Elliot Milco has an excellent post at "Fare Forward" (and a slightly longer version at "The Paraphasic"). The argument is quite good, although interestingly I find myself not agreeing with much of it. That is, I think the argument would be rock-solid if its underlying diagnosis were correct; but I think this diagnosis, while reasonable, is in fact false.
The basic starting idea is that internet debates go sour because we see opponents as floating bits of text rather than real people. This is nothing new, however; we have had textual intermediaries for ages. And while polemic by book and article and letter to the editor has often been heated, the textual wall has in practice always been a control rod in the reaction. Likewise, people are less likely to wrestle with mere projections when dealing with people by text than when dealing with them face to face, for two reasons. First, the evidence of what people have actually said and done is more stable and available, and therefore constantly has to be dealt with rather than ignored, by both parties; thus one notices that people are much more likely to apologize in internet debates than elsewhere, and even when they don't, a heated debate is much more likely to diffuse into something milder, even if still shot through on occasion with irritations. Second, face-to-face interactions involve projection as much as any other kind of interaction, but the lack of an obvious intermediary -- because we don't think of faces and spoken words as intermediaries in the way we still think of texts as intermediaries -- means that people are less likely to consider the possibility that they are dealing with a projection. In both cases we are inferring someone's state of mind from evidence; but people often take offense for much more obviously absurd things in face-to-face interaction than when they interact by text.
I think the illusion that this is not so arises chiefly from the fact that we live in an anti-argumentative culture; there are very few situations in which it's OK to just argue, and those are often of a very restricted type. Democrats and Republicans don't generally argue with each other, for instance; they mostly talk about each other behind the other's back with occasional limited forays for potshots. And politics is more argumentative than the rest of our culture; try to do the Socratic thing and get someone arguing about moral issues outside of a clear political forum, and you will find that people freak out. IRL we like our arguments rare and well-insulated. There are some definite advantages to this for certain aspects of civil harmony, but I think we should not treat it as much more than a matter of cultural taste -- and in fact should recognize that it at times borders on pathological. But the online world is, to a very great degree, an argumentative medium: not little insulated statements of opinion like people are culturally confined to in real life, but responses to responses to responses, and while these need not all be arguments, arguments are inevitable. It is mostly the sheer quantity of argument online that makes it seem like arguments are always going sour. Further, when internet arguments do go sour, it seems mostly to be due to the fragmentary nature of the medium -- arguments are not done all at once, as they are done face-to-face, nor in large chunks, as book and article arguments proceed, nor as rhetorical face-offs, the way debates in the editorial pages have often been done. Rather, they are done by bits and pieces stretched over days by people who are often busy doing other things as well. This is a very difficult way to argue; that people slip under such conditions is not really surprising.
I think, however, that even if this were not so, Elliot's suggestion that "everyone’s insistence on his own position against all opponents can lead to the illusion that moral disputes are fundamentally insoluble" is not generally true. In fact, in most cases it is the reverse: people who have actually argued in moral disputes are less likely to think moral disputes fundamentally insoluble, not more likely. No one who has engaged in heated arguments more than occasionally can possibly avoid having the experience of progress being made, either by being stumped oneself by an excellent argument, or by coming up with a stumper oneself, or by both parties coming up with new arguments. It is stagnation that leads to the illusion of insoluble moral disputes. And this is true elsewhere. People who argue about religion, whatever their position, are not people who think that no progress can be made. The major problem on the internet, of course, is that its fragmentary nature means that only a handful of arguments actually see any serious development. But in all honesty, it probably happens more online than it does elsewhere, for precisely the reason that our culture is anti-argumentative about face-to-face interactions.
Elliot's first two concluding points, that participants in arguments need to make an effort actually to address each other, not just lecturing at each other without listening, and that they need to be making an effort to understand what they are talking about, with the humility to recognize when they don't, are quite right; the closer we get to these the better arguments get. However, while I think the idea behind Elliot's third point is right, I think the formulations is problematic. It sounds well and good to say that pedagogy trumps belligerence, but if you look at the long history of good teachers, Socrates and the like, you find that many of them were quite capable of jumping into a fray swinging. Likewise, it sounds good to say that "Our moral debates have become so unproductive because we treat them as wars over particular conclusions, rather than organic explorations of the truth flowing from first principles," but organic explorations of truth flowing from first principles are things you do in treatises and non-argumentative conversations, not debates. Likewise, it sounds good to say, "The good teacher works by knowing his students, meeting them at their level, and drawing them patiently up to a fuller understanding of things," but in an online debate you are not a teacher except in the sense that everyone is a teacher all the time. A debate online is not your classroom; the kind of patience needed online is very different from the kind of patience needed to teach people in an insulated setting where you have a clear position as a teacher. I've watched professors in the blogosphere for quite some time, and have noticed that many of them have difficulties stemming from this: the habits simply do not always carry over well, and can make things worse. In an internet debate you aren't generally dealing with people who want to be your students, and they certainly have no financial incentive to bear with you in your round-about ways.
And it is true that no one is persuaded by defeat or humiliation, but persuasion is only an aim of argument insofar as argument is purely a matter of rhetoric -- the aim of argument as such is simply to make explicit reasons in their proper place. You should not in fact argue to persuade except under very carefully limited conditions; arguing to persuade as a general rule is what orators and sophists do, and it sounds innocuous but is in the long run not so. You should certainly try to argue in such a way that persuasion is open -- but what kind of attitude and tone that will actually require will vary wildly, and you will often merely be guessing -- but you should instead argue to show what's right and what's wrong, to the best of your ability, so as to refute if you are right (whether refutation yields persuasion or not) and to be refuted if you are wrong. That gentleness in argument is a wonderful quality is not to be denied, and those of us who are too acidic or impatient or coolly ruthless to attain it easily should strive for it. But in a culture like ours one must watch out for the tendency to conflate gentleness with rhetorical pandering. I think Elliot's third point is actually for gentleness precisely as a mean between belligerence and this pandering; and to that extent is right. But we must not slip into thinking that it is our place to win people over: that is for their own reason to do. Rather, we should see our place as making clear what's going on -- providing diagrams, maps, and directions, so to speak, so that they can find the way themselves. And it ends up being a rather different thing.
