« Live video streaming with free... | Home

Ad stramineum hominem

Sun 25 Mar 2018 by mskala Tags used: , , ,

Sometimes I find myself on the receiving end of false accusations of "straw man" argumentation, and it feels like this happens abnormally often to me in particular. It's baffling because when it happens, it doesn't make any sense.

"Straw man" is a technical term in the field of rhetoric and it has a specific definition. It refers to a tactic sometimes witnessed in politics but it's not something I expect to be relevant to more private contexts involving people who are making any pretence of good faith. It's also not hard to recognize when "straw man" applies and when it doesn't. The definition is not complicated or ambiguous. So if someone implacably asserts that I did it, and I did not, then it's not clear how we can possibly proceed any further. Typically they present what they say is evidence in the form of quotes from me that are literally accurate but have no obvious relevance to the accusation, and then they follow simple non sequiturs from there to how I am somehow at fault.

But when someone does make this claim it's clear that they think what they're saying does have a meaning, even though any attempt at elucidating its meaning seems in practice to be doomed. "Straw man" isn't just an opaque insult. It is used as if it were the technical term, even though not in its standard meaning. And these are people who appear to be of at least ordinary, usually significantly higher, intelligence, without obvious cognitive deficits. "I am talking to someone who is just plain nuts!" isn't the most natural interpretation of such a case; and yet often the observed facts of their behaviour are hard to fit into any other interpretation.

I have tried over the years, probably more times than I should have, to discover what such people are really talking about (on the assumption their words have meaning at all) and even to remain open to the possibility that I may really have committed "straw man" unknowingly when I've been accused of it. But that doesn't work. Once this particular false accusation comes out, it's basically game over. After that phrase is on the table the nonsense just comes thicker and faster as long as I attempt to assume good faith and as long as I try to act like I'm talking to a rational person. Every damn time.

In the best such cases there's a chance to drop the line of discussion that was going nowhere and backtrack to some earlier point, avoiding whatever my partner thought I transgressed upon. But that's rare. Usually the best I can hope for is to simply end the discussion immediately. And sometimes they aren't willing to drop it, but keep on trying to engage with me and claim some sort of victory. I've ended up breaking off contact with more than one long-time friend because of repeated and public false accusations of "straw man" argumentation aimed at me. There seemed to be no other way I could end their nonsense.

The scale of consequences from it suggests that this matter is something important, and it would obviously be interesting and valuable to know more about what's going on here: such questions as what people think they're talking about when they accuse me (or anyone) of "straw man" argumentation if the definition does not apply; why they think the behaviour of making false accusations of "straw man" is worthwhile or even in any way acceptable; whether there's anything I really am doing (fallaciously or not) that makes me more likely than others to have such accusations aimed at me; how to effectively deal with them when they occur; and so on.

Although I don't have good answers to all of these questions, I recently made some progress on them which seems worth sharing. I'm indebted to Julia Galef for the following comment posted on Twitter, which shed some light on the situation for me.

The only time you can confidently call "straw man" is if someone's claiming to depict a specific person's argument.

But if they're just describing an arg "some ppl" make, you should hesitate to call straw man. Just bc *you* haven't heard ppl make that arg doesn't mean ppl don't.

She's right - the definition of "straw man" is or should be that it's a false characterization of someone else's (usually your partner's, and necessarily a specific person's) argument, in order to make that argument easier to refute. But when I'm accused of doing it, that definition doesn't apply; instead, it appears that the accuser is trying to use an alternate definition under which an absence of detailed attribution is just as dishonest and unacceptable as false attribution. And very often connected with the "you've committed straw man!" claim is a demand - which I hadn't even noticed was part of the pattern before, because from my point of view it was lost in the ensuing spew of nonsense - for "citation" or "examples" of exactly who said the point I had been addressing, just as if who said it, meant something. "Not you, obviously!" is a good enough answer to that question as far as I'm concerned, because that alone is proof I did not commit "straw man." I did not falsely attribute a statement to my partner that they did not make. But my saying so has never calmed anybody down.

And I would go a step further than Galef does in the last sentence of the quote. She says that "straw man" is only properly applied when someone says some specific person said something that they didn't. I agree with that. And "straw man" is not properly applied when someone says "some people" (not identifying them specifically) say something without naming who those people are, and she says that's because it could well be true that such people do exist even if you don't know who they are. The world is a very big place and many people in it say many different things.

But although true, the fact such people may and probably do exist is not even the reason that this scenario is not "straw man" and not fallacious! Because I don't care whether these hypothetical people exist. "Some people say" is not "straw man" not because such people do exist, but because the question of their existence is not relevant to the argument. I was talking about the thing being said; the question of whether it is a literally true fact that some unspecified person exists somewhere who once said something, is not an interesting question and has no bearing on whether the thing they may have said was itself true. It can be a proposition worth considering even if nobody said it. References to such persons are better regarded as part of the grammar of the language of argument (like the unspecified referent of "it" in "It is raining") than as first-class assertions whose own truth value is worth considering. The truth or falsity of propositions does not depend on who said them, nor on whether anybody said them at all. Or... does it?

Thinking about this some more, I can put it into a larger framework. There are people who have very different ideas from my own about what arguments are, what arguments are for, and therefore what is and is not acceptable conduct in the context of arguments; and from those different ideas it's possible to construct a definition of "straw man" that although nonstandard is more consistent with these people's theory of argument than the dictionary definition of "straw man" would be. This alternate definition and the theory of argument into which it fits count as unacceptable, behaviour that I see as perfectly okay and even admirable, and that behaviour is what I do that people have in mind when they (falsely, under the dictionary definition) accuse me of "straw man."

And there's another relevant pattern, which is that people who are inclined to accuse others of "straw man" are the same people who often commit "ad hominem" and don't seem to see anything wrong with that. My own perspective is that "ad hominem" is nearly unforgivable. It's one of the worst things you can possibly do in an argument. So the combination of "ad hominem" with a false accusation of "straw man" (which accusation itself often becomes an "ad hominem" too) is deadly to my interest in continuing a discussion.

The picture is that I subscribe to a much narrower definition of "straw man" than some people I've talked to apparently think it makes sense to use, I seldom accuse others of doing it, and I don't regard "straw man" as so very big a transgression even when it really has been committed. Meanwhile, there are others playing by very different rules which are hard to reconcile with my own rules and which lead them to commit nearly unforgivable "ad hominem" argumentation with distressing frequency. It's like the two issues of "straw man" and "ad hominem" are complementary. People who care strongly about, and accuse others of, either of them seem not to care so much about the other.

I will try to lay out what's going on on both sides in a way that allows for some possibility of comprehending it. Please understand that to do so I will have to describe some positions I don't actually agree with and cannot defend very far. I am trying to answer the question "Why do people think this?", with the assumption that such people are rational in the sense that they have beliefs and believe the consequences of those beliefs, but I am not trying to convince you that they're right.

Dialog 1

Consider the following sample dialog.

1.1 Plato: Socrates said that the Moon was entirely made of Athenian cheese.

1.2 Plato: The priests of Apollo visited the Moon and brought back samples of its substance.

1.3 Plato: And the samples were profoundly uncheese-like.

1.4 Plato: Therefore the Moon is not entirely made of Athenian cheese.

1.5 Plato: [Therefore Socrates was wrong.]

1.6 Socrates: I did not say the Moon was entirely made of Athenian cheese.

1.7 Socrates: Therefore premise 1.1 is false and the inference 1.5 from it, has not been proved.

1.8 Socrates: Plato has committed the straw man fallacy and is a cad and bounder.

1.9 Plato: Whether I am a cad and bounder has nothing to do with the Moon, and my inference 1.4 follows from premises 1.2 and 1.3.

1.10 Plato: Socrates has committed the ad hominem fallacy.

It seems pretty clear and we might all agree that Plato has committed "straw man" here. This one falls under Galef's definition for a legitimate "straw man" accusation: Plato was claiming to depict a specific person's argument. Assuming, as I will, that Socrates' protest 1.6 is factually true and he really didn't say what Plato said he said at 1.1, then Plato has falsely characterized Socrates' position, and subsequently attacked it, and that's not something Plato ought to do. However, it's worth asking what Plato's point really was. He seems to care most about 1.4: a statement about the Moon, which as he points out is independent of 1.1. Socrates instead reacts to 1.5 - a statement about who is right and who is wrong - and treats that as the important part of the argument.

I put line 1.5 in square brackets on purpose, because I think in a real discussion, Plato would not even say it. It is a logical implication of other things he said, so he cannot legitimately claim it was not part of his meaning. But both because 1.5 is not important to Plato, and because he can guess that it's likely to annoy Socrates, Plato would leave it unsaid. He thinks this whole argument is about the Moon. Socrates thinks this is an argument about who is right and who is wrong. For Socrates to then start trying to talk about 1.5 - a statement that Plato probably did not even make directly, and doesn't consider important - it seems to Plato that Socrates is dragging the discussion off on a random tangent. But it seems to Socrates that no subject matter can be more centrally relevant, because the whole argument right from the start was about who was right and who was wrong.

Socrates then goes on to commit a clear "ad hominem" by calling Plato a cad and bounder. He might defend that action by saying Plato broke the rules first, but accepting such a defence depends first on there being an objective or agreed concept of "the rules," and second on those rules being aimed at some goal of fairness to the participants that overrides goals like determining what the Moon is really made out of. An outsider whose main goal was only to determine the substance of the Moon might not care very much about Plato and Socrates as persons, and their status of having hurt feelings or being fairly treated, as long as it's possible to extract from the record the truth value of propositions about the Moon.

The core of my own point should already be clear: that persons with different ideas of the basic purpose of argument can follow their own ideas in a way that seems legitimate to themselves and is internally consistent, but that leaves them accusing each other of defective argumentation.

Dialog 2

This next dialog is more typical of what actually plays out in real discussions.

2.1 Plato: Many citizens believe that the Moon is entirely made of Athenian cheese.

2.2 Plato: The priests of Apollo visited the Moon and brought back samples of its substance.

2.3 Plato: And the samples were profoundly uncheese-like.

2.4 Plato: Therefore the Moon is not entirely made of Athenian cheese.

2.5 Socrates: Who, exactly, believes that the Moon is made entirely of Athenian cheese?

2.6 Plato: It is said in the agora by many citizens, but their names aren't relevant.

2.7 Socrates: Since no examples have been presented showing that such citizens really exist, Plato has committed the straw man fallacy and is a cad and bounder.

2.8 Plato: Who expressed a belief about the Moon has nothing to do with the fact of its substance, and neither does the question of whether I am a cad and bounder.

2.9 Plato: Socrates has committed the ad hominem fallacy.

Did Plato commit "straw man" this time? He did not falsely characterize Socrates' (nor any specific person's) argument, so it is not the same thing he did in the previous dialog. Calling his acts in both dialogs "straw man" entails applying that label to two quite different acts that may have different ethical status.

Plato's statement 2.1 may well be false. Socrates didn't say the Moon was made of cheese, but short of conducting a referendum in the agora, we don't know what other citizens may believe about the Moon. Galef's comment would suggest that accusing Plato of "straw man" in this case is bad because 2.1 (and its affirmation 2.6) could well be true - maybe many citizens really do believe that. But it is also reasonable to guess that maybe they don't.

Someone sympathizing with Plato could go further: it doesn't matter whether those propositions about what "many citizens" believe are true, calling "straw man" against points 2.1 and 2.6 is bad because the point of the argument was 2.4 and 2.4 does not depend on 2.1 and 2.6 being true. The point of saying 2.1 was only to establish why Plato cares to talk about the Moon at all - he thinks the Moon is a topic of current interest. Discussion of exactly who said the Moon was made of Athenian cheese in particular, whether anybody really did and exactly what they said, and whether Plato is a cad and bounder, are all irrelevant distractions from proposition 2.4 and Socrates' comment 2.7 is, if not "ad hominem," at least a "Chewbacca defense": a rhetorically defective attempt to shift the argument onto irrelevant distractions.

On the other hand, someone who thinks that "straw man" can include attacking the ideas of "some people," might consider Plato's defective argumentation to be worse here than in the previous dialog. In 1.1 he said something false, but easily testable. It is a matter of objective fact whether Socrates really said that about the Moon, and Socrates himself is present to correct any errors. The statements 2.1 and 2.6 are vague, difficult and expensive to test, and maybe not even testable in principle (how many citizens is "many"?). That makes them less useful contributions to any argument. Any individual citizen who says, "I didn't say that about the Moon!" could be conveniently dismissed with "I didn't mean you," and so Plato may never have to face any serious opposition to his position. This all adds up to Plato gaining exactly the unfair advantage associated with "straw man": he can demolish a weak argument instead of a strong one. And that makes it look a lot like an example of "straw man" to anyone who thinks "straw man" should be banned by reason of its conferring an unfair advantage.

Socrates' focus on "examples" in 2.7 highlights the difference and is typical of persons with views of argument similar to his. He might just as well say "citation needed." For Plato to present examples could possibly establish the truth of Plato's statement 2.1; but Plato is unlikely to be willing to put in any effort to produce such examples, because they would at best establish the unimportant framing point 2.1, with no relevance to the point 2.4 that he considers to be the goal of his argument. Plato is especially unlikely to be willing to cooperate with Socrates' wacky random demands after being called a cad and bounder (2.7).

The difference in goals creates a difference in opinion on what Plato's responsibilities are. Both sides may agree that Plato should produce evidence for his claims, but he thinks that means evidence of the Moon's substance (2.2 and 2.3) to support his claim about the Moon's substance (2.4); Socrates thinks Plato should produce evidence about what "many citizens believe" (2.1) and that a foul has been committed because Plato refused to produce any such evidence, and in fact compounded his transgression making the same evidence-free assertion a second time (2.6) after he had been legitimately called out on it.

Dialog 3

Now consider the following variation, which Plato might initiate in a doomed attempt to avoid Socrates' relentless cry of "straw man, straw man!"

3.1 Plato: Nobody believes that the Moon is made entirely of Athenian cheese. Let us consider why not.

3.2 Plato: The priests of Apollo visited the Moon and brought back samples of its substance.

3.3 Plato: And the samples were profoundly uncheese-like.

3.4 Plato: Therefore the Moon is not made entirely of Athenian cheese.

3.5 Socrates: Plato admits that nobody believes the Moon is made of cheese, so refuting such a belief is a waste of everybody's time.

3.6 Socrates: Plato has openly committed the straw man fallacy without even trying to conceal it, and is a cad and bounder.

3.7 Plato: What the Moon is made out of is worth knowing, and whether I am a cad and bounder has nothing to do with that. It also doesn't matter whether anyone believes a given proposition, for the question of whether the proposition is true to be worth considering.

3.8 Plato: Socrates has committed the ad hominem fallacy.

Let us suppose for this dialog that it is factually true, and agreed on both sides, that nobody really believes the Moon is made entirely of Athenian cheese (3.1). Socrates could attempt to contest this point by introducing citizens willing to swear they do believe the Tyroselenian Hypothesis, but after doing so he could not easily accuse Plato of attacking a "straw man," and it's that accusation that interests me.

Given nobody believes the Moon is made of cheese, from one point of view (as Socrates says in 3.5) Plato has openly admitted, right from the start, that he is committing "straw man." The position he attacks is held by nobody, thus unfairly easy to attack, and seeking this advantage surely makes him a cad and bounder. This is not only a "straw man" but the worst "straw man" of all.

On the other hand, there cannot possibly be a relevant misrepresentation here. This is not "straw man" in the form from Dialog 1 which everyone agrees is bad - Plato did not ascribe a position to Socrates that Socrates did not take. This is also not what Plato did in Dialog 2, which some might call "straw man" - he did not ascribe any position to unspecified and unverifiable "citizens." Instead he states a fact to which everyone present agrees, namely that nobody holds a certain ridiculous belief about the Moon. So this is yet a third distinct kind of act; and if "straw man" entails misrepresentation, this is definitely not "straw man." It seems unnatural, to say the least, for Plato to have committed any kind of misconduct, let alone a form of misconduct defined by making false statements, when he stated a fact that even Socrates agrees is true.

Summary thoughts on the three dialogs

In the dialogs, Plato has done three different things. Socrates has called them all "straw man," but exactly what that means is different every time. I've described two different analyses for each dialog. Under one way of analyzing the dialogs, Plato was guilty of "straw man" only in Dialog 1. Even then he may have done it inadvertently, and he considers it not a big deal because it has no relevance to the proposition he is proving about the Moon. And in at least the other two, Socrates was guilty of "ad hominem," because he made false accusations of "straw man." Even in Dialog 1, the accusation of "straw man" although possibly true could easily be understood as saying that Plato was wrong on the point about the Moon because Plato is a cad and bounder, and that would be "ad hominem" too, to whatever extent it can be classified as an attempt to disprove 1.4.

It might be claimed that "Plato has committed 'straw man'" is a neutral, not insulting, objectively true or false proposition, and that Socrates with more care could say it without also saying "Plato is a cad and bounder," and then he would avoid committing "ad hominem." I reject any such claim. Among people who have the kind of intelligence and rhetorical education that leads them to make use of formally described concepts of defective argumentation like "straw man" at all, it is impossible to assert that someone has committed such a thing without its being insulting and without casting a reflection on the person's general credibility in everything else they say. All the more so if the accusation is false. As soon as he says "Plato has committed 'straw man,'" Socrates has also said "[Plato is a cad and bounder.]" at least in square brackets. Much like Plato's statement 1.5, the insult is something intelligent listeners will hear, and Socrates can reasonably expect them to hear it and must take responsibility for having communicated it whether he literally speaks the words or not.

In more blunt terms, if you're going to accuse someone of committing logical fallacies off a fucking Bingo card list of them, and you're in company where the concept of "logical fallacy" is even understood, then you're drawing a weapon and you know it. If you claim after doing so that nobody should be offended, you're a neutral lover of truth (philo-sopher) and it's not you who made things personal and it's perfectly reasonable to draw the weapon of fallacy accusations without expecting to be treated as someone who has drawn a weapon, then you're fooling nobody. It is for this reason that I had Socrates call Plato a cad and bounder in every one of my sample dialogues. It is impossible for him to make the "straw man" accusation in this context without doing so in an insulting way.

Similarly, it might be claimed that Socrates is not guilty of "ad hominem" in any of the dialogs because he only said Plato is a cad and bounder, not that Plato's factual conclusion about the Moon is false because Plato is a cad and bounder. "Not all insults are examples of the 'ad hominem' fallacy." I reject that claim because Socrates has demonstrated he is not primarily interested in the fact of the Moon's substance, but rather in demonstrating that Plato is in the wrong; and that Plato is in the wrong is the point Socrates makes with his insult, entailed by the words he speaks and heard by all intelligent listeners, in each of the dialogs. If he claims that he never actually spoke those literal words and was just commenting irrelevantly about Plato's personal qualities without intending any connection from there to Plato's being in the wrong, Socrates is again fooling nobody.

In Dialog 1, Socrates could have said "I didn't say that" without bringing up "straw man" at all - giving Plato the chance to admit and apologize for his error that was irrelevant to the Moon, and then they could have proceeded to discuss the Moon further. But only if they both agreed that the Moon was the real subject of discussion.

Under the other way of analyzing the dialogs, Plato has committed "straw man" in all three, and certainly is a cad and bounder. Moreover, the "straw man" in Dialog 1 - the only one that is a "straw man" at all under definitions like Galef's - is actually the least objectionable example, because it at least contains an attribution, though false, to a real person. The attribution to Socrates in 1.1 is open to verification, and forces Plato to take responsibility for addressing whatever Socrates really said once that's known. The worst "straw man" under this kind of analysis is the one in Dialog 3, where Plato attacks a position so weak he admits nobody really holds it, and thus avoids responsibility completely.

Which is more important - the Moon, or personal responsibility?

Theory of Argument I

At this point it is reasonable to guess that Plato may believe, based on Socrates' behaviour, that Socrates is a bad faith actor deliberately trying to cause trouble; or at best that Socrates is a raving lunatic whose words are literally without meaning. But what if we assume neither is true?

If Socrates in the sample dialogs is rational and not particularly evil, then he must have some understanding of what is going on these dialogs, and in arguments in general, according to which his behaviour is defensible or even admirable. I call the content of Socrates' understanding of these issues, which leads him to act as he does, his theory of argument. His theory of argument might be thought of as the rules he follows, or tries to follow, and (significantly) that he expects others to follow too - in two senses of "expect," as "I believe it is likely that Plato will act according to this theory of argument, and I'll be surprised if he doesn't" and as "Plato is morally obligated to act according to this theory of argument." If Plato does not act in a way consistent with Socrates' theory of argument, then Socrates will be surprised and likely to conclude that Plato is a bad faith actor (a "cad and bounder" - deliberately doing things forbidden by the rules) or a raving lunatic (unable to mentally engage with the objective truth of the facts embedded in the rules and to know what actions the rules require or forbid).

What theory of argument would cause a person to behave as Socrates does in my sample dialogs? I think it might look a lot like the following.

I.1: Arguments are about persons who may be right or wrong in making claims. The purpose of having an argument is to determine who is right and who is wrong, by evaluating their claims.

I.2: Claims by definition are always the responsibility of specific real persons.

I.3: Within an argument, claims must be attributed correctly to the persons responsible for them, because claims without such attribution cannot possibly be useful to the purpose described in I.1.

I.4: It is a foul to attribute to someone a claim that person did not make, because this action demands an unfair responsibility from that person to stand by the consequences of the claim.

I.5: It is a foul to bring a claim into the argument without accurately identifying the real person who made that claim, because this action unfairly allows the person to avoid responsibility.

I.6: It is a foul to bring a claim into the argument that no real person made, for instance by attributing it to a fictional character, because such claims are not well-formed and not relevant to whether real persons are right or wrong.

This theory defines "fouls" in three distinct places, but all the fouls are named "straw man." Using the same name for violations of three different rules may be justified by the fact that the prohibitions of I.4, I.5, and I.6 although itemized separately are really describing behaviours that are bad for the same reason: all three are crimes against the effort to establish personal responsibility, which is an underlying value and focus of the entire theory of argument. Someone who cares deeply about personal responsibility and sees that as the main point, might well see usefulness in itemizing cases of crimes against personal responsibility, while still assigning just one name to all the cases. Such a person, when accusing another of breaking one of these three rules, also might not see much necessity to clarify exactly which one of the three rules is at issue, nor see any important difference in the severity of the violation depending on which of the three rules was broken.

It is not my purpose to claim that Theory I described above is good or that you ought to subscribe to it. I do claim, however, that it is plausible someone might subscribe to this theory, and that if many persons do subscribe to this theory or something like it, that could explain observed real-world phenomena that otherwise might be difficult to explain.

I've said that Theory I is about responsibility, but note that it is also about fairness, especially fairness to the participants. The unfairness of "straw man" is mentioned repeatedly as the reason for prohibiting it. The focus on fairness underlies this theory's unwritten procedure for handling violations of the rules. In all three dialogs, when Socrates calls Plato a cad and bounder - an action by Socrates which would seem to be counterproductive and inadvisable under almost any analysis that treats "productive" or "advisable" as meaningful concepts at all - Socrates' gratuitous insults occur only after Plato has already broken one of the rules. Socrates is, only after and because of Plato's foul, awarded a free throw. Because that's fair. And the justification for the argument to be conducted according to such a procedure is so obvious that it isn't even written into the rulebook.

Theory of Argument II

What would someone think who did not think that arguments are about persons, responsibility, and fairness? Is there a different theory of argument that would lead someone to behave as Plato did in my sample dialogs, with the same benefit of doubt we gave Socrates of assuming that Plato is not fundamentally evil and is not a raving lunatic?

In every sample dialog, Plato thinks he is talking about the Moon. When Socrates makes his "straw man" accusation, Plato objects that it's irrelevant to the Moon and to the factual proposition Plato was proving about whether the Moon is or is not made of Athenian cheese. He seems puzzled that non-Moon-related questions would ever come up. He rejects discussion of himself, and he doesn't even seem to be offended by Socrates' scandalous insult. Plato's objection to being called a cad and bounder is that that's irrelevant, not so much that it's false or insulting. Plato considers his own status as a cad and bounder to have nothing to do with the Moon; and he considers "does this have anything to do with the Moon?" to be a question worth asking.

So it would seem that Plato's theory of argument must place a great deal of value on factual propositions about the Moon - and if we assume these guys ever have other arguments, it is reasonable to suppose that the real value is on facts in general, not only those about the Moon. Issues related to persons and fairness to them, especially himself and Socrates, are not so important. From that and observations of Plato's other behaviour and what he seems to expect or demand from Socrates, we can infer a reasonably detailed picture of Plato's theory of argument.

II.1: Arguments are about factual propositions which may be true or false and are independent of any beliefs and actions of persons. The purpose of having an argument is for the participating persons to determine which propositions are true and which are false.

II.2: It is a foul to attempt to make an argument be about persons, especially those involved directly in the argument, instead of about factual propositions, because persons are not relevant to which propositions are true.

II.3: Such things as propositions about persons, and even propositions about the persons involved in a given argument, do exist; but their use is dangerous to the entire enterprise of argument because, as a matter of human nature, discussion of such propositions renders the participants unable to perform as impartial judges.

II.4: Factual propositions about participants in a given argument are not to be mentioned when there is any possibility of determining the truth of the main subject without them; and when leaving such propositions unmentioned is not possible, the argument should be handed off to some other set of participants in order to avoid a foul under II.2. A participant who wishes to continue an argument after someone else has introduced factual propositions about the participants, must ignore such propositions as if they had not been mentioned at all.

II.5: Wherever possible participants should avoid identifying who said or believes a given proposition, in order to avoid dragging them into the argument and committing a foul under II.2.

II.6: If it is necessary to associate propositions with individual persons, for instance when presenting arguments in dialog form, it is better to refer to fictional characters, or ancient historical figures, rather than specific real persons alive today.

This theory only defines one "foul": that of making the argument be about persons (II.2). "Ad hominem" would seem to be a case of that, but the classic "ad hominem" of declaring one's partner's facts to be false because he's a bad person, is only one way to transgress on II.2. Showing excessive interest in persons even without insulting them, for instance with demands for citations, also constitutes making it be about persons. And despite defining only the one clear-cut "foul," this theory gives a great deal of softer and vaguer advice about what is and is not good, in the matter of how arguments ought to be conducted.

This theory says (II.1) that it is about "factual propositions," and it is; but other things it is also about that are so deeply a part of its fabric they tend not to be mentioned explicitly but taken for granted, include considerations of what is productive or advisable in conducting argument as a human activity. Fairness to persons and especially to the persons participating in the argument, an important value underlying Theory I, is not an important value near the surface of Theory II. This theory pursues the truth even if doing so is unfair or uncomfortable to the participants, and the idea that Socrates' misconduct in introducing an "ad hominem" might be excused because Plato did something bad first, is completely alien to this theory. Even unexcused misconduct can and often should be ignored under Theory II - not because to do so is fair, but because punishing the offender may be of no utility in determining the truth of factual propositions. But it does indirectly acknowledge, in II.3, the need to keep participants satisfied on points like fairness - not because fairness is so important in itself to the values of Theory II, but because human beings who feel wronged are likely to become ineffective judges of factual propositions.

Relevance is another important thing this theory is about. "Ad hominem" is an antisocial activity that participants ought to avoid on moral grounds because it hurts people, but it is instead prohibited by Theory II just because "ad hominem" propositions are irrelevant. And it is predictable that someone subscribing to this theory would also object to other forms of irrelevance even if they are not explicitly listed in the rules.

One can easily look at these two theories and determine that one is right and the other wrong, or attempt to synthesize a third theory (maybe partaking of parts of these, maybe not) that is better than either. Although I have opinions on those points, those opinions are not my main point here. What I think is more important right here is the idea that both these theories are plausible: someone might believe either of them and act accordingly while being neither a raving lunatic nor solely motivated by evil. And furthermore, it appears that both Theory I and Theory II are popular: for each of these two theories there exists a significant population of real people whose actions are consistent with the theory. Given that such populations exist, it then becomes predictable - and worth predicting - what is likely to happen when persons who subscribe to Theory I and persons who subscribe to Theory II attempt to engage in discussion with each other.

The theories are strongly incompatible

Although these two theories of argument are focused on different values, they are not orthogonal. It is not the case that Theory I is indifferent to the things that are good or bad according to Theory II and vice versa, so that someone could reasonably satisfy the requirements of both theories at once. The disagreement is not passive but active. An innocent bystander (say, Aristotle) entering the dialog between Plato and Socrates cannot participate meaningfully and not take a side. Anyone who behaves as a participant in an argument ought to according to Theory I or II, will be unable to avoid breaking the rules of the other theory.

Consider I.3 and II.5:

I.3: Within an argument, claims must be attributed correctly to the persons responsible for them, because claims without such attribution cannot possibly be useful to the purpose described in I.1.

II.5: Wherever possible participants should avoid identifying who said or believes a given proposition, in order to avoid dragging them into the argument and committing a foul under II.2.

A participant who follows either of those rules will break the other. Similar conflicts exist between other rules - for instance, the soft encouragement of II.6 (which I have heeded in this article) to attribute claims to fictional or historical characters, against the requirement repeated several times in Theory I that all claims should be open to defence by real present-day persons.

But it gets worse, because the conflict is not over technical rule-breaking. The real conflict is in the core definitions of what arguments are about. Theory I says arguments are about persons, but anyone acting as if that were true or important would necessarily be committing the one and only (and therefore worst) transgression to be directly forbidden by Theory II, the "ad hominem" foul.

I.1: Arguments are about persons[.]

II.2: It is a foul to attempt to make an argument be about persons[.]

In the matter of how to handle rule violations, there are important differences between the two theories. As soon as Socrates raises the issue of Plato's having broken the rules, he proceeds to insults and attempting to win the point; that is, from his point of view, not gratuitous but the purpose of the entire exercise. But each time Socrates breaks the rules of Theory II by introducing personal insults, Plato (remaining interested in the Moon) attempts to ignore or minimize Socrates' serious transgression, in a way inexplicable under Theory I. This bizarre action of Plato's in not accepting the offer of a fight, would lend support to any suspicion Socrates might hold that Plato may be staring bonkers; even though by Theory II standards, it is a gracious and admirable way to behave.

Going back to Dialog 1, where I said we might all agree that Plato did something bad, that is true but it deserves further examination. The two theories may agree that something bad was done, but they will disagree on why it was bad. Under Theory I, Plato broke the rules by making a false and therefore unfair statement about Socrates. But under Theory II, Plato broke the rules by talking about Socrates at all - and although Socrates would have been entitled at that point to withdraw in disgust having been wronged, he committed an infraction of his own when he attempted to continue the discussion without ignoring Plato's misdeed. Whether the thing Plato said, about what Socrates said about the Moon, was a true or false statement, is irrelevant under Theory II because it doesn't matter who says anything. This was supposed to be an argument about the Moon. So if he has the chance to apologize, what Plato will apologize for will be the error of having treated a person (Socrates) as important to the argument instead of keeping factual propositions centred - and "I'm sorry I treated you as important" is unlikely to make Socrates less angry!

A person following Theory I is likely to commit "ad hominem" according to Theory II; a person following Theory II is likely to commit "straw man" according to Theory I; and many people follow something like one or the other of these two theories, but it is effectively impossible to believe and follow both at once. That's why the sets of people who accuse others of "ad hominem" and people who accuse others of "straw man" are nearly disjoint.

That is my main point. It remains to consider some related material.

Wikipedia

The worst of Wikipedia's crimes was giving "citation needed" an air of respectability.

It's understandable where "citation needed" came from in the context of Wikipedia. Wikipedia attempts to be a "tertiary source" under its own complex theory of how encyclopedia compilation ought to be done, which partakes of some elements of both my Theory I and Theory II as presented here. Primary sources, under Wikipedia's theory of encyclopedia compilation, state factual propositions and may often be wrong. Secondary sources analyze the facts stated in primary sources, citing them, and attempt to puzzle out what is true. Wikipedia, as not a primary nor a secondary but a tertiary source, attempts to accurately present the analyses of secondary sources, without itself taking a position. That entails another layer of citation. It is a rule of Wikipedia that everything should come from a cited source, or at least be subject to citation of a source. "Unsourced material may be challenged and removed," in the words of the ubiquitous warning templates.

In its best interpretation, Wikipedia's requirement for citations is motivated by values similar to those of Theory II. The Wikipedia editing process is supposed to be about creating an index and overview of human knowledge, not about persons, and it should especially not be about the persons who maintain the Wikipedia articles in question and those persons' individual opinions. Wikipedia's soft admonition against primary sources is part of that: citing a primary source invites an evaluation by editors of whether the primary source reliably reports the truth or not, and that could make the process be about source reliability instead of compilation and summarization. Sticking to secondary sources means that someone other than Wikipedia maintainers will be evaluating truth. The hard admonition in favour of naming ("citing") all sources, despite its likelihood of putting focus on sources as persons instead of on the facts, at least shifts the focus away from the much more dangerous situation of having the process be about the Wikipedia editors who would be responsible for whatever they write. Wikipedia should not contain anything original, only substantive content from elsewhere at most re-worded for style and copyright compliance, in order to avoid any suggestion of Wikipedia editors making it personal. Wikipedia is designed to exclude questions about personal responsibility of Wikipedia editors, structurally eliminating "ad hominem." Anonymity helps keep the focus on facts instead of persons too, even if Wikipedia compromises that principle by allowing the creation of user names and persistent identities.

That's the nice way of thinking about it.

In practice, the fact that someone on Wikipedia can cry "citation needed" without being immediately expelled in disgrace often has negative effects, even in the limited context of Wikipedia. Anyone with a penchant for believing the assumptions of Theory I, which requires citation of every fact, will perceive Wikipedia's similar requirement as validating all the rest of Theory I also. Wikipedia because of its central requirement for citations makes itself look like a space where Theory I and the values of Theory I apply. One need only watch any "edit war" to see that Theory I rules are the rules of the day and the purpose of the activity as understood by the participants is to determine who is right and who is wrong, with facts valued merely as ammunition and citations more powerful for that purpose than facts. And the "citation needed" demand on Wikipedia is in practice just an equivalent of the "straw man" accusation: a weapon for establishing who is right and who is wrong, with the underlying assumption that the question of who is right and who is wrong is the important question.

Most people on Wikipedia have no real exposure to serious intellectual discussion other than on Wikipedia, because the same can be said of most people in the world. High schools in North America have never really taught spoken rhetoric, and colleges and universities there, and high schools elsewhere, no longer do it consistently. The media do not present examples of well-functioning rational argumentation today, and may never have done. Wikipedians learn how Wikipedia works, then having no other examples for comparison think that's how serious intellectual discussion works in general, and they learn not only by reading Wikipedia meta-articles that state high-minded idealistic values, but also by watching the behaviour of other Wikipedians who put their own personal values into practice.

Wikipedia has an air of authority, and it may not be instantly obvious to all observers that Wikipedia's bizarre procedures even in the rare cases where they work well are meant for the context of a collaboratively edited online encyclopedia. Most scholarly pursuits are not encyclopedia compilation. So, someone who sees that "citation needed" is tolerated on Wikipedia, who sees how tactically useful "citation needed" is in shutting down opponents and winning arguments, and who thinks winning arguments is important everywhere because it looks important on Wikipedia, will understandably be led to believe that:

  • calls of "citation needed" are a normal feature of serious intellectual discourse and often used outside of Wikipedia;
  • it is acceptable to use "citation needed" in polite society;
  • intelligent, educated, and admirable people use "citation needed"; and so
  • I should use "citation needed" wherever I can.

None of those statements is true.

Thomas Aquinas and the Scholastics

The 13th Century Christian theologian Thomas Aquinas is an interesting figure for a number of reasons but not least because he seems to be a sitting duck for accusations of "straw man." I'd like to use as example Question 32 (PDF file) from Part 1 of his famous work Summa Theologiae; but I emphasize that I chose that one more or less at random, and his entire body of work is full of pretty much the same kind of thing.

To heavily condense part of Thomas's argument: he's considering here a question about whether it is possible for human beings to know about and be convinced of the truth of the Trinity, purely by "natural reason," or whether reaching such an attainment of wisdom requires exposure to divinely-inspired writings. The Trinity is a doctrine in Christian theology whose details are not relevant here. Thomas treats the Trinity doctrine as a precious jewel, and assumes it is true in this part of the Summa, though he attempts to prove it elsewhere.

Thomas says understanding the Trinity requires divine revelation, and cannot be known by natural reason, but in his usual pattern, he starts out by presenting arguments in the other direction supported by attributions to others. Thomas writes that Aristotle wrote that the Trinity doctrine was true. But Aristotle, having lived long before the relevant divine revelations were given to mankind, could only have arrived at his belief in the Trinity by "natural reason," and indeed presented some natural reason-based support for it; hence there is evidence that to do so is possible.

I rather think that Aristotle probably did not believe in the Christian Trinity and never wrote anything meant to endorse it.

So Thomas's claim about Aristotle, despite being supported with a quotation which I will assume is literally accurate, would be easy to censure as a "straw man." He's attributing a position to Aristotle that Aristotle did not take. There are some subtleties here, in that Thomas is not misrepresenting Aristotle's position in order to attack it - on the contrary, Thomas agrees with Straw Aristotle about the Trinity and says more or less, "The Philosopher said this true thing, but how did he know?", then goes off on the question of how Aristotle knew. Thomas comes to the conclusion (which I share as far as it goes) that Aristotle wasn't really talking about the Trinity doctrine after all, at least not in its elaborately developed form of interest to Thomas.

It remains that Thomas supports his own conclusion by again asserting something about what Aristotle's position really was ("he means to say that the ancients used the number three in their sacrifices and prayers") which doesn't seem well-supported by the quote and may or may not be an accurate characterization of Aristotle's point of view. And by admitting that Aristotle was not talking about the Christian Trinity, Thomas may be said to have admitted that he is guilty of "straw man" himself. This, and many of the other summaries and characterizations of others' positions throughout Thomas's writing - some of which he does disagree with outright - would leave Thomas frequently open to accusations of "straw man."

Note that beside the Summa Theologiae (which is the one I mean when I write just "Summa") Thomas's other most famous work was the Summa contra Gentiles, arguing against positions he said were taken in the Jewish and Muslim religions. Exponents of those religions might well claim Thomas argued against misrepresented, weakened versions of their real positions.

Thomas's writings were controversial in his time, but not because of claims he mischaracterized the positions of others and committed "straw man." On the contrary, one of the matters for which Thomas was condemned was that he was thought to give Aristotle too much of a platform. "Pagan philosophers," it was thought by other Christian theologians in Thomas's time, ought not to be studied and quoted. Compare with the present-day anathema on certain lines of political thought which I dare not name, not because they are said to be factually false but morally wrong, and that because of who advanced them. Any position of one of the bad people is morally wrong and even considering the possibility of its factual truth is a sanctioned activity. Contrariwise, showing that an idea came from a bad person is seen in some present-day circles as a perfectly legitimate way to address and reject it. That is exactly acceptance of "ad hominem" as a valid proof technique.

And as a present-day pagan myself I have sometimes heard a similar objection to that of Thomas's contemporaries, raised in the opposite direction: that good pagans ought not to read or take seriously the ideas of Christian philosophers. I expect to attract some disapproval over the fact that I've spent so much space here on Thomas Aquinas instead of looking for examples and wisdom among "my own people." Ralph Blum, whose writings on Norse runes I quite enjoy, has managed to be attacked from both sides at once for daring to write about "pagan" rune mysticism while being openly Christian himself. To all that I would respond by re-stating my position that the truth of a proposition does not depend on who said it.

But if their application was controversial for reasons unrelated to any accusation of "straw man," the general methods of Thomas were not unique or unpopular in his time. The Summa Theologiae is regarded as one of the greatest examples of the approach to knowledge called Scholasticism - which was the dominant way of learning and teaching abstract ideas throughout Western Civilization for about six hundred years. Everybody did it; it was just the way things were done.

Although a complicated system which cannot be trivially summarized, much of Scholasticism came down to constructing arguments like the one I cited from Thomas. Begin by summarizing views seemingly opposed to your own, often though not always citing them from specific sources. Then address those views, with your own eventually emerging as the correct one. This approach properly applied was thought by the Scholastics to be a valid, and the best, way of proving points. Characterizing the positions of Aristotle in particular might be politically incorrect, but characterizing the positions of previous writers as a general practice, possibly not always employing great care to make the characterizations factually accurate, was the standard. What is the theory of argument embodied in the works of the Scholastics?

Fans of Scholasticism might say that (usually, or ideally, if not always in practice) the Scholastic approach does not attack the works it cites, but rather draws distinctions to show that they are not contradictory to the scholar's thesis after all. Thomas's response to an "objection" is not usually "no"; he says "yes, but." For that reason it might be said that Scholasticism is not entirely, or is not at all, based on "straw man" defective argumentation. Scholasticism is not meant to misrepresent others' positions in order to attack them.

But others who don't like Scholasticism might say that (usually, as applied in practice, even if not in theory) Scholastic arguments do misrepresent sources even if not for direct attack, and also quite often do directly attack the misrepresented positions. Structuring every argument to start with a brief representation of others' positions makes it easy by accident, and tempting on purpose, to have the representation be a misrepresentation. Scholasticism as applied in practice can be argued to be entirely based on something that is dangerously close to "straw man," and makes "straw man" easy to commit.

As well as seeming misrepresentation of specific cited works (I.4), there are also many instances throughout the Summa (e.g. Article 1, Objection 3 in my link) and throughout Scholastic thought in general, of addressing "objections" that are not precisely attributed to specific persons or that nobody has seriously raised before, and these instances may be cause for other kinds of "straw man" accusation (I.5 or I.6).

Was Thomas Aquinas, along with the entire current of intellectual practice he exemplified, a habitual user of defective argumentation? That's conceivable. It seems that his methods would leave him subject to accusations of "straw man" were he to employ them on today's Internet - as I have been myself, whenever I've used a similar approach of starting from and addressing opposing viewpoints on their own merits as they really exist, without solidly proving that someone did hold exactly those views and saying who and charitably interpreting their words in the manner that would make their position look stronger than it is. It is reasonable to suppose that having had almost eight centuries of intellectual progress since Thomas's time, humanity may now have access to better theories of argument than were available to him. Those better theories may even be accessible to, and routinely used by, ordinary people in Internet discussions who are not world-leading philosophers. That is all possible.

But it does not come naturally to me to conclude that idiots on Facebook with their Bingo card logical fallacy lists chanting "straw man, straw man, I win!" are really doing the business of searching for truth in a morally better, more logically valid, and more useful way than the way Thomas did it. Maybe his methods weren't so far wrong after all.

There is a famous quote attributed to Thomas Aquinas which probably has nothing to do with "straw man" as a supposed form of defective argumentation, but it's too good not to mention. It is said that while writing the Summa Theologiae, Thomas received a divine revelation, stopped writing, and was asked why.

"Omnia quae scripsi ego mihi videtur ut palea."

"All I have written seems like straw to me."

Present-day academic disciplines

Present-day academic writing that is not Wikipedia has a tradition of using citations that may have evolved from the use of citations in Scholasticism. It is quite typical, though exactly how typical depends on the discipline, for an academic article to have a section near the start called something like "Related Work," which summarizes, always with citations, things other people have said on the subject of the present work. In some fields this might be considered obligatory and written into journal submission guidelines as a requirement of the format; in others, it may be more of a guideline, or (though this is unusual) not even expected and only occasionally used.

The "previous work" section looks a lot like Thomas's litany of "objections" at the start of each of his chapters; but it differs in that the related works are frequently only mentioned there, and not directly addressed in the body of the article. Very few academic articles are structured as direct responses to things earlier writers said, and even fewer are directly arguments against previous work. More often, an academic article is presented as something new and largely independent, with previous work mentioned out of some kind of vaguely-defined obligation, as reference material per se, or just to illustrate the previous state of the art from which the new article advances. We don't need to tell you the melting point of sodium acetate because you can look it up <here>; or the previous fastest known algorithm was O(n2 log n) as described <here> and now (without claiming that the previous work was in any way wrong) we offer a O(n2 log log n) method.

It's worth digging deeper into that "vaguely-defined obligation" I mentioned. The "previous work" section routinely references many previous articles that are not closely related to the present one; not things it is arguing for or against, but just other material sort of connected. Why bother mentioning that stuff? Brevity is often a high priority in academic writing, but even with space at a premium, including a sufficient "previous work" section is treated as a necessity, the lack of which will result in rejection by the referees. To be so strongly conserved, that section must be serving some purpose.

I think part of the reason for the "previous work" section, and probably the first reason academic writers would think of when asked, is that it's meant to be a service to the reader. Especially bearing in mind that these traditions evolved before the World Wide Web, it can be really hard to find an article on a given topic. Having well-written "previous work" sections everywhere means that a reader who can find an article even approximately relevant to the question they intended to research, may be able to find a closer match mentioned in the "previous work," and after a few such jumps find the article they actually wanted. The policy of having links between articles greatly increases the value of the entire discipline's body of work. Similarly, citations of points from earlier sources as they are used in the article, after the "previous work" section, is a service to readers that allows them to look up more detail on those points, and helps writers keep the length down by allowing them to omit details that can be found in the cited works.

Note that the ideas of citing earlier authors to give them a chance to defend themselves, or to prove that somebody exists who really said this thing and so it's not a "straw man" - the compelling reasons for citation under Theory I - are almost completely unknown in standard present-day academic practice. Those reasons are not why scholars cite previous work.

But there does exist another important reason for the "previous work" section, and for citations in general, beyond service to readers. Scholars live on attention. Making other scholars, as persons, the topic of your article, or introducing irrelevant biographical details about them, would usually be "ad hominem" and inappropriate. But when in an academic article you mention others as the sources of ideas and proceed to discuss the ideas, you are doing the persons mentioned a huge favour. In the present day, some scholars actually earn codified numeric "points" for having their work cited, in a way that factors into management goals and hiring and promotion decisions. Even before we sank to that level, it was thought that an academic worker whose work received a lot of citations was in some important way doing better than one whose work didn't. Hence the vague sense of obligation we feel to cite others' work: a citation is a favour, which we might owe, or hope to have repaid.

And the favour is in a scholar citing another's work at all, not in what we might say about it. It is virtually unknown for serious academic writing to directly condemn a cited source; and it is frequent that the description of a cited source may not be entirely accurate. Neither "straw man" nor "ad hominem" accusations come up as direct call-outs all that frequently - such accusations are a phenomenon of Internet commentators who admire academic methods, more than of real practitioners. What real scholars do is blindly count citations. This behaviour is not entirely explained by Theory I nor Theory II.

There's one other point to think about in present-day academic citation practice, and that is the specific technical form of how citations are done in writing. In many STEM disciplines (such as computer science, my own background), citations in running text are usually done by numbers written in square brackets. The numbers refer to entries in a list of references at the end of the article; variations of this system differ in how they organize the list at the end, and a closely related system uses superscript numbers referring to footnotes on the same page instead of a list at the end. If Plato were a computer scientist, he might write as follows.

It has been reported that the Moon is made entirely of Athenian cheese [23], but recent evidence suggests that that is not the case [5].

Academic disciplines other than STEM tend to use author-year citations, like the following. As in the previous example, the in-text citations can be looked up in a list at the end, which gives more detailed descriptions of where to find the source materials.

It has been reported that the Moon is made entirely of Athenian cheese (Socrates 420 BCE), but recent evidence suggests that that is not the case (Armstrong & Aldrin 1969).

Spot the difference.

It is entirely possible that these conventions for how to write citations are the way they are because of simple historical inertia and they have no other significance. Whoever first published in a field used whatever system was ready to hand, perhaps more or less at random, and then others followed their lead, and today people in any given discipline use their respective citation systems just because that is how people in that particular discipline have always done it and maybe it doesn't mean anything more.

But what if it did mean something? In particular, if someone were a diehard adherent of my Theory I and believed that who said something is always really important, which of these citation schemes would make more sense to that writer? If someone believed deeply in Theory II and thought it was important to emphasize factual propositions independent of who said them, and minimize (as impolite in addition to counterproductive) direct references to persons, which citation scheme would make more sense to that writer? Do citation schemes correlate with theories of argument, and both correlate with other important distinctions between academic disciplines? Does training in using a given citation scheme implicitly affect a writer's approach to presenting arguments in any deeper ways?

Bonus point to ponder: Imperial Chinese civil service exam essays, in which examinees were expected to cite and quote from a defined canon of traditional sources, largely to prove that the writer had memorized those sources. (Cheating by bringing in reference material to avoid memorization was a common occurrence.) This system predated but overlapped with the Scholastic tradition in Western Civilization, and although it influenced Western civil service exams much later in history, it seems not to have influenced Scholasticism directly.

Arguments that are not dialogs

There's another wild card here relating to the form of the argument. Examples given to define what the term "straw man" refers to, including my own examples here, are pretty much invariably presented in the form of dialogs, often implied to be transcripts of oral in-person interaction. One commits "straw man" in a back-and-forth debate with one identifiable partner. But when the knives come out on the Net and someone attempts to call someone else out for "straw man," the situation usually does not look like that at all.

Far more often, one accused of having committed "straw man" will be someone who wrote a Web log article or similar, which was directed to the world at large and not a specific partner, and also not framed as a response to a specific previous source even if one or more sources are cited. Indeed, as I've discussed, the fact that an accused document is not a response to a specific previous source, may itself be put forward as evidence of "straw man" under Theory I. By supposedly concealing the name of who is being attacked (taking for granted that all valid arguments should be attacks on somebody), a document that is not presented as a direct response renders the victim unable to defend themselves and may also create a suspicion as to whether the victim really exists at all (which is assumed to be a question worthy of consideration). Under a different theory of argument, with no partner to mischaracterize "straw man" is just impossible to commit in the non-interactive setting.

But even if we think "straw man" in writing is basically possible, levelling an accusation of a "fallacy" invariably defined by reference to oral in-person conversation, at written communication which is not part of a back-and-forth interactive debate, should give us pause. If we use the concept of "straw man" primarily as a way of understanding non-interactive written communication, why do we need examples from a different realm to define it well? Is it really possible to commit "straw man" under the best definition, in a written article that is not part of a back-and-forth debate?

There is a continuum here, especially evident on the Net. This Web log article is not a response to one specific person or one specific instance of a false accusation of "straw man," but it certainly is a response to the synthesis of many interactive conversations I've had with many different partners and many instances of false accusations. It is a response to the general state of public discussion of this kind of topic. Similarly, this article could easily become part of an interactive discussion as soon as someone writes a comment at the bottom or a response of their own elsewhere.

Postings on the Net in general are often in a sort of grey area between one-way broadcasts and two-way interactions, and the same material can shift between the different states over time. It's not only the Net, either; people were writing responses to previous writings with greater and lesser degrees of specific reference, at least as far back as Plato's and Socrates' time. It can be hard to judge how personal and interactive a given writing was meant to be, and how personal and interactive it may have become at a given moment; and people may reasonably disagree about the position of a given writing on this continuum at a given moment and that may lead to further disagreement on how reasonable it is or is not to bring the concept of "straw man" into play.

But I don't think such ambiguities are the only reason we routinely see false "straw man" accusations made in practice. I think there's something more to the gap between defining the term on oral interaction and applying it to written broadcast media. "Straw man" on the Net is a concept stretched from whatever objectively useful form it might once have had as a tool of rhetorical analysis, into a tactical weapon useful for very different purposes.

The last word

I'll close with a quote from the real Plato. From Phaedrus, 275b-275c, translated by Harold N. Fowler, 1925, emphasis mine:

Phaedrus: Socrates, you easily make up stories of Egypt or any country you please.

Socrates: They used to say, my friend, that the words of the oak in the holy place of Zeus at Dodona were the first prophetic utterances. The people of that time, not being so wise as you young folks, were content in their simplicity to hear an oak or a rock, provided only it spoke the truth; but to you, perhaps, it makes a difference who the speaker is and where he comes from, for you do not consider only whether his words are true or not.

6 comments

Vilhelm S.
I always thought that the main beneficiary of the related work section was not the readers or the cited authors, but the reviewers. I.e. when you evaluate a paper you want to know not only if it is true, but if it is novel. The related work section gives credit to previous work, but more importantly it avoids *taking* credit for it. (And conversely, of course, it provides a place to say how great this additional increment is.)
Vilhelm S. - 2018-03-26 10:36
Michael
I thought that a straw man argument is one in which a claim ('straw man') is created for the purposes of disproving it, for example writing a Socratic dialogue so that you can attack it. Attribution is not necessary for such an argument. Aristotle would organize the universe into such a way that he could demolish a sequence of straw targets to support his own assertions by process of elimination; he does not address my contention that looking at the world through a different Venn diagram would model it better, possibly with different conclusions drawn. Misattribution of fallacious arguments is a separate crime against truth.

There was a time in which 'ad hominem' arguments were the only type with any weight, so long chains of hearsay would be used to attribute claims to trustworthy personages. For example, my 8th grade teacher told my class that Aristotle proved that women were inferior to men because they had fewer teeth (he counted both his and his wife's in support of this argument*). This is not a straw man argument, but it could be an appeals to higher authority or an ad hominem attack on Aristotle.

*: Online sources such as Scientific American claim that he did not actually count the teeth of either of his wives. Personally, I assumed that he had wisdom teeth and they didn't. On the other hand, I believe I'm more highly evolved than most humans as wisdom teeth are obsolete I genetically have fewer adult teeth that almost everyone else.
Michael - 2018-03-26 11:03
Matt
If the related work section is primarily for the reviewers, then why does it get printed in the journal? We do have a custom, at least in computer science and especially for conferences (where page limits are a bigger deal than in journals) of adding an appendix with extra information for the reviewers that will not become part of the final paper when accepted. That's often used for stuff like proofs of lemmata. Lists of citations, if they were only to establish that the paper should be accepted and not also for the final readers, might naturally go there.
Matt - 2018-03-26 16:37
Vilhelm S.
Well, maybe not just reviewers, although they are the most obvious beneficiary, but people who are interested in the paper, as opposed to the subject matter of the paper. It's, like, a meta-section; the other sections are about the the subject matter of the paper, but the related work section is about the paper itself (and how it fits into the rest of the field). I guess that ties in a bit with the themes in your post, because it means that it's more person-related than the other ones, not exactly "Dr. X is right/wrong", but "Dr. X (me!) has hereby made a great contribution".

I guess this is not so different from what you say above, about citation as favor, but I think it's also the case that academic life is a bit more complicated than just blindly counting citations. Maybe the more bureaucratic parts of academia are that crude, but I think if an individual academic, personally, tries to form an opinion about how good someone's research is, they will pay attention to what kind of story they tell. Job talks often have a slide setting out the overall state of the field, and why their particular research fills a critical gap and so on. The introduction and related work sections of a paper (the "meta" sections) do something similar, I think. The running time of your algorithm is an eternal Platonic truth, but the fact that it was you who invented it where others tried and failed is a contingent fact, and your paper wants to convey both...
Vilhelm S. - 2018-03-26 22:03
Kevin C.
You claim up front that "straw man" has a clear, unambiguous definition, but never actually cite an actual definition. I'm working off the Google dictionary definition of "an intentionally misrepresented proposition that is set up because it is easier to defeat than an opponent's real argument.", which appears consistent with the Wikipedia definition.

It seems like you and Galef both think the most important part is that the person who's argument was misrepresented is part of the conversation, whereas to me the most important part is the intentional weakening of the proposition. The examples section of the Wikipedia article appear more consistent with my interpretation, and support the idea it can be applied without specifying who made the argument.

Is it possible the call for "citations" or "examples" are an attempt to see the original (presumed stronger) form of the argument you are addressing? Perhaps so they can point out where you've weakened it?

In dialog 1, the absence of Socrates' original claim makes it hard to evaluate if this is a straw man argument, or just putting unrelated words in his mouth. If it was a straw man, I would expect the original proposition that Socrates made to be something like "The moon is [partially] made of Athenian cheese", which is not actually contradicted by 1.3. Changing the proposition from an implied partially to an explicit entirely replaced it with an easy to disprove proposition.

In dialog 2, demonstrating that actual citizens believe the weak proposition would exempt Plato from "straw man", whereas finding that they actually believe a stronger proposition would make it clear Plato is committing "straw man".

Rule II.5 tends to break down in cases where the individual propositions are too complex to describe in full. Even with II.1 as the goal, it may be necessary to reference real statements of real people because they've done the research necessary to demonstrate some nuanced proposition that is part of the supporting evidence of the overall argument.

Another important consideration is that people are likely operating under *multiple* theories of argument at the same time. For example, they may consciously think they are arguing about the factual propositions, while unconsciously trying to win the argument to increase their sociometric status.
Kevin C. - 2018-03-29 11:08
Bo
Kudos to you for such a consistent argument. Given the length of it, I suspect it took more than one sleep to type it up. Which puts you far beyond all the educated people you are arguing against.

As an educated person myself, all I can say is that I would have a different opinion if I slept on it. And also the next time.

People keep bringing guns to a knife fight. The pleasure of disarming their explosive weapon with a simple blade is hard to deny. But you didn't bring a gun because you're perfectly capable of making an actual nuke.

Keep on fighting the good fight, what else is there to do? Just make sure to keep them breathing.
Bo - 2018-04-02 17:57


(optional field)
(optional field)
Answer "bonobo" here to fight spam. ここに「bonobo」を答えてください。SPAMを退治しましょう!
I reserve the right to delete or edit comments in any way and for any reason.