Related to: Leave a Line of Retreat

Followup to: Talking Snakes: A Cautionary Tale, The Skeptic's Trilemma

"I argue very well. Ask any of my remaining friends. I can win an argument on any topic, against any opponent. People know this, and steer clear of me at parties. Often, as a sign of their great respect, they don't even invite me."

        --Dave Barry

The science of winning arguments is called Rhetoric, and it is one of the Dark Arts. Its study is forbidden to rationalists, and its tomes and treatises are kept under lock and key in a particularly dark corner of the Miskatonic University library. More than this it is not lawful to speak.

But I do want to talk about a very closely related skill: not losing arguments.

Rationalists probably find themselves in more arguments than the average person. And if we're doing it right, the truth is hopefully on our side and the argument is ours to lose. And far too often, we do lose arguments, even when we're right. Sometimes it's because of biases or inferential distances or other things that can't be helped. But all too often it's because we're shooting ourselves in the foot.

How does one avoid shooting one's self in the foot? In rationalist language, the technique is called Leaving a Social Line of Retreat. In normal language, it's called being nice.

First, what does it mean to win or lose an argument? There is an unspoken belief in some quarters that the point of an argument is to gain social status by utterly demolishing your opponent's position, thus proving yourself the better thinker. That can be fun sometimes, and if it's really all you want, go for it.

But the most important reason to argue with someone is to change his mind. If you want a world without fundamentalist religion, you're never going to get there just by making cutting and incisive critiques of fundamentalism that all your friends agree sound really smart. You've got to deconvert some actual fundamentalists. In the absence of changing someone's mind, you can at least get them to see your point of view. Getting fundamentalists to understand the real reasons people find atheism attractive is a nice consolation prize.

I make the anecdotal observation that a lot of smart people are very good at winning arguments in the first sense, and very bad at winning arguments in the second sense. Does that correspond to your experience?

Back in 2008, Eliezer described how to Leave a Line of Retreat. If you believe morality is impossible without God, you have a strong disincentive to become an atheist. Even after you've realized which way the evidence points, you'll activate every possible defense mechanism for your religious beliefs. If all the defense mechanisms fail, you'll take God on utter faith or just believe in belief, rather than surrender to the unbearable position of an immoral universe.

The correct procedure for dealing with such a person, Eliezer suggests, isn't to show them yet another reason why God doesn't exist. They'll just reject it along with all the others. The correct procedure is to convince them, on a gut level, that morality is possible even in a godless universe. When disbelief in God is no longer so terrifying, people won't fight it quite so hard and may even deconvert themselves.

But there's another line of retreat to worry about, one I experienced firsthand in a very strange way. I had a dream once where God came down to Earth; I can't remember exactly why. In the borderlands between waking and sleep, I remember thinking: I feel like a total moron. Here I am, someone who goes to atheist groups and posts on atheist blogs and has told all his friends they should be atheists and so on, and now it turns out God exists. All of my religious friends whom I won all those arguments against are going to be secretly looking at me, trying as hard as they can to be nice and understanding, but secretly laughing about how I got my comeuppance. I can never show my face in public again. Wouldn't you feel the same?

And then I woke up, and shook it off. I am an aspiring rationalist: if God existed, I would desire to believe that God existed. But I realized at that point the importance of the social line of retreat. The psychological resistance I felt to admitting God's existence, even after having seen Him descend to Earth, was immense. And, I realized, it was exactly the amount of resistance that every vocally religious person must experience towards God's non-existence.

There's not much we can do about this sort of high-grade long-term resistance. Either a person has enough of the rationalist virtues to overcome it, or he doesn't. But there is a less ingrained, more immediate form of social resistance generated with every heated discussion.

Let's say you approach a theist (let's call him Theo) and say "How can you, a grown man, still believe in something stupid like talking snakes and magic sky kings? Don't you know you people are responsible for the Crusades and the Thirty Years' War and the Spanish Inquisition? You should be ashamed of yourself!"

This suggests the following dichotomy in Theo's mind: EITHER God exists, OR I am an idiot who believes in stupid childish  things and am in some way partly responsible for millions of deaths and I should have lower status and this arrogant person who's just accosted me and whom I already hate should have higher status at my expense.

Unless Theo has attained a level of rationality far beyond any of us, guess which side of that dichotomy he's going to choose? In fact, guess which side of that dichotomy he's now going to support with renewed vigor, even if he was only a lukewarm theist before? His social line of retreat has been completely closed off, and it's your fault.

Here the two definitions of "winning an argument" I suggested before come into conflict. If your goal is to absolutely demolish the other person's position, to make him feel awful and worthless - then you are also very unlikely to change his mind or win his understanding. And because our culture of debates and mock trials and real trials and flaming people on Usenet encourages the first type of "winning an argument", there's precious little genuine mind-changing going on.

Really adjusting to the second type of argument, where you try to convince people, takes a lot more than just not insulting people outright1. You've got to completely rethink your entire strategy. For example, anyone used to the Standard Debates may already have a cached pattern of how they work. Activate the whole Standard Debate concept, and you activate a whole bunch of related thoughts like Atheists As The Enemy, Defending The Faith, and even in some cases (I've seen it happen) persecution of Christians by atheists in Communist Russia. To such a person, ceding an inch of ground in a Standard Debate may well be equivalent to saying all the Christians martyred by the Communists died in vain, or something similarly dreadful.

So try to show you're not just starting Standard Debate #4457. I remember once, during the middle of a discussion with a Christian, when I admitted I really didn't like Christopher Hitchens. Richard Dawkins, brilliant. Daniel Dennett, brilliant. But Christopher Hitchens always struck me as too black-and-white and just plain irritating. This one little revelation completely changed the entire tone of the conversation. I was no longer Angry Nonbeliever #116. I was no longer the living incarnation of All Things Atheist. I was just a person who happened to have a whole bunch of atheist ideas, along with a couple of ideas that weren't typical of atheists. I got the same sort of response by admitting I loved religious music. All of a sudden my friend was falling over himself to mention some scientific theory he found especially elegant in order to reciprocate2. I didn't end up deconverting him on the spot, but think he left with a much better appreciation of my position.

All of these techniques fall dangerously close to the Dark Arts, so let me be clear: I'm not suggesting you misrepresent yourself just to win arguments. I don't think misrepresenting yourself would even work; evolutionary psychology tells us humans are notoriously bad liars. Don't fake an appreciation for the other person's point of view, actually develop an appreciation for the other person's point of view. Realize that your points probably seem as absurd to others as their points seem to you. Understand that many false beliefs don't come from simple lying or stupidity, but from complex mixtures of truth and falsehood filtered by complex cognitive biases. Don't stop believing that you are right and they are wrong, unless the evidence points that way. But leave it at them being wrong, not them being wrong and stupid and evil.

I think most people intuitively understand this. But considering how many smart people I see shooting their own foot off when they're trying to convince someone3, some of them clearly need a reminder.



1: An excellent collection of the deeper and most subtle forms of this practice of this sort can be found in Dale Carnegie's How to Win Friends and Influence People, one of the only self-help books I've read that was truly useful and not a regurgitation of cliches and applause lights. Carnegie's thesis is basically that being nice is the most powerful of the Dark Arts, and that a master of the Art of Niceness can use it to take over the world. It works better than you'd think.

2: The following technique is definitely one of the Dark Arts, but I mention it because it reveals a lot about the way we think: when engaged in a really heated, angry debate, one where the insults are flying, suddenly stop and admit the other person is one hundred percent right and you're sorry for not realizing it earlier. Do it properly, and the other person will be flabbergasted, and feel deeply guilty at all the names and bad feelings they piled on top of you. Not only will you ruin their whole day, but for the rest of time, this person will secretly feel indebted to you, and you will be able to play with their mind in all sorts of little ways.

3: Libertarians, you have a particular problem with this. If I wanted to know why I'm a Stalin-worshipper who has betrayed the Founding Fathers for personal gain and is controlled by his base emotions and wants to dominate others by force to hide his own worthlessness et cetera, I'd ask Ann Coulter. You're better than that. Come on. And then you wonder why people never vote for you.

New Comment
416 comments, sorted by Click to highlight new comments since: Today at 11:44 AM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

The first rule of persuading a negatively disposed audience - rationally or otherwise - is not to say the things they expect you to say. The expected just gets filtered out, or treated as confirmation of pre-existing beliefs regardless of its content.

Sadly, the unexpected frequently gets translated into the expected, even to the point of explicit denials of a position being ignored repeatedly in a single conversation.

Then say something more unexpected. There's an art to it.

How hard is it to fit a simple denial into a frame? Not hard at all.

This situation is especially troublesome if you've already been market as standard opponent #445, such as when you change your own mind you a new, more subtle position that is similar to your old position.
When I adjust my position in subtle ways like that, I go to pains to point it out, which is rhetorically advantageous and shows that there's a real dialectic going on.
Also, by following their arguments, trying to clarify it and understanding the pieces. Your sincere and genuine attempt to understand them in the best possible light will make them open to your point of view. The smart Christians are some of the most logical people I've ever met. There worldview fits together like a kind of Geometry. They know that you get a completely different form of it if you substitute one axiom for another (existence of God for non-existence of God), much like Euclid's world dissolves without the parallel postulate. Once we got to that point in our conversation, I realized that they we agreed on everything about the world except that postulate, which they were also aware of. I realized that they were neither stupid nor evil, as I had assumed before (a remarkably common, and uncivil, view that atheists have of believers). I still disagree with them. However, I was fine with leaving the conversation with both of our positions unchanged, but understanding each other better.

You've overlooked another way to "win" an argument: To persuade otherwise uninvolved third parties.

Typically, two people arguing are already thoroughly fortified in their opinions. Few people find argument for its own sake enjoyable and thus are unlikely to be lured into a debate they have no emotional stake in; as well, upon rising to the occasion to defend their side, their resistance to acknowledging their opponent's valid arguments will be stronger than ever.

Less-involved bystanders, however, can view the argument with a more impartial eye, and are much more likely to be persuaded. Of course, this is typically the justification made for the style of debating you argue against in this post--especially on the internet, where bystanders are plentiful and social dynamics are strongly subject to John Gabriel's G.I.F. Theory--but it's not at all clear that such an approach is actually effective for this purpose, any more than it is for persuading the opposing party.

As an aside, I can think of at least two other reasons to engage in debate; but neither derives value from actually winning the argument, and thus are irrelevant in this context.


Bystanders may well identify themselves emotionally with one debater or the other, so being "nice" to one's opponent would reduce the defensiveness of the audience as well.

More importantly it helps also the bystanders, be they on your side or the other, come closer to understanding each other. Rather than just converting the undecided among them, which there may not be that many of on some issues.

The science of winning arguments is called Rhetoric, and it is one of the Dark Arts. Its study is forbidden to rationalists, and its tomes and treatises are kept under lock and key in a particularly dark corner of the Miskatonic University library. More than this it is not lawful to speak.

Huh? This can't be the consensus view here. Is it?

Because my opinion has developed over the years to conclude the exact opposite. Rhetoric has always been "the study of how to use language well." What has happened? Wikipedia defines it as "the art and study of the use of language with persuasive effect". Ah, that happened to it. Anyone interested I invite to compare the definitions provided by various dictionaries. Some dictionaries will offer both kinds of definitions, because lexicographers aren't supposed to decide which definition is best.

But I implore others, especially those with a rational bent, to not give up the decent meaning of the term rhetoric. It isn't just "flowery language" that some allege can only be used to obscure thought. It's a whole study, what was once a whole discipline(1). It's the art of how to use language well.

I used to th... (read more)


This post goes hand in hand with Crisis of Faith. Eliezer's post is all about creating an internal crisis and your post is all about applying that to a real world debate. Like peanut-butter and jelly.

If you want to correct and not just refute then you cannot bring to the table evidence that can only be seen as evidence from your perspective. Ie. you cannot directly use evolution as evidence when the opposing party has no working knowledge of evolution. Likewise, a christian cannot convince an atheist of the existence of God by talking about the wonders of His creation. If you picture you and your opponent's belief systems as vin-diagrams then the discussion must start where they overlap, no matter how small that sliver of common knowledge might be. Hopefully, if you and your opponent employ crisis-of-faith properly, those two circles will slowly converge.

For our European readers, I would like to note that what kurige meant by 'Like peanut-butter and jelly' was something like 'they go really well together, and in fact one would probably not put one on a sandwich without the other'. Just try not to picture it; you'll be fine.

Probably most people know this, but if you find yourself needing to mention this again it's vital to add that "jelly" here refers to what we call "jam", because to us, "jelly" is what you call "Jell-O". You can imagine why we're not thrilled by the thought of a peanut-butter and "Jell-O" sandwich!

[insert discussion about the difference between jam and marmalade here]
This is a critical point. One of the reasons arguments seem to exist at all - from what I can understand - is that when people look at the same things in different ways, effectively seeing two different things. A christian might look at the world and see the wonder of God's creation, but a physicist might see nothing but billions and billions of tiny particles interacting. Someone pro-life might see an abortion as a murder, while someone pro-choice might see it as part of a woman's right to her own body. You need to frame the argument so both parties are looking at the same thing for any progress to be made. Otherwise, people just become more and more entrenched in their position, while getting more and more frustrated that the other person doesn't see it their way.
When Wittgenstein wrote his Tractatus, this was basically his only point. If you clearly define your terms, thereby unambiguously fixing the referents for your propositions, then all disagreement will disappear. In later works, he realized that there are a lot of things we do with language other than relating propositions, that you use language before you get definitions, and that things are generally a bit more complicated than he used to think.
4Paul Crowley15y
Of course, in practice this process of definition is more like iterative refining than fixing for all time, but the result is the same: the point is to ensure that your discussion is actually about the world. This is what making beliefs pay rent and tabooing words are all about. (Hmm, we're developing a vocabulary drawn from EY posts - are there more standard terms for these things we could be using?)
that you use language before you get definitions Ah, but that's simply untrue. We use language before we get explicit definitions. The implicit definitions are a necessary precursor for language use.
Maybe. But the point is that implicit definitions are never clearly defined. Indeed, they are hardly definitions- more like an incomplete sense of in what circumstances the use of the word is appropriate.

Yvain, I enjoy your posts, and generally find them useful, informative, and well written.

I also recognize that this view is controversial in some circles, but one thing that would make me enjoy them rather more is if you managed to ferret out the implicit assumption that crops up every now and then that your rationalist protagonists are necessarily male. (Or at least predominantly so, I haven't been back to do an exhaustive stock-take of your gender specific pronoun usage, but I do recall being struck by this at least once before, so I figured it was worth a comment this time.)

Just to clarify, I don't mean Theo here. If you want to use a specifically male example, that's fine. But phrases like "the most important reason to argue with someone is to change his mind" and "[e]ither a person has enough of the rationalist virtues to overcome it, or he doesn't" strike me as problematic.

I'm not for a moment suggesting that you're being consciously sexist here. In fitting with the theme of this post, I spent a fair while rejecting others' calls for gender neutral language under the mistaken (largely emotional) impression that agreeing with them would have be an admission of some deep moral flaw in me, rather than merely a small and relatively painless step towards inclusiveness - and ultimately better communication.

I'm glad you brought that up.

I've thought about this a few times, and I agree with you that it promotes sexism and is bad, but I just really hate using the phrases "he or she" every time I have to use a pronoun. A sentence like "A rationalist should ensure he or she justifies his or her opinion to himself or herself" is just too awkward to understand. And I am too much of a grammar purist to use "them" as a singular.

I used to use the gender-neutral pronoun "ze", but people told me they didn't understand it or didn't like it or thought it sounded stupid. And I tried using "she" as the default for a while, but people kept getting confused because they weren't expecting it, and trying to figure out where I'd mentioned a female.

I'm willing to accept whatever the common consensus is here. Maybe Less Wrong-ers are open-minded enough to accept "ze" where the average reader isn't.

(I've heard some people here use "ve" a few times, but from the context I gathered it was more of a way to refer to aliens/AIs/transhumans than a normal gender-neutral pronoun. Is this true?)

I think I remember reading that the plural used to be conventional grammar and was then deliberately suppressed in favor of "he".

I use the plural. It grows on you surprisingly quickly and isn't at all obtrusive. Anyone who doesn't already have the info stored "Oh, Eliezer uses the plural" after reading my writings for months is a case in point thereof.

Use of the plural form also has the advantage of being the stylistic direction the language is trending to. English is a mass hallucination anyways, why stand in futile defiance of its whims?

The grammatical value of "they" used as a singular has been discussed frequently at the inestimable Language Log, including citations of the form used by such disreputable, notorious abusers of the noble English tongue as William Shakespeare and Winston Churchill. A good post on the subject, though by no means the only one, can be found here.

Maybe next time, we can all argue over splitting infinitives, ending sentences with prepositions, or other happy chestnuts of wholly-unfounded prescriptive grammar pedant absurdity.

Indeed! I pay attention both to gender pronouns and to Eliezer's writing patterns, and I never noticed this. (Eliezer_2000 used "ve" a lot though.) I had previously decided on "he" in order to optimize for flow, but I am happy to accept this well-made point and switch to "they'.

Hofstadter has made an excellent argument on this topic called "A Person Paper on Purity in Language".

Iain Banks, in "Player of Games", also expressed this sentiment pretty well:

"Marain, the Culture's quintessentially wonderful language (so the Culture will tell you), has, as any schoolkid knows, one personal pronoun to cover females, males, in-betweens, neuters, children, drones, Minds, other sentient machines, and every life-form capable of scraping together anything remotely resembling a nervous system and the rudiments of language (or a good excuse for not having either). Naturally, there are ways of specifying a person's sex in Marain, but they're not used in everyday conversation; in the archetypal language-as-moral-weapon-and-proud-of-it, the message is that it's brains that matter, kids; gonads are hardly worth making a distinction over."

My preferred sex-neutral pronoun is "they".

3Paul Crowley15y
Yes, I also prefer "they".
It's unobtrusive and it has a decent chance of actually catching on, unlike any alternative I've ever heard of. There's something to be said for practicality.
Thanks for the thoughtful response. I do appreciate that this inevitably opens up a can of worms, and that no solution is really ideal. I agree that "he or she" awkward in many, if not most situations. For whatever it's worth, my preferred solution is to use the plural (they/them/their) in any situation where it's unambiguous enough to function effectively, and to otherwise use she/her. If people are confused by feminine pronouns... well, that kind of just illustrates the problem, and making them think about that at least serves some sort of purpose.
I'm using this disclaimer from now on. Nearly as hilarious as it is awesome.
Instead of gender neutrality, try to go for gender balance? I use alternate "he" or "she", and occasionally, semi-intentionally contradict myself [for example, in a talk about Bayes, I explained what I meant by "overconfidence" with an example -- the specific numeric example used a name, Sally, and the general definition used "he". For underconfidence, I used "Barry" and "she" respectively". I believe Eliezer used to physically flip a coin for he vs. she.
6Eliezer Yudkowsky13y
Eliezer still does.
I guess that's fairer than switching (there might be an unfair on/off pattern), but would take me out of my writing flow, which is why the strict-alternation compromise is what I adopted.
Idea: every once in a while just flip a coin or otherwise generate a bunch of random bits. Save them and load up the file or get out the piece of paper you wrote the results down on or such when you're ready to start writing. Then simply start peeling the bits off each time you need a new randomly assigned gender.
That doesn't fix the "flow" issue. When I'm writing, the last thing I want to do is to be flipping through my files, looking for the bit file, etc. etc...
Could you use whether the minute on your clock is odd or even?
It still means I need to break my typing to look at an external stimulus. Honestly, so far I've not seen many instances where strict alternation worked badly, so I'm not motivated to solve this non-problem.
It would be better to flip a coin at the beginning of a document to determine which pronoun to use when the gender is unspecified. That way there is no potential for the reader to be confused by two different pronouns referring to the same abstract entity.
Or we could flip a coin once for all of the English-speaking world, so that we aren't confused when we go from one document to another. Or we could just standardize on the male pronoun, which has backward-compatibility advantages.
Why, on your view, would going from using "he" to refer to a hypothetical person in one document, to using "she" to refer to a different hypothetical person in a different document, be confusing? (Not, mind you, that I intend to do this. I've been using the gender-neutral third-person plural pronoun consistently in these situations for years and see no reason to stop. )
Hypothetical people, or people of unknown gender, have no gender in reality I can refer to. If I have to treat them as gendered anyway, surely it is easier to have a default gender to fall back on, rather than having to keep track of the particular nonce gender of this particular hypothetical person/person of unknown gender.
Interesting. For my part, if I'm being told a story about an actual person, whom I don't know, who is referred to as "him" or "her," I don't find it especially confusing to subsequently keep track of their gender. Nor do I find it significantly more confusing if they are hypothetical instead of actual. I hadn't previously realized there were people who differed from me in that regard. That's useful to know: thanks for clarifying.
I'd be very curious to see a study seeing if this did actually impact what gender people think of examples by default. Note that there have been studies showing that kids are more likely to think of a "fireman" as male than a "firefighter" and for similar roles, but I'm not aware of any such study for pronouns. I suspect you'd have the same result.
I'm not a statistically significant study, but given "The agent's husband stood up from the table," I would expect pretty much everyone to assume without much effort that the agent was female, but given "The agent led his husband onto the dance floor," I'd expect most people to become confused, and some to assume a gay male agent, and very few to assume a female agent. That suggests that the "his" gets treated as evidence of the referent's masculinity strong enough to override a strong prior in the other direction.
My predisposition to assume that an agent is male is stronger than my predisposition to assume heteronormative relationships. My immediate reaction to the sentence, "The agent's husband stood up from the table" was to suppose a male agent with a male spouse. But I'm probably unusual in this regard.
I agree with your analysis but I'd like to see some form of formal study confirm it.
I mean, if you go "I am about to write, so I'll load up the random male/female file right now" (though I admit, I haven't tried this and it may very well also be disruptive to quickly tab to that file, check the next random gender and then delete it). Oh well, if that doesn't work, then... next idea then. (I don't have the "next idea", though, so you or someone else will have to come up with it. :))
Couldn't you just default to "he" when writing, then when finished, flip a coin (or refer to whatever randomized gender generator you prefer), and go back and change the gender if need be? It wouldn't interrupt the work flow; it would just be a little work after to revise.
For myself, I use the generic feminine wherever possible in writing, but that's just me. In natural speech, I use they, like everyone else.
Sometimes I flip a coin for each hypothetical person I invoke.
Regarding being a grammar purist, it should be noted that being offended at using 'them' as a singular indefinite is a relatively recent trend. 'ze' and 've' are aesthetically unpleasing, but using them more is likely the only way they would become less so. You won't find me doing it anytime soon though. It should be noted that until recently, 'man' was gender-neutral in English. John Stuart Mill found himself just on the cusp of that, and tried to argue for women's suffrage in England on the basis that the law referred to 'man' and so included women. (he lost). Common archaic equivalents to todays' 'man' and 'woman' are 'were' and 'wif', where 'man' meant the whole species (though commonly that only considered males). 'She' isn't that confusing, and radical feminism isn't the pernicious beast it was in the 90's, so it seems like 'she' is the best bet for a gender-neutral personal pronoun. Personally, I prefer to invent a subject for such a thought experiment and then use the appropriate pronoun for the person's gender - which is what you did here with Theo.
The problem with inventing a subject is that people may notice a (unintentional or even nonexistent) trend to always cast one gender as the brave, smart, rational protagonist and the other gender as the cowardly, stupid, silly antagonist. Personally, I don't care what technique is used (fictional subject, always "he", always "she", "he or she", "they", invented pronouns, etc.)
2Paul Crowley15y
Flip a coin?
A lot of people object to "he or she" on grounds of euphony; but clarity of meaning should always take priority in our considerations over sound. The fact is that "he or she" is what we actually mean. Granted, like any phrase, it is inelegant in certain contexts, and can become tiresome if repeated. So one has to use workarounds. Luckily, "they" (always perfectly acceptable in spoken conversation) is also available for judicious written use. "the most important reason to argue with someone is to change his or her mind" sounds just fine. ("Their" could also be substituted.) "Either a person has enough of the rationalist virtues to overcome it, or he or she doesn't" is bad, mainly because of the "or" preceding "he or she". "He/she doesn't" is better, but "they don't" is probably the best (certainly in a comment; maybe a post should be more formal?). Invented pronouns are just too strange and should be avoided.
Agreed. Sound is deeply important though. Most of us on a forum like this spend our days navigating seas of words. To give no consideration to the sound of those words is exceptionally bad fun theory.
Still sexist, for the reason "whe or ble" is still racist. Also, down with the gender binary. Do we actually mean that we should argue with men and women to change their minds, but not with genderqueers?
FYI: I use this Chrome extension to gender-neutralize what I read:
Yes - and this is the problem. People shouldn't think that a female pronoun is weird... just because it's female. ...and you shouldn't be afraid of using it just because people might think it unusual and get confused for all of two seconds. If you, and the other more post-prolific and respected members of the community used female pronouns more frequently (ie on average: as often as male ones) then eventually it would become commonplace and people would eventually figure it out.... that it's just a pronoun. Just like the other one... only female. Alternatively, a lot of people these days are just fine with "they"/"them". Yes we have twisted English into a way it never was used before and it sounds weird to those of us "brought up better"... but this is what happens with languages. Especially English. I'm sure I could go through your last post and pick out half a dozen things that, in their time, were considered weird and "not correct English"... until everyone that used to hate them died off and it became just part of common language. AFAICT, it's currently the most widely-accepted gender-neutral pronoun. You can fight the tide... or not. :)
Please stick with "he". I agree that it's imperfect, but inelegance matters.

If inelegance is your primary concern, then "she" seems at least as good, and probably a lesser evil for other reasons.

I find using she exclusively offensive.
I find it very hard to consider that anything but nitpicking. Although that's probably because my native language is Finnish and it doesn't have separate third person pronouns for different genders. I don't think that distinction is worth making. Then again, since English does have he and she, perhaps one can't avoid it.
I agree. And after studying Japanese, I started to find it silly that English (like most Western languages) makes the distinction between 'singular' and 'plural'. Like whether we're talking about exactly 1 thing or any number other than 1 is information important enough to encode with every noun, but it's usually not worth mentioning what the particular number is. ETA: exactly what Nebu said below.
I feel like mentioning that English seems to be quite tolerant of not making the singular/plural distinction. When borrowing from languages that don't make this distinction (in my experience, Japanese and Lojban), it seems that people simply use the existing form for both singular and plural: "This gismu is different from all other gismu in that instead of taking just one sumti or finitely many sumti, it can take infinitely many sumti."
Doesn't even have to be non-english words: "this sheep is different from other sheep in that it thinks that it is a fish unlike these fish that think they are sheep" /contrived_example
In everyday life, the difference between one and several often is important enough to mention, but it would be too complex to create special grammatical categories for individual numbers. I'm amazed that ancient people put enough emphasis on past/present/future to justify having irregular verbs. They must have had a very strange conception of time. But then I'm also amazed that Russian doesn't have a definite article...
I think what Thomblake would like (and which is how I understand Japanese to work) is to be to use a noun without specifying whether or not it is plural, and have extra (not necessarily "grammatical categories") contructs for adding the extra information of whether it is plural or not. E.g. * "What did you do yesterday?" * "Oh, I hung out with {friend}." * "Really? Were there a lot of people?" * "Nope, just one {friend}." / "Yes, many {friend}." / "Well, it was three {friend}." So it's not new grammatical categories (as long as you don't consider just prefixing the word "three" to be a new grammatical category). The way English works, there's no way to use a noun while leaving the "1 vs not 1" information ambiguous. If you leave off the "s", you must be referring to exactly one instance. If you put the "s", you must be referring to a non-1 instance (possibly zero instances).

Most of the comments in this discussion focused on topics that are emotionally significant for your "opponent." But here's something that happened to me twice.

I was trying to explain to two intelligent people (separately) that mathematical induction should start with the second step, not the first. In my particular case, a homework assignment had us do induction on the rows of a lower triangular matrix as it was being multiplied by various vectors; the first row only had multiplication, the second row both multiplication and addition. I figured it was safer to start with a more representative row.

When a classmate disagreed with me, I found this example on Wikipedia. His counter-arguement was that this wasn't the case of induction failing at n=2. He argued that the hypothesis was worded incorrectly, akin to the proof that a cat has nine tails. I voiced my agreement with him, that “one horse of one color” is only semantically similar to “two horses of one color,” but are in fact as different as “No cat (1)” and “no cat (2).” I tried to get him to come to this conclusion on his own. Midway through, he caught me and said that I was misinterpreting what he was saying.

The secon... (read more)

You might find enlightening the part of the TED talk given by James Flynn (of the Flynn effect), where he talks about concrete thinking.
Hah! I thought of the exact same thing before I saw your comment: the interviews by Luria with Russian peasants where the peasants refuse to abstract in any way. Shalizi provides an example :
Not to deny that this is an example of someone who does not think abstractly -- I agree that it is -- but there's also a Gricean interpretation: From the subject's point of view, if the experimenter already knows that there is literally not even one camel in all of Germany including all of its cities and villages, then the experimenter would not ask whether there are camels in a specific city in Germany, therefore the experimenter must mean, "There are no camels in Germany not counting the cities", or "There are extremely few camels in Germany".
That doesn't wash given the dialogue, rather than a single question. A single question might reasonably elicit a Gricean answer to a different question, but repeated questioning on the same point?
Mathematical induction using the first step as the base case is valid. The problem with the horses of one color problem is that you are using sloppy verbal reasoning that hides an unjustified assumption that n > 1. If you had tried to make a rigorous argument that the set of n+1 elements is the union of two of its subsets with n elements each, with those subsets having a non-empty intersection, this would be clear.
Induction based on n=1 works sometimes, but not always. That was my point. I'm not sure what you mean. I thought I stated it each time I was assuming n=1 and n=2.
In the induction step, we reason "The first horse is the same colour as the horses in the middle, and the horses in the middle have the same colour as the last horse. Therefore, all n+1 horses must be of the same colour". This reasoning only works if n > 1, because if n = 1, then there are no "horses in the middle", and so "the first horse is the same colour as the horses in the middle" is not true.
The problem was that your ultimate conclusion was wrong. It is not in fact the case that "mathematical induction should start with the second step, not the first." It's just that, like all proofs, you have to draw valid inferences at each step. As JGWeissman points out, the horse proof fails at the n=2 step. But one could contrive examples in which the induction proof fails at the kth step for arbitrary k.
I don't think I ever got to my "ultimate" conclusion (that all of the operations that appear in step n must appear in the basis step). I was trying to use this example where the proof failed at n=2 to show that it's possible in principle for a (specific other) proof to fail at n=2. Higher-order basis steps would be necessary only if there were even more operations.
In The Society of Mind, Marvin Minsky writes about "Intellectual Trauma": This seems to fit the anecdote very well--your interlocutor could not find a fault in the reasoning, noticed it led to an absurdity, and decided that this intellectual area is dangerous, scary, and should be evacuated as soon as possible.
You didn't actually prove that n+1 horses have one color with this, you know, even given the assumption. You just said twice that n horses have one color, without proving that their combined set still has one color. For example consider the following "Suppose every n horses can fit in my living room. Add the n+1 horse, and take n out of those horses. They can fit in my living room by assumption. Remove 1 horse and take the one that’s been left out. You again have n horses, so they must again fit in my living room. Therefore, all horses fit in my living room." That's not proper induction. It doesn't matter if you begin with a n of 1, 2, 5, or 100 horses, such an attempt at induction would still be wrong, because it never shows that the proposition actually applies for the set of n+1.
.... The first n horses and the second n horses have an overlap of n-1 horses that are all the same color. So first and the last horse have to be the same color. Sorry, I thought that was obvious. I see your point, though. This time, I was trying to reduce the word count because the audience is clearly intelligent enough to make that leap of logic. I can say the same for both of my "opponents" described above, because both of them are well above average intellectually. I honestly don't remember if I took that extra step in real life. If I haven't, do you think that was the issue both people had with my proof? I have a feeling that the second person's problem with it was not from nitpicking on the details, though. I feel like something else made him angry.
You need to make this more explicit, to expose the hidden assumption: Take a horse from the overlap, which is the same color as the first horse and the same color as the last horse, so by transitivity, the first and last horse are the same color. But why can you take a horse from the overlap? You can if the overlap is non-empty. Is the overlap non-empty? It has n-1 horses, so it is non-empty if n-1 > 0. Is n-1 > 0? It is if n > 1. Is n > 1? No, we want the proof to cover the case where n=1.
That's exactly what I was trying to get them to understand. Do you think that they couldn't, and that's why they started arguing with me on irrelevant grounds?

And the point that I am trying to get you to understand, is that you do not need special rule to always check P(2) when making a proof by induction, in this case where the induction fails at P(1) -> P(2), carefully trying to prove the induction step will cause you to realize this. More generally you cannot rigorously prove that for all integers n > 0, P(n) -> P(n+1) if it is not true, and in particular if P(1) does not imply P(2).

Sorry, I can't figure out what you mean here. Of course you can't rigorously prove something that's not true. I have a feeling that our conversation boils down to the following: Me: There exists a case where induction fails at n=2. You: For all cases, if induction doesn’t fail at n=2, doesn’t mean induction doesn’t fail. Conversely, if induction fails, it doesn’t mean it fails at n=2. You have to carefully look at why and where it fails instead of defaulting to “it works at n=2, therefore it works.” Is that correct, or am I misinterpreting? Anyways, let's suppose you're making a valid point. Do you think that my interlocutors were arguing this very point? Or do you think they were arguing to put me back in my place, like TheOtherDave suggests, or that there was a similar human issue that had nothing to do with the actual argument?
To butt in, I doubt your interlocutors were attempting to argue this point; they seem like they were having more fundamental issues. But your original argument does seem to be a bit confused. Induction fails here because the inductive step fails at n=2. The inductive step happens to be true for n>2, but it is not true in general, hence the induction is invalid. The point is, rather than "you have to check n=2" or something similar, all that's going on here is that you have to check that your inductive step is actually valid. Which here means checking that you didn't sneak in any assumptions about n being sufficiently large. What's missing is not additional parts to the induction beyond base case and inductive step, what's missing is part of the proof of the inductive step.
Your hindsight is accurate, but more than just recognizing the claim as true when presented to you, I am trying to get you to take it seriously and actively make use of it, by trying to rigorously prove things rather than produce sloppy verbal arguments that feel like a proof, which is possible to do for things that aren't true. This is accurate, and related, but not the entire point. Distinguish between a proof by mathematical induction and the process of attempting to produce a proof by mathematical induction. One possible result of attempting to produce a proof is a proof. Another possible result is the identification of some difficulty in the proof that is the basis of an insight that induction isn't the right approach or, as in the colored horses examples, that the thing you are trying to prove is not actually true. The point is that if you are properly attempting to produce a proof, which includes noticing difficulties that imply that the claim you are trying to prove is not actually true, you will either produce a valid proof or identify why your approach fails to provide a proof. No, your interlocutors were not arguing this point. Their performance, as reported by you, was horribly irrational. But you should apply as much scrutiny to your own beliefs and arguments as to your interlocutors.
The case of two horses is special here because the sets 1..n and 2..n+1 don't overlap if n+1 = 2, and not because of some fundamental property of every induction hypothesis, but that -- along with some arbitrary large n, and maybe the next case if I'm using any parity tricks -- is one of the first cases I'd check when verifying a proof by induction.
The case of P(n) -> P(n+1) (i.e., the second part of the induction argument) that fails is n=1. (In other words n+1 = 2). The second part of the induction argument must begin (i.e., include n >= n0) at the value n0 that you have proven in the first part to be true from 1 to n0. In this case n0 = 1, so you must begin the induction at n = 1.
I have edited my comment to avoid this confusion.
You're right, of course. I was trying to describe the flaw in the set-overlap assumption without actually going through an inductive step, on the assumption that that would be clearer, but in retrospect my phrasing muddled that. I'll see if I can fix that.
Why didn't you drop the "horses" example when it tripped him up and go with, I dunno, emeralds or ceramic pie weights or spruckels, stipulated to in fact have uniform color?
I suspect that I lost the second person way before horses even became an issue. When he started picking on my words, "horses" and "different world" and "hypothetical person" didn't really matter anymore. He was just angry. What he was saying didn't make sense from that point on. For whatever reason, he stopped responding to logic. But I don't know what I said to make him this angry in the first place.

Leaving aside the actual argument, I can tell you that there exist people (my husband is one of them, and come to think of it so is my ex-girlfriend, which makes me suspect that I bear some responsibility here, but I digress) whose immediate emotional reaction to "here, let me walk you through this illustrative hypothetical case" is strongly negative.

The reasons given vary, and may well be confabulatory.

I've heard the position summarized as "I don't believe in hypothetical questions," which I mostly unpack to mean that they understand that hypothetical scenarios are often used to introduce assumptions which support conclusions that the speaker then tries to apply by analogy to the real world, and that a clever rhetoritician can use this technique to sneak illegitimate assumptions into real-world scenarios, and don't trust me not to sneak in assumptions that make them look stupid or manipulate them into acting against their own interests.

I don't know if that's a factor in your case or not, but I have found that once I trigger that reaction, there's not much more I can do... they are no longer cooperating in the communication, they are just looking for some way t... (read more)

You know, come think of it, that's actually a very good description of the second person... who is, by the way, my dad. This hasn't ever occurred to me, but I'll try it the next time a similar situation arises.
"No. Just an example. Lies propagate, that's what I'm saying. You've got to tell more lies to cover them up, lie about every fact that's connected to the first lie. And if you kept on lying, and you kept on trying to cover it up, sooner or later you'd even have to start lying about the general laws of thought. Like, someone is selling you some kind of alternative medicine that doesn't work, and any double-blind experimental study will confirm that it doesn't work. So if someone wants to go on defending the lie, they've got to get you to disbelieve in the experimental method. Like, the experimental method is just for merely scientific kinds of medicine, not amazing alternative medicine like theirs. Or a good and virtuous person should believe as strongly as they can, no matter what the evidence says. Or truth doesn't exist and there's no such thing as objective reality. A lot of common wisdom like that isn't just mistaken, it's anti-epistemology, it's systematically wrong. Every rule of rationality that tells you how to find the truth, there's someone out there who needs you to believe the opposite. If you once tell a lie, the truth is ever after your enemy; and there's a lot of people out there telling lies."

I am curious about the large emphasis that rationalists place on the religious belief. Religion is an old institution, ingrained in culture and valuable for aesthetic and social reasons. To convince a believer to leave his religion, you need not only convince him, but convince him so thoroughly as to drive him to take a substantial drop in personal utility to come to your side (to be more exact, he must weigh the utility gained from believing the truth to outweigh the material, social, and psychic benefits that he gets from religion).

For rationalists' attention, there are myriad more important and relevant issues where human irrationality has an effect on the world. In addition, these issues are normally easier to change people's beliefs about.

People have been believing in God for 500,000 years. People have been believing unsupported things about Global Warming for 30. I would rather teach people how to be skeptical and cautious about modern policy debates than have Yet Another God Conversation.

I was scarred by religion growing up. I understand the impulse to despise it and oppose it. But there came a time in my life when I realized that it was going to be around for as long as humanity, though its fortunes may wax and wane. It's time to move on.

American Rhetoric is an incredible site and there are some real gems that aspire to rational persuasion with some flair.

Malcolm X's "Ballot or the Bullet" navigates the fact that he is black, widely regarded as dangerous and Muslim all at once while urging people to put these things aside and think about his plans and their outcomes. He does a first rate job of tailoring his rhetoric to increase your emotional desire to think and not react.

Milton Friedman is another person to watch live. He ge... (read more)

Sometimes the harsh approach has surprisingly good results. Example.

But Christopher Hitchens always struck me as too black-and-white and just plain irritating

Tangential, I know, but this surprises me. Hitchens, with his literary background, strikes me as a very nuanced thinker, attuned to the various shades of gray. (For example, he's by no means unmoved by religion's contributions to art and culture.) Maybe you're thinking of his talent for devastating rhetorical flourish, as in his infamous comments on Jerry Falwell?

8Scott Alexander15y
I accept your correction. I've only seen Hitchens on TV a few times and never read his book. My introduction to him was in fact his Jerry Falwell comments. If you say he's more nuanced in his writings, I believe you.
4Paul Crowley15y
I've read God is not Great and seen footage of him in debate, and while I admire him in many ways, in other ways think he's a total arse and an embarrassment, and I don't think Yvain's picture of him is all that unfair. There isn't anyone prominent who thinks that the Sistine Chapel is bad because it's religious. (I should add that there's almost nothing he has to say about religion that I actually disagree with, I just wouldn't use his turns of phrase to my religious friends)
You use different rhetoric to energize the base than to sway the undecideds. Hitchens often acts as a cheerleader. All of the big four do from time to time -- well, maybe not Dennett.
1Paul Crowley15y
Hitch has the strongest line in work-the-base rhetoric and so provides a good example of the sort of rhetoric you shouldn't use when trying to sway the undecideds, which was Yvain's point.

This post was actually pretty enlightening. I've had the typical religious debate with a theist before, and I use to go ahead with the 'kill them with arguments' method, and I did notice that it left people more convinced of their beliefs than they were before, even if I 'won' the argument.

In Rhetoric, they call this technique "Concession".

My goal going into arguments is not to crush them or convince them that I am right. I try to keep a more open mind and understand their arguments. If you start out with a goal of crushing them you won't be in a state of mind to admit if their arguments are stronger.

My sense is that the most rational argument is the one that gets you closest to your true goal. The theme of my most persuasive arguments is usually something like, "Once we sort things out I bet we'll see that we're really in substantial agreement." Dale Carnagiesque or not, to me rhetoric doesn't have to be a "dark art," full of manipulation and cant. Instead, I try to give respect and keep in mind what I really need to take away from the transaction. I also conceptualize what my erstwhile opponent needs and why. This way, minds may yet meet and all parties have a line of social retreat.

But the most important reason to argue with someone is to change his mind.

It feels so quaint to see the generic "he" in an Yvain post. (I think he later used novelty gender-neutral pronouns for a while and now uses the singular "they".)

This might have been the last Yvain post to do so, because of this comment further down the page. Looking at the next ~5 of them in chronological order, I couldn't find any generic pronouns, so I couldn't easily test it.

The inventors of the original form of rationalist virtue AND rhetoric sure didn't think that the latter was a dark art. Rationalists should WIN!

Rationalists should shouldn't deny themselves the utility of rhetoric. Any rational rationalist can see that rhetoric is the path to winning, a kind of social theatre that lubricates decision-making with irrational or intermittently rational groups. If a group needs to be convinced of a position within a finite amount of time, bare reasoning isn't always the best option.

Maybe that is too Machiavellian to be "really" rational, but it is the winning path.

I think I am using "rhetoric" in a different way than Aristotle. For Aristotle, it was the art of speaking clearly and eloquently to communicate a position. I am using it more in the way people use when they say "empty rhetoric" or "political rhetoric". "Unless you give up your rights, the terrorists have already won" is my idea of an archetypal rhetorical technique. That may not be fair to the field of rhetoric, but I need some word to describe it and I can't think of a better one, so "rhetoric" it is.

Rhetoric is a technique that may be useful to rationalists, but it's not a rationalist technique. Compare the use of force. I may, as a rationalist, decide the best way towards my goal is murdering all who oppose me, in which case I'll want to know techniques like how to use an assault weapon. But there's still something fundamentally shady about the technique of killing people; it may just barely be justified on utilitarian grounds for a sufficiently important goal, but it's one of those things that you use only as a last resort and even then only after agonizing soul-searching. I feel confident saying that the technique of murdering... (read more)


Be careful about using the "rationalists should win" slogan too literally. Martial artists should win too, but that doesn't mean they should take an AK-47 to their next sparring match and blowing their opponent's face off. Martial artists place high value on winning honorably. I see no reason why we shouldn't emulate them.

I disagree. The problem with using dishonest rethoric to win in a debate isn't that it's winning dishonorably; it's that it's winning at the wrong game - on a game that you wouldn't consider the most important if you looked at it closely.

To continue with the martial arts analogy, imagine say a Chinese kung fu master in World War 2 Nanjing that knows that Japanese soldiers are coming over to kill off all of his family. Should he try to win the fight honorably? Or just try to win using every dirty trick in the book (including running away)? If he focuses on winning honorably, he's lost sight of his main goal (save his family) in favor of a secondary one (win honorably).

Similarly, if you foxus on "winning the debate", and as a result push people into a corner that will make them dislike you and become more attached to their identity as a believer in whatever - you focused on the wrong subgoal, and lost at the one which was important to you.

6Scott Alexander15y
I'm a precedent utilitarian. I try to maximize utility, except when doing so would set a bad precedent that would lower utility later. Precedent utilitarians are usually good about restraining from force. Yes, killing a rich miser and distributing her money to the poor might increase utility. But it sets the precedent that anyone can kill someone if they think of a good enough reason, and most people won't be smart enough to limit themselves to genuinely good reasons. Therefore, precedent utilitarians generally respect the rule of not killing others. But in certain cases this rule breaks down. In the WWII example you mention, it doesn't seem particularly dangerous to set the precedent that you can use force against invaders coming to kill your family. I try to use the same thought process when evaluating when to use rhetoric. If anyone can use rhetoric any time it furthers a goal that they consider genuinely good, then there's little incentive to use rational argument except on the rare hard-core rationalists who are mostly resistant to rhetorical tricks. I want to be able to condemn a demagogue who uses rhetoric without being a hypocrite. If I needed to use rhetoric in a situation where I couldn't blame anyone else for using rhetoric, like trying to save my family, I'd do it. (the problem with precedent utilitarianism is that the calculations are impossible to do with real math, and mostly just involve handwaving. But I hope it at least gives a sketch of my thought processes)
Yvain: "I'm a precedent utilitarian. I try to maximize utility, except when doing so would set a bad precedent that would lower utility later." I think this is an odd thing to say. Any utilitarian ought to be declining short-term gains that result in long-term losses. So why the need for this specific disclaimer?
Yvain seems to be using the term to mean a utilitarian (in the pure sense) who scrupulously considers the force of his example. The implication is that many don't - we're not talking about perfectly rational beings here, just people who agree with the principle of utility maximization.
Um, isn't it kind of rhetorical to compare rhetoric to force and murder? Also, all your articles here that I recall -- likewise those of Eliezer on Overcoming Bias -- are masterful applications of rhetoric. So I'm kind of confused here. Is this one of those "do as I say, not as I do" things?
5Scott Alexander15y
If you mean the articles here are clear or well argued, thank you. I have no objection to clarity or good argument; see the first paragraph of the comment above. If you mean that I'm using dirty tricks like the "terrorists win" example, then I'd like to know exactly what you mean so I can avoid doing it in the future. When I compare rhetoric (meaning "empty rhetoric", as mentioned) to force and murder, I'm not saying they're equally bad, or doing one leads to the other or anything like that. Just that they're bad for the same reason. Both are potentially "useful" techniques. But both prevent rational argument and if used too frequently lead to a world in which rational argument is impossible.
But that is precisely the sort of "dirty trick" you claim to be against. By using murder as an example, you're setting off a "boo light" (opposite of applause light) and linking it to the thing you want people to dislike. That's rhetoric, and emotional manipulation. And it's neither a good thing nor a bad thing, in itself. Used to strengthen a valid argument, it's fine. Arguing that it's bad in and of itself is a misunderstanding... and another "boo light" (e.g. "empty rhetoric", "dirty tricks"). Emotional manipulation is unavoidable, by the way. Boring presenters and neutral presentations are just manipulating people's emotions either towards boredom and not caring, or to "respect", "status", and "seriousness", depending on the audience. It's best to deliberately choose what emotions you want to create, in whom, rather than leaving the matter to chance.
I think the point is that you do a little of both; loosely speaking you are guilty of being fairly eloquent--presenting your ideas persuasively and engagingly, in a style that is inherently likely to increase acceptance. It is an unavoidable facet of human communication that the same idea can be more or less persuasive depending on how it is presented. Over on OB, Robin uses a far more neutral (or at times even anti-persuasive) style, and if memory serves me he and Eliezer have argued a bit about such use of style.
Except, of course, for all those aspects of martial arts which we shouldn't emulate.

We're running up against the equivocation at the core of this community, between rationalists as people who make optimal plays versus rationalists as people who love truth and hate lies.

rationalists as people who make optimal plays versus rationalists as people who love truth and hate lies It's only possible for us to systematically make optimal plays IF we have a sufficient grasp of truth. There's only an equivocation in the minds of people who don't understand that one goal is a necessary precursor for the other.
No, I think there is an equivocation here, though that's probably because of the term "people who love truth and hate lies" instead of "epistemic rationalist". An epistemic rationalist wants to know truth and to eliminate lies from their mind. An instrumental rationalist wants to win, and one precursor to winning is to know truth and to eliminate lies from one's own mind. However, someone who "loves truth and hates lies" doesn't merely want their own mind to filled with truth. They want for all minds in the universe to be filled with truth and for lies to be eliminated from all minds. This can be an impediment to "winning" if there are competing minds.
Rationalists should WIN! Rationalists have better definitions of "winning". They don't necessarily include triumphing in social wrestling matches.
Actually, I think "Rationalists should WIN" regardless of what their goals are, even if that includes social wrestling matches. The "should" here is not intended to be moral prescriptivism. I'm not saying in an morally/ethically ideal world, rationalists would win. Instead, I'm using "should" to help define what the word "Rationalist" means. If some person is a rationalist, then given equal opportunity, resources, difficult-of-goal, etc., they will on average, probabilistically win more often than someone who was not a rationalist. And if they happen to be an evil rationalist, well that sucks for the rest of the universe, but that's still what "rationalist" means. I believe this definitional-sense of "should" is also what the originator of the "Rationalists should WIN" quote intended.
There is a bit of a problem here in that the list of the greatest rationalists ever will be headed by people like Genghis Khan and Prophet Muhammad.
People who win are not necessarily rationalists. A person who is a rationalist is more likely to win than a person who is not. Consider someone who just happens to win the lottery vs someone who figures out what actions have the highest expected net profit. Edit: That said, careful not to succumb to maybe Genghis Khan really was one of the greatest rationalists ever. I've never met the guy nor read any of his writings, so I wouldn't know.
Even ignoring the issue that "rationalist" is not a binary variable, I don't know how in practice will you be able to tell whether someone is a rationalist or not. Your definition depends on counterfactuals and without them you can't disentangle rationalism and luck.
I assume that you accept the claim that it is possible to define what a fair coin is, and thus what an unfair coin is. If we observe some coin, at first, it may be difficult to tell if it's a fair coin or not. Perhaps the coin comes from a very trustworthy friend who assures you that it's fair. Maybe it's specifically being sold in a novelty store and labelled as an "unfair coin" and you've made many purchases from this store in the past and have never been disappointed. In other words, you have some "prior" probability belief that the coin is fair (or not fair). As you see the coin flip, you can keep track of its outcomes, and adjust your belief. You can ask yourself "Given the outcomes I've seen, is it more likely that the coin is fair? or unfair?" and update accordingly. I think the same applies for rationalist here. I meet someone new. Eliezer vouches for her as being very rational. I observe her sometimes winning, sometimes not winning. I expend mental effort and try to judge how easy/difficult her situation was and how much effort/skill/rationality/luck/whatever it would have taken her to win in that situation. I try to analyze how it came about that she won when she won, or lost when she lost. I try to dismiss evidence where luck was a big factor. She bought a lottery ticket, and she won. Should I update towards her being a rationalist or not? She switched doors in Monty Hall, but she ended up with a goat. Should I update towards her being a rationalist or not? Etc.
Hm, OK. So you are saying that the degree of rationalism is an unobservable (hidden) variable and what we can observe (winning or losing) is contaminated by noise (luck). That's a fair way of framing it. The interesting question then becomes what kind of accuracy can you achieve in the real world given that the noise level are high, information available to you is limited, and your perception is imperfect (e.g. it's not uncommon to interpret non-obvious high skill as luck).
Right, I suspect just having heard about someone's accomplishments would be an extremely noisy indicator. You'd want to know what they were thinking, for example by reading their blog posts. Eliezer seems pretty rational, given his writings. But if he repeatedly lost in situations where other people tend to win, I'd update accordingly.
But what about the other case? People who don't seem rational given their writings but who repeatedly win?
Possibly he's just extremely lucky. There are seven billion people in the world - one of these people is almost certain to be luckier than all of the rest. Possibly he is being looked after by a far more competent person behind the scenes; a spouse or a parent, perhaps, who dislikes being visible but works to help that person succeed. Possibly that person really is more rational than you are, but his methods of success are so alien to you that your first instinct is to reject them out-of-hand. Possibly his "writings" are actually being ghost-written by someone else. Possibly he doesn't much care about what he writes, going for low-effort writing in order to concentrate on winning. Possibly he's found one exploit that really works but won't work if everyone does it; thus, he keeps quiet about it. Possibly he's deliberately writing to obscure or hide his own methods of success. Possibly he's found a winning strategy, but he doesn't understand why it works, and thus invents a completely implausible "explanation" for it. ...have I missed anything?
If I understand the Peter Thiel doctrine of the secret correlectly that should be the case in many instances.
Some people are rich and can afford valuable things even if they don't spend their money wisely. Some people might win because they have a lot of resources or connections to throw at problems.
If you define rationality as winning, why does it matter what his writings seem like?
I can't directly observe Eliezer winning or losing, but I can make (perhaps very weak) inferences about how often he wins/loses given his writing. As an analogy, I might not have the opportunity to play a given videogame ABC against a given blogger XYZ that I've never met and will never meet. But if I read his blog posts on ABC strategies, and try to apply them when I play ABC, and find that my win-rate vastly improves, I can infer that XYZ also probably wins often (and probably wins more often than I do).
Well, if what you want to accomplish is motivating large groups of people into supporting you and using them to conquer a large empire, you should study what they did and how they did it.
Now that you mention it, I actually don't.

I didn't end up deconverting him on the spot, but think he left with a much better appreciation of my position.

Indeed. I've had similar thoughts to you, and take care to leave a social line of retreat. In the case of religious people, it doesn't work. In other cases, it works great, though I can't think of any off the top of my head. I just have a general recipie in my head "to convince someone, make it so that they'll look and feel good if they agree with you"

I've had similar thoughts as well. One problem, it seems, is the amount of effort it requires to argue in this way. If you don't know your opponent's position backwards and forwards, it can be very difficult to come up with a line of retreat for them to follow. If you don't actually grok your opponent's position, pretending to do so is unlikely to be effective. Ultimately, it probably pays to specialize a bit in a couple of areas so you know everyone's argument well enough to find those lines of retreat and to help them intuitively understand your argument.

Definitely a great point offered up here and a well-thought out technique that I'd like to try in subsequent debates. Though I think it would be nice to see an example of how leaving a line of retreat would play out. For the example given:

"Let's say you approach a theist (let's call him Theo) and say "How can you, a grown man, still believe in something stupid like talking snakes and magic sky kings? Don't you know you people are responsible for the Crusades and the Thirty Years' War and the Spanish Inquisition? You should be ashamed of yourself!"

This sugg

... (read more)

Cool Article, a nice and useful reminder.

Possibly just a quick question.... I remember some years back there being some sort of post on "Why does it seem like all the rationalists here are male?" and.... all the hypothetical people in your post are guys. In particular "Either a person has enough of the rationalist virtues to overcome it, or he doesn't."

Would it be useful to use more gender neutral pronouns? Especially when trying to give examples of persons who expresses rationalist vitures, assuming them to be Male seems.... like it might inadvertently act to exclude people.

I thought it to be a nice illustration: Dawkins vs. Tyson This is a 2-minute-excerpt of "Beyond Belief", where Tyson accuses Dawkins of "the first type of winning an argument". (But his answer is no more than "You're right. But some people are worse".)

But there's another line of retreat to worry about, one I experienced firsthand in a very strange way. I had a dream once where God came down to Earth; I can't remember exactly why. In the borderlands between waking and sleep, I remember thinking: I feel like a total moron. Here I am, someone who goes to atheist groups and posts on atheist blogs and has told all his friends they should be atheists and so on, and now it turns out God exists. All of my religious friends whom I won all those arguments against are going to be secretly looking at me, trying as

... (read more)

Bad, bad idea. There's no way to avoid losing an argument, because most of the time arguments are social wrestling contests / displays of influence and status.

The only thing you can do is to always make sure you're supporting the right side. That doesn't guarantee that you don't lose, if losing is defined as not coming out as the social victor and failing to convince your opponent or your listeners.

You can't control the responses of others. You can force them to be rational. All you can do is be correct.

As a rationalist, that's all you should really care about anyway. But precious few of us are rationalists.

Surely you don't mean that simpliciter. There are other things that one should care about, perhaps even while wearing a 'rationalist' hat. It seems that being able to argue in a way that supports the truth without alienating one's friends seems like a worthwhile endeavor.

TOMIN And you think I'm crazy for believing in the Ori? VALA Not crazy, Tomin. Just...wrong...

TOMIN There are still so many things about it that mean a great deal to me. VALA I don't doubt that there's morality and wisdom in it. That's what made it such a powerful lure for so many people. I think in principle, the idea of bettering what it's all really about


Well watching people argue, ways to "win an argument" - to give everyone the impression that your points are better - include

a) not listening to the other person b) intimidating them by claiming or implying they are stupid for not agreeing c) making better points and counterarguments

c) usually gets lost in the messy nature of arguments. But you can make the better points, and because of the limits of human knowledge, they still tend to be 'the best guess we have is that X is true', so could still be wrong

In debating technique, I spotted a while ... (read more)

d) Make it look like the other person is saying something the nearest available stupid or objectionable thing to what they actually said. (Alternately, make it seem like they are saying the worst available stupid or objectionable thing within the constraints of what your knowledge of the social context suggests you will be able to get away with.) The above tactic seems to be the go to strategy of practical argument. People - particularly those who consider themselves higher status - do it without thinking about it or trying.

Libertarians, you have a particular problem with this.

You killed our minds. I'm dead now.

[This comment is no longer endorsed by its author]Reply

evolutionary psychology tells us humans are notoriously bad liars

Does it? I genuinely don't know, since I haven't really studied the subject yet, but it strikes me that if lying didn't work, we wouldn't have developed this whole arms race of deception and social games. You might as well say that the Dark Arts in general don't really work, that humans are notoriously bad at them. Yet my own impression is that if you find something to say that's nice or what the other person wants to hear, regardless of whether you mean it, a lot of people will lap it up if you aren't too crude about it. As has been said elsewhere on lesswrong, it takes effort to detect falsehood, whereas acceptance is natural.

Hard to answer. Compared to what exactly? We do know that we have all sorts of microexpressions that reveal our emotional state and that most people give off indications that they are being deceptive. And this isn't just failing to hide dishonestly - there is a whole part of the brain in there actively making muscles move to give indications to others that include telling them we're lying. So whether or not we are good liars I don't know but we are certainly worse liars than we could be if we weren't actively shooting ourselves in the foot!
Bear in mind that the most effective lies are the ones you yourself. So people are very good at convincing themselves of positions which will turn out to be useful if others also believe them. I think it's fair to say that few people are good at conscious lying.
Or we keep plugging away futiley at it because the rewards are so high: cf gambling.

If you believe morality is impossible without God, you have a strong disincentive to become an atheist

That idea is distinct from whether or not God exists. In arguments about evolution I have made the point that it is compatible with both the existence and non-existence of God. I did not say that to "leave a line of retreat" for the Christian anti-Darwinist I was talking to, but because I believed he held an incorrect notion of what Darwinian evolution is. The notion of leaving a line of retreat for others seems less rationalist, for it implies v... (read more)

P(evolution|God) is much lower than P(creationism|God), so even if you are leading them to the right conclusion about evolution, they still aren't really reasoning properly if they still hold their belief in God. In fact, one might argue that they are doing worse. That depends on your motivations. The line of retreat is also useful for improving their reasoning capabilities. Changing one's mind in the face of emotional resistance is hard and takes practice to master.
I'd write that as P(evolution|Creationist God) < P(creationism|Creationist God). One can easily conceive of many variants of God which are not only compatible with evolution, but for which evolution is the most sensible explanation for what we observe around us. E.g. a God which sets the initial conditions of the universe, starts the big bang, and then does not interfere from there on. The other benefit of writing it that way is that it more clearly highlights the tautological nature of the argument for creationism given the existence of (a creationist) God.
however P(evolution | moral God) would be rather low. ie, think about all the nasty stuff that had to happen to, well, give rise to all the currently existent beings. Evolution is a nasty process. A god that is sufficiently intelligent and powerful that it COULD have engineered the species it wanted right from the start, rather than just setting stuff in motion and waiting for something interesting to evolve, allowing all that suffering to happen in the process, well... that would seem to exclude such a being from having anything resembling human morality, right? So when taking into account the often associated "god is good" claim, well, the whole thing completely implodes, no?
I'm not sure you need that. If you're willing to grant an omnipotent being, it seems like spontaneous creation of whatever he wants would be more likely than anything else. If miracles are possible, they're always the simplest explanation for everything. Which might itself provide a methodological reason for denying them.
Well, now we're getting into a discussion about the nature of God. An omnipotent, but non-omniscient God might be compatible with evolution, especially if the universe is large (as it seems to be), because perhaps God, while capable of spontaneously creating stuff, just isn't paying attention to anything going on in our particular neighborhood of the universe. An omnipotent and omniscient, but disinterested God would also be compatible with evolution, as he could, if he wanted to, create new species, but just doesn't bother, as he has other things to worry about. Etc. What about if miracles are possible, but extremely improbable? Which I think exactly describes the universe we are currently in, assuming you are willing to accept "new organisms of a new specie spontaneously coming into existence via random quantum effects" as a possible but extremely improbably miracle.
Nope. by 'Miracle' I mean God goes *poof* and things happen. If you've got an omnipotent, omniscient being with his grubby little paws in everything, then he provides the simplest explanation for any phenomenon.
Not true -- you would still model God as some sort of cognitive entity. Miracles which are parsimonious given the temperament revealed by his previous miracles would be simpler. For example, given a Judeo-Christian God, if you discovered that gay men were living longer happier lives than straight men, this would not be easily explained as a miracle.
Hmm... I think we are talking about the same territory. It's just that in your map, you've labelled the territory as "God" and in my map, I've labelled the territory as "random quantum effects".

"the most important reason to argue with someone is to change his mind."

I don't think I ever try to change anyones mind. If I am involved in a discussion I am only trying to clarify my own thoughts. What other people think is up to them. How can I know, what the other person should think?

The most important reason to argue with someone is that someone should change their mind.
Not sure why Marshall is being voted down here; I agree with him completely. The main reason I "argue" with someone is to seek truth. Perfectly rational agents with the same information should never disagree. So if I disagree with someone, either one (or more) of us is not rational, or one (or more) of us has information that the other one doesn't. If I argue with someone, I am doing them a favor by expending effort to provide them with more information or helping them see their irrationalities. If someone argues with me, they are doing me a favor by expending effort to provide me with more information or helping me to see my irrationalities. When I argue with people who are relatively rational (e.g. most of my friends), this works well. Usually one of us learns something new. When I argue with people who are less rational (e.g. most people in general), this does not work very well, and I run into the problems described by Yvain here.

The psychological resistance I felt to admitting God's existence, even after having seen Him descend to Earth, was immense. And, I realized, it was exactly the amount of resistance that every vocally religious person must experience towards God's non-existence.

I'm amazed. I totally can't understand this kind of thinking (which you believe to be human nature).

Me, I don't believe that God exists. In fact, I hold the belief in God for little less than a mental disease. That is because there is virtually no evidence to support the existence of God, and a... (read more)

If you really believe that then you can test this with hallucinogens; in a non-trivial fraction of users (in good settings), they induce mystical or religious experiences and so there's a good shot they would do so for you. Have such an experience and still maintain your atheism, and maybe I will credit your claims to be atheistic based on purely rational grounds. Otherwise, you just look to me like, say, SF author John Wright: a strident atheist until he had some hallucinations after surgery and immediately flipped his views to become a strident theist. Seriously. Standard hallucinogens like psilocybin or LSD are easily obtained, cheap, and safe for at least a few doses. What's stopping you? Don't you believe your beliefs why you don't believe?
Wait... are you suggesting that psilocybin-induced hallucinations are proof that God exists?
I'm saying that if you have a psilocybin-induced hallucination of God and then become a theist, that's a darn good piece of evidence that stuff like the argument from evil or argument from silence weren't why you were an atheist. (And so if you were claiming previously that they were, you were either lying or badly mistaken.)
Ah. Yes, agreed with this.
I don't think my being an atheist has anything to do with the argument from evil or the argument from silence. (I can explain more if anyone's interested.) I am an atheist because, based on my current knowledge, the hypothesis that God does not exist seems far more likely to be true than the hypothesis that God exists. That's all there is to it. I assume that hallucinogens cause hallucinations, that is, distort my perception of reality. Why should I want to do that? If I were hallucinating and perceived something that convinces me that God exists, I would start believing that God exists. However, I assume that the effects of the drug would wear off sooner or later. When that occurs, I would recall the experience I had and give the "proof" I saw a serious thought. It is likely that I would realise that the perception was not real, I was merely hallucinating. So I would change my mind back to the belief that God doesn't exist. I am not atheist in the sense that I so badly want the God not to exist that should I see any evidence that He exists, I would reject it. I am an atheist in the sense that I consider it reasonable to base my actions on the assumption that God doesn't exist, and I refuse to start believing in God without sufficient evidence that He exists. The author of the article, though, seems to have some psychological problem with the possibility that God exists. That's what my comment was about.
That's pretty much the question. Wright could have reasoned the exact same way... and he didn't. Would you - really?
Wright's pre-conversion writing gave me the impression of someone who really wants to base their life on unyielding and absolute moral axioms, so he's not working that well for me as an ”it could happen to anyone” case. More as an example that the sort of people who like engineering and for some reason become dogmatic hardcore libertarians, communists or religious literalists can dramatically change allegiance after suitable neurological insult.
Mm, I'm not sure that group doesn't embrace LWers as well. We may claim to be open-minded and uncertain, but are we? We have plenty of libertarians here, after all. (I think that would be testable, though; IIRC, there are a number of psychological questionnaires measuring dogmatism or need for certainty/closure (from the old research into authoritarianism). Administer along with some sort of religious questionnaire before psychedelic use, see whether the high scorers on one become higher on religion afterwards as compared to the low scorers, and especially the high scorers who report a specifically religious psychedelic experience. Too bad the drugs are so controlled and there will probably never be any real studies on this...)
I've been wondering whether an unusual number of smart people these days are ones that were libertarians in their early twenties and have become less so later on. Possibly similarly as in an earlier generation an unusual number of smart people were communists in their early twenties and became less so later on. There's definitely a lot of background assumptions sympathetic to libertarianism on LW, but I haven't seen much of the sort of absolutist first-principles stances I associate with the group of people I'm thinking of in grandparent comment. It's the difference between thinking that free markets are a good starting metaphor for thinking about arranging human affairs and insisting that a strict adherence to a few easily listed axioms like absolute property rights can be pretty much the only thing you need to successfully run a human civilization.
Interesting datapoint: MacLean et al 2011 did a RCT of psilocybin. Those reporting a mystical experience saw a rise in their Openness, while those reporting no mystical experience show, if anything, a fall. See the graph on pg6. (It's almost like they're updating on their experience.)
Psilocybin can also induce suicidal despair in a non-trivial fraction of users. I would highly recommend against its use by anybody who isn't extremely emotionally stable to begin with.
Cites or numbers for non-trivial? I looked for info on psilocybin and suicide, and the only review I found cited listed, after I jailbroke a copy, just one suicide and few deaths. I haven't looked into psilocybin in as much detail as LSD, but I had assumed it was considered very safe since it seemed to be the hallucinogen of choice in recent American research.
Personal experience. Psilocybin trips vary wildly, but everybody who uses it regularly eventually encounters an episode of extreme despair. (It's not as bad as LSD, wherein you can very easily get caught in a mental loop - which if you're thinking negative thoughts will send you into an emotional deathspiral - but it's definitely a tangible risk) It's not that it induces despair, per se, but it heightens emotional response to stimuli considerably. (Particularly in the very high dosages necessary to induce hallucinations.) It's necessary to strictly control your environment.