[<< | Prev | Index | Next | >>]

Friday, April 28, 2006

On Debating

This is an excerpt from an email I wrote today re. a long and so far fruitless debate going on in a mailing list I'm on. (This is a small list of very rational people, so this is not your usual pointless, frothing mailing-list banter--hence there is actually some reason to be surprised it has been fruitless.) It begins with some backstory unrelated to that particular debate:

Some guy emailed me out of the blue, having found my journal or something and deciding that I would be an interesting philosophical sparring partner. He posed me, essentially, Zeno's paradox, though I realized eventually he wasn't aware it already existed as such. I answered him straightforwardly, and he replied arrogantly and condescendingly that my answer was crock, he had THE answer, and perhaps I would like to try again? He provided numerous complaints about my answer, and I addressed them all, one by one. His reply in turn was to provide numerous complaints to each of those, and so the whole thing was just blossoming exponentially into a useless mess, mostly (and rapidly) off topic, even though in some sense every point under debate was seemingly a necessary prerequisite to resolving the original question. He eventually provided his answer, which I immediately dismantled (it really was crock, and I have no doubt you two would heartily agree), but of course he found as much fault in these explanations as in my defense of my original answer to his "riddle". All in all, the experience was very much like this [other] debate, with both sides finding--and providing as far as the other side was concerned--endless cascading nonsense.

Most people would have just walked away fairly early on, but I was finding it a fascinating study of people's ability to rationalize. He was, I would say, a Mensa-level thinker, with all that [choice of wording] implies. In this particular case (unlike the [other] discussion) I had such confidence in my position, I felt comfortable focusing entirely on the meta level of the discussion, purely manipulating the manner in which I interacted with him, and specifically seeking to eliminate the sources of noise that were allowing a very concrete and easily resolved matter to blossom and multiply like an invading fungus.

One thing I realized is that it was absolutely a mistake to answer all of his points, whether to agree or disagree. Each of his letters was long and full of claims and implications, some with obvious flaws, some correct but irrelevant, and some both flawed and irrelevant. In particular, often his letters would contain at least one glaringly bogus line of reasoning that begged to be debated -- but often this was not the most important point to the original issue. To address all of his points, or even many of them, was just inviting an exponential growth of noise.

I resolved to pick from each of his replies what I believed to be the single more central fallacy, and to simply ask him about it. It was with great restraint that I did not answer Two or Three or More points, but literally just one--the one that I thought was truly most important. Not the one that was the most wrong, nor the one I had the most to say about, but the one that was the most important to the central question.

And I absolutely minimized my interpretations of his words, drew no conclusions on his behalf, and only put A and B together for him explicitly, with reluctance, and wherever possible using his own words or as close to is as possible--being in particular very careful to phrase things in a way which he would agree with.

And so I would send him one or two line emails, and he would reply with big, long emails shooting off in ten different directions, and I would carefully analyze them for the most central fallacy, and reply to that with another one or two line question, occasionally pulling in a central point from two separate emails of his so that I could ask him about how these two things he claimed fit together...

And fairly quickly in this cycle he replied for the first time with a letter I found to be downright mysticism! And at first I was disappointed that it had so quickly ended here... but he seemed like such a (wants-to-be) rational person, I decided to press on anyway. And that letter was the hardest to reply to because it took a great deal of analysis to discern how he was rationalizing this in his head, and took a great deal of discipline to remain objective and non-dismissive about, but at last I brought the appropriate question forward, and it seemed to break through that particular stone wall and the debate continued more as before for a few more cycles until he finally declared, from his own reasoning, essentially the same answer I had given him at the start, and he acknowledged, reflecting on his own argument, where his fallacies had been, and he apologized for being an arrogant prick. And of course, I haven't heard from him again since.

The most interesting aspect of this to me is that I am absolutely certain that had I addressed even multiple of his points at once, let alone all of them, it would have never ended. Given any excuse to do so, people will evade the point most centrally undermining to their own, and not with conscious intent. I believe, quite simply, that people search in their minds for responses, and that this search turns nothing up where there is real conflict--and this is a non-event like not seeing an elephant that isn't there, and so never crosses conscious awareness.

Meta commentary possibly of interest to both of you: I am ever more convinced that human cognition is synonymous with perception, and that perception amounts to model-based synthesis limited by immediate percepts-- i.e., that the way we think about everything is to make it up based on what we already believe, with fabrication choices made in accordance with what will make us synthesize what we are seeing. That is, I believe people are wholly incapable of seeing what they don't already believe, with the only exception to this rule being Learning, and that Learning is highly constrained to only happen within a certain tight boundary around the frontier of our existing knowledge, thus again limiting it to things essentially touching what we already believe, and furthermore that it doesn't take effect until LTP sets in--a minimum of eight hours--which means that people are essentially incapable of understanding, on-the-spot, anything that violates their existing structure.

What this means in practical terms is that once someone's beliefs have diverged sufficiently from the evidence (or indeed just from someone else's beliefs)--and "sufficiently" is a *surprisingly* small distance--this gap becomes a filter that omits further evidence from being statistically integrated at the intuitive level (or, that is, causes it to integrate differently for the two people--at least one of whom, by the way, is objectively flawed at that point, since when things are working properly this gap should never form in the first place; but all it takes to create it is sufficiently narrow experience for a sufficient duration of time). In other words, *no amount of evidence will change someone's mind* unless sufficient evidence can be presented in a compact enough format in time and space to be held and consciously considered in short-term memory, and even then this process (whereby they acknowledge the rational conclusion given the evidence) will have to be repeated many times before it has a hope of competing with prior faulty statistical integrations. The only exception to this, ironically, is anything that can be slipped into presuppositions of anything they are willing to entertain as intuitive truth--even if just a work of fiction which is indulged through willful suspension of disbelief. And conversely, most forms of debate will only *firm* someone's intuitive convictions (i.e., internal biases--even if wholly incorrect to begin with) because of the constant and intentional positioning, out of ego defense, of the status-quo as correct and the challenge as faulty. I.e., reverse psychology has much deeper and longer-lasting impact than any conscious argument, and I suspect this is often what drives a wedge into initially small gaps and turns them into dogmatic abysses.

For this reason, I believe the only, or at least best, way to make any progress in a debate around complex issues is to work hard each and every cycle to distill down to the most central point, and to keep claims compact and concise, and wherever one does present evidence, to take the time to gather enough of it and with an acceptably objective method, so that in one shot it can be construed as compelling, or at least meaningfully debated as such. In particular, I do not believe in addressing every point in an email, I believe in addressing THE point to your best ability to discern it. It is better to address one point and have the opponent declare that that was not the point (then you just try again) than to address a tangent and have them do the same, ad infinitum.

I have been endeavoring to do this more lately myself, but I am far from making it a habit, and it takes a great deal of self-restraint to ignore tantalizing tangents. Certainly I believe it's polite and necessary to answer a direct question amongst few or any that has been asked twice, but I don't consider all questions amongst many to be asked with equal import--mine or others'. (Many questions in debates such as the [other] one are ambiguously rhetorical.)

In short, I think we should be addressing fewer of eachother's points, not more. But we should choose those points to be as truly central as we can make them--not simply choosing the easiest point to win--and we should distill our address down to the least presumptive form we possibly can.

It should be apparent from all of the above that this declaration is first and foremost a critique of my own style and a reminder to myself, but I would also like it applied toward me as well as by me.

The hardest remaining part for me is I have a very hard time figuring out what to do with it when someone presents something they see as clear evidence or a clear explanation of XYZ and I fail to see the connection at all. This is like when that guy came back with what I called his mystical reply -- sometimes there seems simply nothing to address because whatever discrepancy there is in premises is so far removed from the immediate claim that there is simply no common ground upon which to evaluate it. One is tempted to seek the underlying premise discrepancy here, but this in itself can become a tangent to the core question. I still don't know how to resolve this.

[<< | Prev | Index | Next | >>]

Simon Funk / simonfunk@gmail.com