better discussion systems could counteract polarization
It’s an understatement to call our political system polarized, but I don’t think we should just “blame the politicians”, or even the system of primaries and elections that selects them — our public discussion itself has become much more polarized in recent years; to some extent politicians are just reflecting that.
But that should give us optimism — if the root problem lies with “we the people”, and especially if it’s related to the recent rise of web-enabled social media (defined broadly as all systems for online exchange of opinion), maybe we can fix it — modified social media should be easier to design and bring about than a constitutional amendment, and we can keep trying various changes until we find things that seem to help.
current systems artificially enhance polarization
A big part of the problem, I think, is “mutual misunderstanding” — we estimate the views of our political opponents (among the populace, not politicians) as being more extreme and less reasonable than they actually are. (Or at least, that’s my hope and belief. I don’t really think most Trump supporters are racists, or most US “liberals” are against political free speech, though both those fears are prominent in online commentary. Personally I fear a president Trump much more than Clinton, but I don’t disrespect people who guess differently. The future is always uncertain, and it’s not surprising if good people with very different experiences and worldviews prioritize problems differently, and guess differently about which potential bad policy choices would be more likely or worse under different leaders, given the certainty that great and difficult change is coming.)
Specifically, I think current systems for selecting and promoting online commentary overrepresent simplistic and provocative opinions, compared to their actual prevalence. That is, I think the average person with a given political belief is much less extreme and unreasonable than the average visible online comment supporting that belief. On any issue with two prominent “sides”, this leads to a “positive feedback loop”, in which people on one side see too many unreasonable comments from the other side, and think of them as representative, even if they would have been considered too extreme by most people on the originating side. Eventually we get “demonization” of ordinary people on each side, in the minds of ordinary people on the other side.
small groups in person show we can do better
Is there anything that could fix this?
Our experience with small groups of people, working together on practical problems in the “real world” — even when they mostly can’t choose who to work with (e.g. in families, workplaces, or juries) — gives me hope that there is. We may disagree, and initially have less than full mutual respect, but in most cases we can come to understand each other’s opinions, find common ground to build on, and eventually reach some consensus or compromise about how to proceed.
The key is partly that the relationships are personal, in the sense that we work with the same individuals long enough to understand their opinions and personalities, and we have an incentive to keep working with them smoothly. To get the benefits of this online, what’s relevant is not “real names” — what matters is fostering continuing relationships with individuals.
Online communities are different, since they take on a character broader than individuals, and they’ll almost always be voluntary associations. So they might be more analogous to conferences or public meetings or even government bodies, which can indeed be more contentious in practice than small groups. But we may be able to construct them to get the best of both worlds.
a chain of respect
One key idea is a “chain of respect”. I know some people (acquaintances and/or bloggers) who I think are worth listening to on certain topics, even when I don’t agree. They know other people in turn. If we form a directed graph of all person/topic pairs on which “person A thinks person B is somewhat worth listening to regarding topic T”, it will usually be mostly connected (that is, have paths on which I can see comments from most people), over people with sensible things to say about a given topic. We can use it to determine what I’m more likely to see. (In practice the edges would have user-chosen numerical weights (between 0 and 1) about “relative worth of attention”, which would multiply along paths, to determine my priority for receiving or seeing a given comment. And the topics could be arbitrarily subdivided and connected.)
I won’t see much spam or trolling this way, since there will be no path from me to its author on which everyone has high respect for the next person’s commentary. But as long as I know some open-minded people, I will see intelligent commentary from the “other sides” of the various topics I read about. There will still be a danger of “bubbles of like-minded opinion”, but as long as the rules and data are open (unlike in current centralized social networks), there will be easy ways to detect and mitigate this for people who want to.
(I’ve written more on this in a blog post called “A Social Network for Ideas”, explaining how it would support not only a graph of people, but a graph of all their individual comments and how they’re related. I sometimes call that a “graph of all ideas”. There are other systems with related goals and ideas, mentioned in that post and its comments. Several people are discussing how new systems like that might be realized, in those comments and in a Google+ community called “collaborative design software”.)
how to use a “graph of all ideas”
We could use a “graph of all ideas” for much more than just seeing more representative or respectable comments:
- we can try to trace out logical arguments, and understand the prevalence of various chains of reasoning that lead people to different conclusions.
- we can ask for the best few arguments on either side of an issue, where by “best” we mean “most persuasive to people whose opinions we respect”.
- we can find other people or communities who think about related sets of topics or projects.
As for political polarization: if this kind of idea-browsing system became popular, it could have a long-term direct benefit, by improving people’s understanding of other people’s ideas, and the development of better ones.
But it could also have a nearer-term indirect effect, even if only a few people used it, provided they were people trying to come up with ways to improve things, if a system like this could help them exchange ideas about the complex phenomena causing problems, and what might be done about them.
