Online course review: How to Reason and Argue

Not quite two years ago, I took a couple of online courses from the new program/organization/school/site Coursera, which were quite interesting. The idea of open-access college-level courses is tricky; while it reduces the costs of education and makes it accessible to loads more people, the ability to accurately test participants and eliminate cheating is problematic right now. One of the courses that I took was on genetics and evolution, and was immensely informative. I scored lower than I wanted (or thought I would,) and it appears this might have been an issue with a lot of people, since they changed the grading structure almost immediately upon seeing the results of the class, making it a little more lenient. These are, after all, the first, experimental versions of these courses, so there are some teething pains. Curiously, though, that particular class is now available as a legitimate, college-credit course, meaning you can apply it towards a degree.

I’m going to talk about the other one, though, titled, Think Again: How to Reason and Argue. I figured this would be right up my alley, but ended up dropping the course after only a couple of weeks, and it was entirely due to the structure. In fact, I tried again, thinking that I should give it another chance, and re-enrolled for the class as it started again on August 25th of this year (it’s still going on as I type this) – I didn’t even make it that far, getting supremely fed up by the middle of week two. Seriously.

First off, it was interesting how different the class structures were. While both operated as online video lectures and presented weekly quiz questions, the genetics course was quite detail-oriented and required no small amount of problem-solving – it was clear you were expected to work hard on the class. The reasoning class, on the other hand, proceeded much slower and reiterated things, as far as I was concerned, way too much – I got the impression the students were expected to be a lot slower to grasp the concepts, and the presentation is actually condescending in its delivery.

Even if this could be ignored, there was the approach. For some reason, instructors of language courses seem to believe that diagramming sentences – preposition, verb, active noun, and all that horseshit – leads to a greater understanding of language, and it was this structure that the professors of the reasoning class adopted. From my standpoint, this is the way you teach a computer how to ‘understand’ language, but it’s an inherent part of how people learn to talk and write and doesn’t gain anything from being diagrammed – I couldn’t tell you what a predicate verb is and have never in my life had the faintest reason to determine one. When learning another language, it may be useful insofar as sentence structure is different from what someone learned in their youth, but that’s a translation thing.

But let’s be real. No one who is trying to learn how to reason and argue is going to sit down and diagram a fucking sentence – there’s no point to knowing how the premise relates to the conclusion (especially not in labeling the goddamned thing,) and this is especially pointless and unwieldy in conversation. There is the barest value in being able to construct a viable argument yourself, but most people have already learned sentence structure in grade school, and it certainly does not require repeated exercises, even just demonstrated within the lecture, of partitioning off sentences. Because the solid, useful part of reasoning and arguing (I lean towards words like “discussion” and “debate” myself) is presenting a solid, unflawed line of reasoning for ourselves, while spotting the flaws in other people’s points. So one doesn’t need to know whether an argument is inductive, deductive, or conductive – they need to know how to spot the subtly misleading aspects, the flawed premises, the assumptions, and the logical leaps.

Let me give an example. I recently came across the statement, on a forum, where someone argued that finding mundane explanations for the various Loch Ness Monster sightings (logs, lake sturgeons) does not mean a monster doesn’t exist. Someone else called that illogical, but that’s incorrect; it’s perfectly logical. But it sucks as an argument all the same. The flaw is in treating ‘Loch Ness Monster’ as an entity, rather than as a cultural artifact that requires much more extensive evidence (like a carcass) to establish as something beyond folklore. Worrying about the logic in the sentence does nothing to reveal the error in the approach, which lies with attitudes, assumptions, and ignoring the weight of probability. Another demonstration of how useless logical structure is within arguments is in an example that I used previously: “We have no evidence for gnomes, therefore gnomes do not exist.” This is an illogical statement – the conclusion does not follow from the premise – but it tells us nothing about the existence of gnomes either way. The functional way of dealing with the topic is to simply ask what evidence we have to demonstrate that gnomes exist in the first place.

[I have to insert a brief elaboration on this aspect. Logic and reasoning cannot ever be considered proof, of anything – sorry, philosophy majors, but it’s true. Our history is loaded with examples of logical arguments and even mathematical equations that, quite simply, fell flat when they encountered the raw facts. Logic is only as good as the information it predicates upon, and that’s always imperfect. That’s why we look for hard evidence, everywhere.]

Moreover, the course contained no mention of the kind of things that I would have thought would come first, right out in front, such as the difference between persuasion and competition. Many, many discussions, debates, and arguments are only forms of competition, and to be blunt, such an attitude is unlikely to produce anything positive – even if your case is overwhelmingly compelling, your ‘opponent’ isn’t going to concede the point, because that’s admitting defeat. Good discussions have to be free of animosity and competition to the greatest extent possible (which is often not very far, but that’s mankind for you,) and this takes a very specific, very deliberate approach. Carl Sagan was marvelously accomplished at this, in that he almost never tried to prove a point, but instead asked pertinent questions, leading down a path that demonstrated the flaws without ever accusing someone of missing them.

Another contribution by Sagan, now adopted quite widely among skeptics, is a list of debating fallacies often called the Baloney Detection Kit. While I have rarely ever broken down an argument by structure, I have constantly used portions of this list – I certainly wouldn’t consider it all that someone would need, but the points therein are encountered so frequently that not using it is putting anyone at a distinct disadvantage in a debate. Feel free to put this down to a difference in opinion, but rather than spending weeks studying basic structure, I’d bring up common debating fallacies on day one, and revisit it constantly.

Yet another key aspect of debate is being able to find the emotional triggers that cause people to have such strong opinions in the first place. Most arguments have an emotional bias – that’s the way humans are, we attach feelings to ideas. But we very often fail to recognize this for what it is, and believe every opinion that we hold is the result of rational consideration. If we arrive at a decision by thinking it is the best conclusion given the information we had at hand, we (by all rights) should have little difficulty changing that decision given further pertinent factors – yet this is rarely seen, is it? However, decisions based on how they fulfill some emotional need are incredibly hard to change, and standpoints arrived at without rational process are very unlikely to be changed by rational process – think about such topics as vegetarianism, abortion, and religion. Addressing these usually requires the ability to demonstrate the lack of a rational process, or to locate the incorrect assumption that was built upon. Vegetarians may insist that it’s wrong to harm animals, but who determined this, and why? Everything dies, and in ‘the wild,’ this is very frequently not pretty. Moreover, humans are not an introduced species, anywhere on the planet, so there is nothing unnatural about our actions, regardless. Such points reveal a few of the assumptions that formed much of the bedrock of the arguments – once dismissed, the arguments are inherently weaker.

To be sure, perhaps at some point later on in the course, the instructors manage to address these, and other factors such as manipulative phrasing, appealing to emotion or ego, circular arguments, confirmation bias, and so on. But, I’m skeptical. In two other courses that I’ve taken, (genetics and cosmology, both from Coursera,) the instructors were able to cover fundamentals very quickly, and were involved in fine details within a few videos – less than an hour of class time. There was no condescension, no pointless reiteration, and no time wasted on establishing extremely basic information. I find it hard to believe that something as simple as debating effectively could possibly require far more setup than these very specific, elaborate sciences. I have my suspicions as to why the difference is so marked, but they remain only suspicions; regardless, the poor approach and the painfully long time to get past certain simple points make this course a complete miss for me. There are far better, more efficient ways to learn how to debate, persuade, and produce cogent arguments.

*     *     *     *

I haven’t yet decided if I’m going to link to this review in the course forum, or perhaps send it to the instructors. I imagine that if I do either, there may be some challenge over whether I know more about the topic, or could do a better job or whatever. I’ll let anyone decide that one for themselves – here are links to several previous posts, any of which could be read in about the same time as the average video lecture provided within Coursera.

Proverbial thinking

The exception proves to rule

Hooray! I scored a “Not Negative!”

Fear of the knowable

For a given value

Unevidence

There are skeptics, and then there are skeptics

And, for giggles, a couple of examples that I’m particularly pleased with:

Dealing with the real world

Too smart to be intelligent

Comments are closed.