Online meta-discussion is full of ideas and terms from debate — accusations of “moving the goalposts”, citations of logical fallacies. When people withdraw from discussion, they get accused of not really believing in their own ideas — arguing in bad faith, or fearing that their claims won’t stand up in “the marketplace of ideas”. How did unstructured communication with strangers become tinged with the expectation that it would follow the rules of our most elaborately artificial rhetorical spectator sports? And why do we, when communicating with strangers, default to considering such communication an argument?
The west likes debates. In particular, anglophone enlightenment thinkers like those who initially engineered american cultural norms inherited ideas about the use of rhetoric in the search for cosmic or metaphysical ground truth from Plato and Descartes, adversarial collaboration in countering errors in science from Bacon and Occam, adversarial argument in law from English judicial arrangements, and competition as a factor in driving down prices from Adam Smith — so Americans of an intellectual bent (especially educated libertarians, many of whom have approximately the same set of ideas about how the world ought to be as Thomas Jefferson minus the support of slavery, and some of whom also share Jefferson’s ideas about slavery) are especially prone to going all-in on this intellectual form of trial-by-combat.
Of course, we know from experience that regular trial-by-combat is not a great way to arrive at justice. Generally speaking, the stronger party will win, regardless of who’s right. If the parties are evenly matched, luck plays at least as big a role as whatever one might gain in confidence or motivation from seeing themselves as right. And, we wouldn’t even be getting to the stage of formalized combat without both parties thinking they were right (or, more rarely, thinking they could get away with being wrong)! Informal debate has similar problems, which is why none of the traditions of adversarial search for truth mentioned above are very much like it.
Plato (and his sock-puppet Socrates), when they present their own ideas, are not convincing to modern readers; nevertheless, when they tear down other people’s arguments, they are fairly compelling — why? Because the socratic method does not actually involve attack or competition. Instead, it involves identifying key gaps in understanding and asking them to be filled. A perfectly consistent model (regardless of whether or not it happens to correspond to reality) will only seem stronger after the socratic method is applied to it, while inconsistent models (regardless of whether or not specific individual parts match reality well) will fall apart. In other words, this method is not adversarial at all!
Likewise, the scientific method. To the socratic method, we add the demand that claims be consistent with reality (by testing them). The only time this becomes adversarial is when someone has an irrational emotional attachment to a flawed model (which happens all the time, but is not really relevant to the mechanism at hand).
The cartesian method of doubt does not involve a second party at all (although ideas from it have been applied to adversarial or pseudo-adversarial contexts, sometimes strengthening their ability to come to the truth but often merely justifying stonewalling tactics like descending into sub-arguments about definitions). Here, the single participant finds gaps in his own certainty. Descartes’ bar for certainty was untenable even for Descartes, and is certainly untenable for us (since we cannot put our faith in some benevolent God who would never trick us despite creating optical illusions and dreams, nor can we necessarily take cogito seriously), but at the same time, we ought to be comfortable with accepting and acting upon whatever information is most solid.
Smith’s ideas about economics are fine, within the very narrow domain in which they apply, but we all know both from personal experience and rigorous analysis that even minor inequalities at any stage tend to grow and warp subsequent stages — in other words, this is an ‘adversarial’ game that only produces net benefits when collusion and inheritance are abolished.
The closest thing to an informal debate here is a jury trial — but, keep in mind that this, too, has key differences. In a trial, while there are two opposing factions, they are represented (usually) by impartial professionals (lawyers) — who are required to show each other all their research beforehand and provide time to come up with opposing arguments. These impartial professionals are also involved in selecting and dismissing jury members (the initial pool of which is chosen by lot). There is a moderator (the judge) and a set of complex rules of order, a long period of deliberation, and judgement is often determined by consensus rather than simple majority — consensus with terms established by the judge. Despite all these precautions, the individual charisma and rhetorical ability of lawyers predicts outcome better than the actual truth of claims, and the complexity of these rules leads largely to folks cutting the gordian knot with plea bargains — in other words, preferring to simply go to jail rather than deal with the burden of the mechanisms that nudge this system closer toward truth-finding.
Informal debates like those we have on social media have no moderator, no standards of evidence, a self-selected jury of already-interested parties (many of whom have already made up their minds), and, generally, no already-agreed-upon subject of inquiry. Participation is fluid, topics change, and the interchange generally begins before any of the arbitrary number of debaters really know any of the others’ positions. The debaters often have substantially imbalanced personal power, and what’s worse, the ‘jury’ tends to be composed primarily of people with an existing relationship to one of the debaters. In other words, this kind of structure is ill-suited to finding truth but extremely well-suited to spiraling out of control without convincing anybody of anything.
On the other hand, outside of an adversarial framework, none of these attributes of online discourse are a problem. Some of them are even a boon to specific forms of non-adversarial discussion — for instance, online social spaces are extremely well-suited to socratic dialogue (something that used to be called ‘trolling’ until the ubiquity of the adversarial stance led to that term getting misapplied to flaming, griefing, disinformation, spamming, and all other forms of online antisocial behavior).
Without standards of evidence and with a drifting topic of inquiry, non-adversarial discussion is free to become exploratory and lead to all parties involved developing previously-unconsidered ideas. With fluid participation comes a diversity of thought and knowledge — folks can drop in whenever they have something relevant and interesting to say, even (perhaps especially) when it complicates matters.
Since nobody is supposed to ‘win’, there is no incentive to hide or dismiss the complexities of the topic — which after all will probably never be truly amenable to one side or another of some binary opposition decided upon before discussion began.
When popular personalities engage in nonadversarial discussions in public, their fanbase gets exposed to new and interesting ideas, rather than being turned into a militia for the defense of old and boring ideas. After all, these folks are generally interested in the discussion because of their interest in its participants rather than previously-existing knowledge about the ideas in question, so non-adversarial discussion can turn the parasocial tendencies of unsegmented social media landscapes into a force for intellectual growth rather than conservative traditionalism. What’s more, the influence of participants will tend to grow based on their ability to say interesting things, regardless of its original level: while debating a popular figure is a great way to get SWATted, non-adversarial discussion with a popular figure is a great way to get exposure.
Non-adversarial discussion does not mean withholding judgement indefinitely, or failing to call people out on bullshit, or failing to correct misapprehensions. Any wide-ranging discussion will involve doing these things. What it means is that these ‘gotchas’ are not the point of the discussion, and there is no particular incentive to find and expose them beyond ensuring clarity.