Western Orientalism on MetaFilter

February 15, 2019 6:45 AM

From discussion on the Big Data AI language model, some folks innocently mentioned the Chinese Room, a thought experiment about a language we don't know.

The Chinese room is from 1980, from a white dude who was looking for a language he, and his students, wouldn't understand. So of course he chose a language spoken by the numerical majority of people on Planet Earth.

Orientalism is also responsible for other near epithet idioms and nouns like "Chinese Whispers" (which started out as "Russian Whispers"), "Chinese Checkers" (which isn't Chinese - Chinese stand here for "odd", "weird", "foreign", "different", from, you know, normal checkers).

I'm wondering if it would be worth it for folks here on left-leaning, progressive-leaning, but still quite white MetaFilter to consider dismantling this linguistic micro aggression rather than promote it.

posted by kalessin to Etiquette/Policy at 6:45 AM (148 comments total)

Especially since we do have established, non-loaded versions of these terms: Turing Test and Game of Telephone.
posted by Weftage at 7:05 AM on February 15, 2019 [7 favorites]

I think educating folks on some of these etymologies is one of the best ways to combat this - it wouldn't have occurred to me to wonder about the history of Chinese Checkers, as I haven't played it or thought about it since I was about eight, but hey, turns out it's a German board game named Sternhalma, rebranded by an American toy company for no very good reason!
posted by restless_nomad (staff) at 7:07 AM on February 15, 2019 [26 favorites]

I think it's an excellent suggestion, but an alternate name is needed because the underlying idea is too useful to abandon entirely. Perhaps some fictional and/or constructed language would be better for avoiding unfortunate cultural associations. "Heptapod room" (for the aliens in Arrival/Story of Your Life)? "Esperanto room?"
posted by DevilsAdvocate at 7:12 AM on February 15, 2019 [6 favorites]

I had no idea Sternhalma was a thing!

The main thing to remember when we talk about education is that we can't ask people to do the emotional labour of educating us when they are the ones being hurt by the language we are using. Nonmods can't see flags, how could someone "flag" this type of thing so that commenters know they've said the wrong thing without causing further injury?
posted by wellred at 7:15 AM on February 15, 2019 [1 favorite]

This is over the top hypersensitivity. The name of the thought experiment is the Chinese Room. There's nothing "orientalist" about it, Searle simply chose a language that he didn't speak.
posted by Spacelegoman at 7:18 AM on February 15, 2019 [28 favorites]

"Chinese stands here for "odd", "weird""

I'm grateful for this post because without it I never would have thought to question the name of that idiotic teenager stopped-at-a-red-light activity Chinese Fire Drill.

I mean not that I've thought about it in decades, maybe it's not even a thing kids do any more. Regardless, it had never occurred to me to examine the name and now it makes a lot more sense.
posted by komara at 7:18 AM on February 15, 2019 [6 favorites]

I completely agree that we need better, and more sensitive, terminology.

I disagree with Weftage's point: the Turing Test ("can a machine approximate a human well enough that a human can't tell?", that is, an engineering question) isn't quite a drop-in replacement for The Chinese Room ("can a bunch of symbolic manipulations be considered 'thinking'?", that is, a philosophical question). And the Game of Telephone (a children's game concerning signal degradation) is unrelated to both.

I also disagree with Spacelegoman's point: I think it's unreasonable to tell people how to feel. If words hurt, they hurt, and we should try to help.

A further question: it's all well and good to avoid making well-intentioned-but-harmful jokes (which was the case in the thread referenced here), but what if we were to actually be talking about the history of AI and philosophy, say, and wanted to link to prior work in the field (which universally uses the harmful terminology)? Is it better to avoid such references entirely even at the cost of discussion, or is there a sensitive way to include such references?
posted by ragtag at 7:21 AM on February 15, 2019 [15 favorites]

This is over the top hypersensitivity. The name of the thought experiment is the Chinese Room. There's nothing "orientalist" about it, Searle simply chose a language that he didn't speak.

I think it's reasonable to look at it as sensitivity that just wasn't brought to bear originally because the stakes weren't there for the originator. This isn't some lone example of mild exoticism, and it doesn't require burning computer science or information theory to the ground or assuming explicit ill intent on Searle's part to stop and reconsider it and say "huh, yeah, that kinda has a crappy patina to it, maybe we should find some alternate jargon going forward".
posted by cortex (staff) at 7:24 AM on February 15, 2019 [40 favorites]

The name of the thought experiment is the Chinese Room. There's nothing "orientalist" about it, Searle simply chose a language that he didn't speak.

"Searle wasn't intentionally being orientalist when he chose Chinese" may well be true, but that doesn't mean we should perpetuate unintentional orientalism, now that we recognize it as such.
posted by DevilsAdvocate at 7:26 AM on February 15, 2019 [36 favorites]

"Nonmods can't see flags, how could someone "flag" this type of thing so that commenters know they've said the wrong thing without causing further injury?"

As a practical matter, you can certainly flag-with-note and say "hey, this is orientalist, would you mind asking the thread not to use this term?" and we can do that for you if you don't want to deal with it yourself.

Or -- if you feel you want to -- it's perfectly fine to say, "Hey, 'Indian style' isn't great -- today people prefer 'cross-legged' or 'tailor's seat' or 'criss-cross applesauce.'" Doing that in a thread usually goes best if the comment is informational and contains just information about the problematic term and better alternatives. Make a second comment to argue about the main topic of the thread -- otherwise people conflate the arguing with the informational correction and get emotional about the informational correction. A brief informational comment like that doesn't tend to derail the thread and people tend to respond like, "huh, I hadn't realized that, but whoa, yeah, I gotta stop that." (Plus if people start arguing "but I've said Indian style since kindergarten!" it's easy for us to delete those derails because they'll just be responding to the correction, and not mixing that in with a whole substantive comment about the main topic.)
posted by Eyebrows McGee (staff) at 7:39 AM on February 15, 2019 [16 favorites]

I agree with the principle of why these shouldn't be the terms for things. But--Metafilter does not, itself, have the social clout to create change on these fronts, does it? It isn't going to be formerly known as that anywhere else. I would agree with the idea that the history of these names should be acknowledged as a part of the conversation, but coming up with new names for things that aren't going to be used anywhere else seems like it's just going to create a Metafilter-specific jargon for things we don't often actually discuss. The actual people doing the work are the ones who need to figure out the new terms, if there isn't already a convention for a replacement.

If there is a conventional new term for something, switching to that is obvious. But I'm not sure that's the case for the Chinese room, for example. If we invent a new name for it but you can't Google that new name and none of the people doing serious research use that new name, that just seems bound to create a very unhelpful level of confusion.
posted by Sequence at 7:52 AM on February 15, 2019 [6 favorites]

I agree with the principle of why these shouldn't be the terms for things. But--Metafilter does not, itself, have the social clout to create change on these fronts, does it?

We're not going to turn the whole tradition of reference on its head, no. But we do have the capacity to do better with it here, in our space, and to in doing so model and normalize a better approach that could find its way into broader usage over time.

Even if all we accomplish is being a little more thoughtful among ourselves, that's a good outcome. We don't have to have a fix for the whole world to try and do a little better locally; talking about stuff like this where we could within this community do a little better, be a little more aware, is a good thing.
posted by cortex (staff) at 7:57 AM on February 15, 2019 [21 favorites]

Good heads up, I hadn't noticed it before, but fortunately I don't experience much "Chinese" as "weird" language around here. Maybe an actual "communicating with extra-terrestrials" room? Coming up with cute new terms for ideas that make your writing incomprehensible to new readers seems to me to be about half of what philosophers do, but I'm pretty thinly read.

> the underlying idea is too useful to abandon entirely

[Daniel Dennett bursts into the room]
posted by lucidium at 8:02 AM on February 15, 2019 [4 favorites]

Sequence, social change is usually accretional. Even tiny discussions on Metafilter can add to an ultimate effect. I would contend that it's worth doing, even if ourselves may not see the results directly.

The actual people doing the work are the ones who need to figure out the new terms

...and some of those actual people are here on Metafilter, and may inform & be informed by the discussion.

(And I see, on preview, that cortex has preceded me. Let me then underscore his comment.)
posted by Weftage at 8:04 AM on February 15, 2019 [2 favorites]

This does not come from a vacuum. In the US we grew up playing a game called "telephone", in which a secret was quickly whispered once into the ears of a chain of people, and what came out suffered entropy that often resulted in amusing mutations of the original phrase. In the UK this is apparently called "Chinese Whispers", which always makes me balk. But then I hear folks from the Japanese side of my extended family self-identify as "Oriental" here and just have to bite my lip.

There apparently was a recent-ish episode of Red Dwarf where the question "Is the name 'Chinese Whispers' racist?" itself mutates and creates a comedy of errors. So at least this isn't something nobody in the UK finds uncomfortable.

Now my dad was a classicist and Greek Orthodox, and he hated the phrase "it's Greek to me" for similar reasons. Whenever anybody said that he'd just start speaking to them in Greek (but being a classicist, it usually had the style of the New Testament, which was a bit like breaking out into Shakespearian English). The origin of that phrase comes from Shakespeare's Julius Caesar, Act 1 Scene 2:

CASSIUS

Did Cicero say anything?

CASCA

Ay, he spoke Greek.

CASSIUS

To what effect?

CASCA

Nay, an I tell you that, I’ll ne'er look you i' th' face again. But those that understood him smiled at one another and shook their heads. But, for mine own part, it was Greek to me.

In this scene, Roman characters speak Greek to conjure airs and erudition, similar to the way Latin is used in the UK. But the phrase has come to be a stand-in for "a foreign language that sounds like babble to me."
posted by rum-soaked space hobo at 8:07 AM on February 15, 2019 [12 favorites]

Always odd in this global day and age when "foreign" languages are presented as being unfathomable, and on MetaFilter of all places.
posted by JamesBay at 8:12 AM on February 15, 2019 [2 favorites]

By the way, the name for the drawing-driven version of "telephone", where you whisper something to someone and they draw it and pass the drawing to the next person, is called ECYP for "Eat Cat You Poop". I think a genuinely absurdist name fits the game more appropriately than trying to find a living language to compare it to.
posted by rum-soaked space hobo at 8:13 AM on February 15, 2019 [1 favorite]

I think the request is entirely reasonable due to the history that accompanies the reference, but it is also an example of a use having historical significance that carries a descriptive value in adhering to context of its origin.

In that sense, to me, there is some similarity to other terms still used despite being out of date for today's language, the Negro League in baseball, for example. I'm sort of torn between seeing the specific reference carrying some need to keep its historical context as much because it shows the mindset of the times while agreeing the term shouldn't be used any more broadly than in referencing events or ideas of past importance.

I'm not going to argue for its use, but I'm wary of altering quotes as well since that feels like it distorts history. I guess I'd say the difference would be in re-using the idea of the test as a new hypothetical where the use of "Chinese Room" would be compounding a bad choice versus speaking of Searle's argument specifically where its use is bound to Searle and might be better maintained.
posted by gusottertrout at 8:31 AM on February 15, 2019 [1 favorite]

For what it is worth, I am willing to bet, the game reference is actually an even less erudite reference to the Japanese game of Go, with round, omnidirectional moving pieces. I kind of think it meant more clever than checkers, because it definitely is.

I worry this will become ageistic since older Mefites grew up in the hate stew, and there are a million casual references linking back to and through, many cultures various fears and differentiating expressions. I am making no attempt ot excuse any hurts. We are almost all possessed of various ethnicities, I try neither to war inside myself, or with others outside myself based on the offense my parts may take.
posted by Oyéah at 8:35 AM on February 15, 2019 [1 favorite]

I think it's not necessarily a problem that these things become another bit of erosion in the generation gap. I constantly think about the stories told in the comments of this FPP about a revolting slogan intended to celebrate privilege. A lot of the people who encountered it say an older person used it gleefully and then realised how horrible it was based on the scowls they got back.

These things will also be clung to by racists as "symbols of our stronger past that overly sensitive people tried to censor". In the UK there was a type of stereotypical doll that became an epithet for black people. Before Brexit overwhelmed us here, the BNP proudly sold these dolls to show they were standing up for their racism but then not standing behind it themselves, saying "they're just dolls. I had one as a kid and loved it! It's not about hate..." etc.
posted by rum-soaked space hobo at 8:45 AM on February 15, 2019 [6 favorites]

Especially since we do have established, non-loaded versions of these terms: Turing Test and Game of Telephone

Whilst "Telephone" is clearly the same game as"Chinese Whispers", the Chinese Room & the Turing Test are two different philosophical though experiments as ragtag points out.

I can’t think of an alternate name for it that would make the referent obvious to the reader unfortunately. Suggestions?
posted by pharm at 9:01 AM on February 15, 2019 [5 favorites]

I'm wary of altering quotes as well since that feels like it distorts history.

I don't think anyone here is suggesting altering direct quotations. But even if they were, there are established ways of doing so without distorting history.
posted by DevilsAdvocate at 9:03 AM on February 15, 2019

Is there anybody actually hurt or offended by the name of the experiment “Chinese Room” or is this all hypothetical sensitivity?

“Chinese” in the experiment does not mean unfathomable. It specifically means communication which has meaning but is not understood, and an important part of choosing Chinese is that it has a different written form with enough complexity that you can avoid arguments that the man in the room could just learn parts of it by reading the messages.
posted by the agents of KAOS at 9:03 AM on February 15, 2019 [24 favorites]

I agree that this is a very specific concept which we'd need to come up with an entirely new term for because there isn't a substitute in common usage (or even limited usage, for that matter) which I'm aware of.

But I've got to say I would really welcome a new term because I have cringed every time I've ever heard its name or had to use it myself.
posted by XMLicious at 9:12 AM on February 15, 2019

"Orientalist" is subtly but significantly different from the usual colloquial meaning of "racist". The "Chinese Room" term may not be the latter, but it's certainly the former.
posted by tobascodagama at 9:14 AM on February 15, 2019 [6 favorites]

As I said in the initial thread, Searle's thought experiment benefits from Western Orientalist perceptions in that it invokes the spectre of Chinese-language translation as happening without intention or conscious understanding. It is a (perhaps mild) example of a very real and damaging thread of Western prejudice regarding Asian people, languages, culture, and thought.
posted by gauche at 9:15 AM on February 15, 2019 [3 favorites]

I don’t remember the idea that this translating creation was in some way /bad/ being part of the thought experiment gauche. Do you have examples of it being used in that way?
posted by pharm at 9:20 AM on February 15, 2019 [1 favorite]

I questioned specifically the term “Chinese Room”, thank you.
posted by the agents of KAOS at 9:24 AM on February 15, 2019 [6 favorites]

And gauche: that reads to me like a fundamental misreading of the experiment, which is saying that this mechanistic translation is not equivalent to the human thought process of understanding Chinese.
posted by the agents of KAOS at 9:25 AM on February 15, 2019 [3 favorites]

I don’t remember the idea that this translating creation was in some way /bad/ being part of the thought experiment

It doesn't require the concept of translation, or of the chosen language, to be itself bad for the ready-to-hand choice of Chinese as an example of something inscrutable to feel kinda yech. That's the whole thing with stuff like microaggressions; they can exist as a nasty friction point in a system of injustice without being some bright-line example of intended offense.

There's a long history in the United States certainly and more broadly in white western culture of treating various Asian cultures as exotic, and often interchangeably so (consider how often "Chinese" has historically been a stand in in US pop culture for "I dunno, some sort of Asian?"), and that is the context in which an otherwise defensibly "just picked a language at random" can take on a weird, not-great feeling. It's difficult even with provably the best of intentions to not have that sort of loaded choice land uncomfortably. Responding to that cultural context and that discomfort in an empathetic way is worth doing.

So pinning whether "Chinese room", as the presentation for that thought experiment, on a question of whether the thought experiment itself, or Searle himself, or the idea of translation itself is intentionally being orientalist is missing the point. A totally neutral thought experiment proposed by someone with no ill-intent using an out-group culture as a placeholder example can still come off as weird and of-a-piece with a less-neutral historical context, because the world is messy and we don't get to indemnify ourselves or the things we've learned or habitualized in the past from that messiness.
posted by cortex (staff) at 9:33 AM on February 15, 2019 [17 favorites]

The experiment is about translation without intention, as a way of thinking about what it means to know that someone or something else is conscious. How I recall it is as a criticism of the Turing Test as a means of identifying consciousness. The Turing Test, you will recall, postulates that if you can have a conversation over a terminal with an intelligence, and you cannot tell whether that intelligence is human or artificial, than it is conscious.

Searle comes in and says, but wait a minute, suppose the conversation were being conducted in Chinese, a language that you don't speak, and you were alone in a room and had with you a really extensive and foolproof lookup table such that when you received input in the form of Chinese words and sentences, you could refer to the lookup table and produce output such that your interlocutor would think that you were fluent in Chinese. We would not say that you "knew" Chinese in such a case. Why is that not what is happening in the Turing Test?

Why I understand this to be problematic is that it plays into and evokes already-present Orientalist ideas of the Chinese language being inscrutable and mysterious and "other." Kalessin's point about it being "a thought experiment about a language we don't know" is, for instance, decentering and othering of the experience of millions of people who do know Chinese. Who is the "we" in this instance? Not speakers of Chinese, that's for sure. There is no reason for the language to be Chinese specifically and indeed I can think of several languages -- any dead language, for starters -- which might be better used in its stead.

None of these things had to be Searle's conscious intention when he went reaching for "a language we don't know" and landed on Chinese, but that does not mean that these meanings are not present anyway in the use of the term.
posted by gauche at 9:35 AM on February 15, 2019 [11 favorites]

To me, the phrase

the spectre of Chinese-language translation as happening without intention or conscious understanding

seemed like a value judgement on the actual thought experiment itself, separate from the idea that using "Chinese" as a placeholder for "incomprehensible" is something that one should avoid (which seems completely reasonable to me, but a little more difficult in this one particular case due to the specificity of the referent).
posted by pharm at 9:39 AM on February 15, 2019 [1 favorite]

(that was a reply to Cortex’s criticism of my previous comment, not a response to gauche.)
posted by pharm at 9:42 AM on February 15, 2019

how searle introduces his gedankenexperiment:

Suppose that I'm locked in a room and given a large batch of Chinese writing. Suppose furthermore (as is indeed the case) that I know no Chinese, either written or spoken, and that I'm not even confident that I could recognize Chinese writing as Chinese writing distinct from, say, Japanese writing or meaningless squiggles. To me, Chinese writing is just so many meaningless squiggles.

the paper: Minds, Brains and Programs.
posted by 20 year lurk at 9:45 AM on February 15, 2019 [6 favorites]

To me, "meaningless squiggles" is still pretty othering. The language associated with my heritage uses a very different alphabet from the one English uses, it makes me feel very weird to think of someone referring to it as meaningless.
posted by wellred at 9:50 AM on February 15, 2019 [17 favorites]

I have the same reaction to these threads just about every time one is made:

Step 1: What the fuck? Are they serious? This is a thing they're offended by?
Step 2: No, really, there's a perfectly legitimate reason why that thing is the way it is. It isn't intentional racism. Let me explain...
Step 3: I mean, sure, they're being pretty reasonable about how it isn't, like, super evil racism just a low level microagression it would be nice to avoid, but come on, can anyone really expect us to eradicate all of this from our vocabulary? What are the alternatives, honestly?
Step 4: No, this is an issue that is worthy of respect and I need to just deal. Let's figure out better vocabulary.

I aspire to get to the point where I can just arrive at Step 4 without mentally processing Steps 1 through 3. I sometimes manage to completely skip step 1 and I am getting pretty good at keeping steps 1 and 2 to myself, but I stumble on step 3 and say that shit out loud way more than is okay.
posted by jacquilynne at 9:56 AM on February 15, 2019 [32 favorites]

Ha, jacquilynne, same! I was all ready to burst in here with my Searle-splainin and now have arrived at "ok this seems reasonable" instead.
posted by Pyrogenesis at 10:00 AM on February 15, 2019 [1 favorite]

I can’t think of an alternate name for it that would make the referent obvious to the reader unfortunately. Suggestions?

I suspect that these days many references to the thought experiment could be replaced with phrases like "a Google Translate-like system" or "a really good neural network model" or, in the context of the FPP, "a GPT-2-like system." Looking at how it came up in the original thread, "What is the world but an interconnected series of [ML models]?" maybe flows well enough and seems to make a similar point. You could also just call it something like "Searle's picture of AI" if you needed to be more specific about concomitants/circumstances of the argument.
posted by Wobbuffet at 10:11 AM on February 15, 2019

I can’t think of an alternate name for it that would make the referent obvious to the reader unfortunately.

"Searle's Room"
posted by Etrigan at 10:12 AM on February 15, 2019 [52 favorites]

This feels Searle-splainy, but isn’t that the entire point - that he wouldn't personally know a heart from a house? He obviously knows that they carry meaning for those who recognise them, but it’s not a meaning that he personally is privy to.

If it had been called “Searle’s Room” from the start, then would that have been better? Yes, it centres the non-Chinese speaking philosopher, but it is his thought experiment.

(On preview...hah! jinx.)
posted by pharm at 10:14 AM on February 15, 2019 [1 favorite]

And, for what it is worth, even setting aside what Searle said or meant by the thought experiment which gave rise to the term, the term has come to have a meaning which plays exactly into the problems that kalessin has raised.

It means a thing which seems like and behaves like a mind but isn't one.

Like jacquilynne upthread, I didn't wake up this morning thinking this term was problematic, but it definitely is. I think this community would do well to avoid it.
posted by gauche at 10:25 AM on February 15, 2019 [2 favorites]

So, are we still allowed to say Chinese checkers ?
posted by Pendragon at 10:32 AM on February 15, 2019

I remember playing with a Chinese checkers set at my grandparents place when I was a kid that was super racist in design and packaging. We're talking stereotypical "Chinese" calligraphy font English and cartoon characters with exaggerated teeth and slanted eyes and everything. It was probably made in the 50s.

I remember that the whole thing felt off and wrong to me, and beyond and the packaging the game itself didn't seem to be particularly Chinese and I've always wondered about this weirdness.

And until today I had no idea about the actual origins of that game. Or why it was even called Chinese checkers.
posted by loquacious at 10:35 AM on February 15, 2019 [3 favorites]

A quick Wikipedia look shows you that "Chinese checkers" is German, and is called "Sternhalma." Let's learn that, eh? It also says it is a modern and simplified version of an American game called "Halma," so for me, I will never use the phrase again, cause it's so very off from the truth.
posted by agregoli at 10:41 AM on February 15, 2019 [5 favorites]

I love the Chinese Room thought experiment and I agree it'd be useful to have a more neutral term, too. I went looking. Couldn't find anything on Google. Asked on Twitter and got no replies other than one half-serious suggestion of "Hypercard Room". "Giant pile of coupled logistic functions running on a GPU Room" would be relevant to current tech, but perhaps a bit of a mouthful.

I agree with agents of KAOS's point that part of Searle's construction is rooted in the ideographic nature of Chinese writing. There's only one written language in the world with that property. I don't think it's essential to Searle's argument, but it's a part of it.

What do they call the Chinese room in China?

(Related: Greek to Me in many languages. I remember reading somewhere if you walk the graph of "incomprehensible language" in all languages, you tend to end up at Chinese as the most inscrutible language. Chinese languages in turn reference the supernatural, "ghost script" or "heavenly language".)

Hell, Searle's alive, anyone know him well enough to ask him what he thinks?
posted by Nelson at 10:41 AM on February 15, 2019 [2 favorites]

The "Chinese Room thought experiment" is a great example of cultural bias and white supremacy in human-made artificial intelligence.
posted by JamesBay at 10:49 AM on February 15, 2019 [10 favorites]

What do they call the Chinese room in China?

中文房间
Zhōngwén fángjiān
posted by the agents of KAOS at 10:54 AM on February 15, 2019 [7 favorites]

There's only one written language in the world with that property.
Ancient Egyptian?
posted by soelo at 10:57 AM on February 15, 2019 [4 favorites]

There's only one written language in the world with that property. I don't think it's essential to Searle's argument, but it's a part of it.

It's also, as noted above, the most commonly used primary language in the world, and second place isn't close. If you're picking something that's uniquely inscrutable, then it seems like something that a billion people use every single day ain't it.
posted by Etrigan at 10:58 AM on February 15, 2019 [13 favorites]

Agents of KAOS: it would help to give an actual literal translation of that Mandarin, as Google Translate inevitably translates it as "The Chinese Room"!

(I think it means something like "The Language Room" or "The Writing Room" but, like Searle, my knowledge of written Chinese is non-existent so that’s purely an inference made from looking up the individual characters online.)
posted by pharm at 11:01 AM on February 15, 2019

My point about Chinese being a unique written language isn't that it is inscrutable. Clearly it's not. My point is that it is logographic, in that there are 60,000+ different symbols. That's a piece of the argument about how the magic homunculus inside the Room does its job. It's working with word-sized symbols. The argument still works with a phonetic alphabet like English, but maybe not as clearly. There's also a bit of frisson about how written Chinese is a writing system for multiple different spoken languages. In some sense, the way Chinese is used today as a written language already is a consciousness-less translation system.

Wikipedia does list a few other logographic writing systems; Ancient Egyptian, as soelo mentions. So maybe the "Hieroglyphic Room"? There's about 1000 symbols IIRC, and "hieroglyphics" has a meaning of "inscrutible" that's perhaps not so problematic since it is an extinct writing system. ("Mayan room" would be more offensive than "Chinese Room" IMHO.)

Seconding a request of the literal translation of 中文房间. Google Translate tells me 中文 is "the Chinese Language," so perhaps it is literally "Chinese" + "Room"?
posted by Nelson at 11:07 AM on February 15, 2019 [2 favorites]

(which started out as "Russian Whispers")

While confirming to myself that it's also Russian Scandal or Russian Gossip, I found an 1899 article, Polyglot Russian Scandal in Thackeray's Cornhill Magazine which folks might find interesting.

In it, George Somes Layard conducts an experiment in translation. Prefiguring the modern trend of running text back and forward through layers of machine translation to see what pops out at the end, he assembles a sequence of scholars to translate a short verse back and forwards between English and a different language. It goes from English to Latin to English to French to English to Greek to Italian to English to German to English to Farsi, and that's where things get a little weird. Layard basically throws his hands up and says Well, guess it's in Persian now! Oh well.

…my acquaintanceship with Oriental scholars being limited, I find that, having got into the language of Omar, it is quite another thing to get it out again in the form required.

He leaves it as an exercise for the reader to find a Persian friend of their own. I'll leave it as an exercise to the reader to draw their own conclusions about George Somes Layard's Victorian scholarship.
posted by zamboni at 11:08 AM on February 15, 2019 [1 favorite]

That's a piece of the argument about how the magic homunculus inside the Room does its job. It's working with word-sized symbols.

Sure, but there's nothing actually essential to Searle's proposition that requires the language in question to be logographic, or purely so, and there's a whole passel of writing systems that are both not purely logographical and also not very morphologically similar to latin alphabetical characters. Japanese outside of kanji uses a set of syllabaries; Korean uses a kind of accretive logography built of out smaller characters; Arabic uses a complex set of ligatures and contextual forms that render words as connected units that differ visually from the sum of their alphabetic units.

Which, I don't think there's anything particularly meaningful on its face about Searle's choice of Chinese as an unfamiliar language, cultural yech of canonizing that Chinese-as-inscrutable trope in an information science thought experiment aside. But I also just don't think there's anything mechanically necessary or interesting to say in defense of that choice. Searle's experiment works just as well with any abstract notion of an unfamiliar language, and would have been as clear with a number of other choices as with Chinese; I think what is fairly easy to criticize about the cultural context from which that particular choice was made is that Chinese is for a western audience an especially familiar stand-in for the unfamiliar.

It's that cultural context in which it was coined and then carried forward without blinking for decades that's mostly at issue, but Searle's choice of example for it exists within that context and I don't think there's any satisfying argument that his choice of example was so essential to the thought experiment that it's worth defending specifically on that basis.
posted by cortex (staff) at 11:19 AM on February 15, 2019 [4 favorites]

Wikipedia does list a few other logographic writing systems; Ancient Egyptian, as soelo mentions. So maybe the "Hieroglyphic Room"? There's about 1000 symbols IIRC, and "hieroglyphics" has a meaning of "inscrutible" that's perhaps not so problematic since it is an extinct writing system. ("Mayan room" would be more offensive than "Chinese Room" IMHO.)

As I understand it, part of Searle's Room is that there also needs to be people outside the room who are fluent in the language in question.
posted by zamboni at 11:21 AM on February 15, 2019 [5 favorites]

It wouldn't surprise me to learn that there's a difference between the experience and sensitivities of Mainland-born Chinese immigrants in the US and Chinese-Americans, such that what hurts or offends a Chinese immigrant might not offend someone who is 4th generation Chinese-American, and vice versa. It seems particularly cruel to take a reasonable request (incidentally, from someone who has identified themselves as "personally, emotionally, psychologically, which sums up to materially hurt by use of Chinese in an Orientalist fashion") and dismiss that as "white people falling over themselves" and kind of funny. If you don't think the request to think more carefully about use of Chinese as a mystical inscrutable other is a worthwhile one, I guess keep at it. But like, you don't have to willfully be an asshole about it.
posted by ChuraChura at 11:30 AM on February 15, 2019 [45 favorites]

We saw similar conflicts over whether wearing Kimono was cultural-appropriation or not IIRC.
posted by pharm at 11:38 AM on February 15, 2019 [4 favorites]

Yeah, it seems like there's a difference between saying "I am Chinese; this is what I think" and "My wife is Chinese and I will use her words to speak with authority about an experience I do not have."
posted by ChuraChura at 11:38 AM on February 15, 2019 [10 favorites]

Mod note: One comment removed. I'm real fuckin' tired of "what this thread needs is an edgy joke!" as a response to discussions and this isn't the first one of those from you in recent memory, Cool Papa Bell, so I'm entirely done.
posted by cortex (staff) at 11:53 AM on February 15, 2019 [15 favorites]

It's racist to me cause I'm Asian American so must be one of those overprivileged Westernized ABC's who know nothing about the issue...

In reality I've directly experienced marginalization as a Chinese American, via benign linguistic microaggression, once by my thesis advisor and another time while having lunch with a well-known theoretical computer scientist.

In those instances, I was held to this weird standard of looking Chinese thus knowing Chinese and thus able speak on linguistic matters, until they decided I didn't know enough.

I'm not saying the questions were scientifically fascinating, they were. But the approach and the way my identity mattered in a way that didn't for the askers only worked to impede my ability to participate in an socially constructed learning experience.

It's funny cause the Searle question is the same. "Chinese" is irrelevant; Searle's own exposition is triggering and distracting and keeps reminding me of his Whiteness, and a room in which a person has a menial role in a room with such a label is simply bad optics however we'd like to deny any logical connection.
posted by polymodus at 11:57 AM on February 15, 2019 [12 favorites]

I like the term "Searle's Room" as a replacement name for it. If you knew of the "Chinese Room" thought experiment then you probably already know that Searle formulated it and may even already refer to it as "Searle's Chinese Room" so we just remove the word "Chinese" and the thought experiment is still easily identifiable. You could probably slip "Searle's Room" into conversation without anyone batting an eye.
posted by any portmanteau in a storm at 11:58 AM on February 15, 2019 [15 favorites]

Came in to suggest "Searle's Room" as a perfectly legible substitute; happy to see others got there first. This is an incredibly easy request.
posted by a snickering nuthatch at 12:42 PM on February 15, 2019 [4 favorites]

Here to boost "Searle's Room"
posted by salt grass at 1:23 PM on February 15, 2019 [1 favorite]

literal translation of 中文房间

It is in fact "Chinese room," if my undergrad Mandarin is serving me well--those first two characters are center/China and then language, if translated on their own.

That being said, why does it matter what Chinese people in China think about something that English-speaking people are doing, when Chinese-diaspora English-speaking people who deal primarily with Anglos and other white people speaking English talk about what stings to them?

This is often something I hear when people of a particular diaspora express distaste for the way their home culture is treated in their adopted nation--the conflation of the experiences and feelings of people in the diaspora with people who live in the home culture, and it's silly. I'm Irish-American: that means I have a set of specific cultural experiences, including veneration of St. Patrick's with alcohol rather than prayer and the association of corned beef and cabbage with my people, which folks who are Irish and stayed in Ireland don't have. Asking them for their opinion of whether it's okay for me to do X (or for X to be done to me) in my current nation--me being the fourth generation born here!--is just plain silly.

The same thing applies to other diasporic experiences, including the Chinese diaspora. No one thinks that Ghanans get to dictate the African-American experience, do they? The same thing applies to people who left their old homes by other means, carrying their experiences and their roots alongside them. The experiences are different in the new place, and the sore spots and chafed pains are different there, too. It pays to listen specifically to the groups of people who are talking, not to similarly-identified people talking about something else half a world away.
posted by sciatrix at 1:42 PM on February 15, 2019 [17 favorites]

So, are we still allowed to say Chinese checkers ?

Searle's checkers
posted by paper chromatographologist at 1:51 PM on February 15, 2019 [6 favorites]

Was Searle being Orientalist? Absolutely. I pointed this out years ago in my page on Searle, and purposely reversed the story, making it about a Chinese person executing rules to process English. This is logically equivalent, and it's no accident that Searle prefers the form where he is contemplating "mysterious squiggles". It helps him beg the question: he wants to prove that computers couldn't possibly "understand language", so he imagines a situation where a human (himself) can't possibly understand the input.

On the other hand, the Room isn't folklore, like Telephone; it's a specific thought experiment developed by Searle, responding to a specific moment in AI. It seems weird to want to polish up a man's work to remove his biases. If you're going to correct his errors, why not go all he way and remove his blithe equation of the CPU with the program? Only then, of course, the entire thought experiment falls apart.
posted by zompist at 2:00 PM on February 15, 2019 [7 favorites]

I totally agree it is bizarre and ridiculous that white people make such a big deal over this. Why can't we just accept an extremely modest request to just consider avoiding this term? There's no request to excise the phrase from the English language or profess it to be a terrible slur. Just go ahead and give a real person's presumably good faith assertion that it might be hurtful to some people, and decide to accept that and move on. Then it doesn't have to be this big crazy thing.
posted by skewed at 2:27 PM on February 15, 2019 [7 favorites]

Seems like a silly objection to me - the form of almost all physical laws named after someone have been "polished up" so to speak for their modern presentation to increase their exactitude and remove erroneous misconceptions as our understanding of the topic increases. It's actually quite interesting going back to older scientific papers and seeing the different language and notation that were used to originally present the concepts.

And I was feeling like this was one of the grumbly things where it's not perfect, but it's also hard not to reference in the philosophical discussion of things, until someone pointed out that you can just call it Searle's room.
posted by Zalzidrax at 2:33 PM on February 15, 2019 [3 favorites]

But--Metafilter does not, itself, have the social clout to create change on these fronts, does it?

Widespread, immediate change, maybe no. I’ve said this before, though, that being a member here and taking part in these discussions has changed my vocabulary, has changed the my word choice and the way I speak. It continues to do so, and those changes aren’t limited to my time spent here. I take those changes with me into my daily life, and yeah, I’ve had conversations with friends who have no connection to this site about language and reasons for maybe not using terms that, through metafilter, I’ve come to understand as offensive or problematic. It’s not a great societal awakening happening instantly, but there is change, and what we do and say and learn here can affect the world outside of here.
posted by Ghidorah at 2:53 PM on February 15, 2019 [9 favorites]

It seems to me that the nasty part of him choosing Chinese is that, ha ha, what are the chances that he, a white guy, could be expected to understand Chinese? So it's a Chinese Room, and not Russian- because Russian is a language he might be (somehow) suspected to understand. He's using Chinese as a stand-in for "something he could not plausibly understand", not just something he happens not to understand. But of course the thought experiment is just as valid if it was the "Russian Room." Unless you know Russian, written Russian is as incomprehensible to English speakers as Chinese is.
posted by BungaDunga at 3:17 PM on February 15, 2019 [2 favorites]

Frankly, I think Searle's choice of Chinese led him (and others) astray in his experiment.

If he had instead imagined a Chinese-speaking person who knows no English in the room, and if he had imagined that person receiving English messages, looking the messages up in a book, and transcribing responses, then I think people would more quickly say, "Hey, wait a minute? How does that book understand English? How can it have a correct English response to any English message?"

They would see that the existence of this book renders the thought experiment meaningless. It doesn't matter whether the person transcribing messages understands them; the question would be whether the process going on in the book "understands" them.

The only way Searle's Room seems to make a convincing point is if the orientalism gets you to ignore the fact that those Chinese messages have real linguistic content and any "book" that could provide a response to any given Chinese message is itself passing the Turing test in Chinese, and therefore the question is not what's going on in the room, but what's going on in the book.

The orientalism isn't just harmful to Chinese people, it prevents English-speaking people from thinking clearly.
posted by straight at 3:26 PM on February 15, 2019 [19 favorites]

So really, it's not just the name of the experiment that is orientalist. The logic of the whole thing falls apart if you take out the orientalism.
posted by straight at 3:32 PM on February 15, 2019 [1 favorite]

Nah, the Russian alphabet is an alphabet. It's even made up of similar characters to English, unlike e.g Thai. It's quite straightforward for literate English speakers to memorize a few words in Cyrillic and to begin seeing patterns etc in the words if they are exposed to enough, but not so much for Chinese because, as mentioned, they are characters and not an alphabet.
posted by the agents of KAOS at 3:47 PM on February 15, 2019 [2 favorites]

They would see that the existence of this book renders the thought experiment meaningless. It doesn't matter whether the person transcribing messages understands them; the question would be whether the process going on in the book "understands" them.

The only way Searle's Room seems to make a convincing point is if the orientalism gets you to ignore the fact that those Chinese messages have real linguistic content and any "book" that could provide a response to any given Chinese message is itself passing the Turing test in Chinese, and therefore the question is not what's going on in the room, but what's going on in the book.

Not really. The thought experiment is a response to the idea that a process that produces the same apparent results as a conscious process must itself experience consciousness - so that a computer that can converse like a human must be conscious like a human.

I've always thought that idea was silly, but a lot of people seem to believe it - often, I think, as a misunderstanding of the Turing Test, which Alan Turing formulated as a test of intelligence rather than consciousness. The point Searle was making was that these are not the same thing, and the thought experiment does a good job of demonstrating that.
posted by A Thousand Baited Hooks at 3:50 PM on February 15, 2019 [4 favorites]

(Also, straight, perhaps you should check with all the Chinese speakers working in the field before announcing your findings).
posted by the agents of KAOS at 3:51 PM on February 15, 2019

The thought experiment is a response to the idea that a process that produces the same apparent results as a conscious process must itself experience consciousness - so that a computer that can converse like a human must be conscious like a human.

It seems to me that the setup of the thought experiment simply moves the question to whether a book like that could exist without somehow being conscious. And it seems to me that would be obvious without the linguistic obfuscation. Imagine an English speaker mechanically transcribing responses from the book to English messages and you can see whether the person in the room understands the messages is irrelevant to the thought experiment's question.

But, the agents of KAOS, you are right it is presumptuous and possibly insulting for me to ascribe that entirely to orientalism rather than the thought experiment being muddled and overcomplicated. I apologize.
posted by straight at 4:26 PM on February 15, 2019 [3 favorites]

> It's quite straightforward for literate English speakers to memorize a few words in Cyrillic and to begin seeing patterns etc in the words if they are exposed to enough

For the purpose of the thought experiment it doesn't matter. The room is able to converse in Russian immediately after I start operating it, whether or not I might eventually start to understand Russian after careful study. For the first week, the room is definitely conversing in Russian without any plausible help from me.
posted by BungaDunga at 4:27 PM on February 15, 2019

It's quite straightforward for literate English speakers to memorize a few words in Cyrillic and to begin seeing patterns etc in the words if they are exposed to enough, but not so much for Chinese because, as mentioned, they are characters and not an alphabet.

Well, this threatens to descend into a discussion of Searle, but let's not repeat his mistakes and his orientalism.

You're Searle in the room. You get a message like 我喜欢马。If you don't know Chinese, then staring at the message probably won't help. Though neither will staring at Я люблю лошадей.

But. You have a rulebook next to you. The first rule in the rulebook is this:

"If you see 我, write down I/me."

Do you still maintain that, with a bunch of such rules, you would never be able to understand the text?

This is precisely why Searle's argument is so infuriating... he puts everything down to whether the CPU (himself in the room) "understands" anything, and downplays the rulebook where the understanding (if any) really lives. The orientalism is being used in the service of the bad analysis.

(To be clear: the CPU never understands anything, and no one in AI thinks it does. Focusing on what's understood by the man in the room— the CPU— is a colossal bit of misdirection. If you really want a name for the thing, maybe try Searle's Red Herring.)
posted by zompist at 4:33 PM on February 15, 2019 [3 favorites]

It's kinda fun watching people re-discover all the arguments sophomores in college make when discussing this thought experiment. It's not clear that any individual Metafilter commenter is sentient, but perhaps the collective of us all on this server exhibit sentience.
posted by Nelson at 4:39 PM on February 15, 2019 [15 favorites]

I refuse to be a partial photon contributing to server sentience.
posted by clavdivs at 4:43 PM on February 15, 2019

For me the language choice of Chinese is probably the best that could be made for the thought experiment itself. One side is English (I'd assume) which is Indo-European and alphabetic. To maximize the hardness of the problem you need the most distant in problem space language that you can come up with that meets some requirements. It must be living and not dead (dead languages to not change so the problem space is fixed in size). It must be natural and vast in content and speakers (constructed languages would fail because they're constructed or they're no native-only speakers). Chinese hits all of the hardest points. Two large populations of native-only speakers. Totally different writing systems (one has ~26 characters, the other untold thousands). They are diametrically opposed as far back as we've been able to track. Indo-European vs Sino-Tibetan. It would be possible to be testable (you can find a person who can speak both, two people who can only speak one.

Hindi is Indo-European, Arabic is Afro-Asiatic (but with a relatively small set of characters).

But I'm pretty happy with Etrigan's "Seale's Room" proposal because my only real gripe was that if you change the name of the experiment (say all through the original thread in question and this post itself) and someone read that and was intrigued... They'd be at a loss of how to find out more about it themselves. This is akin the the evil game where parents tell kids that this thing is called this name and they find out that nobody else uses that name for that thing. There's a compromise between changing names while allowing it to still be searchable and accessible. Should an interested party head to their library or Wikipedia or Google and look for 'name of thing' they should find that name. If they pick up a book and flip to the index, they should find that name.

At least until we can manage a global search and replace across all humanity.
posted by zengargoyle at 4:46 PM on February 15, 2019 [4 favorites]

The first result on Google for "Searle's Room" will take you right to a description of the thought experiment, so it seems like that substitution would work fine.
posted by straight at 4:54 PM on February 15, 2019 [4 favorites]

But does Google *understand* what it's referring you to?
posted by uosuaq at 5:09 PM on February 15, 2019 [12 favorites]

The relative distance of the language doesn't matter. It just matters that the operation of the room is purely automatic- the operator, though human, is just following instructions provided as a lookup table. Any apparent understanding exhibited in the answers is coming from the mechanical operations, not the human. As long as the human legitimately doesn't understand the inputs, it's fine. (What if Searle secretly learned Chinese? It's possible!) A wall of Chinese text is not less comprehensible than a wall of cursive Cyrillic, for me personally or for Searle himself, probably.

You could do it with a translator who translates English to an incomprehensible conlang and back again, with the room operating entirely in the conlang, as long as you trust the translator not to make stuff up.
posted by BungaDunga at 5:25 PM on February 15, 2019

(To be clear: the CPU never understands anything, and no one in AI thinks it does. Focusing on what's understood by the man in the room— the CPU— is a colossal bit of misdirection. If you really want a name for the thing, maybe try Searle's Red Herring.)

The thought experiment was a response to a "Strong AI" position which held that the CPU really does understand in this situation - that a combination of "computer that does not understand a language" and "program that allows the computer to simulate a conversation in that language" results in a system that does understand the language in a way equivalent to the way that a human might. I think this position is called functionalism, and it's more of a philosophical thing than an AI/computer science thing.

Searle reduced this position to a thought experiment about a person in a room to show how it breaks down, because it's pretty clear that a person in a room shuffling papers covered in symbols that they can't read doesn't gain a conscious understanding of those symbols, equivalent to a person who can read the symbols and produce the same output by actually understanding them, just because they have a (very complex) set of rules for manipulating the symbols.
posted by A Thousand Baited Hooks at 5:34 PM on February 15, 2019 [4 favorites]

I'm white: if a Chinese American person says my use of the phrase Chinese Room has Orientalist subtext and they find that hurtful or offensive, that's enough for me.

There is a wide range of ways I can engage with the idea: I can go read about the history of Orientalism (which to me seems important), or the writing of Searle (which might be super interesting). Or, I could not engage at all. But the simplest question I can ask myself is: Is using this phrase so important to me, that I want to hurt this person's feelings?

If I want to talk about the thought experiment going forward, it seems easy enough to just say, "Searle's famous thought experiment" or to use the name he picked and footnote it, as Kalessin suggests. Honestly, what's the big deal with that?
posted by latkes at 5:43 PM on February 15, 2019 [12 favorites]

I think part of what's hard about these discussions for people on the power end of racial inequality is that we don't want to feel like we're bad guys. It comes with a lot of shame and resentment. And it's hard to move past those strong feelings to be open to new perspectives.

On the latest Code Switch podcast, Gene Demby interviewed an author who writes about the minstrel roots of Mickey Mouse. He talked about how people find it so hard to accept that a beloved figure, Mickey, could have roots they find embarrassing, shameful or wrong and they think acknowledging Mickey's roots means you have to throw out something you love. He didn't have time to go much farther, but for me, I took it as a way to think about how frankly, our culture is steeped in racism, because racism was the power that literally built our country (from enslaved Africans to exploited Chinese laborers to murdered indigenous communities). So necessarily, much of what we value is tainted with racism.

One goal for white people could be to find a way to have curiosity and openness about the less positive meanings our beloved practices, images and phrases have to people who aren't white, without getting overwhelmed with negative feelings. I mean, maybe you love The Mouse, for one example, and no one is taking that away, but that doesn't mean we can take away the reality of the influences that made him. Ideally, we could own the past in an honest way, and in a way that elevates the people who have experienced disproportionate harm.
posted by latkes at 5:56 PM on February 15, 2019 [18 favorites]

(I meant to add - if the thought experiment seems kind of pointless, that's probably because here in 2019 it's much easier to think of computers that can produce superficially human-like behaviour while clearly not experiencing human-like subjective states than it was in 1980.)
posted by A Thousand Baited Hooks at 6:02 PM on February 15, 2019 [1 favorite]

The thought experiment was a response to a "Strong AI" position which held that the CPU really does understand in this situation

No. I'm sorry to insist, but this is key to understanding Searle's misdirection. The CPU (Central Processing Unit) is a part of a computer, the bit that does most of the processing, but does not include the memory or disk drive. All the more so, the CPU does not include any programs the computer can run. The CPU never learns anything: it is entirely unchanged by the operations it performs.

Searle is quite clear that he likens his role in the room to the CPU and not to the program. He writes: "The rule book is the "computer program." The people who wrote it are "programmers," and I am the "computer." The baskets full of symbols are the "data base," the small bunches that are handed in to me are "questions" and the bunches I then hand out are "answers."

Pretty much everyone in AI and computer science who has replied to Searle believes that if anything understands, it's the system: the CPU plus program plus database. Searle's mistake is to insist on taking the one part of the system that absolutely does not understand what it's doing, the CPU, and having it speak for the system.
posted by zompist at 6:18 PM on February 15, 2019 [9 favorites]

But what do you mean by "understand" in this context? Do you mean that the system is able to produce a response that's the same as a human who speaks the relevant language would? Or that the computer system as a whole has a subjective experience of understanding equivalent to the human's?
posted by A Thousand Baited Hooks at 6:27 PM on February 15, 2019

I don't think we need to go deep into Searle's thought experiment in this thread, which is about the term "Chinese room" and finding alternatives! Definitely someone could make a post about that if they wanted to, though.
posted by Eyebrows McGee (staff) at 6:28 PM on February 15, 2019 [8 favorites]

floam: My wife was born in mainland China, went to college and graduate school in the US. I showed her this thread. She doesn't understand white people falling over themselves in this way and thinks it's kind of funny.

I'm fully East Asian in appearance, and culturally settler-Canadian (ie not indigenous and largely not Chinese either, so kinda culturally "white" aside from being a target of racism since kindergarten) going back 3 generations. I'm sure you didn't intend to invoke any deep cultural negativities with the above contribution. So the following is only for reference that readers may or may not care to factor in, when composing future commentaries re racial nuances (or, if readers were already aware, it's just a reminder):

A common way that many white people react, when we speak up about microaggressions, is to
1. invoke, as definitive, the dismissals and amusement of Asian people born and raised in countries where people who looked like them were the norm. It makes them judge and jury on matters that we, not they, have experienced from birth.

This invocation
2a. simultaneously brackets together 2 groups of people for no reason, except superficial similarities of appearance. (I've been speaking up about this on this site since 2009, and IRL for years before and to people who were assholes about it. Versus Mefites who generally, since 2007-and-later etiquette discussions, have tended to be thoughtful re being able to empathize with unfamiliar perspectives. So I hope that that impression of Mefites will continue to hold true in this thread)

....anyway, to continue, b. this sort of reaction places Asian-Asians' dismissals and amusement as superior, relative to our "Hey how about reconsidering this terminology?" Or if not intended to be cast as superior, these remarks seem to me to be thrown in as, at minimum, equally authoritative...

...on a subject (microaggressions, Othering, etc) that we, not they, have experienced from birth.

I'm also with kalessin on this point: in my experience, many Asians-from-Asia already tend to look down on people like me as inferior. Looks Chinese! But can't speak Chinese or even understand it except for a few words! Useless! Even now there are those, including full-blood Asian-[Americans, Canadians, whatever] who especially look down on people like kalessin. Again, I'm sure nobody in this thread would mean to evoke or reinforce racial hierarchy bullshit like this. I'm just mentioning it because it's easier to avoid perpetuating when you know it exists. If anybody's so inclined.

On a lighter note! At Christmas, I and my half-white half-sister were talking to our dad about whiteness and racism and Chinese-ness. She asked if there was a Chinese term for people like her. Dad considered, then said a phrase.

Both of us simultaneously: "What's that mean?"

He looked uncomfortable. "Well...half a loaf?"

She's a huge Harry Potter fan and is very grounded, so I hissed at her: "Muuuuudblood!" We laughed our asses off.
posted by cybercoitus interruptus at 6:47 PM on February 15, 2019 [27 favorites]

It seems to me that the nasty part of him choosing Chinese is that, ha ha, what are the chances that he, a white guy, could be expected to understand Chinese?

This kind of projection of motivation onto people without basis has always been one of the weirder aspects of MeFi style social justice arguments for me.

If Searle had spoken Chinese, then he would have chosen another language: Korean maybe, or Arabic (which has the advantage from the POV of the thought experiment of having a complex written form even though it is much closer to English than Chinese is.) But he doesn’t speak Chinese & it’s his room that has him in it, so he chose Chinese.

The move to generality from there makes it Orientalist, especially combined with the unfortunate collision with the pattern "Chinese X" being used for othering ideas (some explicitly racist) about Chinese people.

Kalessin has (perfectly reasonably) requested that, given the above, the name not be used. Instead we get a bunch of people deciding that what’s /really/ important is having a massive fight over whether Searle was a racist or not. Isn’t that besides the point?
posted by pharm at 1:58 AM on February 16, 2019 [4 favorites]

We called it "Star checkers" by the way. It's a star, after all. And "Stern" is just German for "star".
posted by crush at 8:29 AM on February 16, 2019 [1 favorite]

Searle's Room. Do we all agree that term is better to use? Hooray!
posted by Nelson at 9:27 AM on February 16, 2019 [1 favorite]

Searle Room is the obvious choice, no possessive. The language is Searle.
posted by Going To Maine at 9:58 AM on February 16, 2019 [1 favorite]

anem0ne: "they see symbols and squiggly lines i write and then they look it up in their rulebook, and then they decide it's just so much noise because i don't look like them"

I think stuff like this is why the original post got so much pushback -- it's misinterpreting the thought experiment to cast it in a racist light that isn't there. Searle didn't choose Chinese and describe it as "meaningless" (to him) because Chinese is so weird and foreign and who can even read that stuff except weird AI and mysterious foreigners, amirite? He chose it because the experiment requires an interpretable language that he doesn't speak, and Chinese, as a widely understood language that's also very different from Searle's English, fits the bill perfectly. On the contrary, the argument hinges on Chinese being rich with human meaning, as it posits a a theoretical system of algorithms so advanced that it can respond to Chinese input with output that sounds absolutely human even though the only human actually involved with producing the output doesn't understand it himself (which is the whole point of the thought experiment).

At a stretch, you could call the argument Anglo-centric because it assumes an English-speaking central character that also doesn't speak Chinese, but it's not exactly racist to posit a hypothetical that doesn't match everyone's life experience.
posted by Rhaomi at 10:04 AM on February 16, 2019 [8 favorites]

... you know what? Calling it Searle's Room, or a Searle Room problem is brilliant - it's a far better description of the idea, which was just an idea dreamed up by some person. That he called it a 'Chinese' Room is barely relevant, he could have called it a Finnish Room or Quebecois Room or take your pick -
Most importantly - the new name removes the distraction of associating the idea with 'Chinese' (first time I saw this I wondered what exactly did tonal languages have to do with AI?)
posted by From Bklyn at 11:33 AM on February 16, 2019

Can any explain again more clearly how being unintentionally exclusionary with life experiences not a form of microaggression? And how pushback based on this rationale is not itself racist?
posted by polymodus at 11:37 AM on February 16, 2019 [1 favorite]

On a lighter note! At Christmas, I and my half-white half-sister were talking to our dad about whiteness and racism and Chinese-ness. She asked if there was a Chinese term for people like her. Dad considered, then said a phrase.

I tell my kids that no one defines their identity. They are not "half" if they choose not to be. They are "both" if they want to be. But it's up to them.
posted by JamesBay at 11:38 AM on February 16, 2019 [2 favorites]

I mean, if people want to talk about misinterpreting, can I just mention the overwhelmingly white-centric structures in American yet world-class philosophy and computer science departments, and point out how lack of representation reproduces bias in an ideologically neutral appearance? So you get anachronized ideas like, assuming my fellow colleagues aren't Asian or Asian American in real life academic work which today makes zero sense at all?
posted by polymodus at 11:46 AM on February 16, 2019

I would be interested in Rhaomi's detailed justification for what was said, because that's a username I've seen on this site and I am just really surprised by the opinions held in that comment above.
posted by polymodus at 11:50 AM on February 16, 2019

anem0ne: "they see symbols and squiggly lines i write and then they look it up in their rulebook, and then they decide it's just so much noise because i don't look like them"

Rhaomi: I think stuff like this is why the original post got so much pushback -- it's misinterpreting the thought experiment to cast it in a racist light that isn't there.

I read anem0ne's comment as relating to some white people's intransigence or indifference when anem0ne has suggested using accurate terminology, rather than "kabuki," to describe political situations that don't involve actual kabuki. (Kayfabe seems like a winner for actually accurately describing ongoing Republican political theatrical horrors.)

Another common reaction from many white people in discussions like these is to dissect originators' intent, as if that intent exists in a vacuum. Those of us who say "This language is problematic," say it because we do not have the option to experience it in a vacuum. There's ongoing Orientalism that's still largely acceptable in this society for white people to perpetuate. To people who have been the targets of this bullshit, white people vigorously insisting on using original terminology looks real cozy with ongoing racist/Orientalist bullshit. Especially when there's a perfectly accurate term (Searle room) that doesn't rub shoulders with any of the bullshit.

Repeating, "No original bad intent" without acknowledging that a term has things in common with ongoing problematic Orientalism... It looks to me like being 1. at best, uninformed about, 2. uncaring about, or 3. colluding with (using plausible deniability), ongoing White-centric bullshit. (I don't think #3 is operative in this thread but I could be wrong.)

If people wish to convey something different than those 3 options, they could perhaps consider expanding beyond a hyperfocus on the originator's innocuous intent. Wider lenses are available.
posted by cybercoitus interruptus at 12:23 PM on February 16, 2019 [8 favorites]

Anyone who initially inferred that this Searle guy who they'd never met can't have even been unconsciously racist when he formulated his chosen way of explaining the concept of the Room, since otherwise there'd have been mention of that racism when the concept was introduced in a computer science class or philosophy class or in the course of prodigious independent reading or wherever, should really in 2019 be going back and reevaluating that initial inference. Because you probably drew a faulty conclusion which informed and defined your conception of racism, Ouroboros-like, if you can pronounce with certainty and a straight face that this 1980 utterance was as pure as the driven snow and untainted by even the shadow of a biased disregard for the ethnicity employed as a tool to illustrate the idea.

Like even if you personally knew the guy that would be "a stretch" but just sort of inductively vouching for him based on his reputation is a much more absurd thing to do than positing any of the characterizations articulated in the OP.
posted by XMLicious at 12:27 PM on February 16, 2019 [7 favorites]

Thank you, floam! <3
posted by cybercoitus interruptus at 7:22 PM on February 16, 2019

polymodus: "I would be interested in Rhaomi's detailed justification for what was said, because that's a username I've seen on this site and I am just really surprised by the opinions held in that comment above."

Hi polymodus -- sorry for not responding right away; I just got home from a long day at work and couldn't get back to you appropriately till now (I posted that first comment on my lunch break).

FWIW, I got no qualms about referring to the experiment as "Searle's Room" and agree it's a more neutral and universal term for what's become a globally significant term of art. I was just responding to the multiple comments both here and in the other post that seemed to heatedly ascribe a more overtly bigoted or Orientalist valence to the experiment than felt warranted -- kalessin's "Serle fucking chose "Chinese" because he doesn't understand Chinese. The epitome of Western "faultless" exoticism" for example, and anem0ne's first comment here (which is referring more to floam's comment than the experiment directly but still riffing on the experiment as an artifact of intolerance -- e.g., "they decide it's just so much noise because i don't look like them").

Full disclosure: I love the ideas underlying Searle's experiment, and the repeated digs at it in this vein were bugging me -- as 20 year lurk pointed out, Searle formulated the argument with himself as the test subject, so he was describing the intelligibility of the languages from his own point of view, not judging the value of either or suggesting that his lack of understanding is somehow superior or the norm, just what his experience would be in the scenario he described (which required the use of a modern language that was unknown to him). I mean, if Searle had posted his argument as a MeFite now instead of in a paper decades ago, would it be reasonable to call him out here in these fairly harsh terms just because he discussed a hypothetical in which he doesn't understand another language that other people do? I don't think it would be, and wouldn't want that to be the standard here. (I also wanted to emphasize the encouraging fact that Searle's logic is founded on Chinese being a richly meaningful and human language, which doesn't support the idea that he was implying, even unconsciously, that it was objectively nonsensical/alien/weird outside of his own inability to read it).

To be clear, I acknowledge that it doesn't make sense to continue using a name anchored in Searle's POV to refer to what's become a landmark idea in AI science (especially since it sounds similar to the offensive phrases kalessin mentioned up top), and agree that in the broader context outside his original paper using a name that doesn't assume what languages the person in the room knows is totally appropriate. I just felt like some of the critiques here were inaccurately and unfairly maligning an innocuous and worthwhile idea by overlooking why it used the languages it did and thus casting it as more inherently offensive than it was, and wanted to highlight that (IMHO) mischaracterization as a likely source for some of the resistance kalessin was getting early on (as opposed to just thinking he was being overly sensitive or whatever, which one user had already stated at that point).

I'm sorry if my first comment made it sound like I was dismissive of anybody's discomfort with the "Chinese Room" phrase or opposed to an alternative name -- I think the critique of the name as it's used today makes sense, and I'm glad folks were able to find a suitable replacement. I just don't think that the original thought experiment deserves the censure that it was getting from some of the comments here.
posted by Rhaomi at 10:11 PM on February 16, 2019 [3 favorites]

anem0ne, I admit that I didn't recognize the full extent of the rhetorical point you were making when I first responded, which is my fault for not reading closely enough at the time. While I do think the metaphor contributed to the dynamic I mentioned above -- likening floam's initial dismissal of your concerns to Searle's inability to read Chinese, implying both came from a racially insensitive place -- I see now that this wasn't the point you were intending to make. I'm sorry for making you feel unheard -- and I gotta say, I (ruefully) admire the way you've woven your feelings so elegantly into the topic at hand.
posted by Rhaomi at 10:44 PM on February 16, 2019

"If he had instead imagined a Chinese-speaking person who knows no English in the room, and if he had imagined that person receiving English messages, looking the messages up in a book, and transcribing responses, then I think people would more quickly say, 'Hey, wait a minute? How does that book understand English? How can it have a correct English response to any English message?'"

I think you're right about this and it partly explains why I've never thought the choice of language was salient and therefore why I have not until now recognized this as orientalist (which I think I'm at least a little bit woke to): I've always been a functionalist and I've always thought this thought experiment was incredibly obtuse. It never held up to scrutiny for me and my response has always been "How does that book 'understand' [x]?"

Maybe I'm mistaken, but I think that you're correct in your judgment that, for those to whom Searle's argument is compelling, the orientalism of it is doing some/much of the work. (And, to be sure, Searle thought it was a slam-dunk.)

I think this should be generally instructive: leveraging a person's impulse to "other" can do a great deal of (misleading) rhetorical work, hidden in plain sight. Even about something as apparently abstracted as Searle's argument.
posted by Ivan Fyodorovich at 5:43 PM on February 17, 2019 [7 favorites]

For the purposes of the thought experiment, the book is more an ends-justify-the-means stand-in for behaviorism/functionalism than something that could plausibly exist.

Even in the '80s, the concept of the Turing Test was 30+ years old, and early chatbots like ELIZA were sounding remarkably intelligent. Searle was arguing against the functionalist premise that any such program which could convincingly emulate human response, regardless of its internal processes for doing so, was ipso facto conscious, and did so by positing a scenario where the mechanisms carrying out the program (ink on paper, human rotely carrying out instructions) were obviously lacking conscious awareness of the convincing output that functionalists value.

You could try to make his scenario more plausible by replacing the book and person with a whole network of people with "0" and "1" cards physically carrying out a low-level binary simulation of a computer running an advanced chatbot AI. But criticizing the practicality is like taking issue with Hilbert's Grand Hotel because there literally aren't enough resources in the universe to construct and populate a hotel with infinite rooms. How Searle's rulebook produces humanlike responses (regardless of language) can be handwaved away, because how it works isn't the point -- just that the mere fact that it does work would not in itself be proof of consciousness.

(Of course, one counterargument is that the system as a whole constitutes a conscious being, even if the individual parts don't. But that raises all sorts of further questions -- what part of the system experiences consciousness? At what point is an arbitrary representation of such a system intricate enough to achieve that awareness? Could you use such an arbitrary representation to simulate a human mind? Etc.)
posted by Rhaomi at 9:25 PM on February 17, 2019

I think the problem I have with the setup is most easily seen if you imagine Searle in the room mechanically transcribing answers to English questions according to some rule book that tells him exactly what to write.

Does Searle understand the messages? Yes.
Does this seem relevant? No. He's just a secretary taking dictation. If there is consciousness behind the answers to the questions it's somewhere else.

Or imagine two rooms. One spits out intelligible answers to questions. The other spits out nonsense. In both cases the guy in the room is just transcribing answers according to a set of rules. Does the difference between the room that appears to provide conscious answers and the one that doesn't seem to have anything to do with the guy transcribing the messages?
posted by straight at 10:19 PM on February 17, 2019 [1 favorite]

As already mentioned, this isn't really the place to discuss Searle's argument. My point was only that because in my opinion Searle's argument is risible (I really have no patience with it, just as I find Anselm's ontological argument similarly begging the question via obfuscation), the choice of language vis a vis both Searle and his audience never even caught my attention.

But if you're receptive to the argument, which in my opinion means you're likely persuaded by your unexamined intuition (in this case, looking for where "understanding" and "consciousness" is and believing you'll know it when you see it, as well as the argument from ignorance fallacy), then the argument's persuasiveness is likely much strengthened by the evocation of orientalist othering by Searle from his intended audience because it, too, is all about unexamined intuition. I think the argument itself is a sophistry, a sleight of hand, and the choice of language is a purposeful part of the misdirection.
posted by Ivan Fyodorovich at 11:02 PM on February 17, 2019 [5 favorites]

Well that was a much more polite way of saying what I was winding up to say. Up above EMcG suggested that anyone wanting to discuss the Searle's Room concept itself could make a post.
posted by XMLicious at 11:18 PM on February 17, 2019 [1 favorite]

But... we can't really, can we? That would be deleted as chatfilter.

I was trying to explain the way the experiment works to help address a misunderstanding that kept on coming up, and seemed to be contributing to the confusion: that Searle's original orientalist framing was somehow concealing a flaw or some kind of underhanded trick in the argument (which seemed almost to be turning into ... and if you're fooled, maybe there's something wrong with you, too ...). But we're not allowed to talk about that, so I'll just say that "Searle's room" is a pretty good alternative.
posted by A Thousand Baited Hooks at 3:39 AM on February 18, 2019 [1 favorite]

re: the effects of culture on perception of the room Ivan mentioned, I don't see how Searle's particular language choice affects the persuasiveness of his argument. You could swap the two as straight did upthread, or even use a non-linguistic form like hydraulic pipes or plastic balls to implement the room's conversational program, and it doesn't alter his point, which is that an arbitrary computational system can (in principle) implement a program that responds in an intelligent way without any part of that system having the capacity to be consciously aware of what it is saying -- ergo, intelligent response is not by itself proof that a computer is self-aware. (Having the rulebook produce English instead of Chinese might make English speakers more inclined to question whether said rulebook is even possible, but as I said earlier the IRL plausibility of the room isn't the point -- it's just a metaphor for the "strong AI" definition Searle was trying to refute -- so it doesn't aid his argument to "obscure" the work the book does for English readers by making it a different language.)

As to Orientalism, the "otherness" of Chinese in his example is only evoked (from Searle's POV) in terms of his literal inability to speak it, in order to show that despite being a conscious agent himself he wouldn't (couldn't!) consciously understand the content he'd generate in the room, even as that content itself seemed to be the product of a conscious mind. Beyond that purely functional lack of understanding (which is essential to the argument), any perceived cultural "exoticness" of the language isn't helpful or even relevant to the hypothetical.

Kalessin was right, though, to point out that continuing to use Searle's frame of reference by default in naming and describing the experiment going forward contributes to Orientalist thinking by inviting the listener to imagine themselves (or a generic person) as the human in the room who obvs can't understand Chinese, despite that not being true for most people on Earth (including many in computer science). And the ensuing discussion found a great alternative to that. So I'm glad this thread happened, despite any friction.

Also, feel free to take this point to Memail, Ivan, but I'm curious why you find the whole thing so unconvincing independent of this language discussion. I enjoy it because it raises so many interesting questions, but the actual argument he's making (an intelligent-sounding digital program would not necessarily be self-aware) seems fairly narrow and straightforward as far as these things go. I'd also love to hear your take on similar thought experiments like Mary's Room/qualia and p-zombies.
posted by Rhaomi at 7:25 AM on February 18, 2019 [2 favorites]

But... we can't really, can we? That would be deleted as chatfilter.

Of course you could. Some of my favorite FPPs are things I made because a related discussion came up in a conversation on the Blue or the Grey annoyed me and I wanted to create a space to talk about a concept that was inherently tangential to the topic of that conversation. They're both framed to stand on their own--and they do, I am pretty sure that you'd have to do quite a bit of work to figure out what irritated me on the site enough to go "let's talk about these things" if you weren't in the same conversations I was at the time, years later--and neither discussion particularly references the discussions that created them. All you have to do is think about your framing, find a few good links, and create a salon for people who want to talk about that topic to sit and talk about it in.

This discussion is about the Orientalism behind Searle's choice of naming and framing his thought experiment, not whether or not his thought experiment was relevant to computing or whether he was right about AI. If those are the things you want to talk about, it would be easy enough to create a topic about Searle's room on the Blue, given a little time to find a decent link to introduce folks who haven't been following along here to things and listening to the requests of the community here so as not to irritate more people about it, and there you have it! A place to talk about computing history!

But right here, we are talking about the impacts that this experiment has had on the English-speaking descendents of Chinese immigrants, who routinely experience many small equations of their people and their home culture to inscrutable stereotypes. That shit? It sucks. We're not really talking about whether Searle or any specific person meant well or not, we're talking about what the aggregate impact of Searle's decision to call his room had on the people who are most affected (which is descendents of the Chinese immigrant diaspora).

Let's think about what our choices can do in the aggregate, and how we might change the way we approach these topics in the future so as not to invoke those centuries of racism, shall we? It's not about whether or not the past did things well: it's about how we, this community, can do things better in the future, and how we can learn to know what to look for so that we don't accidentally invoke this deep and vicious history without meaning to.
posted by sciatrix at 8:44 AM on February 18, 2019 [6 favorites]

Those of you arguing about the thought experiment itself rather than engaging with kalessin's point are coming across as dismissive and deflective. I echo the request to move it to MeFi.
posted by lazuli at 9:03 AM on February 18, 2019 [4 favorites]

Arguing about how the choice of "Chinese" specifically might be relevant to the experiment is pretty directly responsive to kalessin's request to not use that word. Either the entire request was met by just changing the name of it while still using Chinese in the setup, or the goal is to change the description as well. You can't replace that without figuring out what it's doing.
posted by the agents of KAOS at 12:03 PM on February 18, 2019 [1 favorite]

That's a good point, KAOS, and helps illuminate for me why it felt necessary to explain why I thought Searle's original formulation of the experiment was not itself problematic, even though uncritically repeating his framing (and naming it after that framing) might be.

As I understand it, the ask here is to stop referring to the Room in a way that makes Anglo-centric assumptions about which languages are understandable and which aren't. Renaming it is an easy enough step, but describing it takes a little more thought. One option is to emulate the wording of the original paper by putting yourself in the room instead of the listener or a generic "default" person, and explicitly saying that the input/output language is only meaningless from your POV. You could also get creative with the circumstances and replace the unintelligible I/O language (and external interlocutors) with aliens or something, but this might have a greater effect on how the language functions in the metaphor. (Posting on my break atm so I can't explore this idea fully, but I'd be interested in hearing other perspectives).
posted by Rhaomi at 12:58 PM on February 18, 2019

It does help to explain the experiment using a real, known language, because there has to be the possibility that the person in the room understands the symbols being manipulated. It also helps if it's a language which uses a script that the person in the room can't read. It still makes sense if you take those away, but it gets a little harder to follow. Here's a bit from the original paper (PDF) where it's made clear that the language is specifically one that Searle doesn't understand:

I want to block some common misunderstandings about "understanding": In many of these discussions one finds a lot of fancy footwork about the word "understanding." My critics point out that there are many different degrees of understanding; that "understanding" is not a simple two-place predicate; that there are even different kinds and levels of understanding, and often the law of excluded middle doesn't even apply in a straightforward way to statements of the form "x understands y"; that in many cases it is a matter for decision and not a simple matter of fact whether x understands y, and so on.

To all of these points I want to say: of course, of course. But they have nothing to do with the points at issue. There are clear cases in which "understanding" literally applies and clear cases in which it does not apply; and these two sorts of cases are all I need for this argument. I understand stories in English; to a lesser degree I can understand stories in French; to a still lesser degree, stories in German; and in Chinese, not at all. My car and my adding machine, on the other hand, understand nothing: they are not in that line of business. We often attribute "understanding" and other cognitive predicates by metaphor and analogy to cars, adding machines, and other artifacts, but nothing is proved by such attributions.

posted by A Thousand Baited Hooks at 1:25 PM on February 18, 2019

Oddly, my understanding of the experiment Searle proposed doesn't really require a real, known language.

Yea, mine either. Klingon would, IMO, be a good substitute (although it wouldn't have been at the time, afaik). To discuss any reasons why I might choose that instead of, say, a randomized selection of colors, would clearly be indulging in some kind of thing that you are guessing at but bizarrely unwilling to say.
posted by the agents of KAOS at 2:06 PM on February 18, 2019

...but of course the propositions under discussion are explicitly that Searle was influenced by conscious or unconscious racism or Orientalism, that the mechanism of conveying the concept leveraged racism or Orientalism intentionally or unintentionally, and that regardless of whether any of those scenarios are true the unhesitating use of the term “Chinese Room” is experienced by some people as a microaggression related to their ethnicity.

So there's no need whatsoever to get into hypotheticals about any fictional person's bizarre unwillingness to name their real objection to the selection of comprehending the Klingon language as a vehicle over color qualia.

Nor any need to be linking to the Wikipedia entry about what the Turing test is on the tenth occasion of it being mentioned in the thread, since neither involved technical details nor basic stuff like that is going to inform a conversation about any of the aforementioned propositions. This, I think, is what lazuli meant about the fixation of the discussion on details related to the internal reasoning or significance of Searle's concept seeming dismissive and deflecting.

The impression foremost in my mind has been, it's like turning a discussion about an unarmed black guy getting shot by the police into a discussion about the intricacies of firearm technology. It seems like exaggerated avoidance of the nominal topic—racism or a similar type of ethnicity-related bias—in a fashion that could be readily explained by unconscious or conscious racism itself, next to what to be honest seem like rather cursory assurances of how unsullied by the pervasive racism of his society Searle's original articulation was.

It's the sophistication of the discussion and analysis of racism and Orientalism we should be turning up to 11 in this thread rather than ascending to dizzying heights of philosophical or computer science inquiry. And understanding why a space where these kinds of biases are under discussion needs to remain focused on those societal issues is an important part of understanding the issues themselves.
posted by XMLicious at 3:17 PM on February 18, 2019 [7 favorites]

The impression foremost in my mind has been, it's like turning a discussion about an unarmed black guy getting shot by the police into a discussion about the intricacies of firearm technology.

No. It's like turning a request to not use the word firearm when discussing someone shot by police getting turned into a discussion about firearms.
posted by the agents of KAOS at 4:01 PM on February 18, 2019

No. It's like turning a request to not use the word firearm when discussing someone shot by police getting turned into a discussion about firearms.

Notice how you removed any context of racism out of your rephrasing there.
posted by XMLicious at 4:09 PM on February 18, 2019 [1 favorite]

Let's not go into finding analogies to police violence. Just go ahead and make the point you want to make without the analogy.
posted by LobsterMitten (staff) at 4:09 PM on February 18, 2019 [1 favorite]

Specifically, Orientalism as demonstrated by the AI / philosophy experiment that you called out in the post. Which, despite the lengthy thread, I still don’t see (in contrast to the other idioms you mentioned). But happy to avoid the term here as requested, given the strength of feeling (yours and others).
posted by inire at 5:40 PM on February 18, 2019

By which I'm saying the gambits in play right now at the end of this discussion are super unoriginal, super unsurprising, and very non-novel.

They're not even novel to this particular thread by this point.
posted by tobascodagama at 6:48 PM on February 18, 2019 [4 favorites]

I think a larger issue that is inevitably broached simply by considering this specific case in the first place is whether it's even possible to conclusively exonerate a given bit of speech or an action as not racist, anyways. In contemplating that question I have been constantly amazed by how frequently and quickly my past self and some of my fellow white people here in the U.S. are capable of arriving at a conclusion of “definitely not racist” rather than “maybe racist.”

Like you would kind of think, in a nation like the U.S. where racial segregation and separate drinking fountains and black people being unable to buy a sandwich in some places exist as living memories, and where George Takei can relate his recollections of being confined to a concentration camp as a child with his family, a reasonable heuristic would be to err on the side of “maybe.”

People expect that there's going to be something like a Euclidean reasoning process by which you can check off most speech and actions in the “not racist” column when the question arises.

But there aren't many things in life, especially for members of minorities and women it would seem, where you can just apply a simple set of rules and then not have to worry that you won't face any possible consequences for even the appearance of how you speak or act.

The very expectation that there's going to be some simple test you can apply and hence definitively clear any given statement or action as not racist—because it would be huge pain in the ass if you had to run through a complex process with many competing and shifting and sometimes contradictory concerns, right?—that expectation, that consideration of racism can't be anything beyond the most minimal burden or complication and still be valid, would seem to me to itself be a major facet and mechanism of white supremacy, serving to produce lots of false negatives.
posted by XMLicious at 7:08 PM on February 18, 2019 [8 favorites]

"We need to finish hashing out this very important piece of philosophical work before we can adequately address whether this piece of philosophy may contain problematic bits" is exactly the process that reinforces White supremacy.
posted by lazuli at 8:20 PM on February 18, 2019 [7 favorites]

kalessin: "I wonder if, after white folks have finished flogging the AI/philosophy issues they need to hash out, we might return this thread to the original topic: Orientalism."

Your thread is about the Orientalism of the "Chinese Room" as it was mentioned on the blue. There seems to be consensus that the name is problematic and should be changed, but there were differing opinions as to whether the argument itself is Orientalist, benefiting from Orientalism, or playing on Orientalist bias to paper over weak logic. Focusing on this isn't a distraction from the problem of Orientalism -- hashing out these questions helps those without an intuitive understanding of Orientalism recognize what specific forms of Orientalist thinking you want to discourage, getting people on the same page as to what rhetorical aspects (if any) should be altered or avoided in invoking the Room and metaphors like it here in the future. Doing this requires some measure of engagement with the underlying mechanics of the argument, insofar as they interact with cultural biases, to interrogate how they are problematic.

I understand and agree with the premise that naming the thought experiment after Searle's Anglo-centric point of view and making that "English speaker not grokking Chinese" set-up the default in generalized discussion of the matter alienates people who do speak Chinese and contributes to thinking of Chinese as "other". I also see how the phrase "Chinese Room" echoes the other phrases kalessin cited, unintentionally reinforcing the othering effect. What I fail to see so far is how Searle's selection of Chinese in his original metaphor is itself problematic. I question this not to defend Searle's honor or distract from a frank discussion of implicit racism, but to try and develop an understanding of whether, how, and why his metaphor is Orientalist in order to discourage any such specific harmful patterns going forward, both in discussions of the Room and more broadly in the use of any metaphor that mentions Chinese or China-related subjects.

To boil it down, let's say Searle had been a poster on the blue, and posted the following comment in a thread about AI:

Just because a computer can respond in a convincingly human way doesn't mean it understands what it's saying, or has any awareness at all. For the sake of argument, imagine that such a computer existed and were conversant in a language that I do not speak or even recognize, such as Chinese, and I were put in a room and tasked with physically carrying out the operations that constitute its program by manipulating cards with Chinese characters according to rules written in English. Now, the logical operation of the program I enact is in principle the same as the intelligent computer, and a Chinese speaker could input language and receive convincingly human replies, but nothing in the room -- ink on paper and my monolingual ass -- is capable of understanding the content of those replies. So where is the awareness?

Does this act of using Chinese as an example of a language one doesn't speak themselves, for the purposes of a thought experiment about themselves, constitute Orientalist thinking? I'm leaning towards no (as long as it's clear the commenter is only speaking from their own POV instead of acting like Chinese is inherently unintelligible), but I'm open to hearing other arguments.

Or how about we sever this from the AI jargon entirely -- consider a comment like this:

After years of orienteering, my sense of direction is pretty well honed. Like I don't speak a word of Chinese, but you could randomly drop me in the middle of China without the aid of an English map or signage and I could still find my way around the terrain from sight alone.

That seems innocuous to me, whereas a commercial boasting about a GPS app that can help you navigate "even in the middle of China!" would be more clearly offensive, because it Anglo-centrically assumes that nobody can interpret Chinese maps/signs/people.

At the moment, characterizing comments like the first two as Orientalist/harmful/problematic implies to me that any metaphor that references a marginalized/non-privileged subject from a privileged POV, even incidentally and non-judgmentally, is inherently insensitive, which strikes me as overly broad. But again, if I'm misinterpreting or exaggerating the definition or harmfulness of Orientalism here, please correct me -- I'm just trying to understand what people think is acceptable or not about statements like these.

On preview: Not asking for a General Theory of Racism, just the specific ways in which an argument like this is problematic. If there's something inherent to Searle's thought process that's harmful, then just changing the name won't address it because folks will still be potentially recapitulating the same harms under a different title. Basically, is calling it Searle's Room sufficient to "dismantl[e] this linguistic micro aggression" as kalessin says in the OP, or is a more critical re-evaluation of the argument itself necessary?
posted by Rhaomi at 8:35 PM on February 18, 2019 [3 favorites]

I think it's helpful to ask oneself, "Does assuming that no one to whom I'm addressing my argument speak any Chinese dialect make sense?" when one is presenting any sort of linguistic argument. Beyond that, it reads like people splitting hairs in order to continue making Orientalist arguments.
posted by lazuli at 8:40 PM on February 18, 2019 [3 favorites]

In that light, I suppose you could characterize the harm at work in my examples as "it's kind of self-centered (literally) to employ a metaphor tailored to your own point of view, even if it's valid for you, because not everyone listening shares that point of view and can relate to your metaphor in the same way -- especially online." And that this selfishness is made more fraught when the metaphor is characterizing a marginalized subject from a privileged point of view, even if that characterization is incidental or non-negative or not implied to be universal. Does that sound fair?
posted by Rhaomi at 8:55 PM on February 18, 2019 [1 favorite]

From my perspective, yes, that's fair, but I'm basically being stubbornly obnoxious because I can afford to be, so make sure you have buy-in from people more directly affected, too.
posted by lazuli at 8:59 PM on February 18, 2019

I'm persuaded by others and my own reasoning that the choice of language¹ is doing some rhetorical work with Searle's intended audience, whether or not it was deliberate. It seemed to me quite relevant to counter the strong and frequently made claim here that the choice of language is irrelevant to the orientalism² of the thought experiment. Especially because "the language doesn't matter" was my own initial reaction. This seems a convincing justification for my discussion of the merits and particulars of Searle's argument. (It was not, actually, and I'll get to that in a moment.)

To be sure, my comment was provocative in that it implies some activated latent orientalism in those who find Searle's argument persuasive, and I agree and apologize for this because at the outset of my comment I emphasized that this isn't the place to argue the particulars of the argument. Those who felt criticised by that rightly likely feel that this was underhanded.

But set all that aside. Sadly, I'd forgotten my own advice from a previous Meta on the similarly fraught topic of cultural appropriation (bolding added):

In my view, there's two key things going on: privilege fragility, and what I'll tentatively call privilege insularity.

The defensive reaction that characterizes privilege fragility is a very recognizable and ubiquitous thing and it inevitably wrecks the thread.

Likewise, that abstract, aloof, impersonal analysis that's available to the privileged, a stance that treats the conversation as an intellectual exercise or a high school debate competition, is also recognizable and ubiquitous and wrecks discussion.

My own comments demonstrate that "privilege insularity" in spades. I regret and apologize for this.

What is salient here is that someone from the affected community has pointed out that the argument is problematic. Its name, derived from its choice of language, evokes some orientalism. We've found a good substitute with "Searle's Room", but much more important are the larger points kalessin has been making.

When a complaint such as this is lodged, those of us who are insulated by our privilege will always, in the end, have to decide for ourselves how persuaded we are and whether or how much we will alter our behavior in response. The merits of this particular case are very important in that sense, but not with regard to how we hear and respond to the fact the complaint was made at all. Far too often, critiques of the particular complaint provide cover for witting and unwitting privileged behavior that is emblematic of the general problem. I work very hard to be self-aware about this, but I have more work to do. As a community, we all do.

1. If I seem to be avoiding "Chinese", it's because I have some objections to the orientalism implicit in how the written language and great diversity of spoken topolects are all thought of as "Chinese" by those of us who know very little or nothing on the topic. I know very little, so I'm not engaging on the matter, but I'm uncomfortable and thus avoiding the cavalier use of "Chinese".

2. Surely someone has already linked to the Wikipedia entry on Said's Orientalism? If not, then so.
posted by Ivan Fyodorovich at 9:06 PM on February 18, 2019 [3 favorites]

This is mostly for those who don't see the Orientalism. (And if you still don't, please consider that maybe there is not an essential need for another hot take from a white person denying that something is Orientalist. That does not make the issue go away for those who do see it.)

First: why does it have to be Chinese at all? As gauche pointed out, the story would have worked fine if it was cuneiform. There is a long tradition of viewing China and Chinese as exotic and impossible to understand. This gets tedious for someone Chinese, or even for someone who loves Chinese.

Second: look at how Searle talks about it. He gives an example of a rule from the rulebook: "Take a squiggle-squiggle sign from basket number one and put it next to a squoggle-squoggle sign from basket number two." (That's from the Scientific American version of his argument.) That's honestly pretty appalling; it's pure othering of the writing system, to the point of mockery.

(Quick Chinese lesson: what do you think 一 二 三 mean? Not a trick question, it's dead easy. And you can still see the picture in 木 tree, 山 mountain, 口 mouth. Squoggles my 屁股.)

Finally, it's not accidental that Searle is rhetorically playing up the incomprehension of the man in the box, using something he assumes his readers think is "mysterious squiggles" to do so. That is the whole game he is playing: pretending that the experience of the man in the box is the only thing of importance in the room. To the same end he dismisses the rulebook-- which in any Turing-Test-passing program would be unbelievably enormous-- as "bits of paper". (That one is from the original BBS paper.)

Admittedly, all this is easier to see for people who don't like Searle's argument! If you do find it convincing, well, why not change the languages at least? Create a Martian Room whose inputs and outputs are in English.

(But really, if you confront the exoticism in Searle, I think you have to ask: what happens in the man in the box owns a dictionary? Does the argument depend on the writing being "squiggle-squiggles", rather than a language an English speaker can learn? It doesn't even violate his supposed axioms-- a dictionary is purely 'syntactic' in his terms. Yet it seriously weakens his argument.)
posted by zompist at 10:22 PM on February 18, 2019 [2 favorites]

Rhaomi, I want to point out that your basic approach seems to be to assume a complete transcendence on Searle's part of the endemic racism of American society and proceed from there. I imagine I don't have to go on about the Chinese Exclusion Act and Dr. Seuss's WWII propaganda cartoons of Japanese people or the round-up of Asian-Americans into camps and other elements which would have been present in his formative environment and in the U.S. of 1980.

But this isn't a court of law; we don't have to use a “beyond a reasonable doubt” standard, but instead could err on the side of reality and take Searle's place, time, and social context as warranting a presumption of influence of racism, or at least as warranting a presumption that maybe he was influenced by racism, and so we could require the persuasive evidence we seek to be of his transcendence of his society's racism.

You mention harm; I think it's easy to see, even without an “intuitive understanding of Orientalism”, that the practice of generously granting benefit of the doubt with regards to racism, in a society where we collectively participate in tons of structurally racist stuff all the time and you can't throw a stick without hitting some racism, will be prone to enable racism and create a tendency to overlook it. Especially when we're all only human and may not always hand out these sorts of generous assumptions of personal virtue equally.
posted by XMLicious at 11:09 PM on February 18, 2019 [4 favorites]

Notice how you removed any context of racism out of your rephrasing there.

Yes! Thanks for noticing my deliberate and careful editing.

Back on topic: My impression of the thread so far is that calling the experiment Searle's Room is easily accepted, and that the goal of the thread was not to discuss the details of the experiment or what language it uses, or whether Searle was being Orientalist in his choice at the time. I'm not sure I understand what else the thread is meant to be about. Can anyone give an example of what they'd hoped would be discussed?
posted by the agents of KAOS at 11:41 PM on February 18, 2019

My context has been to imagine this in a modern academic setting at an American university. My feeling would be, using the original name and formulation of Searle's Room is not okay, given the globalized audience. If I were giving a seminar on it, I would not use it (I'd probably give a short explanation of why I would diverge from the original name since that's the one people know). And probably, I might choose English, i.e. suppose the operator didn't speak English. CS and logic people are pretty good at such hypotheticals.

Also, I found a Medium article about the Chinese Room experiment and it used an Orientalist image, the Rickshaw font. Also, this bi-monthly's MIT tech review changed one of their article titles, originally it was something like "locked out of the Chinese Room." These are easily googled.
posted by polymodus at 12:25 AM on February 19, 2019 [2 favorites]

kalessin: "If Rhaomi and KAOS would care to spend the effort to scroll up and revisit this post's title and subject matter, they might wish to.

I'm not here to be told, inaccurately, and with a clear agenda (to correct me from criticizing white people and white culture in the context of the defense of white supremacy amid my accusation and criticism of Orientalism), and classically (and as I noted, profoundly unoriginally), with a soupçon of patronization, what the post I composed is about, nor am I interested in having my original motives explained to me.

Though I certainly understand wanting to respond to continued effort by Rhaomi and others, who are attempting to repurpose this discussion to their own ends, I, personally, will entertain only novel gambits going forward."

Kalessin, I feel like I've tried repeatedly to ground my comments here in terms of better understanding the scope of your original ask, including directly quoting your request, in order to "be mindful as [folks] proceed into the future," as you said. (Be mindful of the phrase "Chinese Room" alone, or something deeper about the underlying argument? That's what I've been trying to hash out). Also, since my initial comment here rather embarrassingly misinterpreted what somebody else was getting at, I've tried to be more explicit about what (and why) I'm thinking and what I understand other people to be saying, to avoid talking at cross-purposes. I apologize if that comes off as patronizing or telling you what you really mean -- I'm not trying to put words in your mouth, I'm just trying to make clear what I understand your request to be, so that any misapprehension on my part can be noticed and corrected.

I'm disappointed to hear you find at least some of these efforts (how much mine vs. KAOS is unclear) inaccurate, agenda-driven (in defense of white supremacy), unoriginal, patronizing, and repurposing. That is the polar opposite of what I'm trying to do here. And it's especially surprising because I feel like at this stage I've come to agree with what I understand your viewpoint to be.

In that spirit, here are the lessons I've drawn from participating in this thread, linking directly where relevant to where I'm getting it from, along with what I believe is a good way to avoid that harm in the future:

• the name "Chinese Room" is offensive because it assumes a Western POV and echoes racist phrases, and should be called the more neutral "Searle's Room"
• by uncritically repeating Searle's individual Western POV in general terms and making it the default formulation of the argument in pedagogy, modern descriptions of Searle's Room constitute unconscious systemic bias against Chinese speakers. This can be corrected by making Searle's unintelligible language an invented one, and more broadly by working to include more non-Western perspectives in the field, so that it's harder to overlook those perspectives.
• the original formulation of Searle's Room and similar arguments that reference marginalized subjects from a privileged POV are insensitive because they hinge on a metaphor that is only valid from that privileged POV, and implicitly ask marginalized listeners to think of themselves as "other." This can be avoided by being more mindful of how one refers to other cultures, how those references might sound to a diverse audience, and to try to make those references as inclusive as possible.
• Searle's usage of Chinese specifically has Orientalist valence because Chinese has historically been used as a stand-in for "complicated" or "not understandable." This can be avoided by no longer referencing Chinese in this manner.
• As a privileged white guy in 1980 America, Searle probably harbored plenty of racist views, so arguments that read as defending him personally sound more dubious. (FWIW, I agree with this, and my references to Searle were more to his original argument than to him as a person.)
• Also, in a meta sense, conversations about stuff like this work better and more respectfully by not derailing on to unrelated subjects, instead centering on listening to the perspectives of those raising the issue to begin with.

That is, to the best of my reckoning, all the aspects in which the usage of "Chinese Room" kalessin called out is objectionable, and how to effectively address such usage going forward. If I forgot/minimized/misinterpreted anything, please let me know so I can improve my understanding. (And if this attempt at summarizing my understanding and inviting correction is one of those "gambits," then I apologize in advance for falling short, but I'm doing the best I know how to do here.)
posted by Rhaomi at 11:55 AM on February 19, 2019

I think we may be reaching the point in here where less is more. That is, Rhaomi, I get that you're trying to engage intensively as a way of taking this issue seriously. But I think it may end up having the opposite effect from what you intend, and for now it's probably better to ease off.
posted by LobsterMitten (staff) at 12:19 PM on February 19, 2019 [3 favorites]

Also, some of those bullet points were arrived at in response to subsequent discussion in the thread, not kalessin's original post alone. So I acknowledge it's fully possible kalessin only wanted to halt the use of the phrase and finds this deeper examination distracting and unnecessary. I hope it's clear I engaged in this line of thinking not to derail kalessin's thread, but to develop a better understanding of what was being asked for and how to most effectively carry that out.
posted by Rhaomi at 12:19 PM on February 19, 2019

It is clear, and I appreciate the summing up of one rather involved aspect of the thread - thank you.
posted by inire at 1:46 PM on February 19, 2019 [1 favorite]

I would also be interested in hearing any recommendations kalessin and others with similar concerns might have for discussing this topic more considerately in the future.

It seems pretty easy and a good idea to use the term "Searle's Room" when just making a passing reference to the thought experiment. But when actually discussing Searle's ideas, would it be more considerate to to paraphrase descriptions of the scenario using a fictional or unspecified language rather than Chinese? Or would it be better to preface with a discussion of Searle's original terms and why they might be a problem?
posted by straight at 2:49 PM on February 20, 2019

From my perspective, I felt like there was a general tone (which is obviously hard to pinpoint) from several commenters of "Searle's thought experiment is important and I am going to make arguments about why you're wrong; prove me wrong before I believe you," rather than coming from an assumption of, "OK, I trust you and your judgement on this issue. The Orientalism is a problem. Can we reframe the useful parts of this thought experiment in non-othering ways, or is it a wash-out?"

As jacquilynne pointed out, it would be helpful if we could start the discussion at step 4.
posted by lazuli at 3:51 PM on February 20, 2019 [4 favorites]

There's always that slight danger, though, of "getting a PoC friend" who doesn't actually agree with your take on things. If we skip straight to step 4, we're surrendering our own judgment to this qualified person. How to handle the contradictions? Assume the worst?
posted by uosuaq at 2:47 PM on February 21, 2019

At which point, I'll agree that it's worth having a thread of this length about a philosophical argument which, despite having majored in philosophy, I'll probably never refer to again? I'm sorry to be antagonistic, but you're still assuming that after my education I'll wind up agreeing with you about this thing that...clearly annoys you, but...well, your advice is good. Read more, learn more, maybe even make some friends. PoC friends that is. I'll give Craigslist a shot. ;)
posted by uosuaq at 3:07 PM on February 21, 2019

If “assume the worst” is intended to stand in for “assume racism when someone is referring to Chinese as ‘squiggle squiggle’ and ‘squoggle squoggle’, and require clear incontrovertible evidence of it not being racism rather than holding back and waiting for prosecutable smoking-gun evidence before you'll respond to extremely feasibly possible racism, informed by the types of sources of information kalessin proposes” then yeah that seems like what is actually necessary in a society in which we do not at all behave as though overt racism is “the worst.”

Imagine how the history of responses to #MeToo behavior might have gone differently in the last few decades if people had “assumed the worst” and simply actually consistently criticized guys and held them accountable for blatantly sexist and harassing behavior, and not put up with it when others reached for exonerating “locker-room talk” rationalizations first.
posted by XMLicious at 3:36 PM on February 21, 2019 [2 favorites]

Okay, so:
"White dude seeks PoC friends from various backgrounds, who will speak their minds, for personal edification. Several of each type (Asian/Latinx/veteran etc.) needed."
Am I getting close?
posted by uosuaq at 3:46 PM on February 21, 2019

I'm not British, so no. I feel like you're actually telling me to go out and "make new friends" based on their demographics, for my own enlightenment. Let's just say that approach rubs me the wrong way.
posted by uosuaq at 3:58 PM on February 21, 2019

Mod note: uosuaq, you've got a history of digging in for no particularly good reason in threads sometimes and this is feeling like one of those times. Give it a rest in here, and let's let this exchange drop.
posted by cortex (staff) at 4:01 PM on February 21, 2019 [1 favorite]