or, Social Skills 101


Why debate at all?
Pray for a good enemy
What’s wrong with flamewars?
Some caveats to start
Positive advice
The big 2: Abuse, disdain, sarcasm
More annoyances
Minor peeves
On gang-bangs
Evidence: WANT

Good arguments are fun, but bad arguments are horrible. This page is the result of many years pondering what the difference is.

Why debate at all?

Pray for a good enemy

Someone (unfortunately I forget who) mentioned that the best thing an academic can have is a good enemy. Your friends will, perhaps, read through your papers and make a few comments. But only an enemy will read through an entire book, for free, finding every error and questionable statement.

Some of the best pages on my site— the anti-libertarian page and the essay on chance resemblances— owe their quality to some very annoying critics. They didn’t change my mind, but they pointed out gaps in my knowledge or logic, and inspired more information and more clarity.

I think this is what Proverbs 27:17 had in mind: “Iron sharpens iron; so a man sharpens the face of his friend.”

As a corollary, if you’re arguing on-line, your opponent is doing you a favor by their criticism. Respect the time and effort they put in by responding civilly.

What’s wrong with flamewars?

If arguments are so good, why complain at all? Because once an argument becomes a flamewar, the good elements disappear. It becomes a waste of time.

Anything that isn’t a discussion on the issues wastes time:

Not only do these things make people mad and make the conversation unpleasant, they kill forward progress. If you’re talking about each other, you’re not talking about the issues any more and no one will learn anything.

Some caveats to start

Positive advice

A lot of this page will be about things going wrong, so let me start the other way. Here’s some guidelines for good arguments: OK, now let’s talk about things going wrong!

The big three: Abuse, disdain, sarcasm

Personal abuse

This is the worst sin, and yet for that very reason, the easiest handled.

You don’t have to engage in a conversation that starts with name-calling. If someone is signalling up front that they’re going to be a jerk, they’ve disqualified themselves from civil discussion. Tell them so, or ignore them. But life is too short to engage with such people.

That’s my policy with e-mail: if it starts with abuse, I throw it out.

If a person becomes abusive on a public forum, there are usually clear rules they’ve violated. Contact an admin.

(Back on Usenet few newsgroups were moderated— and most filled up with garbage. In that case the best response to abuse was *plonk* ... the sound of the abuser hitting the bottom of your killfile.)

Abuse can occur in any field— Geoffrey Pullum once proposed an annual award for Withering Invective for linguists— but it’s become regrettably commonplace in politics. That doesn’t make it acceptable. Abusive people are clowns; ignore them.

Treating opponents as idiots

I hope it’s clear that this is rude; but even more importantly, it’s itself stupid. Disdain is how smart people make themselves stupid.

If you think your opponents are idiots, you don’t expend much thought on your arguments, or theirs. So you lose all the benefits of argumentation I outlined above.

A good example is Searle’s Chinese Room; Searle’s disdain for his critics keeps him from seeing glaring holes in his argument.

It’s also common in political arguments... otherwise intelligent people often become brute dualists when it comes to politics, treating the other side as entirely demonic.

A rule of thumb: If an otherwise smart person says something dumb, assume you misunderstood. Maybe they were unclear, maybe they didn’t mean a particular word the way you took it, maybe you didn’t follow them, maybe they left out something because they thought it was obvious or it was a side point or would take too long to explain.

Wrong response: “Moron. Pyros do not have a win button.”

Better: “Wait, do you really mean that pyros are easy to play? How often are they the top scorers, then?”

It’s probably unwise to offer many counter-arguments— if you’re not really addressing his points, it will only waste time if he has to bat away your counters.


Sarcasm is fun. But.

In years of online discussion, I’ve met a lot of cranks: in linguistics, in physics, in software design, in politics. To be a crank you really can’t have much of a sense of humor— it gets in the way of the terrible earnestness. The one type of near-humor they all share, however, is sarcasm. Cranky screeds are usually puncuated by ferocious bursts of snark.

As a result it’s rather spoiled for me as a rhetorical device. “Well, I’m sorry the US tried to bring freedom to people who were trying to kill us.” Where to start in explaining the misconceptions? Why bother?

Sarcasm isn’t clever; it’s easy and cheap, just one step above calling people names.

To me, it signals that the fruitful discussion is probably over. The other person isn’t really thinking any more; all they want is a flamewar.

A small number of people have a gift for verbal abuse— but even with them, it just has entertainment value for their friends; it’s of no use in a debate.

More annoyances

The big three tend to come up all the time. The following may or may not be present, but they’re all bad signs.


It’s not bad if a debate gets rough at time, so long as there’s still content going on. Once the meta-discussion starts, however, it rarely lets up. If someone starts complaining about the person they’re arguing with, I conclude that they’re more interested in their own dignity than in the subject.

Worse than that: just as spelling flames are very likely to contain at least one spelling error, some karmic law guarantees that meta-discussion is always self-incriminating.

I wish computers came with a little device like a virus checker that would reach out and administer an electric shock when someone’s about to make complaints about behavior that they themselves are guilty of.

A few keywords it would check for:

The shock should be dialled up if any of these are described as willful.

People hate unfairness, and they will be angry like big shaved bears if they’re accused of the same things they see you doing.

Personally— despite the impression that may be created by this page— I’m willing to tolerate some roughness. But not if the other person can’t take the same roughness back. If you’re insulting, accept that you’ll be insulted back.

But again, the main problem is that meta-discussion ends the actual discussion.

“You don’t understand!”

This is perhaps the most common bit of meta-discussion. The frustration is natural— as I suggested, people want to be understood even more than they want to be agreed with.

Much of my working life consists of designing user interfaces; when I get home, I’m a writer. Wearing either hat, I think it’d be pathetic to blame users or readers for having difficulty with my work. It’s my job to make myself understood.

Instead of whining about “misreading”, take responsibility for your words. You were unclear. If you think this way, you’ll have a much better time in debates.

And believe me, your writing isn’t as clear as you think. I pride myself on my writing skills... but in the heat of an argument I may not notice that something can be taken several ways, or I may be more forceful than clear.

As well, the best and clearest exposition often occurs to me in the course of and because of an argument.

Another thought: if you feel your point wasn’t understood: maybe it was, but it just wasn’t considered very interesting, or the other person wanted to rush on to make a different point. Or maybe person A understood it but B didn’t.

If what you’re explaining is something big— quantum mechanics, the comparative method, deconstructionism— don’t expect it to go down easily. Explaining things for outsiders is a skill, and the longer it’s been since you learned the material, the harder it may be to take an outsider’s point of view.

Rhetorical telepathy

Some debaters want to talk about you; and not knowing you, they resort to fantasy. Common tactics: These things don’t even make sense as a debating tactic— they can never be convincing, because an on-line stranger doesn’t know more about my brain than I do.

A more subtle form: trying to hold people accountable for your deductions from your description of their position. Maybe you think that Republican, or Democratic policy will be the ruination of the country. That doesn’t mean that your opponents have the actual intention of ruining the country and that you can ask why they want to do so. Who knows what’s on their minds; but a good guess is that they think they’re helping it.

Helpful tip: people respond better to questions about their beliefs, even hostile ones, than to statements about them.


For some reason, in a failed argument people behave like Lewis Carroll characters, focussing minutely on word choices and surface meanings. “You said that was unforgivable. Fine, since you won’t ever forgive, there’s no use talking.” More than once I’ve got into an argument on a thread, and when I finally figure out what the problem is, it’s the precise wording of the very first paragraph I posted, wording I’ve half forgotten since I’ve stated my point better since.

If you’re one of the people who does this... er, well, don’t do that. Do-overs are allowed. If you watch carefully, you’ll often find that people’s positions converge to some extent in the course of an argument, as people add missing qualifiers and adapt to attacks.

Argument by adjective

Some people seem to think that adjectives are damning. “Typical statist/dualist/Chomskian response.”

There are sophisticated forms of this as well. I’ve met people who show real glee when they can fit you into a two-term category... “So you’re a Keynesian Situationalist.” It’s as if they think they have your number, that henceforth you must follow the behavior they’ve assigned to you.

First problem: such statements doesn’t actually show anything. Even if correct, it’s just an assignment to category X, and doesn’t prove a thing unless you accept the assumption that category X is bad— and this being an argument, you don’t.

The deeper problem: as a linguist, I’m not impressed. Words and classification systems are human inventions, mental devices to help us look at the world. If you assign people to a category, it tells us next to nothing about them, though it may tell us much about your set of boxes.

Labelling political positions is especially silly; no one can be reduced to a one-word slur, or even a short phrase. It certainly can’t be used to predict an individual’s other opinions; if you try, you’re likely to be wrong and waste time as the individual has to correct your guess.

Corollary 1: it’s not important exactly what word to apply to a situation. Most people are Whorfians when it comes to politics— other people’s politics. But loaded terms are usually used as irregular verbs: “I’m retaliating for outrageous atrocities; you’re practicing indiscriminate aggression; he is a terrorist. I’m spiritually enlightened, you’ve found religion, he’s joined a cult.” This sort of thing isn’t insight, it’s simply judgment. And there’s rarely any reason to ramp up the rhetoric on your judgments; they’re usually obvious anyway.

Corollary 2: any time you’re tempted to dismiss a class of people as not knowing what they’re talking about, or as evil, or just not very smart, you are the one in intellectual trouble.

Prejudice is another way smart people make themselves stupid. You never know where wisdom will strike. Once I met an unrepentant Stalinist who dominated a discussion (it was about the attempted coup in the last year of the USSR) with his calm and superior insight.

Straw men

A straw man is a parody of your opponent’s position (usually delivered with sputtering sarcasm). It’s not only an infuriating technique, it’s dumb. The parody doesn’t affect your opponent in the slightest, because it’s your invention. All it does is make him mad, and waste time as he swats it down.

An example: “You say that all languages are equally complex? I guess you never saw the 50 pages of conjugations in my Russian textbook. You can explain English verbs in half a page. So in your world does 50 = 1/2?”

(If that piques your interest, see the sci.lang FAQ.)

Why does it make people mad? Again, people want to be understood. A straw man is a declaration that the person doesn’t want to understand.

Ways to avoid building straw men:

Some tricky points, though:

Minor peeves

These don’t come up as much, though they’re annoying when they do.

Opponent conflation

If I intervene in a thread, it’s often to respond to some single point that catches my eye. If I respond to X’s post, it does not mean that I endorse everything anyone has said up to that point against X, nor am I responsible for what other people are saying earlier or later in the thread.

It may seem a lot to ask to figure out what’s the position of everyone in a large debate, but most forums come with a rather helpful feature: people’s words are listed under their names, so you can easily see what each one said.

The malicious non-response

Sometimes I don’t respond to something, and someone pounces on this, as if it’s so telling that I conveniently passed over this apparently devastating point.

As a public service, here’s five more likely reasons I might not have responded to every item in your 400-line post:

Grudge hoarders

On a public forum you may well argue with the same people over and over, perhaps on the same topic, perhaps on many different ones.

A natural tendency is to hoard grudges from thread to thread, and it soon makes all encounters acrimonious and the forum unpleasant. Somebody makes an innocuous post, and they’re immediately attacked— “There you go again!”— starting at the emotional peak of the last flame war.

I try to follow a simple rule in such cases: in the next thread, we start afresh. Take a break, forget the gnawing grievances, flush your buffers, do whatever it takes to approach the next thread as if you’d never seen the person before.

It helps to meet people offline or have threads for getting to know people and having fun. It’s easier to have good arguments with people you feel good about.

Topic police

This may be the least of the problems described on this page, because it usually doesn’t work anyway. But some people, especially if they started a thread, feel that they can control its contents. But people discuss what they find interesting, and a thread can veer off unpredictably from its origins.

Some boards, like mine, allow admins to branch off side topics, so that can be a good idea if people are talking about two very different topics at once.

On gang-bangs

Sometimes you get a minority viewpoint and a whole bunch of people arguing against it. This requires a little extra sensitivity.

Allow the lone person (if they’re a regular) a little extra leeway. It’s not much fun to have half a dozen people jumping on you; it’s easy to get angry.

Don’t be a bully. Being in the majority doesn’t mean that you get to mock those who aren’t.

It’s not necessary to have six people making the same objection. Read the thread to see if somone already made your point.

It’s a very human thing that when someone is being extra dense, people feel that one devastating rejoinder will break through and fix them. At the same time the embattled person may feel, also very naturally, that he has to respond to every post. Put these tendencies together and you have a flamewar that can last forever and take over the forum. As an admin, I usually shut down such discussions when it’s clear that they’re going nowhere.

Also see Cranks and Ideologues below.

Evidence: WANT

If you have no evidence, you’re not debating, you’re giving an opinion. We all do this when talking with family and friends, of course... but in a public online forum, you’re no longer bullshitting with your buds. No one has to buy your opinion just because you’re you.

It’s not an imposition to ask for evidence. This is the Internet; you have access to your books and Google. Be prepared to back up your claims.

Not all evidence is created equal, however.


For some reason, Usenet was full of cranks. These are people who gleefully go against the mainstream, whose alternative methods are entirely unfalsifiable, and rely in improbable leaps, personal dicta, or impenetrable obscurity.

They can be lucid at times, or when off their pet idea; this can encourage you to get into long discussions with them. But they’ll never follow your demonstration of the weak points in their theory, and they have little time for this anyway— they want to go back to exposition, or to ranting against their opponents.

(How do you diagnose crankery? The surest telltale is unfalsifiability. Cranks are often tireless researchers; but they only accumulate positive evidence. They cherry-pick mainstream sources for factoids that support their position and ignore everything else. The linguistic ones, armed with dictionaries, amass mountains of ‘evidence’. But they never contemplate a test that might reject their hypothesis. A skeptical professor I know has a simple test for people who want to argue with him: “What would it take to prove you wrong?” For a crank the answer is “Nothing could.”)

Now, arguing with cranks can be fun, even useful— again, my chance resemblances paper benefitted greatly from a few run-ins with them. And in some ways it’s a public service to expose their errors and explain mainstream science.

But, don’t count on them ever changing their minds.


That’s a cover term for anyone with an elaborate, often minority point of view: fundamentalists, militant atheists, extreme libertarians, Esperantists, wingnuts.

Some of these folks are cranks, but many are not. They may have perfectly good reasons for their beliefs, and many are smart and well informed.

So have fun arguing, but be aware that they’re not going to admit a point, however well argued, that would invalidate their ideology. A reasonable Esperantist might agree with many of your points and deplore the excesses on his own side, but he’s not going to cease to be an Esperantist.

(Yes, people do change ideologies. But don’t count on it.)

As I mentioned, I think is is possible to argue across an ideological gap; but only by finding common ground. If the gap is large enough, that can be a challenge. (The gap need not be political or religious. There are significant gaps between philosophy and linguistics, or between formal and cognitive linguistics.)

Both sides should be patient with the process of identifying common ground. Talking with people who share our ideology, or our academic field, we know what we all agree on, what’s obvious, and what’s not. With outsiders we don’t. Don’t get upset if someone carefully explains something obvious— it wasn’t obvious that it was obvious. And if they go on and on based on something you don’t accept, maybe they didn’t realize that; point out the non-shared assumption and let them start over.

Text © 2009 by Mark Rosenfelder.

The illustrations are screen caps from Team Fortress 2, taken either by myself, Chris Livingston, or other Mefites.