You can't invoke social science factors that should apply to all political rallies as an explanation for why McCain rallies specifically have become so nasty. Moreover, it is really questionable whether the crowd at a political rally interacts sufficiently for these type of factors to apply.

It's much more likely that the extremism at the rallies reflects the nastiness of the McCain campaigns advertising; McCain and Palin's nasty speeches; and the ongoing right-wing whisper campaign to smear Obama. The reason people at McCain rallies think Obama is a secret muslim and a terrorist is that that is what they have been told over and over again.

I'd say they do interact pretty fluidly, in their boos and cheers. Sure they need to be primed beforehand, but when they all get together in the same place, then you get the self-reinforcing effect.

You could probably say the same thing about ultra-partisan websites. Luckily this has resulted in the GOP not even being able to run a competent negative campaign this time around. They are lost in their paranoid conspiracy-theory End Days melodrama, talking only to their fellow inmates, and haven't managed to get traction beyond the die-hard social conservative base, which is apparently all that's left of the GOP.

They're going for the emotional hit instead of trying to convince people who don't share their obsessions. Satisfying in the short term, but it'll probably lose them the election. What a damn shame.

"Interacting" at a political rally can involve something as simple as a low-level hmmmm, if it is picked up by enough people. To deny that sort of non-verbal/semi-verbal interaction would be as myopic as trying to deny that a performer in a live presentation doesn't respond to the audience with changes to their delivery and timing.

I'm not saying that the crowd doesn't interact at all. But the article suggest that discussion with like minded individuals can promote extremism and posits this is because there are social benefits to holding views stronger than those of the group mean. At a political rally you really don't have an opportunity to assess the group mean.

I guess it's possible that the guy at a McCain rally yelling "kill him" is trying to out do all the other people who are merely shouting "traitor". Certainly the crowd feeds off its own energy. But I think it ultimately reflects the tone that is coming from the campaign itself. Presumably the crowd at an Obama rally is equally like-minded but they don't exhibit any similar level of blood lust.

The article doesn't go into mechanisms, which can be anything and everything from simple nods and 'amen, brother' to what you had in mind - actual discussions. The point of the studies and of this article is not that people of like mind get together and reason together to achieve a more extreme point of view but rather by coming together in any semi-structured form - a political rally for instance - the result, when the points of view start of fairly homogeneous, will often result in stronger, more extreme views simply from the heterodyning effects of group interaction. Not group discussion, just simple and often nonverbal interaction.

You: "The article doesn't go into mechanisms..."

The Article: "Social scientists have proposed several reasons for why like-minded groups tend to polarize. Two have survived scrutiny. The first is that homogenous groups are privy to a large pool of ideas and arguments supporting the group's dominant position. Everybody hears the arguments in favor of the group's belief, and as they're discussed, people grow stouter in their beliefs.

The second ... has more to do with how we see ourselves. We are constantly comparing our beliefs and opinions to those of the group. There are advantages to being slightly more extreme than the group average. It's a way to stand out, to ensure others will see us as righteous group members."

Both the proposed mechanisms require more interaction (e.g., discussion) than you would get at a political rally.

The second reason is a broad label that includes social feedback that is not explicated in the article but which includes everything I have mentioned and more.

Do you really think each individual in a rally walls themselves off from all the rest, that group dynamics play no role in what they refer to as 'how we see ourselves'?

Think of religious gatherings - before any sermon begins music, songs, reading and response - none of which constitutue discussion (including the sermon) yet all of which are designed, just as a political rally is designed, to bolster, feed and instill a feeling of community, that is, the very 'like minded'ness that is at the center of this article.

We're talking about different things. Clearly there is a crowd dynamic at a political rally (or a religious service or a rock concert) that creates a sense of community and makes you more open to the content of the event. Everybody knows that.

What this article is saying is that people become more extreme when they gather in groups of like-minded people, for the reasons cited above. My point is that at a political rally you have no way of knowing the mind of other crowd members, except through their response to the message that is coming from the stage. So the content of the message put out by the campaign is what is being amplified by the crowd effect, not group opinions. I'm guessing that if you took the crowd from a McCain rally to a rock concert or a football game instead they wouldn't come away with more extreme political views.

Remember, the point is to explain why McCain rallies in particular have become so nasty. It's not the crowd effect, which should operate at Obama rallies just the same. It's the fact that McCain is running a nasty campaign.

"...rock concert or football game..."

That would be where the 'homogeneous' or 'like-minded' factor comes into play. When the crowd is heterogeneous - a mix of views - such as a rock concert or a football game, nope, no change in views. That was the point of the article, that when the views are homogeneous, like-minded, people pick up cues - not necessarily verbal cues, though they can play a role too - and the emotional reaction - as opposed to a thoughtful decision based on discussion - is to become more extreme than the average. The reason for this, again, is emotional not rational. The desire is a tendency to establish what has been perceived what the group view as a sort of 'even better' attitude.