It's not like first and second semester students deploy every possible logical fallacy. They actually tend to stick with just a few: Straw man, slippery slope, and post hoc ergo propter hoc.
And they're not even mistakes qua mistakes -- as errors or accidents -- but as expressions, I think, deep expressions, of our assumptions and common presuppositions about the way the world works. Niklas Luhmann talks about the space of rationality, the span in speaking and in society where something is taken as or counted as rational, and says "logical" is defined, for example, but the outer limits of paradox and tautology, that is, the space between these two places, the limits where logic twists in on itself (warps, crumples, inwardly crushes or implodes, like under the forces of gravity, but with a spectrum similar to color) and it is still logic even then, but mangled, and a good border for the range of what we call logic. I think that's interesting in the context of common student essay errors, in that sometimes, actually, the reason this bit or that bit of logic is wrong is built exactly on the reason or the assumption or the belief that also leads to a right bit of logic being right.
Put it another way: If you dig down past the error, look past why, precisely, the claim in the third paragraph or that particular conclusion was wrong, and try to come up with why these fallacies are common and constant and sometimes seemingly intractable, you come to something common in all of us. I think you come to the same assumptions that give us any space of rationality at all.
Like straw man. Underneath, below the flawed abstraction of fake opposition, isn't the assumption that being right is connected (somehow) with being good? And isn't this an assumption we all make, and should make, and need to too -- the assumption that makes working to be right and trying to really think through and discern what's right seem important?
And slippery slope: Isn't the basic kernel of an idea that little bad things lead to big bad things, and small, theoretical errors lead to public and very actual disasters, and isn't that essentially right, and also essential for a functioning society? Isn't this an assumption I want people (like say, those who build oil drilling rigs in the ocean, or those who decide what can be sold and how on Wall Street) to make?
Those two fallacies are actually pretty easy to correct. Or, maybe what it is, is, it's not too hard to show the students how the correct assumption about reason can be separated pretty cleanly from the particular error of a lack of charity, or apocalyptic extrapolation. They seem to get that, once it's explained once or twice, and correct themselves and their arguments.
The third one is harder though. More insistent. And I wonder if that's because it's not so easily distinguished from the supposition I have to and need to make for the sake of reason and argument and rationality. It's true that too the Latin tends to intimidate. It can communicate threatening complexity and "you can't understand this." Probably I should call it something else, like the "lucky shirt fallacy", like begging the question seems easier to understand or feel like you understand than petitio principii, instead of just sticking with the long Latin name and translating it and trying to hammer hard the question to ask themselves, "how does this cause -- cause cause cause cause cause -- that?"
But isn't it, underneath and at root, an unanalyzed belief about narrative? Isn't it the basic idea that meaning unfolds in time? That previous and prior events cause and create and explain later ones? Note that no one makes the mistake of causally connecting co-occurant events. Or, they do, but only by asserting that the one happened first and then the other. In a really common example of the error, for example, you say "there's a woman with an umbrella, a dog, and then a dump truck, but did the woman cause the dog and the dump truck," which assumes a sequential relationship even in the effort to undermine the assumption of a sequential, causal relationship. The first time I heard post hoc ergo propter hoc explained (by a preacher on tape) it was with the example or peanut consumption in D.C. as the water level of the Potomac, which seem to have an inverse relationship, as every time sales of salted peanuts rises, the water in the Potomac drops, and when the sales slack off, the water rises again. Of course the connection is only incidental, and it's not that peanut-thirsty Washingtonians are drinking up the river, but that the river goes down in summer, when it's hot and doesn't rain, which is baseball weather, which is when the most peanuts are eaten in the nation's capitol. The error, here, can be conceived of almost entirely as a question of the misarrangement of events. The intuition that chronology equals causation, or, really, reason and order, which we take as causation, seems unshakable.
Even when there's a good argument to be made, for example, about how a later event caused, or, more loosely, created or constructed an earlier event -- and since I identify with poststructuralism and as I've been reading some of the historography of Hayden White and the cultural studies of Stuart Hall, I think there is, in fact, a strong (though maybe limited) case to be made that the past is never the past as such, but is a construction and perception and an artifact of the present -- that seems so counter-intuitive as to require a kind of suspension of belief. It seems so wrong and obviously wrong that it serves as a caricature of silly postmodernists who reject capital-T Truth.
There's a moment in Stephen Hawking's Brief History of Time where he suggests that time is connected with the expanding universe and that as that expansion might eventually slow down and pause and then reverse itself and start to collapse and contract, time might run in reverse. I'm pretty sure he suggests that time won't actually run in reverse but that that's a good metaphor for Big Bang cosmology expectations (or anyway that's how I understood in my freshman astrophysics class), and in the film there's a scene where everyday scenes are shown in reverse, so a smashed cup leaps back onto the table and reassembles itself. What's so freaky and sci fi about this of course is the idea of time in reverse, of meaning going backwards and causation going that way too.
And even the kind of language I've used here like "underneath" and "at root" participates in that same assumption of chronology and narrative, that idea (a reflex) that earlier events explain later ones.
I wonder if this doesn't mean, though, and if the intractability and commonality of the fallacy doesn't mean that narrative is logically "prior" to logic. Or that logic is always involved in a context of narrative. That logic itself isn't a kind of strange or strained version of narrative. The idea of meaning unfolding in time certainly asserts (itself) equally in the valid and invalid logical moves. So that "underneath" it all are narratives, and the form of narrative as explanation, as the form something has to take to count or be considered as an explanation. We treat this central space and this place where we operate as if it were the site of rationality and even, à la Jürgen Habermas, a consensus of reason and what will be considered as reasonable, but it seems that we can easily lose logic, we can shake it, as if it were an armature spy on our tail, but narrative and that form or framework sticks with us.
If logic is the space between tautology and paradox, isn't it true that both of those borders are still understood even if they are the outer edge of logic because they're still well within the space of narrative and those assumptions of the form that that's meaning? It's normally true, too, that what is called non-narrative or anti-narrative is still very much narrative (as my uncle said, even a sentence is the unfolding of meaning in time, and consider as an example the three narratives in the sentences in the line in Oz: "Medicine is not narrative. Blue trashed sky. Firemen on ladders into the smoking night.") This structure is pervasive and certainly pervades even our grammar, which is how I would explain, for example, hypotaxis.
This is, of course, an old point for me -- that we are embedded in stories. Which is not to say they're natural or formless or unconstructed, or that they can't be criticized or analyzed, but just that is there is no outside of narrative, there's no translation from this to something else, there is no encoding and decoding as there's no non-narrative code that we could know. Even within logic and when logic breaks down, the structures of story are still there. Logic is also embedded within the assumptions of narrative. Hayden White notes in Tropics of Discourse that "Our discourse always tends to slip away from our data towards the structures of consciousness with which we are trying to grasp them," which gets much more complicated when we realize that the "data" isn't free and floating separately from the structure either, and he note that tropes, for example, are fundamental even in valid syllogisms:
"The move from the major premise (All men are mortal) to the choice of the datum of to serve as the minor (Socrates is a man) is itself a tropological move, a "swerve" from the universal to the particular which logic cannot preside over, since it is logic itself that is being served by this move. Every applied syllogism contains an entheymemic element, this element consisting of nothing but the decision to move from the plane of universal propositions (themselves extended synecdoches) to that of singular existential statements (these being extended metonymies)" (italics original).The idea is, then, that we are radically trapped within narratives, within that range of forms and those assumptions of meanings. Which isn't, I don't think, as bad as it might sound. We still, for example, can reject as wrong the errors of post hoc ergo propter hoc, but maybe it would work better if I was making the case not by an appeal to the rarified and limited space of logic, and instead tried to explain it in terms of the aesthetics of story.