Friday, September 09, 2005

Attacks and Arguments

Sometimes people make false accusations. Let's define the "fallacy" fallacy as being when someone mistakenly rejects an argument on the basis of an alleged fallacy that it does not in fact exhibit. (I suppose this process could iterate: If I falsely accuse you of making the "fallacy" fallacy, then I thereby commit the ""fallacy" fallacy" fallacy. And so on.) Anyway, I suspect the most common form of the "fallacy" fallacy involves a false accusation of argumentum ad hominem. In a delicious twist of irony, many of these false accusations actually amount to genuine ad hominems themselves. Here's the typical pattern I have in mind: Person A makes an argument sprinkled with insults. Person B objects, "That's an ad hominem!", and refuses to address the substance of A's arguments.

The problem is that people commonly object to any form of insult as ad hominems. But this is mistaken. An ad hominem fallacy is when you reject your opponent's argument because of some characteristic of the advocate that is irrelevant to the content of the argument made. In general, what matters is the argument, not who makes it. (I will mention some exceptions below.) But not all "personal attacks" take this fallacious form. Rather than saying "you suck, therefore your argument does", one might instead provide an adequate counterargument, then append: "your argument sucks, therefore you do". Such gratuitous insults may be unwise, but the counterargument doesn't depend upon them, so it's a mistake to object to the counterargument (and ignore its substance) on that basis.

Further, I think insults aren't always inappropriate. Creationists and homophobes, for example, tend to make revealingly bad arguments. In critiquing these arguments, we first aim to show why the conclusion doesn't follow. But we are also in a position to draw a conclusion about the character of the person advancing the argument. Many arguments are so bad that they could not be honestly made by any informed rational person. Thus anyone who makes them must be either stupid, ignorant, or dishonest. In the course of criticising such an argument, a partisan may wish to point this out, just to emphasize how incredibly bad the argument really is. I don't think it's necessary wrong to do so. Some positions are so lacking in rational warrant that they deserve our scorn.

For an example of the "fallacy" fallacy in action, just look at political blogs. Liberals tend to insult Bush a lot in the course of their arguments. Conservatives then respond by dismissing them as "Bush-haters". But in fact it isn't the "Bush-haters" that were committing ad hominems - after all, they had arguments to back up their insults. On the contrary: the conservatives are effectively arguing, "You hate Bush, therefore I don't need to address your substantive argument." - and that is a clear case of argumentum ad hominem. As Leiter notes:
This guy sure is pissed, not without reason. The Bush shills and loyalists have coined a phrase as a defense mechanism against this kind of increasingly common reaction: they refer to "Bush haters," as though this kind of response is irrational and inexplicable. It doesn't occur to them that there are reasons people hate Bush, that people are responding to events, to facts, to stupid things the man and his Administration have done.

(I note that Leiter himself is commonly targeted by people committing the "fallacy" fallacy. People are put off by his abrasive tone and plain-spoken insults, and thereby conclude that he's committing "ad hominems", no matter the substance of his arguments. Of course, one may question whether it's the best way for him to persuade anyone who doesn't already agree with him -- but Leiter has explained that that is not his purpose anyway.)

Okay, so far we've distinguished "personal attacks" based on whether they form premises or conclusions of counterarguments. Only the former are fallacious -- and not even all of them are. I can think of two general exceptions. The first type is when the person being attacked will be responsible for carrying out the action being debated, and hence their character is relevant to assessing the likely consequences of the action -- I explain this in more detail over at Prior Knowledge, in response to a "fallacy" fallacy from Simon Clarke.

The second type is more complicated, and it was brought up in our political philosophy class last semester, where G.A. Cohen (I think it was him) uses it in relation to the incentives debate over tax. Anyway, the basic structure is this: a kidnapper has stolen your child; you will only get your child back if you pay the ransom, therefore you should pay the ransom. This seems like a fair enough argument, e.g. if made by the police who are helping you. But the normative force of the argument is completely different if it is instead made by the kidnapper himself. After all, he is responsible for making the unfortunate premises true. As such, his use of the argument is reprehensible blackmail, rather than helpful advice. That doesn't necessarily make it unsound. But it does mean that the person making the argument (i.e. the kidnapper) can be blamed for doing so. Cohen argues that when capitalists argue that they will work less hard if taxes are raised, this is a similarly reprehensible "threat" of an argument.

Anyway, the point to note here is that arguments are not simply the free-floating propositional structures that philosophers are trained to see them as. Sometimes it makes an important difference who is making the argument. As such, it is not always illegitimate to draw attention to this fact.


  1. His mistake is that it is partly an ad hominem - but it is also partly a valid point.
    Having said that if you dont pause arguments with people whouse ad hominems how wil lthey ever learn not to do it?
    rather like using war to prevent states invading eachother.

  2. I continue to be amazed at the general disrepute with which Ad Hominem (AH) arguments are held among educated people. Of course, we use them - validly - all the time and our society would not function without them. Most of the time (as with the other so-called Fallacies), their use is not fallacious.

    For example, lawyers rely on them in court. If you want to demolish statements made by the witness of an opposing side, you show that the witness is not reliable, you attack his/her character or person or background, and allow the jury to infer that the statements of the witness are not to be trusted. Conversely, if you want the jury to believe an expert witness, you seek first to establish the credentials and expertise of the witness, before asking for his/her expert opinion.

    And none of this usage of AH arguments is necessarily fallacious, as the philosophers of argumentation have long known. They enable us to reach mostly-valid conclusions readily under conditions of limited resources (eg, of information, time, or mental processing capabilities).

    AH arguments are related to epideictic arguments -- those where we judge the content of the argument on the basis of its form. One politician puts her case clearly and succintly, another rambles and digresses. We therefore tend to believe the former rather than the latter. If you think this is unwise, imagine choosing between two medical specialists advising you about a course of cancer treatment - of course, most of us would choose the doctor who spoke clearly and succintly about the treatment, and not the verbal rambler, even when (or especially when), the proposed treatments were the same.

    Nothing different in principle here from choosing between two scientific theories on the basis of the characters of their proponents. Perfectly rational and justifiable behavior.

  3. Kofi,

    I don't think I agree with you. If the purpose of the dispute is after facts and the factual substance of the argument then the character of the arguer or proponent is largely irrelevant. However, as Richard quite clearly stated, if we have to measure the argument with an eye towards the character of the arguer or proponent then clearly we ought to evaluate whether the character of the arguer affects their argument. You did not I think read the post carefully.

    Anyway, Richard, a nice post and blog. I look forward to your column on



  4. Kofi - I think part of what you are pointing out is the fallibility of our judicial system - ie that we will find it easier to convict people we dont like. the other part is the relevant aspect to adhominem - which should be seperated out and could probably be stated in a non personal sort of a way.

    I note that at times ad hominems may be the only information we have and at times we might choose the doctor/politician with the most personal insults or the one with the least insults depending on other factors - for example a doctor that is constantly insulted but never in a way relevant to his work is quite likely to be a very good doctor.

  5. My response is somewhat mixed here, Richard. Speaking from experience, on one hand, it is never legit to disregard the arguments on the basis of the source. A total jerk can still be the messenger of truth, so to speak, and to reject their argument because they are a jerk is sort of like killing the messenger.

    On the other hand, I have found that when people start spewing ad hominems, they are not really interested (at least at that point in time) in rational argumentation. They are interested in winning, showing the other as an ass, no matter what. Continuing the argument -- whether in personal life or in philosophy -- tends to be counter productive. I think "tit for tat" works excellently well in those situations -- insult them back, and tell them that you are willing to return to the argument as soon as they are. :-)

  6. Richard, since an ad hominem fallacy is a fallacy of irrelevant, any insult is an example of an ad hominem fallacy if it is (1) put forward where an argument is needed; and (2) is not such an argument. It does not matter whether the position deserves scorn or not.

    The problem with attributing ad hominem fallacies is the same as that with all fallacies of irrelevance: people can disagree about the relevance. (Indeed, this is usually what is really at issue in cases where Leiter has been accused of ad hominem; he claims it was strictly relevant to his argument for his point, and his opponents, Althouse or whoever, deny that it was.) Indeed, they can reasonably disagree about it.

  7. Re Cohen's argument, isn't he confusing an expression of intent (on the part of the kidnappers) with a mere prediction (on the part of capitalists)? That is, capitalists aren't really threatening they will work less as a direct response to higher taxes, but rather predicting that higher taxes will shift their opportunities for realizing their values in such directions that working so hard will cease to be their best alternative. But the kidnappers are threatening, not predicting.

  8. This all seems pretty much in order, but bear in mind that much discourse doesn't clearly mark off premises and conclusions. Arguments peppered with insults (as political arguments often are) are often plausibly read (and, I suspect, a lot of the time correctly read) as containing this IMPLIED argument: my opponent is a fool, opportunist, fascist, liar, pointy-headed lefty, etc., etc.; THEREFORE, you needn't take his arguments seriously.

    Is this implied ad hominem never present in Brian Leiter's full-throated invective? Beats me; but someone as clearly angry as he is would have to made of pretty stern stuff never to indulge such a tactic.

    Another comment on Professor Leiter: he may be too pessimistic about the futility of trying to persuade the unpersuaded with a weblog. I actually find myself agreeing with much more of what he writes than I did a year or more ago. This I am convinced is in part due to his arguments, which are more often than not quite cogently constructed. An anecdotal observation, I realize; it might be interesting to look further into the question of how many others have found themselves influenced by some sharply argued website or other.

  9. Yeah, I too am more optimistic about the power of reason, but I can certainly understand others' cynicism.

    Aristomedes - insofar as capitalists have control over their own actions, and could choose to act for the best interests of all rather than sticking to their own "best alternative", I think this "prediction" amounts to a threat in the sense that Cohen requires. (Though it obviously is not so unreasonable or blameworthy as the kidnapper's threat.)

    Brandon and Vera - fair points.

  10. Reason and logic can be used to justify anything, as philosophers have often repeated, logic says nothing about "the world." You can only begin to form arguments and assess an argument's validity when you agree about the truth of the premises., These premises are statements about the "how the world is." The problem here, simply stated, is that validity does not imply soundness (i.e. simply because an argument is valid or "truth-preserving" does not imply that it is "actually true."

    This is where the disagreement stems. The solution to such disagreements is, ideally, a community of inquirers working in good faith. Determining who gets to determine "actual truth" is where the real problem lies. Insults are a way an arguer attempts to undermine the individual as a source of truth, i.e. X can't be trusted they are a big fat liar as evidenced by Y"

    As for Brian Leiter, the problem he faces is that whereas his critiques of the "reactionary right" are often apt; Leiter himself has made insulting and often bizarre remarks which lead one to question whether he has an undistorted view of the world from which he could form arguments which are both valid and sound.

    The point Leiter makes is that the reactionary Right wing politics are the product of a certain unbalanced psychology consisting largely of what psychoanalysts call "reaction formations." This may be true, but Leiter himself has said things which make many question his own status as a balanced and impartial purveyor of truth. It's one thing to be irreverant for the purpose of conveying a shocking truth, it's another to just be rude. When you groundlessly insult others, it isn't that your argument is invalid, it's that I surmise your account of "how the world is" (i.e. the premises you begin from) is wrong or distorted (perhaps for psychological reasons) and therefore you may form a valid argument, but a nonetheless unsound one.

    [Comment moderated to remove gratuitous personal attacks. Sorry Kosta, you can take it elsewhere, I'm not about to host such remarks -- RC.]

    Following from this, to answer the big question, how do we determine whether our community of inquirers has operated in good faith and come to the truth? Simple, their conclusions breed good results (i.e. when you act on what they say, it works!)

  11. Well, yes: Using Brian Leiter as an example only highlights the problem with someone who is all too ready to attribute evil or irrationality to his opponents, no matter what the issue. Yes, he's not technically committing an ad hominem, but Leiter's vituperation is so over-the-top that it simply detracts from his credibility. He seems more interested in generating excuses to hurl insults at people than in making any sort of genuine intellectual contribution to anything. (That goes for his blogging, not his scholarship, needless to say.)

  12. Another comment on Professor Leiter: he may be too pessimistic about the futility of trying to persuade the unpersuaded with a weblog. I actually find myself agreeing with much more of what he writes than I did a year or more ago. This I am convinced is in part due to his arguments, which are more often than not quite cogently constructed. An anecdotal observation, I realize ...

    But, it also applies to me as well.

  13. One must be a complete moral retard to think that the "wrongness" of sodomy is somehow a central and unrevisable moral truth. is just an insult, not an attack against the argument by attacking the man.

    Great example in your blog. Though the implication is that the guy is such a retard that you can't credit his argument, which is why, I think, there is so much confusion by the general public.

  14. Several people have brought up the issue of credibility. Kosta makes the obvious point that sound arguments require true premises, but goes on make the more contentious claim that you can assess the truth of a claim based on the character of the person who makes it. This is an important issue which I discuss further here (as I already mentioned in a previous comment).

    P.S. Guys, please take care to ensure that you are merely talking about personal attacks, as opposed to engaging in them yourself. I will moderate comments if necessary.

  15. I think that the *interests* of speakers are often relevant to assessing their contributions to debates. That anyway always seemed like a key factor in reading newspapers to figure out what is going on. If someone endorses an empirical claim when their interest goes contrary to the recognition of that claim, that is reason to give it more credence that the same claim made by someone who has a reason for wanting you to believe it. I often see people who point out the interests of people making claims accused of making ad hominem arguments. But for the above reason I think this is a mistake.

    FWIW, I tend to think that most "fallacies" which get labeled as informal fallacies are really just abbreviated arguments that are not all that bad. And part of my reason for believing this is the way people label arguments of this sort as fallacious. But that is another issue and it is late -- late enough that I'm doing a bad job spelling words that I know.

  16. Erik says: "If the purpose of the dispute is after facts and the factual substance of the argument then the character of the arguer or proponent is largely irrelevant."

    Of course, this is not the case. If our only means of determining the facts of an argument is via witness testimony, then the character and reliability of that witness is precisely relevant. This is why AH arguments are admitted in courts, and why, as I said, AH arguments are not in general fallacious.

  17. It seems that often the confusion arises in part due to uncertainty about the shape of the ad hominem reasoning. Is the point of the attack to show that some proposition is false or that some testimony doesn't provide us with sufficient justification for acceptance?

    You have to get pretty creative to find plausible instances of virtuous ad hominem's in which the facts about the speaker's character serve as justifying reasons for believing some proposition. It takes fare less creativity to think of cases where facts about the speaker's character give us reasons to worry that the fact that the speaker asserts p does not significantly raise the chances that p is true. You might think that both forms are fallacious in that there is no generally good inductive inference from the facts about the speaker's character to the existence of either an undermining defeater or a rebutting defeater.

    When someone points out that there are instances in which there are connections between facts about character and certain kinds of defeater, these appear to be exceptions to a general rule and you surely cannot say that something is a good inductive inference because on rare occassions it works.

    That being said, functional, it would be nice to see an example in which BL's attribution of irrationality or depravity is unwarranted. I'm a pretty regular reader of his and I cannot think of a single example. Maybe I need to think or look harder but it's kind of silly to cite him as an example without citing an example of misconduct.

  18. I think a distinction has to be made between personal attack, and substantive disagreement.

    I was making the obvious claim, that neither Leiter's scholarship nor his experience has substantiated or given warrant to statements he's made about other philosophers. That's not, by any means a personal attack, it's an obvious point which can be easily argued, and if wrong, refuted. Are Leiter's attacks on fellow philosophers warranted? I believe not. I've read nothing by Leiter which represents clear criticisms of those he's impugned. For example, in the case of Isaiah Berlin's reccomendation letter for HLA Hart, Leiter comments:

    "There are many ironies in this letter, not least of which is that, with the benefit of hindsight, philosophers now recognize Berlin as the "slender bread-knife" (perhaps a plastic one!), not Hart!"

    Now while Berlin was in a position to provide candid personal criticism of Hart, Leiter is not in a similar position to provide such criticism of Berlin. In any case I am aware of no "philosophers" who share Leiter's view of Berlin.

    Unless, you have some personal familiarity with the individual in question, you should restrict your criticisms to substantive issues in their philosophy. This is what I've attempted to do, although I can see where my comments could be interpreted as personal attacks due to some unfortunate phrasing and analogies I drew between Brian Leiter and Pauly Shore. Incidentally, I did like "In The Army Now." I feel about Leiter as I feel about Pauly shore, I don't always agree nor like all of his work, but he's occasionally funny.

  19. Clayton --

    If you're a regular reader of Leiter's, surely you're familiar with how he'll call someone a "fascist" or "Taliban" at the drop of a hat.

    Now if you're not familiar with his practice in that regard, you can't be much of a "regular reader." Or if you think that all his uses of such terms are warranted, then there's really no point in discussing that point any further. In my experience, the willy-nilly and unthinking use of such derogatory language makes a person less credible, even if it's not a true ad hominem.

  20. I've no idea who Kosta is, though he has posted
    extremely ignorant remarks about philosophy on my blog in the past, and reprises his performance here. In fact, the view I expressed of Berlin is a view widely shared by philosophers--but to know this, you'd have to know some philosophers, and know something about philosophy, of course.

    The anonymous "functional deviancy" misrepresents, as anonymous posters so often do, my blogging practices. (Thanks to Clayton for pointing this out.) What s/he presumably means to say is that s/he has a substantive disagreement with me on certain political and moral questions and so thinks the vituperation is unwarranted. Everyone allows that vituperation has its place, and so the only question (undiscussed, of course) is one on the merits: e.g., is this political view really a case of proto-fascism? are these individuals really aspiring theocrats? and so on.

    On a positive note, I'm glad to see that Richard seems to have educated a few of these folks as to the fact that I almost never employ ad hominem arguments.

  21. While Brian Leiter claims that I misrepresent his blogging practices, he effectively concedes that he uses vituperation/insults with great regularity -- i.e., that what I said was precisely true. His real beef with my post is that he believes (obviously) that his use of vituperation, insults, etc., is warranted. (If Leiter is claiming: "Functional said that my use of vituperation is unwarranted, but in fact it IS warranted" -- well, that's a matter of personal and subjective opinion, one that is not susceptible to being described as a simple factual "misrepresentation.")

    Everyone allows that vituperation has its place, and so the only question (undiscussed, of course) is one on the merits: e.g., is this political view really a case of proto-fascism? are these individuals really aspiring theocrats?

    Sure: The term "Taliban" has its uses, e.g., to describe the actual Taliban. But when such terms are deployed on a hair-trigger at the slightest provocation, they do detract from one's credibility. For example, someone who believes in good faith (not that Leiter will admit that such a thing is possible) that there are various scientific problems with evolution that should be discussed in the classroom (i.e., the Cambrian explosion), Leiter isn't content to point out that these individuals are incorrect on the merits. Instead, he immediately cries "Taliban" or "theocrat" or some similar insult.

    In this context, the term "Taliban" is wildly inappropriate and hysterical. There is no genuine resemblance whatsoever between (1) a person who thinks that there are scientific problems with evolution, and (2) a system of government that forces religious adherence on everyone, stones or otherwise kills people for minor transgressions, forces women to wear burkas, bans the playing of music, explodes historical statues, etc., etc. To use such a term merely serves to detract from one's reputation for level-headedness (if such a reputation ever existed in the first place).

  22. Last effort: you misrepresent my blogging practices when you say that I use terms like fascist and theocrat "at the drop of a hat" or "on a hair-trigger." This is false, and shows you do not read my blog regularly or carefully. I use these terms, in conjunction with extended documentation and discussion of the merits, and I use them because they are warranted. The sincerity of a false belief is neither here nor there; and when the sincerity is explained by religious delusion, and those suffering from the religious delusion also want to impose other aspects of their religion on others, then talk of theocracy and the "Texas Taliban" makes good sense. (The actual Taliban were sincere too, you know.) But to know this you'd have to read and discuss what I actually write, instead of hiding behind anonymity and sanctimony. (I have discussed the meaning of "Texas Taliban" at length on the blog--anyone interested can read the response to idiotic remarks like "functional's" there. Apparently our "functional" illiterate doesn't understand metaphorical uses of language.)

    I'm on the road now, so I guess we're done.

  23. I suspect that this debate is so intense because analysis of AH arguments exposes one of our great myths about ourselves. It suits us (liberal, educated, westerners at the third millenium) to believe that scientists and other educated people generally make decisions on the basis of the arguments or the facts, not on the basis of allegedly-extraneous issues such as the character of the proponent of an argument. But this is a myth, a delusion we hold, which is shown not to be true with any close study of how science is actually done, and how scientific judgments reached.

    Scientists adopt or reject theories on the basis of all sorts of reasons, as Feyerabend showed, including AH ones, and so do we. And damn right, too! These other issues are not extraneous at all.

    None of us can know the future or the past. Indeed, we can barely know the present (eg, the vast majority of us are not in any position to replicate major scientific experiments). Witness testimony is our only access to the "facts" of these worlds. And that being so, the character of the witness should be at the centre of relevance, not at the margins, in assessing that testimony.

  24. To repeat: Leiter thinks that his use of the term "Taliban" is "warranted." This is his subjective and idiosyncratic opinion. But it is simply juvenile to accuse someone of "misrepresent[ation]" and "false[hood]" simply for disagreeing with a subjective opinion. I might as well say that "Leiter is lying when he uses the word 'Taliban,' because his opponents do not even remotely resemble the Taliban in any respect, except in the highly abstract sense that they all allow religion to affect public policy, a characterization that is true of almost all American politicians." That wouldn't be fair: Leiter isn't "lying" about an objective truth here. He's merely using inflammatory language to express his political disagreement.

    And it's amusing to see Leiter pretending that his use of such terms is "in conjunction with extended documentation and discussion of the merits." What "extended" analysis of any sort is found here, or here, or here, or here, or here, or here, or here, or here, or here. There are plenty more posts that I could cite. Leiter's use of the term "Taliban" is reflexive, knee-jerk, and obviously intended to be inflammatory.

  25. Functional,

    I guess I can't fault you for failing to provide examples. Examples included: an attempt to ban books written by homosexuals from public libraries; the introduction of creationism into the public class room; a group pushing for the abolition of social security and the use of Genesis in the science classroom; more creationist nonsense; a broken link to Falwell; more creationist nonsense; more creationist nonsense; and a broken link to what I suspect is more creationist nonsense.

    What level of extended analysis is required for justifying the use of terms such as 'fascist', 'theocrat', or 'Taliban'?

    Does Leiter bother explaining why we might be warranted in calling a group bent on teaching our kids creationist nonsense in public schools '[insert name of state] Taliban'? Does he bother explaining why creationist nonsense is nonsense? No. He's not a biology teacher and we shouldn't want him to cut and paste a biology textbook into a blog entry every time some uneducated mob decides to take their case to the schoolboards. There is no extended analysis at either of these levels. Is any needed? If it reflexive and knee-jerk? Maybe. Is any careful thought required for using a term of abuse for religious fanatics hell bent on denying well established scientific claims? No. If you can't dismiss these people reflexively at this point, what is the point of careful thought and analysis. Are the remarks inflammatory? Maybe. Remind us why that is wrong?

  26. Substituting one bad argument for another, Brian Leiter comments here: "In fact, the view I expressed of Berlin is a view widely shared by philosophers--but to know this, you'd have to know some philosophers, and know something about philosophy, of course." Who are these philosopher I would have to know? I assume

    As for my previous "ignorant" remarks on his blog, I merely stated, correctly, that Anglo-American philosophy has shown little regard for history and the history of philosophy. Brian Leiter refuted my argments with the following: "Kosta Calfas has an exceptionally superficial knowledge of what is actually going on in philosophy in the English-speaking world these days." See, now THIS is Ad Hominem. You can read the whole exchange here (

    Fortunately, I'm not alone in having an "exceptionally superficial knowledge of what is actually going on in philosophy" when I state there is little appreciation for history in philosophy. As the late Sir Bernard Williams of Oxford put it:

    "Indeed, a lot of philosophy is more blankly non-historical now than it has ever been. In the so-called analytic tradition in particular this takes the form of trying to make philosophy sound like an extension of science... The American philosopher who stuck on his office door the notice 'Just say NO to the history of philosophy' was probably riding on the idea that the same could be said of philosophy."

  27. Are people making the distinction between (i) an unwarranted insult, and (ii) an insult that instances the ad hominem fallacy? This is also important.

    Suppose some genteel philosophical person provides an argument for motivational internalism in metaethics. She gives a clear statement of her argument, unmixed with anything else. If it helps, assume that the argument is pretty interesting and worth taking a look at. Then I present a bona fide criticism of her argument, with the sentence "You are such a fucking dipshit" at the end. My insult is completely uncalled for. Yet I have not committed the ad hominem fallacy.

    Moral: Just because an insult is completely uncalled for doesn't make it an instance of the ad hominem fallacy.

    So there are (i) warranted insults, (ii) unwarranted insults, and (iii) insults that instance the ad hominem fallacy.

    Also, Brandon writes:

    "Richard, since an ad hominem fallacy is a fallacy of irrelevant, any insult is an example of an ad hominem fallacy if it is (1) put forward where an argument is needed; and (2) is not such an argument"

    That doesn't follow. An ad hominem fallacy (at least typically) must be an argument from X's personal characteristics to some conclusion on the argument X has presented. If I simply say "He sucks doo-doo!!", I am not presenting an argument from his personal characteristics to some conclusion on the argument he has presented. I'm simply saying he sucks doo-doo, or expressing my scorn for him, or something. (I suppose there might be some wild scenario where my saying "He sucks doo-doo!!" conversationally implicates my endorsement of such an argument, but those are few and far between).

    Now suppose I'm in a situation where an argument is needed--I really do need to give an argument. Unfortunately, I don't recognize this need, so I let loose with "He sucks doo-doo!!" Surely I haven't committed any fallacy, much less the ad hominem fallacy. After all, I haven't presented any argument, fallacious or otherwise.

    This isn't to say I'm above criticism. At least I've failed to recognize the fact that an argument was needed. That was a failure on my part. Probably I've done much worse than that, by failing to recognize that "He sucks doo-doo!!" was absolutely uncalled for. But at no point have I committed a fallacy.

  28. I'll also add, re the example from Cohen, that there is a distinction between (i) an argument, and (ii) a presentation of an argument. One is a free-floating propositional structure, evaluable for validity and soundness and the like. The other is a social act, evaluable for appropriateness and courtesy and the like.

  29. So there are (i) warranted insults, (ii) unwarranted insults, and (iii) insults that instance the ad hominem fallacy.

    All trivially true (although not recognized often enough).

    But what of it? Are we to conclude that category i can be used without hesitation simply because it is neither a fallacy nor unwarranted?

    No. It might be true, for example, that the women I brush against in the street is fatter than a rhinoceros, but that does not make it polite or useful to point out that fact. Or more to the point here, the line between categories i and ii (warranted and unwarranted) is not at all clear. This is because whether an insult is warranted or unwarranted depends on judgments that are extraordinarily subjective, personal, tied up with moral premises that one cannot rationally expect others to share. In practice, this means that when I think an insult is warranted, other people may with perfect plausibility and rationality think it unwarranted.

    All of which is to say, a wise and thoughtful person will not conduct himself by letting fly with insults at every possible opportunity.

  30. Sorry to post yet again, but I like this example:

    Suppose I notice how ugly you are and I present to you the following argument (assume you've done nothing to deserve this):

    1. You are extremely ugly.
    2. For any given extremely ugly person x, odds are, hardly anyone is willing to have sex with x.
    3. Therefore, odds are, hardly anyone is willing to have sex with you.

    The argument itself might be unimpeachable. But the fact that I presented it to you makes me a rotten bastard. I am definitely worthy of contempt, and perhaps worthy of being spat on.

    Moral: Just because an argument is unimpeachable doesn't make it OK to present the argument to others.

  31. Sorry, functional, we posted at the same time. And making the same point.

    But I'm not sure if I want to agree that questions of warranted insult are the kind that we can't expect others to share. Just I expect people to share my basic views on torture and rape (so much so that I will condemn them for disagreeing), there might be some insults for which I expect others to share my view that they are warranted to or unwarranted (so much so that I will condemn them for disagreeing).

    For example, I expect people to share my view that calling a stranger monstrously fat is unwarranted, no matter how true or how artfully presented. If they don't share my view, I will condemn them as callous bastards. Are you not with me on this?

    (And whether moral questions are subjective or non-rational seems to me beside the point. I think my opposition to torture and rape to be completely non-rational, and that there aren't any wrongness facts out in the world, ready to be described. Nevertheless, if you don't share my opposition to torture and rape, I will condemn you without a second's thought)

  32. Two minor points. Contrary to what you suggest, functional, it would seem to follow that an insult is warranted if it is neither fallacious nor unwarranted. I would have thought the warranted ones just were the ones that were not unwarranted or, which comes to the same thing, not without warrant.

    I think I see what you were getting at, though. It seems that there is a difference between an epistemic warrant and a moral one. Perhaps the epistemic warrant is necessary for the moral one but not sufficient as the above examples seem to establish.

    It is a mistake, however, to think of an insult as an instance of an ad hominem since insults are not arguments. Elements of arguments might be insulting. The offering of an argument might be a clever way of insulting someone but 'You're a pinko' does not make for an argument. It does that only once set along side further elements in reasoning towards some conclusion.

  33. Yeah, in the sense of "warranted insult" I was working with, I doubt saying "you're fat" is warranted just in case the target is indeed fat. I meant a morally warranted insult, the kind of insult that's worth making, or that the person to deserves to have directed their way.

    Some milquetoasts might think that insults are never morally warranted, but I took it that everyone in this conversation agreed that insults are morally warranted at least sometimes.

  34. Cole writes:

    "I'll also add, re the example from Cohen, that there is a distinction between (i) an argument, and (ii) a presentation of an argument. One is a free-floating propositional structure, evaluable for validity and soundness and the like. The other is a social act, evaluable for appropriateness and courtesy and the like."

    Your second sentences does not follow from your first, because (a) "a free-floating propositional structure" is a REPRESENTATION of an argument, not necessarily the argument itself. We could represent arguments by other means (eg, by diagrams); and because (b) Insofar as it is a representation of an argument, a propositional structure only represents a PROPOSITIONAL argument.

    Since Frege and Wittgenstein focused everyone's
    attention on propositions, we've forgotten that there are lots of other locutions (eg, commands, requests, promises, etc), which can be used in arguments. If I exhort you to "Study hard!", where's the proposition?

    Finally, your structure assumes the content and the presentation of an argument are independent of one another. If they are not, as when a witness in court tells us what he saw on some past occasion at which we were not present, then it behooves us to assess the presentation in order to make a rational assessment of the content. To do otherwise, is irrational.

  35. If you exhort me to "Study hard!", where's the argument? "Non-propositional argument" is an oxymoron.

    Also, we should distinguish between arguments and testimony. In testimony (e.g. witnesses in court), we are asked to believe something merely because the person is asserting it. In such cases, their character and reliability is obviously of central importance and relevance. Not so for arguments. There the authority is external -- I ask that you believe my conclusion, not because I say it, but because rationality compels you (given that the premises are undisputed, etc.).

    If a premise P is disputed, then I might present a sub-argument which establishes P as its conclusion. And so forth. At no stage do I ask you to "trust me". That's entirely unnecessary. A good argument will force you to accept the conclusion on the basis of premises (or sub-premises) you already accept. No testimony required. Hence my character is strictly irrelevant, and ad hominems would be genuinely fallacious and irrational.

  36. Kofi: The way "proposition" is typically used in contemporary philosophy, it isn't a type of locution alongside commands, requests, promises, etc. A proposition is something like an abstract bearer of semantic value (content, truth-value, etc.) Perhaps the locution-type you were thinking of is assertion?

    In any case, an argument can be represented by a diagram, by a series of Latin characters, by a series of Cyrillic characters, by an utterance, by a series of hand signals, etc. But the argument itself (the thing being represented) is a free-floating propositional structure.

    Hope this helps.

  37. Actually Richard, there has been literature describing "visual arguments." While I'm skeptical about the existence of such a thing, it begins with the premise that language and propositions facilitate reasoning, but are not a sine qua non for it.

    Antonio Damasio, the famous neurologist has done studies which suggest reasoning to be linked to the capacity for emotion and cognition located in the right pre-frontal cortex rather than language centers of the brain. Still, it's an open question whether reasoning itself can externally manifest in a medium other than propositions. His finding is that you cannot reason without emotion.

    As for dispensing with "trust" and relying on the force of the better argument, this is the really contentious thing. A "good argument" cannot "force" you to accept anything precisely because the persuasiveness of an argument is independent of its validity (and sometimes its soundness depending on whether you have a pragmatic definition of truth i.e. the argument is true because it works in persuading).

    Getting back to your claim about the independence of an argument from "character" evaluations and testimony. We often poo poo arguments that appeal to emotions rather than reason, and the question is whether this dichotomy is false, i.e. the argument that reaches a conclusion that makes me "feel better" is the better argument.

    Ad Hominem arguments work because they appeal to often tacit dislike of the person you are arguing against. You are attempting to persuade your reader by invoking similar sentiments and this is where the "force" of the argument lies. Is the Ad Hominem argument fallacious? Yes! Is it irrational? Not necessarily, it depends on your concept of reason!

  38. Also, as an example of the testimony / argument point Richard made: Suppose I see a trail of cake crumbs leading to a cake-lover I know to be unscrupulous with icing smeared all over his face, nervously saying, "I did not eat the missing cake!" The circumstantial evidence makes it reasonable to conclude that what he says is false. But I'm not evaluating any argument he has presented. If he had cited a number of facts and tried to show how they strongly support the conclusion that he didn't eat the missing cake, then there'd be an argument to evaluate. But then the circumstantial evidence would be irrelevant to the strength of the argument.

  39. Cole said,

    "That doesn't follow. An ad hominem fallacy (at least typically) must be an argument from X's personal characteristics to some conclusion on the argument X has presented."

    As this stands, this can't be quite right; if it were true, then poisoning the well, which is certainly a typical case of ad hominem, would never be fallacious -- it's pre-emptive by definition, whereas your formulation requires that ad hominem always be reactive.

    I think I understand, though, your point about how expression of an insult where an argument is needed might not itself be put forward as an argument -- it might be a complete breaking-off of rational discussion. That's fair enough. But supposing a stable context of rational discussion, I can't think of a counterexample to the claim that an insult is a commission of ad hominem when (1) it is put forward where an argument is needed and (2) is not such an argument.

    One thing that seems relevant to this discussion is the principle of charity; since relevance and irrelevance can be slippery, charity would seem to require that we presume people not to have committed the fallacy except in cases where it's fairly obvious that they have.

  40. I didn't even think poisoning the well was a fallacy (i.e., fallacious argument-kind) just much as a nasty way to do business. Suppose I conclude my portion of the debate by saying, "So I'll wrap it up and hand it over to my opponent, who can amuse you with his wild ideas". Have I given an argument? Obviously not, I'd think. Hence I haven't given a fallacious argument.

  41. Brandon: Also, you can have temporary suspensions of rational discourse without a complete breaking-off thereof. For example:

    "Some fools never learn. Recently Mr. X, a fool if there ever was one, argued that since p, it follows that q. Apparently this is what passes for thought in the land of fools. Of course, there's an obvious objection, that p is ambiguous between p', which is obviously true but doesn't entail q, and p'', which does entail q but is obviously false. This point has been made countless times and Mr. X has had plenty of time to address it, but alas! he's more interested in impressing his fellow fools than in the truth."

    Notice how a legitimate criticism of an argument is contained within all the invective. It can be isolated, formalized, and evaluated. So rational discourse is still going on. It's just heavily decorated.

  42. Cole, Richard -- methinks you are confusing the map with the territory.

    Arguments are only ever "free-floating propositional structures"? What a quaint, reductionist notion! How old-fashioned! Funny how contemporary western philosophers have so much trouble accepting as arguments anything other than propositional structures.

    Not everyone is so culture-bound. In 2003, the Australian Native Title Tribunal was presented with a painting prepared by the Ngurrara people of the Great Sandy Desert of Western Australia as the argument in a hearing preliminary to a court case over traditional land rights claims. The fact that the community were able to make this painting (which represented pictorially their traditional relationships to various geographical features in the region) was presented as their case for their claim to the land. Although in no sense of the term "a propositional structure", this looks, sounds and quacks like an argument to me. The Native Title Tribunal agreed.

  43. There's a big difference between arguments in philosophy and arguments in a courtroom. It's a word with many meanings.

    Also, you're being mean for no reason.

  44. Cole, Richard -- With your limited notion of what an argument is, how do you account for arguments for (or agaainst) undertaking some action?

    The head of the argument is an exhortation, or a command, or a request, or a promise. None of these types of statements are propositions, since they do not have truth-values. (What could possibly be the truth value of "Study hard!" or "I promise to study hard" ?) The supporting justification (the body of the argument) may be propositional (eg, statements about the world), and/or they may be appeals to the values of the actor(s) involved. How does one assess the truth-content of an appeal to values?

  45. Cole --

    Sorry, I don't mean to be mean. It's just how I was taught to do philosophy. Apologies if any of this causes offence.

    I know "argument" is a word with many meanings. I am criticizing your use of it as overly-narrow, and ill-matched to how the word is used (a) in the real world (see posts about the law) and (b) by philosophers of argumentation (see my most recent post).

  46. Cole,

    It's true that one can have temporary suspensions of rational discussion; the question, I suppose, would be whether such temporary suspensions aren't just what is most usefully meant by the term 'fallacy'. After all, one could argue that every fallacy of irrelevance is a temporary suspension of rational discussion. If I'm right, for instance, a case of ad hominem doesn't have to be put forward as an argument; it just has to be fulfilling a role in the discussion that needs to be filled by an argument. And this, I think, follows fairly easily from thinking of ad hominem as a particular kind of irrelevance in reasoning; and irrelevance in reasoning need not always take the form of an argument itself. Thus, the "wild ideas" example could be ad hominem, assuming that it isn't just the conclusion of a prior argument.

    (One of the difficulties of pinning down a fallacy of irrelevance like this one is that it is very context-sensitive; in some types of rational discussion insult might be just the idiom in which the argument is expressed, in which case it wouldn't be any more fallacious than making one's syllogisms rhyme. So it depends on what sort of discussion is going on, which is one reason why I think reasonable people can disagree. The "Mr. X" could be a completely rational argument in some contexts; the insult, for instance, could just be part of the idiom of discussion. In that case, there wouldn't be any suspension of rational discussion at all. What wouldn't be rational is to use the labeling of Mr. X as a fool as an irrelevant substitute for argument.)

    I suppose one could argue that poisoning the well isn't a real ad hominem; after all, the best classification in this sort of case will just be whatever classification that's most useful, and if it turns out that poisoning the well is better classified as something other than a fallacy, then that's the way it should be classified. It often is treated as an ad hominem fallacy, though, and I think the fallacy can be understood in a way to make it clear why poisoning the well is treated as one.

  47. Kofi,

    (a) It isn't clear to me how the painting constituted an argument. It sounds more like evidence (e.g. perhaps the lawyers would say, "This painting demonstrates the ties that the Ngurrara had to various geographic features," etc.).

    (b) Practical reason concerns normative facts. I can present an argument for the conclusion that you ought to study hard. But the mere imperative "study hard!" cannot be the conclusion of an argument. If it has no truth value, then there can be no evidence or justification for it, so it cannot figure in an argument. (Can you cite any "philosophers of argumentation" who disagree with this?)

    That arguments are propositional is true by definition. That's just part of what the word "argument" means. If you deny this, you are simply speaking a different language from me. As such, I doubt this disagreement will prove productive.

  48. just browsing (link from brian leiter's blog)
    two thoughts

    1. "study hard!" could be an ought imperative, which has assertive value, it depends on context.

    2. i always thought that ad hominem could be inductively valid and therefore sometimes not fallacious - ie in a court room
    (as pointed out)

  49. Richard --

    I'm contesting Cole's definition of arguments as free-floating propositional structures. You can't simply say in response to my challenge that arguments are this way because the word "argument" only refers to propositions. (In the trade, that's called a circular argument, another one of the styles of reasoning traditionally called a fallacy. Because truth is preserved across the inference step there is nothing fallacious about circular arguments in general.)

    On practical reasoning, of course an imperative can be the conclusion of an argument, as I said in an earlier comment. If not, then how could anyone ever make an argument in favour (or against) doing some action? We hear and make these arguments all the time in everyday life -- just take the last 3 years of public discourse on Iraq as an example.

    Perhaps what you are really saying is that in the model currently adopted WITHIN MAINSTREAM PHILOSOPHY arguments can only have propositions as their conclusions. If so, this just supports my criticism made above (in my post mentioning Frege), that philosophers have been seduced by one model (a model with propositions), had have ignored other types of utterances.

    You asked for philosophers of argumentation: have a look at Doug Walton's books on practical reasoning, or the work of Katie Atkinson on modeling practical reasoning with values.

  50. "how could anyone ever make an argument in favour (or against) doing some action?"

    I told you already! We can make arguments for the conclusion that some action ought to be done. The thing in bold is a proposition of normative fact. This is how practical arguments work. We present reasons or evidence for thinking that something ought or ought not to be done. THIS is what's been going on for the past three years concerning Iraq. You can't have evidence for an imperative, by contrast. That's just nonsensical.

    It might be true that the U.S. ought not to have invaded Iraq. It cannot be true that "Don't invade!"

  51. Richard --

    1. Sincerely, how does one assess the truth value of a statement of the form "some action ought to be done"? I cannot see how such a statement could be called propositional. What is the test one would undertake (even in principle), whose answer is either True of False, which would provide one with the assessment?

    2. It is clear that we disagree about what the conclusion of an argument can be. I allow non-propositional utterances, which you don't. This strikes me as yet another compelling reason against your model of an argument. (I also allow non-propositional utterances elsewhere in an argument, and other forms of argument. These positions put me in the majority of people in the world, even if not in the majority of philosophers.)

    3. Regarding your statement about the Ngurrara painting as evidence rather than as argument: The novelty of the Native Title Tribunal ruling in the Ngurrara case (and why it was widely reported in the press) was that the painting itself was accepted as the basis for the community's claim to proceed to the next stage of the legal process, not some written or verbal argument based on the existence of the painting. I think that makes the painting itself an argument.

    Of course, one could later construe the Tribunal activities in the form of a propositional structure, as you did above. Putting aside the question as to whether the Tribunal were construing their actions this way, or even whether they thought they were, my criticism is not eliminated by such a reconstruction.

    For, if one believed that the painting was evidence for a propositional argument, what is the nature of that evidence? A painting is clearly not a free-floating propositional structure, so we would have (in your reconstrual) an argument whose conclusion is a proposition and whose body (the evidence) is a painting. The problem with the propositional paradigm for arguments just moves one step back; it does not go away. Similarly, with any further reconstrual.

    4. If one tries to stand outside this debate between your good self and Cole (on the one side) and myself (on the other) one may perhaps see a Kuhnian conflict of paradigms:

    (a) You (or rather, Cole) asserted a particular paradigm model for an argument.

    (b) I attempted to demonstrate what I believe is an inadequacy of this paradigm by presenting an example of argument (the Ngurrara painting) which was not in the form permitted by the paradigm (propositional structure). (I also presented another example against, but ignore that for now.)

    (c) You interpreted the Ngurrara example in a way which fits the propositional paradigm (ie, as evidence for a propositional argument). This is perhaps the classic move by a defender of a contested paradigm. (That sentence is not intended as a criticism, since you could have just ignored my example.)

    (d) I've attempted (Point 3 of this comment) to show that this intepretation does not resolve the inadequacy I believe exists with the paradigm.

    (e) You may well respond to my attempt in Point 3, with a further defence of the paradigm. And so on.

    Looking at our debate from this external position, I conclude that resolution between us is not possible (as you indeed suggested above). I don't think this is because we are using one word with different meanings -- such a conflict would potentially be resolvable by adoption of a shared dictionary -- but because we have adopted different paradigmatic models of an argument.

    Resolution is only possible if exactly one side adopts the other's paradigm, or if some uber-paradigm can be found which incorporates both paradigms as special cases. Since you don't accept non-propositional entities as components of arguments, I can't imagine such a uber-paradigm would be acceptable to you (since it would mean accepting examples such as the Ngurrara painting as arguments). I conclude that resolution is not possible.

    5. This debate has strengthened significantly my conviction that mainstream philosophy has a paradigmatic model of argument which is both seriously inadequate for real-world reasoning and which does not incorporate current thinking in the philosophy of argumentation.

  52. ``. . . there is an exemplary anecdote recorded by Sanctorius and repeated by Galileo: an Aristotelian is present at a dissection which shows by ocular demonstration that nerves originate in the brain and not in the heart (as Aristotle had claimed); he then confesses to the anatomist that he had made him see the matter so palpably and plainly that if Aristotle's text were not contrary to his ocular demonstration, and did not state so clearly that nerves originate in the heart, he would be forced to admit that what he had seen was true.''

    Ian MacLean: Logic, Signs and Nature in the Renaissance:
    The Case of Learned Medicine.
    Cambridge University Press, 2002,
    p. 192.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.