peo­ple thought I was try­ing to give an ex­pla­na­tion of the illu­sion in the post. I was not.

• I think some peo­ple thought I was say­ing that the vi­sual sys­tem it­self was stupid or bro­ken and we needed to “fix” our brain to ad­just for its short­com­ings. I was not. I was try­ing to say that our feel­ing of “bizarre” was stupid be­cause we are ex­pect­ing some­thing differ­ent from our vi­sual sys­tem than what the vi­sual sys­tem pro­vides.

• I de­liber­ately wrote this post more ag­gres­sively and con­cisely than I gen­er­ally write. Per­haps this de­graded its clar­ity even fur­ther.

I am half tempted to take this post down, rewrite it, and put it back up, but I don’t know how much that would help.

• Well, don’t do any­thing that takes down the com­ment sec­tion. Many of the com­ments are in­sight­ful and, um, say things that should have been in your origi­nal post.

De­mys­tify­ing op­ti­cal illu­sions, and vi­sual cog­ni­tion in gen­eral, is a very good ex­er­cise in ra­tio­nal­ist re­duc­tion.

• Okay. Do you think it would be valuable to just edit the post in place?

As best as I can tell, these are the trou­ble para­graphs:

To­day I looked at the above illu­sion and thought, “Why do I keep think­ing A and B are differ­ent col­ors? Ob­vi­ously, that is not what my vi­sual sys­tem is try­ing to tell me.” I am be­ing stupid when my eye looks at this illu­sion and I in­ter­pret the data in such a way to de­ter­mine dis­tinct col­ors. That in­for­ma­tion is not be­ing trans­mit­ted and re­ceived. If it were, the illu­sion wouldn’t be an illu­sion.

An op­ti­cal illu­sion is only bizarre if you are mak­ing a bad as­sump­tion about how your vi­sual sys­tem is sup­posed to be work­ing. It is a flaw in the Map, not the Ter­ri­tory. I should stop think­ing that the sys­tem is re­port­ing True Colors. It isn’t. And, now that I know this, I am sud­denly cu­ri­ous about what it is re­port­ing. I have dropped a bad be­lief and am look­ing for a re­place­ment. Once I have found the right an­swer, this op­ti­cal illu­sion should be­come as un­in­ter­est­ing as ques­tion­ing whether 1 is prime. It should stop be­ing weird, bizarre, and in­cred­ible. It should high­light an ob­vi­ous re­al­ity.

Is this bet­ter?:

To­day I looked at the above illu­sion and thought, “Why do I keep think­ing A and B are differ­ent col­ors? Ob­vi­ously, some­thing is wrong with how I am think­ing about col­ors.” I am be­ing stupid when I look at this illu­sion and in­ter­pret the data in such a way to de­ter­mine dis­tinct col­ors. My ex­pec­ta­tions of Real­ity and the in­for­ma­tion be­ing trans­mit­ted and re­ceived are not lin­ing up. If it were, the illu­sion wouldn’t be an illu­sion.

An op­ti­cal illu­sion is only bizarre if you are mak­ing a bad as­sump­tion about how your vi­sual sys­tem is sup­posed to be work­ing. It is a flaw in the Map, not the Ter­ri­tory. I should stop think­ing that the sys­tem is re­port­ing True Colors. It isn’t. And, now that I know this, I am sud­denly cu­ri­ous about what it is re­port­ing. I have dropped a bad be­lief and am look­ing for a re­place­ment. In this case, the vi­sual sys­tem is dis­t­in­guish­ing be­tween [what term goes here?], not in­di­vi­d­ual RGB style col­ors. Now that I have right an­swer, this op­ti­cal illu­sion should be­come as un­in­ter­est­ing as ques­tion­ing whether 1 is prime. It stops be­ing weird, bizarre, and in­cred­ible. It merely high­lights an ob­vi­ous re­al­ity.

• It seems to me that you are still us­ing the word colour in a way that sug­gests you haven’t re­ally grasped the in­sight that makes this illu­sion seem not-bizarre. That in­sight is fun­da­men­tally that the state­ment “this ball is blue” is not equiv­a­lent to the state­ment “a digi­tal photo of a scene con­tain­ing this ball would have pixel val­ues of 0, 0, 255 at pixel lo­ca­tions where light from the ball reached the sen­sor”. It is a much more com­plex (and more use­ful) state­ment than that. The bad as­sump­tion is that ‘colour’ when used to re­fer to a prop­erty of ob­jects in the world de­ter­mined through vi­sual per­cep­tion has any sim­ple re­la­tion­ship with RGB val­ues recorded by a digi­tal cam­era. You still seem to be talk­ing as if RGB val­ues are some­how ‘true’ colours.

• The bad as­sump­tion is that ‘colour’ when used to re­fer to a prop­erty of ob­jects in the world de­ter­mined through vi­sual per­cep­tion has any sim­ple re­la­tion­ship with RGB val­ues recorded by a digi­tal cam­era.

Espe­cially in the case of hu­man tetra­chro­mats.

• I am try­ing to find a way to say what you said with one phrase or word. I feel like I am strug­gling to find a term.

• I think the key for me in un­der­stand­ing this type of illu­sion (and the gen­eral phe­nomenon of colour con­stancy) was to re­al­ize that ‘colour’ in com­mon us­age (“this ball is blue”) is per­ceived as a prop­erty of ob­jects and we in­fer it in­di­rectly based on light that reaches our reti­nas. That light also has a ‘colour’ (sub­tly differ­ent mean­ing) but it is not some­thing we per­ceive di­rectly be­cause it is not very use­ful in it­self.

This makes perfect sense when you think about it from an evolu­tion­ary per­spec­tive—we evolved to rec­og­nize in­var­i­ant prop­er­ties of ob­jects in the world (pos­si­bly fruit in trees for pri­mates) un­der widely vary­ing light­ing con­di­tions. Directly per­ceiv­ing the ‘colour’ (RGB) of light would not tell us any­thing very use­ful about in­var­i­ant ob­ject prop­er­ties. There is enough over­lap be­tween the two mean­ings of colour for them to be eas­ily con­fused how­ever and that is re­ally the root of this par­tic­u­lar illu­sion.

In com­puter graph­ics we com­monly use the term ‘ma­te­rial’ to de­scribe the set of prop­er­ties of a sur­face that gov­ern how it re­sponds to in­ci­dent light. This en­com­passes prop­er­ties be­yond sim­ple colour (“shiny blue ball”, “matte blue ball”, “metal­lic blue ball”). I don’t know if that us­age is well un­der­stood out­side of the com­puter graph­ics field how­ever.

• I com­pletely agree with you. At this point, I am just try­ing to clean up the ar­ti­cle to help clar­ify the an­swer be­hind the illu­sion. Does the phrase, “I should stop think­ing that the vi­sual sys­tem is re­port­ing RGB style col­ors” mesh okay? That is the only lo­ca­tion of RGB as of this edit.

• Yes, I think ‘RGB colours’ is bet­ter than ‘True Colours’ in this con­text.

• Thanks. Do you have any other sug­ges­tions that may help clar­ify the ar­ti­cle? Your ex­pla­na­tions have been very helpful. Learn­ing the terms was ap­par­ently some­thing I never both­ered to do. Oops. :P

• The ar­ti­cle reads bet­ter now. So do you feel the bizarreness has dis­ap­peared now you un­der­stand the phe­nomenon bet­ter?

• Yes. The key point that you men­tioned here:

I think the key for me in un­der­stand­ing this type of illu­sion (and the gen­eral phe­nomenon of colour con­stancy) was to re­al­ize that ‘colour’ in com­mon us­age (“this ball is blue”) is per­ceived as a prop­erty of ob­jects and we in­fer it in­di­rectly based on light that reaches our reti­nas.

This hap­pened some­time this morn­ing. The more I read here, the more I un­der­stand it in the sense that I know the name of the rele­vant field, a whole bunch of new terms, and more de­tails about how we per­ceive col­ors. It gets less and less bizarre as the day goes, which is always fun. :)

• How is “Color gross of light­ing con­di­tions”?

• It sounds like you’re try­ing to come up with a sen­tence or two that cap­tures all of the in­sight on color that the com­menters have given. While I’m a big fan of sum­ma­riz­ing, and a big critic of those who can’t, I don’t think you can get it to work here. In­stead of your fi­nal bolded change (the oth­ers are good), just point to or quote a few good com­ments that show what the vi­sual sys­tem is do­ing, and how the op­ti­cal illu­sions trick it.

• How about:

In this case, the vi­sual sys­tem is not try­ing to dis­t­in­guish be­tween in­di­vi­d­ual RGB style col­ors (more de­tails in the com­ments).

EDIT: I up­dated it with some­thing similar. Hope­fully it was an im­prove­ment. :) Thanks again for your help (which isn’t to say that I wouldn’t mind more help...)

• I am half tempted to take this post down, rewrite it, and put it back up, but I don’t know how much that would help.

Tak­ing this and SilasBarta’s thoughts to­gether: can you ap­ply this same meta-prin­ci­ple to some­thing sub­stan­tially differ­ent in a new post, writ­ten with a recog­ni­tion of these con­fu­sions? That post could cite this post with a “Fol­lowup to:” line, and elab­o­rate on your dis­cov­ery in some way.

• Would it be bet­ter to just re­place the con­tent of this post? I can archive the origi­nal in a com­ment here for fu­ture con­text.

• I would be dis­in­clined to that course, but hard-pressed to jus­tify it more effec­tively than by my idiosyn­cratic gen­er­al­iza­tion of one of a num­ber of prin­ci­ples I have heard—I quote from the post:

Don’t try to rewrite his­tory. Look, we make mis­takes. We all do. Some­times we post an es­say and we get stuff wrong in it. Some­times that stuff makes the whole es­say wrong. Some­times, we put up an es­say in­no­cently and it turns into a firestorm of con­tro­versy we never meant. Some­times, we find our­selves in a cru­cible on all sides.

The temp­ta­tion is to go back. Re­vise. Re­word what we said. Take the es­say down en­tirely.

It is never a good idea. Ever.

I don’t think you have any­thing to be ashamed of in this post. It’s not deep, it’s not ex­traor­di­nary in its con­clu­sions, but it is cor­rect and brief. The com­plaints seem to me best ad­dressed by elab­o­ra­tion and dis­cus­sion—things which re­quire far more than a brief edit placed at the end of the post.

As SilasBarta men­tioned, there’s a lot of com­men­tary on this post that is worth pre­serv­ing, and should be pre­served with the origi­nal post. It would be un­fair to the com­menters to ren­der their com­ments in­com­pre­hen­si­ble—even briefly—by dis­tor­tion of that to which they re­sponded.

And, if I may be frank, if the idea which in­spired this post is in­ter­est­ing, it is prob­a­bly ca­pa­ble of gen­er­al­iza­tion. The idea of my own which I pro­moted to a post I did so be­cause I saw that it was ap­pli­ca­ble be­yond the scope of its origi­na­tion, and in a man­ner which was nat­u­ral, el­e­gant, and in­ter­est­ing. It proved of in­ter­est to a num­ber of peo­ple here, de­spite its un­abashedly alge­braic treat­ment. If you can find a prof­itable ex­ten­sion of your con­cept, it will be likely to be worth re­port­ing in a fol­lowup post (and if you are con­cerned about the ap­pro­pri­ate­ness of it, I—as one re­main­ing up­voter of the OP—will have sent my email to you in a PM, and be will­ing to com­ment on any draft you wish to send).

If you can­not find a prof­itable ex­ten­sion of your con­cept, it is prob­a­bly not worth the time to re­vise. Con­sider your post du­bi­ously suc­cess­ful (it is still in pos­i­tive ter­ri­tory, is it not?) and leave it be.

• I don’t think you have any­thing to be ashamed of in this post. It’s not deep, it’s not ex­traor­di­nary in its con­clu­sions, but it is cor­rect and brief. The com­plaints seem to me best ad­dressed by elab­o­ra­tion and dis­cus­sion—things which re­quire far more than a brief edit placed at the end of the post.

It’s not so much that I am ashamed; I am just frus­trated. The be­hav­ior of this post caught me com­pletely off-guard. It was up­voted to +5 within a few hours and peo­ple started ask­ing ques­tions. After my re­sponses, the post dropped to +1. The karma it­self doesn’t mean much to me, but the feed­back here was ev­i­dence of some­thing greater than a non-in­ter­est­ing or in­cor­rect post.

Peo­ple were will­ing to talk about it, so I stuck it out for as much feed­back as I could. The in­vest­ment was com­pletely worth it. I got sev­eral com­ments worth of ex­tremely valuable in­sights to my writ­ing style and how to bet­ter post here at LessWrong.

I think the post it­self failed, but the whole ex­pe­rience has been a net gain.

As SilasBarta men­tioned, there’s a lot of com­men­tary on this post that is worth pre­serv­ing, and should be pre­served with the origi­nal post. It would be un­fair to the com­menters to ren­der their com­ments in­com­pre­hen­si­ble—even briefly—by dis­tor­tion of that to which they re­sponded.

I agree. My in­tent in the re­vi­sions has been to keep peo­ple from be­ing dis­tracted by my quirks and lead­ing them into a won­der­ful dis­cus­sion in the com­ments. This par­tic­u­lar illu­sion has a lot more his­tory be­hind it than I origi­nally thought; I learned a lot.

And, if I may be frank, if the idea which in­spired this post is in­ter­est­ing, it is prob­a­bly ca­pa­ble of gen­er­al­iza­tion. The idea of my own which I pro­moted to a post I did so be­cause I saw that it was ap­pli­ca­ble be­yond the scope of its origi­na­tion, and in a man­ner which was nat­u­ral, el­e­gant, and in­ter­est­ing. It proved of in­ter­est to a num­ber of peo­ple here, de­spite its un­abashedly alge­braic treat­ment. If you can find a prof­itable ex­ten­sion of your con­cept, it will be likely to be worth re­port­ing in a fol­lowup post (and if you are con­cerned about the ap­pro­pri­ate­ness of it, I—as one re­main­ing up­voter of the OP—will have sent my email to you in a PM, and be will­ing to com­ment on any draft you wish to send).

Thank you very much. I have to sit on the events of to­day and pon­der if there is a next step to take. If a fol­lowup is com­ing I will cer­tainly take you up on your offer.

• An ad­den­dum—as far as my rec­ol­lec­tion of the origi­nal goes, your ed­its ap­pear rea­son­able, al­though I would not have risked them on my own post. I con­grat­u­late you on a suc­cess­ful re­vi­sion, but my offer stands.

• In the spirit of not dis­put­ing defini­tons, may I sug­gest: A and B are the same col­orRGB, but, in­ter­pret­ing the image as a pic­ture (as the eye does), not the same col­orALBEDO.

Edit: Cor­rec­tion—“as the vi­sual sys­tem does”.

• This is per­haps beat­ing a dead horse, but “albedo” is sup­posed to be a ra­tio be­tween re­flected and in­ci­dent lights, and I would bet that the albedo of these two patches of screen is also iden­ti­cal, just as their RGB val­ues are iden­ti­cal.

• Not in the ac­tual 3D scene that your brain in­ter­prets the pic­ture to be of, only in the con­text of a 2D print­out of the image (albedo is not re­ally a rele­vant prop­erty for emis­sive dis­play de­vices like LCDs or CRTs).

• When I said “in­ter­pret­ing the image as a pic­ture”, I meant, “in­ter­pret­ing the image as a pic­ture of a checker­board with a cylin­der cast­ing a shadow on it”—the albe­dos in ques­tion are of the squares A and B on the de­picted board.

• Ah, “in­ferred albedo”. In that case we agree.

• They are the same screen-color, but differ­ent in­ferred-col­ors.

• I iden­ti­fied a use­ful and co­gent point in your post and it was this: When­ever you re­ceive data from any source (your brain, your eyes, a drug study, Less Wrong) you’ve got to be aware of how that data has already been pack­aged. Tak­ing the data at face value—for ex­am­ple imag­in­ing your brain is ac­tu­ally mak­ing a claim about the RGB val­ues of the pix­els—can lead to prob­lems, mis­con­cep­tions, mis­takes.

• Your eye didn’t evolve to re­port trivia like, “Th­ese two col­ors are ac­tu­ally the same.” Your eye is re­port­ing the most use­ful in­for­ma­tion—from which di­rec­tion the light is com­ing, the shaded re­gion un­der it, and the fact that the floor is tiled.

Which is more amaz­ing—this pic­ture, or a pic­ture that some­how tricked the av­er­age per­son into notic­ing two col­ors were the same, but didn’t no­tice the pic­ture also had floor tiling, light di­rec­tion­al­ity, and shad­ing? I’d say this pic­ture is pretty tame in com­par­i­son to the pic­ture that could do that.

• I am be­ing stupid when my eye looks at this illu­sion and I in­ter­pret the data in such a way to de­ter­mine dis­tinct col­ors.

Tell that to your an­ces­tors who es­caped from the saber-tooth cat hid­ing in the shad­ows at dusk.

• A side note: The only rea­son that prime num­bers are defined in such a way as to ex­clude 1 and nega­tive num­bers is be­cause math­e­mat­i­ci­ans found this way of defin­ing them a bit more use­ful than the al­ter­na­tive pos­si­bil­ities. Math­e­mat­i­ci­ans gen­er­ally de­sire for im­por­tant the­o­rems to be stated in a man­ner that is as sim­ple as pos­si­ble, and the the­o­rems about primes are gen­er­ally sim­pler if we ex­clude 1. There is a more de­tailed anal­y­sis of this ques­tion here:

http://​​www.aska­math­e­mat­i­cian.com/​​?p=1269

• if you use a poor defi­ni­tion such as, “Prime is a num­ber that is only di­visi­ble by it­self and 1.”

I have a fond­ness for this par­tic­u­lar defi­ni­tion, and like to think of 1 as a “very spe­cial” prime num­ber. To the ex­tent that I usu­ally give a lit­tle speech when­ever an op­por­tu­nity arises that (ahem) the only rea­son I know of that ‘1’ is ex­cluded from the primes (more of­ten than not) is be­cause al­most ev­ery the­o­rem about prime num­bers would have to be mod­ified with an “ex­cept 1″ clause. But a nat­u­ral defi­ni­tion (any­thing along the lines of “already com­pletely fac­tored”) would in­clude it. If you dis­agree, which defi­ni­tion—or the satis­fac­tion of which the­o­rem—do you think is more com­pel­ling?

(Just in case you per­ceived you were get­ting too much heat about “colour”...)

• I’d have to pretty strongly dis­agree. To me, the “essence” of primes is that you can fac­tor any num­ber into primes in a unique way. That’s the most nat­u­ral defi­ni­tion. They’re the mul­ti­plica­tive build­ing blocks of the nat­u­ral num­bers; ev­ery­thing can be re­duced to them. If 1 were prime, you could no longer fac­tor uniquely.

• Your com­ment and this com­ment were ad­ja­cent in my mes­sage folder which I found amus­ing.

Thomblake wrote:

Car­ing about what’s right might be as ar­bi­trary (in some ob­jec­tive sense) as car­ing about what’s prime, but we do ac­tu­ally hap­pen to care about what’s right.

It’s funny how we do care.

• Hmm… I agree this is com­pel­ling. How­ever, since I’m re­sis­tant to up­dat­ing my world view about 1-the-dis­crim­i­nated-prime-num­ber, I’ll con­tinue to proffer counter-ar­gu­ments:

• the Fun­da­men­tal The­o­rem of Arith­metic is pretty im­por­tant, but may still not be the “essence” of what prime is

• the FTA it­self re­quires the “ex­cept 1” clause: “all nat­u­ral num­bers can be uniquely fac­tored into primes ex­cept 1″ -- which would make some­one thing 1 ought to be prime

• the FTA already as­sumes ‘mod­ulo per­mu­ta­tions’, we could eas­ily throw in ‘mod­ulo 1’

• Wikipe­dia—the first and last au­thor­ity on such things—care­fully writes in an en­tire sen­tence unto it­self, “The num­ber 1 is by defi­ni­tion not a prime num­ber,” sug­gest­ing just how ar­bi­trary this is. (My own em­pha­sis added.)

The best ar­gu­ment I came up with for not in­clud­ing 1 as prime, be­cause I tend to worry about how things are con­structed, was with the seive of Eratos­thenes.

The seive of Eratos­thenes says that you can find the primes by start­ing with all the nat­u­ral num­bers > 1; let 2 be the first prime num­ber, and then be­gin elimi­nat­ing all mul­ti­ples of 2 and the mul­ti­ples of sub­se­quent primes as you find them. If you in­cluded ‘1’ in the first step, then you would elimi­nate all the num­bers in the first step.

• I think you’re re­ally failing to grasp the con­tent of the unique fac­tor­iza­tion the­o­rem here. Firstly we don’t think about fac­tored num­bers as prod­ucts of primes up to per­mu­ta­tion, we think of them as prod­ucts of dis­tinct prime pow­ers (up to per­mu­ta­tion, I sup­pose—but it’s prob­a­bly bet­ter here to just take a com­mu­ta­tive view­point and not re­gard “up to per­mu­ta­tion” as worth spec­i­fy­ing). But more im­por­tantly, you need to take a mul­ti­ary view of mul­ti­pli­ca­tion here, not a bi­nary one. 1 is the empty product, so in par­tic­u­lar, it is the product of no primes, or the product of each prime to the 0th power. That is its unique prime fac­tor­iza­tion. To take 1 as a prime would be like hav­ing bases for vec­tor spaces in­clude 0. Al­most ex­actly like it—if we take the Z-mod­ule of pos­i­tive ra­tio­nals un­der mul­ti­pli­ca­tion, the set of primes forms a free ba­sis; 1 is the zero el­e­ment.

• In­for­ma­tion and ex­per­tise like this is why hang­ing out at Less Wrong is worth the time. I es­ti­mate that I value the in­for­ma­tion in your com­ment at about $35, mean­ing my pre­sent self would ad­vise my former self to pay up to$35 to read it.

So, I get it. My brain is more wired for anal­y­sis than alge­bra; so this isn’t the first time that lin­ear alge­bra has been a use­ful bridge for me. I see that we could have a ‘vec­tor space’ of in­finite-di­men­sional vec­tors where each vec­tor (a1, a2, …, an, …) rep­re­sents a num­ber N where N = (P1^a1)(P2^a2)...(Pn^an)… and Pi are the or­dered primes. Clearly 1 is the zero el­e­ment and would never be a ba­sis el­e­ment.

I should ad­mit here that my back­ground in alge­bra is weak and I have no idea how you would need to mod­ify the no­tion of ‘vec­tor space’ to make cer­tain things line up. But I can already spec­u­late on how the choice of the “scalar field” for spec­i­fy­ing the a_i would have in­ter­est­ing con­se­quences:

• non-nega­tive in­te­ger ‘scalar field’ --> the pos­i­tive in­te­gers,

• all in­te­gers ‘scalar field’ --> pos­i­tive ra­tio­nal num­bers,

• com­plex in­te­gers --> fi­nally in­clude the nega­tive ra­tio­nals.

I’d like to read more. What sub-field of math­e­mat­ics is this?

• I see that we could have a ‘vec­tor space’ of in­finite-di­men­sional vec­tors where each vec­tor (a1, a2, …, an, …) rep­re­sents a num­ber N where N = (P1^a1)(P2^a2)...(Pn^an)… and Pi are the or­dered primes.

Oh! And or­thog­o­nal vec­tors are rel­a­tively prime!

• I’m not sure that the idea of or­thog­o­nal­ity is defined for mod­ules, is it? Is there a stan­dard defi­ni­tion of an in­ner product for a Z-mod­ule?

• I’m not sure that the idea of or­thog­o­nal­ity is defined for mod­ules, is it? Is there a stan­dard defi­ni­tion of an in­ner product for a Z-mod­ule?

Yes; the same defi­ni­tion works. See here.

• Yay! I ac­tu­ally got some­thing right!

• What sub-field of math­e­mat­ics is this?

Num­ber the­ory, ne? Or is that too gen­eral?

• It looks like it’s more ab­stract alge­bra (pos­si­bly ap­plied to num­ber the­ory) that byrnema is in­ter­ested in. Check out Wikipe­dia on mod­ule.)

• Pre­cisely! Thanks also.

• A sec­ond com­ment...

You’ve cer­tainly con­vinced me that ‘1’ should not be in­cluded in the set of things that are used to uniquely fac­tor num­bers. How­ever, how I can I know if this set is the set of “primes”?

I guess I was think­ing that the essence of primes was about their ir­re­ducibil­ity/​atomic-ness. The num­ber 5 would be con­sid­ered prime be­cause you can’t de­scribe it mul­ti­plica­tively in any way ex­cept by us­ing the num­ber 5. Us­ing my preferred no­tion, the num­ber 0 and the num­ber −1 would also be “prime” (as Mr Hen guessed). Is there a differ­ent word for this con­cept?

• See wikipe­dia on nat­u­ral gen­er­al­iza­tions of prime num­bers. In par­tic­u­lar note that most of the defi­ni­tions say “units” in­stead of “1″, like “Irre­ducible el­e­ments are ones which can­not be writ­ten as a product of two ring el­e­ments that are not units.” which rules out 0 for the in­te­gers, +, x and in­cludes the pos­si­bil­ity of mul­ti­ple units (-1 and 1).

I don’t know off­hand of any nice, com­monly refer­enced prop­erty P(S,O) that is: A,x,y in a struc­ture S with op­er­a­tion O: A is P just when if x O y = A then ei­ther x = A or y = A. Which I be­lieve is the gen­eral prop­erty you’re think­ing about?

Edit: with O com­mu­ta­tive I do believe

• Thank you. And yes, that is the prop­erty.

• For some rea­son, I never imag­ined fac­tors this way.

18 = 3^2 2^1
97,020 = 2^2
3^2 5 7^2 * 11

I sup­pose I have seen them printed out that way, but the deeper struc­ture there never clicked. Cool.

• As it hap­pens I’m part­way through “An In­tro­duc­tion to the The­ory of Num­bers” by Niven, Zuck­er­man, and Mont­gomery at the mo­ment. Lots of prob­lems are in­cred­ibly easy to solve given this struc­ture. The ex­am­ple that springs to mind is the very straight­for­ward proof why the com­bi­na­to­rial for­mula n! /​ (r! (n-r)!) always gives you an in­te­ger.

Up­date: Well hav­ing been scored up I feel like I should give a hint on the ac­tual proof: for any prime p and any n, the great­est power of p that di­vides n is

\sigma_{i=1}^{\in­fty} floor( \over{n}{p^i} )

and for any real num­bers a, b, floor(a + b) >= floor(a) + floor(b).

Oh for real TeX markup!

• Do you recom­mend the book? If I were in­ter­ested in the sub­ject, is this good to pick up or can you think of a bet­ter op­tion?

• I’m en­joy­ing it, but it touches on ab­stract alge­bra as an al­ter­na­tive ap­proach rather than lean­ing on it for ev­ery­thing; I’d kind of pre­fer the lat­ter.

• You may be a good per­son to ask this ques­tion:

I was won­der­ing if there was a func­tion f(x, y, z) so that x and z rep­re­sent the left and right sides of com­mon math­e­matic op­er­a­tors and y rep­re­sents the level of op­er­a­tion. So f(1, 2, 4) would be 1 + 4 and f(2, 2, 4) would be 2 * 4. Bet­ter ver­sions of f(x, y, z) would have fewer end cases hard­coded into it.

The rea­son be­hind this is to han­dle op­er­a­tor lev­els greater than ad­di­tion, mul­ti­pli­ca­tion, and ex­po­nents. The ca­sual anal­y­sis from my grade school and un­der­grad level math shows the pat­tern that mul­ti­pli­ca­tion is re­peated ad­di­tion and ex­po­nents are re­peated mul­ti­pli­ca­tion.

My quick at­tempts at com­ing up with such a func­tion are spiral­ing into greater and greater com­plex­ities. I figured some­one else has to have thought about this. Do you know of a place I can start read­ing up on ideas similar to this? Is what I am do­ing even plau­si­ble?

Quick thoughts based on me play­ing around:

• Ad­di­tion may be level 0, not level 1

• The se­quences never re­ally look ex­actly like mul­ti­pli­ca­tion ta­bles, but the pat­terns are similar enough to ap­pease me

• Ideally, ev­ery­thing can be re­duced to the sim­ple con­cept of X + 1 so as to walk along the num­ber line

• In prac­ti­cal terms, I have no idea how to ex­press “nega­tive” lev­els. Divi­sion and roots are un­ap­proach­able at this point in my play­ing around.

• Cool, thanks. It seems like one of my first tries was pro­duc­ing num­bers similar to the Ack­er­mann func­tion. Knuth’s ar­row no­ta­tion es­sen­tially takes over af­ter mul­ti­pli­ca­tion. But those two ar­ti­cles will give me enough to read to keep mov­ing on. :)

Do you know of any that go the other way into smaller and smaller num­bers?

EDIT: I found the right sub­ject name through links on your links. It is called hy­per­op­er­a­tion.

• 1 /​ Ack­er­mann func­tion.

• Sure, that works, but it isn’t ex­actly what I am look­ing for. Is it pos­si­ble to ex­press the di­vi­sion op­er­a­tor in a man­ner similar to how mul­ti­pli­ca­tion can be ex­pressed us­ing ad­di­tion? My in­stinct is tel­ling me prob­a­bly not.

• You can have in­verse op­er­a­tions for the higher op­er­a­tions as well. 4^4 is 256, so you can think of 4 as the “tetrated root” of 256. Also see this

(I’m us­ing ‘tetrat­ing’ as a term for the op­er­a­tion af­ter ex­po­nen­ti­a­tion: in other words, 4 tetrated to the 4th is 4^(4^(4^4))).

Two prob­lems: there may not be a clear way to define tetrat­ing and higher op­er­a­tions to frac­tional amounts, and ex­po­nen­ti­a­tion and up aren’t as­so­ci­a­tive, so you need a con­ven­tion for what to do with the paren­the­ses.

• Hyper op­er­a­tors. You can rep­re­sent even big­ger num­bers with Con­way chained ar­row no­ta­tion. Eliezer’s 3^^^^3 is a form of hy­per op­er­a­tor no­ta­tion, where ^ is ex­po­nen­ti­a­tion, ^^ is tetra­tion, ^^^ is pen­ta­tion, etc.

If you’ve ever looked into re­ally big num­bers, you’ll find info about Ack­er­mann’s func­tion, which is triv­ially con­vertable to hy­per no­ta­tion. There’s also Busy Beaver num­bers, which grow faster than any com­putable func­tion.

• Wikipe­dia—the first and last au­thor­ity on such things—care­fully writes in an en­tire sen­tence unto it­self, “The num­ber 1 is by defi­ni­tion not a prime num­ber,” sug­gest­ing just how ar­bi­trary this is.

A mea­sure of the ar­bi­trari­ness is the his­tory, which is that 1 was con­sid­ered prime up to the 19th cen­tury and was a mat­ter of fash­ion dur­ing the 19th cen­tury. That sug­gests that unique fac­tor­iza­tion is not, in it­self, enough to mo­ti­vate the defi­ni­tion. Per­haps its ex­ten­sion to the gaus­sian in­te­gers or the more rad­i­cal ver­sion for gen­eral num­ber rings prompted the defini­ton.

• This re­minds me. Pre-19th cen­tury it was thought that part of what it was to be a mam­mal was to give live birth, in ad­di­tion to hav­ing mam­mary glands. 1 is the platy­pus of num­bers.

• nat­u­ral defi­ni­tion: “A prime is a nat­u­ral num­ber with ex­actly two fac­tors”

I’m not sure I quite un­der­stand your sug­ges­tion: we should define 1 as prime, but then write “ex­cept for 1” ev­ery time we use the word prime? Wouldn’t it be quicker just to ex­clude 1 in the first place (even if there were some sense in which 1 was prime)?

• But a nat­u­ral defi­ni­tion (any­thing along the lines of “already com­pletely fac­tored”) would in­clude it.

How do you see 0 or −1, us­ing this defi­ni­tion?

• A fac­tor of a num­ber M is a num­ber that evenly di­vides M with no re­main­der. Zero has in­finitely many fac­tors, definitely not prime.

...re­gard­ing −1, I can’t think of any­thing rele­vant that I know about the re­la­tion­ship be­tween nega­tive num­bers and prime num­bers.

Later edit: Then I com­pletely changed my mind… and de­cided 0 and −1 should be prime rel­a­tive to how I would define it’s essence. I note that you in­tu­ited what I re­ally meant by prime bet­ter than I did!

• Yeah, I was just cu­ri­ous. I like toy­ing around with the fun­da­men­tals be­hind the maths and see­ing what hap­pens. :)

• Uh? What do you mean by “ob­vi­ous” in that last sen­tence?

(Post oth­er­wise in­ter­est­ing, and I for one like them short.)

• “Ob­vi­ous” as in not “weird, bizarre, or in­cred­ible.” Would “sim­ple” be a bet­ter word there?

• I’m just not see­ing what ob­vi­ous re­al­ity it high­lights, so ei­ther I’m par­tic­u­larly dense or it’s not in fact ob­vi­ous.

So, rephras­ing: what re­al­ity is be­ing high­lighted by the “illu­sion” ?

Your prime num­ber anal­ogy sug­gests that it’s in fact the “both col­ors are the same” as­ser­tion which is an illu­sion. The per­cep­tual re­al­ity is that the pix­els in these ar­eas are dis­crim­i­nated as differ­ent col­ors. The illu­sion con­sists of look­ing at pixel with iden­ti­cal RGB val­ues and think­ing “Oh, these have the same po­si­tion in col­orspace, I ex­pect my brain to per­ceive them as iden­ti­cal.”

The re­al­ity sug­gested by the “illu­sion” is that this ex­pec­ta­tion doesn’t hold in gen­eral, it’s a stupid model. A smarter model would take more things into ac­count be­fore it pre­dicted what our brain will per­ceive as iden­ti­cal col­ors.

But this is very much non-ob­vi­ous...

• I’m just not see­ing what ob­vi­ous re­al­ity it high­lights, so ei­ther I’m par­tic­u­larly dense or it’s not in fact ob­vi­ous. [...] But this is very much non-ob­vi­ous...

The post is key­ing off of Think Like Real­ity.

Real­ity has been around since long be­fore you showed up. Don’t go call­ing it nasty names like “bizarre” or “in­cred­ible”. The uni­verse was prop­a­gat­ing com­plex am­pli­tudes through con­figu­ra­tion space for ten billion years be­fore life ever emerged on Earth. Quan­tum physics is not “weird”. You are weird. You have the ab­solutely bizarre idea that re­al­ity ought to con­sist of lit­tle billiard balls bop­ping around, when in fact re­al­ity is a perfectly nor­mal cloud of com­plex am­pli­tude in con­figu­ra­tion space. This is your prob­lem, not re­al­ity’s, and you are the one who needs to change.

Look­ing at the image it should be ob­vi­ous that the col­ors do not look the same. This is re­al­ity. We think they should look the same even though it is ob­vi­ous they don’t. Once we find the right an­swer to why they don’t look the same, the illu­sion should stop be­ing bizarre.

If you find an ex­pla­na­tion, re­turn to the illu­sion, and still think the illu­sion is bizarre, than some­thing is wrong. You fall into the cat­e­gory that EY is dis­cussing in Think Like Real­ity.

I am con­vinced that most of what we con­sider to be fancy illu­sions will be con­sid­ered ob­vi­ous to fu­ture gen­er­a­tions. They will look at this image and won­der why we thought it was so fas­ci­nat­ing. When our op­tics sys­tem is solved it would com­pletely ridicu­lous to as­sume that we would look at that image and think that the two squares should look like the same color.

• But your post hasn’t offered an ex­pla­na­tion. And I don’t, in fact, look at that image and think that the two squares should look like the same color.

A and B are in fact differ­ent col­ors, for a value of “in fact” which takes into ac­count that the pic­ture is a pic­ture of some­thing—a checker­board. My vi­sual sys­tem makes the cor­rect in­fer­ence, con­di­tioned on the as­sump­tion that I’m look­ing at a checker­board.

EDIT: what I should say is that I’m still sur­prised, know­ing what I know about my vi­sual sys­tem and how it works, when you tell me that the pix­els have the same RGB val­ues. But that’s not a “re­al­ity is weird” sur­prise, it’s more like the sur­prise of learn­ing some in­ter­est­ing bit of trivia.

To re­ally be to­tally un­sur­prised, I’d have to en­hance not just my knowl­edge of the vi­sual sys­tem, but my vi­sual sys­tem—to in­clude an RGB cal­ibra­tion sys­tem.

• EDIT: Oh, okay, I read your edit and that makes much more sense. I agree that it may be difficult to get to the point of be­ing un­sur­prised. Get­ting there isn’t ob­vi­ous. You know you are there when you are un­sur­prised by the illu­sion. Once Real­ity is un­sur­pris­ing and ob­vi­ous, you are there.

I feel like I have lost the point of this con­ver­sa­tion. What, in the fol­low­ing, do you dis­agree with?

• The image is an op­ti­cal illusion

• The squares marked A and B ap­pear to be differ­ent colors

• In re­al­ity, the squares marked A and B are the same color

• It is more cor­rect to say that A and B are the same color than to say they are differ­ent colors

• The rea­son be­hind the op­ti­cal illu­sion ex­plains why A and B ap­pear to be differ­ent colors

• This rea­son is con­tained some­where in­side of the “vi­sual sys­tem”

• It is bet­ter to not be sur­prised by Reality

• The squares A and B should ap­pear to be differ­ent colors

• We should not be sur­prised when A and B look like differ­ent colors

• It is in­cor­rect to call Real­ity bizarre as per Think Like Reality

• “Ob­vi­ous” is not a good word for the op­po­site of “bizarre”

• I dis­agree with #3 and #4. Also #6, mildly—it’s not just our vi­sual sys­tem that’s at is­sue here, it’s our color vo­cab­u­lary and our “meta” think­ing about our color sys­tem that ex­plains the “illu­sion”—that ex­plains why we might think it bizarre.

• Oh, okay. Well… I guess I don’t feel like de­bat­ing the defi­ni­tion of color since that is com­pletely ir­rele­vant to the point of the post. I wish this was made clear ear­lier.

Per­haps I can an­swer your origi­nal ques­tion this way:

Uh? What do you mean by “ob­vi­ous” in that last sen­tence?

It has noth­ing to do with the ac­tual an­swer to the ex­am­ple illu­sion. What I mean is that once we have the an­swer to an illu­sion, the illu­sion should stop sur­pris­ing us.

• Nei­ther do I feel like de­bat­ing the defi­ni­tion of color.

What I am is dis­ap­pointed. You brought up the “color con­stancy” phe­nomenon as an in­stance where “think like re­al­ity” is ap­pli­ca­ble, and then failed to fol­low through with an anal­y­sis of what is ac­tu­ally go­ing on. You sound as if you are con­tent to know that the phe­nomenon is in prin­ci­ple ex­pli­ca­ble; as if the post has done its job by demon­strat­ing your com­mit­ment to “think­ing like re­al­ity”. I would pre­fer you had gone deeper into the ob­ject-level anal­y­sis and offered your own ex­pla­na­tion of what is go­ing on in this par­tic­u­lar case.

This is a lit­tle bit like par­ents who lec­ture their chil­dren about the im­por­tance of be­ing truth­ful, vs. par­ents who demon­strate be­ing truth­ful—and be­ing OK with con­fronting un­pleas­ant truths.

EDIT: I didn’t mean to sound sanc­ti­mo­nious (I re­al­ize I do sound sanc­ti­mo­nious). My main in­tent is to ex­press a wish re­gard­ing what I’d like to see in fu­ture posts of this type.

• What I am is dis­ap­pointed. You brought up the “color con­stancy” phe­nomenon as an in­stance where “think like re­al­ity” is ap­pli­ca­ble, and then failed to fol­low through with an anal­y­sis of what is ac­tu­ally go­ing on.

The point of this post is not to de­bate, dis­cuss, or an­a­lyze color con­stancy. The point of the post is to talk about illu­sions and how we think of them as bizarre when we shouldn’t.

I have not once de­bated color or the the­o­ries be­hind the illu­sion. All I did was use a word one way when other peo­ple use it an­other way. I am not try­ing to offer some strange, new truth about a pic­ture. I used it as an ex­am­ple be­cause peo­ple rec­og­nize it, not be­cause that par­tic­u­lar ex­am­ple mat­ters.

EDIT: I didn’t mean to sound sanc­ti­mo­nious (I re­al­ize I do sound sanc­ti­mo­nious). My main in­tent is to ex­press a wish re­gard­ing what I’d like to see in fu­ture posts of this type.

I ap­par­ently have com­pletely missed with this post. I have watched its karma swing all over the place in just a few hours and all of the 36 com­ments so far are talk­ing about some­thing I con­sider com­pletely ir­rele­vant to the in­tended point. The same thing hap­pened with my last post, too, so some­thing is very off in my ex­pec­ta­tions re­gard­ing peo­ple’s re­sponses to the post. Some­thing I did caused you to ex­pect some­thing that I had ab­solutely no in­ten­tion of pro­vid­ing. It sucks for you and it sucks for me.

I don’t re­ally mind you sound­ing sanc­ti­mo­nious be­cause I don’t care about our rel­a­tive moral sta­tus. I find it frus­trat­ing that we had to go back and forth so long to end up where we did. I am not fully con­vinced my point ever did get across, but at least now I know how you per­ceived the post. Hope­fully by the next one I will have figured out what went wrong.

• I’m afraid I still don’t fully un­der­stand the point of your post. I hon­estly don’t find this illu­sion bizarre any more be­cause I do un­der­stand what our vi­sual sys­tem is re­port­ing. You ended the post with what sounds like a re­quest for an ex­pla­na­tion that ren­ders this illu­sion non-bizarre. I think be­tween the var­i­ous re­sponses that has been pro­vided. You seemed re­sis­tant to ac­cept­ing that ex­pla­na­tion ini­tially which is prob­a­bly why dis­cus­sion of it took over the com­ments.

It is true that most illu­sions have an ex­pla­na­tion that ren­ders them non-bizarre. This one does and it has been pro­vided. What other point(s) were you hop­ing to make?

• You ended the post with what sounds like a re­quest for an ex­pla­na­tion that ren­ders this illu­sion non-bizarre.

This was my mis­com­mu­ni­ca­tion. I was skip­ping over the ex­pla­na­tion in an at­tempt to cut on length.

You seemed re­sis­tant to ac­cept­ing that ex­pla­na­tion ini­tially which is prob­a­bly why dis­cus­sion of it took over the com­ments.

I was try­ing to avoid talk­ing about the ex­pla­na­tion be­cause in my mind the post was only us­ing that par­tic­u­lar ex­am­ple as an ex­am­ple. I was per­ceiv­ing the en­su­ing dis­cus­sion as nit­pick­ing. (And fully ac­knowl­edge that this was a com­mu­ni­ca­tion er­ror on my part.)

It is true that most illu­sions have an ex­pla­na­tion that ren­ders them non-bizarre. This one does and it has been pro­vided. What other point(s) were you hop­ing to make?

That is it, that is the point. This wasn’t meant to be an awe­some post of amaz­ing new con­cepts. It was just con­nect­ing the dots be­tween two sub­jects I hadn’t con­nected yet. This con­nec­tion was that illu­sions aren’t tricky. We think they are tricky be­cause we were ex­pect­ing some­thing differ­ent from re­al­ity.

• Hope­fully by the next one I will have figured out what went wrong.

My hy­poth­e­sis is that you picked a mis­lead­ing ex­am­ple to make your point. Similar color illu­sions are dis­cussed in, e.g., Den­nett’s “Con­scious­ness Ex­plained”, where he dis­cusses the kind of pro­cess­ing that the brain does to see color: they’re one of the best illus­tra­tions of how mis­lead­ing the idea of so-called “qualia” is.

It looked like you didn’t re­ally un­der­stand your main ex­am­ple. A and B are differ­ent col­ors, at least in terms of how the brain ac­tu­ally per­ceives color, and it sounded like you talked about them as if they were definitely the same. You wrote of the illu­sion as if it was mys­te­ri­ous and un­ex­plained, when it has been used as a canon­i­cal demon­stra­tion of how con­scious­ness works (Den­nett’s book has some­thing similar on the cover in one edi­tion). This con­fu­sion made it hard for me to un­der­stand your main point.

• Some­times it is use­ful to ask read­ers for a blow-by-blow re­sponse to your writ­ing. (See my pre­vi­ous post on com­mu­ni­cat­ing effec­tively, and fol­low the link to Richard Gabriel’s book.) Here is mine:

First para, it may help to know I didn’t fol­low the link at the end of your first para­graph, I kept on read­ing. So when you led off with “illu­sions are cool”, my in­ter­nal nar­ra­tive was “yes, illu­sions are cool”.

Your sec­ond para had me think­ing some­thing like, “He’s go­ing to go past sur­face ap­pear­ances, re­gard­ing the pic­ture he’s posted. OK, good. Let’s keep read­ing.” I prob­a­bly looked at the pic­ture only at that point.

Your third para is a side-trip into num­ber the­ory. It made me a lit­tle im­pa­tient, but I tried to rein that in, as I took in your point about “be care­ful about the defi­ni­tion of prime­ness or you will get sur­pris­ing re­sults”. I’m primed to re­ceive some in­sight re­gard­ing such ap­par­ent “sur­prises”.

Fi­nal para and I’m pleas­antly sur­prised at how short this post is, I may have up­voted it at this point with­out quite read­ing to the end. (Upvote later re­tracted, af­ter some de­liber­a­tion, for the rea­sons I gave.) I’m still not fol­low­ing links, by the way, just barely glanc­ing at the URLs. One I did look at was the “color con­stan­cies”. I am still ex­pe­rienc­ing a feel­ing of ap­proval as I read “dropped a bad be­lief and am look­ing for a re­place­ment”. Yep, yep.

I guess what then makes me un­easy is the string of “should, should, should” in the fi­nal three sen­tences, capped by “ob­vi­ous”. Wait, what ? I rewind to “look­ing for a re­place­ment”. For­ward back to “ob­vi­ous”. You’re mak­ing me feel like I missed some­thing, so I fire off a com­ment—in ret­ro­spect per­haps hastily—about the part of the post that right now bugs me.

In ret­ro­spect, what I was ex­pect­ing that you didn’t fulfill was “I am look­ing for a re­place­ment...”, set­ting up for ”...and here’s how I plan to do that”.

The point of your post (cor­rect me if wrong) is “the ‘cool’ of op­ti­cal illu­sions is an in­stance of the ‘weird’ that is referred to in Think Like Real­ity”. If that’s your en­tire point, it just maybe doesn’t quite de­serve four para­graphs.

You bring up a spe­cific ex­am­ple that has en­gaged your think­ing. I ap­prove of that. You make a “meta” point. I can ap­prove of that, but pro­vi­sional on your show­ing how your ob­ject-level think­ing has benefited from the meta.

• This is ex­tremely valuable to me and I greatly ap­pre­ci­ate it. My thoughts are be­low be­cause I pro­cess with words; they are not cri­tiques of your read­ing pro­cess.

First para, it may help to know I didn’t fol­low the link at the end of your first para­graph, I kept on read­ing. So when you led off with “illu­sions are cool”, my in­ter­nal nar­ra­tive was “yes, illu­sions are cool”.

Hmm… yeah, I didn’t pre­dict that. The whole post was as­sum­ing that you had, at one point, read that post. Note to fu­ture self: Quote the rele­vant part from the pre­req­ui­site link. I skipped it this time to try cut­ting down words but it hurt more than helped.

Your sec­ond para had me think­ing some­thing like, “He’s go­ing to go past sur­face ap­pear­ances, re­gard­ing the pic­ture he’s posted. OK, good. Let’s keep read­ing.” I prob­a­bly looked at the pic­ture only at that point.

Your third para is a side-trip into num­ber the­ory. It made me a lit­tle im­pa­tient, but I tried to rein that in, as I took in your point about “be care­ful about the defi­ni­tion of prime­ness or you will get sur­pris­ing re­sults”. I’m primed to re­ceive some in­sight re­gard­ing such ap­par­ent “sur­prises”.

Note to fu­ture self: Keep sec­ond ex­am­ples shorter. When ad­dress­ing a par­tic­u­lar sub­ject be­yond sur­face di­a­log check the lan­guage the ex­perts use.

Fi­nal para and I’m pleas­antly sur­prised at how short this post is, I may have up­voted it at this point with­out quite read­ing to the end. (Upvote later re­tracted, af­ter some de­liber­a­tion, for the rea­sons I gave.) I’m still not fol­low­ing links, by the way, just barely glanc­ing at the URLs. One I did look at was the “color con­stan­cies”. I am still ex­pe­rienc­ing a feel­ing of ap­proval as I read “dropped a bad be­lief and am look­ing for a re­place­ment”. Yep, yep.

I went for short this time be­cause I thought the point didn’t de­serve a lengthy post. I think a lot of the ini­tial up­votes I re­ceived were ev­i­dence that I was on the right track with this goal. I sus­pect that most of them were re­tracted later.

I guess what then makes me un­easy is the string of “should, should, should” in the fi­nal three sen­tences, capped by “ob­vi­ous”. Wait, what ? I rewind to “look­ing for a re­place­ment”. For­ward back to “ob­vi­ous”. You’re mak­ing me feel like I missed some­thing, so I fire off a com­ment—in ret­ro­spect per­haps hastily—about the part of the post that right now bugs me.

In ret­ro­spect, what I was ex­pect­ing that you didn’t fulfill was “I am look­ing for a re­place­ment...”, set­ting up for ”...and here’s how I plan to do that”.

That makes sense. Notes to fu­ture self: Finish the pri­mary ex­am­ple. Don’t ex­pect the read­ers to as­sume you did it offscreen.

The point of your post (cor­rect me if wrong) is “the ‘cool’ of op­ti­cal illu­sions is an in­stance of the ‘weird’ that is referred to in Think Like Real­ity”. If that’s your en­tire point, it just maybe doesn’t quite de­serve four para­graphs.

Where could I trim? I would guess the num­ber ex­am­ple was ex­tra­ne­ous. Your sum­ma­tion of the point works.

You bring up a spe­cific ex­am­ple that has en­gaged your think­ing. I ap­prove of that. You make a “meta” point. I can ap­prove of that, but pro­vi­sional on your show­ing how your ob­ject-level think­ing has benefited from the meta.

Okay. This is the big point I get to take away from your anal­y­sis. Again, thank you.

• In other words, this is the vi­sual ver­sion of “if a tree falls in the for­est...”, ex­cept that we already defined ‘color’ as qualia rather than wave­lengths, right?

• Since you men­tion it, that’s some­thing I should have brought up in one of the Mitchell_Porter con­scious­ness threads: the col­ors you see are not ac­tu­ally matched up one-to-one with the wave­lengths hit­ting your retina. Rather, the vi­sual sys­tem does some­thing like sub­tract­ing away the av­er­age color.

Mean­ing, the color that you ex­pe­rience see­ing de­pends on all the col­ors in the scene, not just the wave­length of the light com­ing off each spe­cific ob­ject.

Some peo­ple were talk­ing as if you were get­ting di­rect knowl­edge of (some­thing equiv­a­lently ex­press­ible as) wave­lengths, which is un­for­tu­nate, since part of the path to de­mys­tify­ing qualia is un­der­stand­ing this kind of pro­cess­ing.

• Um, “we already defined”—the refer­ent(s) of that phrase are very am­bigu­ous, I’m afraid. Who’s “we” and where was that defi­ni­tion ?

I definitely agree that color dis­crim­i­na­tions in the brain (the pro­cesses that even­tu­ally end up with color words com­ing out of our mouths) are about way more than wave­lengths. I’d pre­fer the term “dis­crim­i­na­tion” to “qualia”, the lat­ter car­ries philo­soph­i­cal bag­gage that I’d rather do with­out.

• Um, “we already defined”—the refer­ent(s) of that phrase are very am­bigu­ous, I’m afraid. Who’s “we” and where was that defi­ni­tion ?

It’s a royal ‘we’ in this case: Some sub­set of the group of com­menters here at LW, and that sub­set doesn’t in­clude me. It was dis­cussed at some length in the re­cent dis­cus­sion of con­scious­ness. I wasn’t pay­ing much di­rect at­ten­tion to the con­ver­sa­tion, though, so I can’t be more spe­cific than that. (I’m not even sure that the rele­vant bits are all in one post’s com­ments.)

• I think what you’re try­ing to say is that the phe­nomenon be­comes highly ex­pected, not aber­rant with re­spect to your world model.

• There’s a rather awe­some colour con­stancy op­ti­cal illu­sion in this Amer­i­can Scien­tist ar­ti­cle—click on the en­large image link on the ru­bik’s cube image. I’ve mir­rored the image here in case the link goes dead. The blue tiles in the left image are the same shade of grey in RGB terms as the yel­low tiles in the right image. H/​T to this ar­ti­cle.

• If I look at the pic­ture with­out fo­cus­ing (i.e. thereby see­ing it as 2D), A and B look the same color.

If I look at the pic­ture and fo­cus but cover the green cylin­der with my hand, they can look ei­ther the same or differ­ent (de­pend­ing on whether or not I no­tice the shadow in that case.)

I agree with Andy Woods: this is no illu­sion at all, ex­cept in the sense that it is an “illu­sion” that the board is three di­men­sional.

• I re­mem­ber not re­ally “get­ting” these illu­sions when I was a kid. I just didn’t find them in­ter­est­ing, it looked too straight­for­ward.

The idea of a “2D screen in­side our head” is not our nat­u­ral in­tu­ition. Be­fore learn­ing about these things, I just felt that I sim­ply per­cieve the en­vi­ron­ment around me. I don’t see a flat pixel grid in front of me when I walk around, I rather have a model of the en­vi­ron­ment that I con­tin­u­ously up­date and I per­cieve the ob­jects “from where they are”, just like I feel leg pain as if it were “in my leg”, de­spite the fact that pain ac­tu­ally hap­pens in the brain. I see ob­jects where they are in the 3D model, not where they are on a vir­tual screen.

The screen and pix­els anal­ogy may be so preva­lent in mod­ern times be­cause of the TV, pho­tos or even ear­lier re­al­is­tic paint­ings. But early art was not re­ally re­al­is­tic, which I think ei­ther shows they were

• not skil­led enough to draw re­al­is­tic art with per­spec­tive dis­tor­tions and shad­ing, or

• they didn’t think of vi­sion the way we do to­day, they were more fo­cus­ing on the ob­jects and their pro­to­typ­i­cal shapes, rather than their po­si­tion in the vi­sual field and the “ac­tual col­ors”.

The sec­ond ex­pla­na­tion seems more plau­si­ble to me.

Th­ese illu­sions are only illu­sions if you take the “2D screen and pix­els” view of vi­sion. Now that view is also im­por­tant for tech­nolog­i­cal ap­pli­ca­tions, and it’s also biolog­i­cally rele­vant (retina cells are sort-of pix­els), I’m just say­ing it’s not re­ally an illu­sion against builtin in­tu­ition.

• I don’t see a flat pixel grid when I walk around, ei­ther; I see a 3D scene (gen­er­ally only where I’m cur­rently look­ing; I mean, I can re­call where things are when I’m not look­ing at them, but they’re not in my cur­rent vi­sual model, that mem­ory has to be stored el­se­where).

And yet, a lot of op­ti­cal illu­sions work for me; be­cause (as in the case of the illu­sion in this ar­ti­cle) the draw­ing is close enough to what the re­al­ity looks like to fool my “scene re­con­struc­tion” mod­ule in my brain, and I re­con­stuct the rele­vant 3D scene when I look at it. Some op­ti­cal illu­sions (such as this one ) work by be­ing able to fool my scene re­con­struc­tion mod­ule in two differ­ent ways...

• Some­what re­lated: I think we do have a 3D map of the en­vi­ron­ment even for things that we aren’t look­ing at at the mo­ment. For ex­am­ple I feel as if I had a de­vice in my brain that keeps track of which peo­ple are in which parts of the house right now (or where some emo­tion­ally-loaded ob­jects are). I don’t have to ex­ert con­scious effort speci­fi­cally for this.

Another thing: it’s in­ter­est­ing to think about why we can see dots and lines and shapes at all. By this I mean, why do these low-level things reach our con­scious aware­ness? You aren’t con­sciously aware of your blood sugar level or hor­mone lev­els. You do feel a sort of ag­gre­gated well-be­ing feel­ing con­sciously but the de­tails don’t reach the con­scious level. It’s a strange and bizarre thing to think about what vi­sion could be like if our con­scious­ness didn’t have ac­cess to dots and shapes and col­ors style low-level image data and we only “felt” the gist of it, for ex­am­ple by only feel­ing our cur­rent 3D model in some way. (It could be similar to blind­sight.)

One an­swer could be that our vi­sion is so com­pli­cated that the un­con­scious parts just can’t cope with it fully, they can’t an­a­lyze it suffi­ciently, and con­scious pro­cesses (evolu­tion­ar­ily re­cent brain parts) need ac­cess to the ba­sic “pixel-data” like things as well.

But again, maybe when we in­ten­tion­ally try to look at spe­cific dots (as if look­ing at pix­els, in­ter­pret­ing the vi­sual field as a screen), we maybe aren’t re­ally look­ing at the low-level in­put but rather a re­con­struc­tion. Maybe we are see­ing lines, cor­ners and other ge­o­met­ric prim­i­tives laid on top of one an­other, like an SVG image, not like a BMP image. Maybe we don’t re­ally have con­scious ac­cess to the low-level vi­sual sig­nals, we just have ac­cess to a re­con­struc­tion.

I don’t think neu­ro­science has found out these things already, but it should be pos­si­ble to read off of con­nec­tions of brain ar­eas.

• Some­what re­lated: I think we do have a 3D map of the en­vi­ron­ment even for things that we aren’t look­ing at at the mo­ment.

I do not ap­pear to have that—or, at least, I don’t get much use out of it if it’s in here. While I can keep track of who is where in the house, I do so more in the form of a list of Last Known Lo­ca­tions, not in any sort of map (2D or 3D).

Pos­si­bly re­lated—I am no­to­ri­ous for get­ting lost eas­ily while driv­ing, and can get very badly turned around if I am merely a short dis­tance away from where I should be. I tend to nav­i­gate by mem­o­ris­ing a route from A to B, as a list of di­rec­tions (turn left at the third cor­ner, then it’s the fourth street on the right...) and then I get into trou­ble if I can’t fol­low that route. (Nowa­days, I tend to lean heav­ily on GPS when go­ing to new places).

• Since the top­ics are re­lated (but I’ll ad­mit I’m bi­ased to­ward see­ing that) - maybe a “Re­lated” link to the “Adap­tive bias” post would make sense.

• “Why do I keep think­ing A and B are differ­ent col­ors? Ob­vi­ously, that is not what my eyes are try­ing to tell me.” I am be­ing stupid when my eye looks at this illu­sion and I in­ter­pret the data in such a way to de­ter­mine dis­tinct col­ors. That in­for­ma­tion is not be­ing trans­mit­ted by my eye. If it were, the illu­sion wouldn’t be an illu­sion.

I think this is a bit mis­lead­ing. To the ex­tent that your eyes can be thought to be try­ing to tell you any­thing, “these are differ­ent col­ors” is ex­actly what they are try­ing to say. It’s fal­la­cious to as­sume that there is some sort of “true” in­put that one’s eyes are re­port­ing, and which the post-pro­cess­ing stages cor­rupt. (I’m not say­ing you com­mit­ted this mis­take, but peo­ple might get that im­pres­sion.) In­stead the mean­ing emerges purely from the post-pro­cess­ing stages. In fact, they’re do­ing a pretty darn good job, as one might note from the fact that op­ti­cal illu­sions nearly never no­tice­ably dis­tort our judg­ment in nat­u­ral con­di­tions.

• Yes, I com­pletely glossed over the finer points of eye-brain in­ter­ac­tion. I did not think it was needed to get the point across. I sup­pose I also used a bit of lin­guis­tic sleight-of-hand by im­ply­ing the eye “tells” me things. I sort of just called ev­ery­thing from the oc­cur­rence of light leav­ing an ob­ject to me per­ceiv­ing the light leav­ing the ob­ject as “The Eye.”

If you can think of a bet­ter way to say it with­out adding a heck of a lot of com­plex­ity I am more than will­ing to edit the post.

• I’m not sure if the eye-brain in­ter­ac­tion is the rele­vant part. Even if you change “eyes” to “the vi­sual sys­tem”, the point of “these are of a differ­ent color is the very thing your vi­sual sys­tem is try­ing to tell you” re­mains true.

• What would you call it? The point re­mains that the col­ors are not the same. If we think the col­ors are the same, we are in­cor­rect.

• They are not the same, but our vi­sual sys­tem is try­ing to tell us they are the same—and you can’t re­ally say it’s wrong to make that judg­ment, as do­ing so leads to cor­rect re­sults the over­whelm­ingly vast ma­jor­ity of the time. (Ba­si­cally, I’m say­ing the same thing as AndyWood’s com­ment and the re­sponses to it.)

• They are not the same, but our vi­sual sys­tem is try­ing to tell us they are the same.

Yeah, this makes sense. The trick is ask­ing, “The same what?” The an­swer, “Color” is not de­scrip­tive enough.

I never meant to say it is right or wrong for the vi­sual sys­tem to do what­ever it is do­ing. I mean to say it is wrong to ex­pect some­thing differ­ent from the vi­sual sys­tem than what it is do­ing.

• “my op­tic sys­tem”? “my vi­sual sys­tem”?