Why using quotation marks doesn’t cancel racism or sexism. With a brief response to Agnes Callard

Would you show an ISIS video, depicting a brutal killing of hostages, to the survivor of their murders? Of if you prefer a linguistic medium: would you read Breivik’s Manifesto to a survivor of his massacre? – Asking these questions, I’m assuming that none of you would be inclined to endorse these items. That’s not the point. The question is why you would not present such items to a survivor or perhaps indeed to anyone. My hunch is that you would not want to hurt or harm your audience. Am I right? Well, if this is even remotely correct, why do so many people insist on continuing to present racist, sexist or other dehumanising expressions, such as the n-word, to others? And why do we decry the take-down of past authors as racists and sexists? Under the label of free speech, of all things? I shall suggest that this kind of insistence relies on what I call the quotation illusion and hope to show that this distinction doesn’t really work for this purpose.

Many people assume that there is a clear distinction between use and mention. When saying, “stop” has four letters, I’m not using the expression (to stop or alert you). Rather, I am merely mentioning the word to talk about it. Similarly, embedding a video or passages from a text into a context in which I talk about these items is not a straightforward use of them. I’m not endorsing what these things supposedly intend to express or achieve. Rather, I am embedding them in a context in which I might, for instance, talk about the effects of propaganda. It is often assumed that this kind of “going meta” or mentioning is categorically different from using expressions or endorsing statements. As I noted in an earlier post, if I use an insult or sincerely threaten people by verbal means, I act and cause harm. But if I consider a counterfactual possibility or quote someone’s words, my expressions are clearly detached from action. However, the relation to possible action is what contributes to making language meaningful in the first place. Even if I merely quote an insult, you still understand that quotation in virtue of understanding real insults. In other words, understanding such embeddings or mentions rides piggy-back on understanding straightforward uses.

If this is correct, then the difference between use and mention is not a categorical one but one of degrees. Thus, the idea that quotations are completely detached from what they express strikes me as illusory. Of course, we can and should study all kinds of expressions, also expressions of violence. But their mention or embedding should never be casual or justified by mere convention or tradition. If you considered showing that ISIS video, you would probably preface your act with a warning. – No? You’re against trigger warnings? So would you explain to your audience that you were just quoting or ask them to stop shunning our history? And would you perhaps preface your admonitions with a defense of free speech? – As I see it, embedded mentions of dehumanising expressions do carry some of the demeaning attitudes. So exposing others to them merely to make a point about free speech strikes me as verbal bullying. However, this doesn’t mean that we should stop quoting or mentioning problematic texts (or videos). It just means that prefacing such quotations with pertinent warnings is an act of basic courtesy, not coddling.

The upshot is that we cannot simply rely on a clear distinction between quotation and endorsement, or mention and use. But if this correct, then what about reading racist or sexist classics? As I have noted earlier, the point would not be to simply shun Aristotle or others for their bigotry. Rather, we should note their moral shortcomings as much as we should look into ours. For since we live in some continuity with our canon, we are to some degree complicit in their racism and sexism.

Yet instead of acknowledging our own involvement in our history, the treatment of problematic authors is often justified by claiming that we are able to detach ourselves from their involvement, usually by helping ourselves to the use-mention distinction. A recent and intriguing response to this challenge comes from Agnes Callard, who claims that we can treat someone like Aristotle as if he were an “alien”. We can detach ourselves, she claims, by interpreting his language “literally”, i.e. as a vehicle “purely for the contents of his belief” and as opposed to “messaging”, “situated within some kind of power struggle”. Taken this way, we can grasp his beliefs “without hostility”, and the benefits of reading come “without costs”. This isn’t exactly the use-mention distinction. Rather, it is the idea that we can entertain or consider ideas without involvement, force or attitude. In this sense, it is a variant of the quotation illusion: Even if I believe that your claims are false or unintelligible, I can quote you – without adding my own view. I can say that you said “it’s raining” without believing it. Of course I can also use an indirect quote or a paraphrase, a translation and so on. Based on this convenient feature of language, historians of philosophy (often including myself) fall prey to the illusion that they can present past ideas without imparting judgment. Does this work?

Personally, I doubt that the literal reading Callard suggests really works. Let me be clear: I don’t doubt that Callard is an enormously good scholar. Quite the contrary. But I’m not convinced that she does justice to the study that she and others are involved in when specifying it as a literal reading. Firstly, we don’t really hear Aristotle literally but mediated through various traditions, including quite modern ones, that partly even use his works to justify their bigoted views. Secondly, even if we could switch off Aristotle’s political attitudes and grasp his pure thoughts, without his hostility, I doubt that we could shun our own attitudes. Again, could you read Breivik’s Manifesto, ignoring Breivik’s actions, and merely grasp his thoughts? Of course, Aristotle is not Breivik. But if literal reading is possible for one, then why not for the other?

The upshot is: once I understand that a way of speaking is racist or sexist, I cannot unlearn this. If I know that ways of speaking hurt or harm others, I should refrain from speaking this way. If I have scholarly or other good reasons to quote such speech, I shouldn’t do so without a pertinent comment. But I agree with Callard’s conclusion: We shouldn’t simply “cancel” such speech or indeed their authors. Rather, we should engage with it, try and contextualise it properly. And also try and see the extent of our own involvement and complicity. The world is a messy place. So are language and history.

18 thoughts on “Why using quotation marks doesn’t cancel racism or sexism. With a brief response to Agnes Callard

  1. Hi Martin,

    I think you’re right in pointing out that “the idea that quotations are completely detached from what they express strikes me as illusory”. On the other hand, when we choose to talk about some set of insults/slurs by way of description rather than quotation (“gendered slurs”, say, or “the n-word”), don’t we also understand *that* in virtue of understanding real insults? I think that both ways of addressing the phenomenon (of insulting/hate speech) might be distressing/triggering to some — the question is whether one of the ways is, in general, less so than the other; or whether one of the ways, in general, signals greater distance than the other. (And I think I might lean towards what I take to be your position here; namely, that by describing rather than quoting we distance ourselves more.)

    Liked by 1 person

    1. Thanks, alienaid, great point! Yes, I agree that we understand both quotations of slurs and descriptions in virtue of understanding the real thing. (I’m not excluding that descriptions might open up extra paths, like e.g. analogies, that aid the understanding.)
      So does this mean that descriptions will trigger someone in the same way? I’d agree again that the distance to exposure to the real phenomenon might be greater. However, ultimately I’d defer to someone who underwent the experience in question.

      Looking for examples now:
      – I would hope, for instance, that my opening examples work as (minimal) descriptions and would hope they affect the reader much less than quotations would.
      – Conversely, my three-year-old daughter and I recently listened to the audiobook Pippi Longstocking. Right at the beginning, I flinched several times: I had completely forgotten how often the n-word is used, and hearing it triggered a physical reaction. (Back in the day, it didn’t worry me; now it does. Even though I’m not part of what I take to be the target group, the insult seems to carry over.)

      Liked by 1 person

  2. I don’t think the comparison of Aristotle’s philosophical writings to the manifesto of a terrorist is apt.

    Even if philosophical theorizing and half-baked, plagiarized, post-hoc rationalizations of antecedent hatred exist on the same spectrum, that doesn’t make them comparable. Black and white are similar in that they exist on the same spectrum, but serious claims of similarity should end there, since black and white exist on opposite sides of that spectrum. Aristotle’s philosophical writings and manifestos exist on the spectrum of writings, but that’s pretty much all they have in common.

    Isn’t it precisely the genre, intentions, purposes, and context of Breivik’s manifesto that make it impossible to “read literally” in Callard’s sense? The manifesto is less like an object of contemplation than an incitement to harm. No matter the evil of slavery, I’d be hesitant to draw lessons about (reading) Aristotle’s philosophical writings from considerations about (reading) Breivik’s manifesto.


    1. Thanks! You raise what I think is a crucial question for this issue. And it’s probably not going to be decided quickly. Let me make two points in response:

      1) Initially, I didn’t think of the Breivik example as a comparison to Aristotle. I was looking for a piece of text that, contrary to a categorical understanding the use-mention distinction, would convey why texts can carry threats etc and terrorise even if they are ‘merely quoted’. Usually, examples of hate speech get quickly dismissed and discussants advise others “to toughen up” or “stop behaving like snowflakes.” I was hoping to find an example that can’t get dismissed so easily. My claim is that such texts have harmful or hurful effects even if merely quoted.

      2) A. Callard claims that we can exempt texts from those having such effects by reading them literally. Against her, I deny that such a literal reading works. Much should be said about this, but my main reason is that, even if the speaker were just entertaining pure thoughts, the *hearer* might not be able to turn off their own attitudes. So my claim is about the reception or interpretation of such texts.

      You seem to argue that certain texts (qua being intend as objects of contemplation?) don’t even have such effects, and thus, in those cases, a literal reading is possible. (Do you mean the texts themselves? Or Aristotle?)
      Whatever you say, my response is this: Independently of the kind of text and independently of the writer’s intentions, the hearer or reader can be affected in several ways. For instance, I can find the Aristotelian idea that women are misformed men hurtful, even if Aristotle would (with Callard’s help) lay claim to the idea that he was just observing nature. Whatever else we say about Aristotle, I think the hearer has a legitimate reason to feel hurt by the text. Thus, we don’t have a literal reading, and legitimately so. The moral claim that I see following is that it would be wrong to deny the legitimacy of such a response by saying “just toughen up; Aristotle is an alien”.

      A further question arising from your worry and Callard’s claims is this: If Aristotle is an alien, but Breivig is not really on the same spectrum, who is then more of an *alien* for us? By Callard’s reasoning Breivik is closer to us. But by your reasoning it seems that Aristotle loses his alien status.

      In any case, I think your question is important and I’m not sure how to answer it sufficiently. But I do think the comparison – however apt – might reveal some things.


      1. Thanks for your reply.

        “Initially, I didn’t think of the Breivik example as a comparison to Aristotle. I was looking for a piece of text that, contrary to a categorical understanding the use-mention distinction, would convey why texts can carry threats etc and terrorise even if they are ‘merely quoted’.”

        I see. I misunderstood. I think I agree that some texts can carry their hurtful content even if merely quoted or mentioned.

        But you still seem to be implying — correct me if I’m wrong — that if the Breivik manifesto can carry its hurtful content in this way, so can Aristotle’s philosophical writings. And I would reply that that conditional claim presupposes similarities that aren’t obvious — at least, not to me.

        I also think it’s plausible to think that one and the same content can be hurtful if used or mentioned in one context and benign if used or mentioned in another. I can mention one of Marx’s calls for revolution to rile up an already violence-thirsty crowd or I can mention it on a slide in an afternoon lecture when half of my class is falling asleep. In which context is it more plausible to think hurtful content was carried? As you contend, it depends on audience attitudes, sure. But it depends on lots more besides.

        “Independently of the kind of text and independently of the writer’s intentions, the hearer or reader can be affected in several ways.”

        It’s a nice question, the question of the extent to which the hurtfulness of content is constituted by audience uptake.

        But if it’s incumbent upon the presenters of certain content (to attempt) to take into account the attitudes of the audience, why is it not incumbent upon the audience (to attempt) to take into account the intentions of the writers of the content? Not that this will automatically erase hurtfulness. But a principle I see you putting in play here is something like: Be sensitive to the points of view of those involved. Shouldn’t that apply to the audience too?

        “You seem to argue that certain texts (qua being intend as objects of contemplation?) don’t even have such effects, and thus, in those cases, a literal reading is possible.”

        I’m not sure what I was arguing there, to be honest. I’m quite taken by your point that understanding mentionings piggy-backs on understanding straightforward uses. So, to my delight, you’ve probably changed my mind about the possibility of literal readings. So I grant that quoting Aristotle and quoting Breivik have that in common. There just seems to be something about quoting hurtful things from a philosophical treatise in an academic context that is importantly different from quoting hurtful things from an obviously intentionally hateful screed in a political context.

        Liked by 1 person

  3. At some point in the last twenty years analytic philosophy absorbed this obnoxious people that are tacitly forcing the continental notions about race, gender, and “society” to the discipline. This results in terrible analytic philosophy with hyper dogmatism. This one long piece of rubbish that exemplified this trend.


  4. Interesting, but I’m not sure about this: “understanding such embeddings or mentions rides piggy-back on understanding straightforward uses”.

    Imagine that I have just returned from a foreign country, let’s call it Manderia, whose language has an unusual alphabet. I write to you: ‘I learned that “~*^**~” is a word in Manderian’. It seems to me that you in all likelihood understand my embedding/mention without understanding the embedded morpheme (perhaps I don’t even understand it; I just copied it from my tour guide).

    So I think there is a clear distinction between use and mention, though it may not be *categorical* in the following sense: a speaker can use by mentioning. For example, I could quote a passage from Aristotle in such a way as to condemn it: ‘As Aristotle, stupidly, says, “S”.’ This is a kind of use of S – we could call it a meta-use – and may be dependent on its first-order meaning, but need not: perhaps I don’t understand S (it’s in Greek) but take it on someone’s word that it’s a bad statement.

    Of course, there are ways to set up the mention of something insulting as a way of just making the insult. For example, suppose that one is conversing with a woman and utters: ‘Well, as Aristotle said “S”,’ where S is a demeaning sentence about women. This could be a way of using S literally as a put-down while just pretending to mention it.

    As I see it, however, it is not that mention is determined by or dependent on use so much as that the distinction is recursive: a phrase can be embedded in a mention context, which can itself be embedded in a use context, and so on.

    Thought-provoking post; thanks.


    1. Many thanks, Joshua (if I may), for this very intriguing consideration! I realise I should have been more clear, and I also think that I moved rather quicky when running quotations and mentionings etc. together in the way I did. As you probably know better than I do, the literature on this issue is highly technical and I want my point to be rather neutral with regard to the theoretical options. So I really welcome your suggestions. Let me just make a few remarks:

      I happily grant that your foreign language case would be a good counterexample to the claim that, for each mentioned expression E, understanding “E” requires understanding the use of E.

      But i guess that we employ different notions of understanding. For my discussion, “understanding” would involve something like “being able, in principle, to use the mentioned expression”. You might object that this borders on circularity. But my point is that people are hurt etc. by such expressions because (1) they would, potentially, understand customary forms of use and (2) the distinction between use and mention isn’t clear cut. (In the same vein, I’d say that one cannot ‘entertain’ the content of a belief without understanding what it means to hold that belief. So I can’t merely entertain the idea that it might be raining without understanding what it means to believe that it’s raining. In this sense, quoting others, understanding ascriptions of beliefs or mentioning what someone says requires an understanding of what it means to actually hold a belief or use an affirmation.) If I am not even potentially able to use (or understand uses of) “~*^**~”, then neither the use of “~*^**~” would hurt me, nor – a fortiori – the mention of that word. In other words, foreign languages or languages unknown to the hearer form a special case.

      The “meta-use” case is more to the point. And here I would reply what I said about descriptions in response to alienaid above. The third case is indeed closest to what I have in mind, with the caveat that we do not need to ascribe the intention of insulting to the speaker.

      So to accommodate your objection, I should add that my discussion rests on a more comprehensive notion of understanding, where understanding mentionings involves at least the potential understanding of uses. Or to exploit a possible analogy to the ISIS video example: I’d treat the case of your foreign languguage like a case in which the viewer is not able to see the video, either because the viewer is blind or the screen doesn’t work. All the viewer understands is *that* a video is being shown.


  5. Thanks for this blog post. I had read the Callard NYT piece with some appreciation, but do think you raise some important questions. I think though, for myself, your point is overstated and obscures what may be a more indeterminate hermeneutical problem. For how to distinguish between immoral or hate speech or speech that incites to hateful or immoral action, from speech that is expressive or evocative of ideas we may find anachronistic, or wrong factually or morally, or even ideas that challenge our own assumptions and beliefs? This is where questions of what authority or respect we give the author may come in. This is also where social-historical context come in, as well as the context of the authors writings as a whole. I do not think this is just a matter of quotation marks, or to be fair whether or not we should read a writer as an alien would. It remains more unclear. Why read aristotle at all? is he just a bigoted dead white racist sexist male? but, if we do choose to read him, how can we read his notions of slavery and women in the context of his times, in the context of his writings on ethics and politics and human nature, and in such a way that we can both understand why he thinks these things, why they are perhaps wrong, and hopefully in the process think more clearly on why we think today the things we do.


    1. Many thanks for your comment! I completely agree that there are two relativelely distinct approaches involved in our treatments of past authors:
      – one is, as you suggest, explaining texts in their “social-historical context”,
      – the other is evaluating their “authority or respect”.

      Although they are not always neatly distinguished, these approaches are different in kind and method: the former presents specific hermeneutical problems, but it would be ‘moralising in advance of established facts’ to focus on the question of bigotry etc (thus, goin for the second approach). I’d also say that the former approach is more focused on the text and context, whereas the the latter is more focused on our own responses to what we take to be the author’s intentions. (By the way, I also think you can analyse current debates on free speech with the former approach. Perhaps worthwhile.)

      Now I think your impression of ‘overstating the case’ arises, when someone is professing to engage in the first approach but ends up going for the second issue. Am I doing this?
      When I wrote this text, I was initially interested in the current effects of (hate) speech. Then Agnes Callard’s indeed interesting piece came along, and I thought this is worth including. Using your distinction, I’d say she makes a similar move. She tells us whe should read Aristotle historically (literally), as an alien, not like someone involved in contemporary politics. And that’s fair. But by suggesting that there is a current question of whether Aristotle should be “cancelled”, she places her endeavour firmly in the second approach, where “authority and respect” are at stake.


      1. You start this off by saying I “don’t want to harm anyone in the audience,” but there’s a lot of hidden meaning there. Much of the fight of cancel culture is the demand for ever-increasing sacrifices in order to avoid a claim of subjective harm, coupled with debates about the level and honesty of that harm. In other words, the fight is about the distinction between (as you put it) courtesy and coddling.

        So, for example, it may well be that the claimed subjective harm of finding out that Aristotle was a sexist (without appropriate trigger warnings) is a “harm” which many people, including me, are in fact willing to impose.


        1. Thanks for your comment! Yes, fair enough. I guess depending on what is at stake, we all have to balance values and make choices. And then we might end up admitting that we are prepared to harm others to some degree (and perhaps warn them accordingingly). But that’s quite different from inventing reasons to argue that our choices are not harmful or morally superior (e.g. because of, ahh, free speech).


        2. I don’t know what you mean by “inventing reasons”. Can you clarify?

          I think that many of these free speech actions are fine, because they are a) not definitionally harmful or b) justifiable and outweighed by benefits, if you choose to define that as harm. This seems the most consistent approach, most obviously because it would otherwise require me to adopt a definition of “harm avoidance” which would require everyone to limit their conversations with folks who are universally acknowledged as being on the wrong side of morality. After all, telling David Duke he is immoral would presumably cause him subjective harm under your approach.


        3. I used the term “hurtful” instead of “harmful,” because I have worries similar to yours. (But also because the OP used it.) I’m not sure if I’m drawing a distinction here that makes no difference, but I’m imagining that claims of being harmed are subject to public demonstrability in a way that claims of being hurt are not. Hurt seems indexed to individual perspectives and attitudes and personalities in a way that harm is not; harm seems to be indexed to the encultured human being qua encultured human being. So I’d grant that if some uses and/or mentions carry hurtfulness, those same uses or mentions don’t necessarily carry harm. And, like you (or so I gather), I’m inclined to think the conversation should focus more on harmfulness than hurtfulness (granting, of course, that there’s a difference like the one I claimed there is).

          Liked by 1 person

  6. Just to be entirely clear about where we diverge:

    Would you show an ISIS video, depicting a brutal killing of hostages, to the survivor of their murders?

    Yes. Of course, I wouldn’t *seek them out and surprise them* but if I was teaching a course on “ISIS propaganda videos”, or was asked “why do you say the ISIS video showed ___, can you prove it?” or anything relevant, then I would play the video.

    would you read Breivik’s Manifesto to a survivor of his massacre?

    Yes. Similarly, I wouldn’t *seek them out to do so* but if I were discussing the manifesto I would absolutely quote it. The fact that someone is a survivor grants them no particular comprehension of the manifesto, nor any particular standing regarding its philosophical interpretation.

    FWIW, I also support the rights of the Nazis to march and speak in Skokie, or the right of someone to quote from Mein Kampf while explaining its problems, even although plenty of my Jewish ancestors were murdered by the Nazis.

    Asking these questions, I’m assuming that none of you would be inclined to endorse these items. That’s not the point.
    Well, if you’re operating from such a strange assumption it is very much a relevant point, don’t you think?

    My hunch is that you would not want to hurt or harm your audience. Am I right?
    No. Well, yes in that I wouldn’t “want” to, I take no particular joy in hurting anyone. But I would not feel any unusual obligation to avoid it, and I think you’re using “would not want to hurt” in its colloquial (as opposed to logical) sense.

    Well, if this is even remotely correct, why do so many people insist on continuing to present racist, sexist or other dehumanising expressions, such as the n-word, to others?

    Because, among other reasons
    a) we think what they said is important–not necessarily good but merely important
    b) we think it is best to quote what they said. Mark Twain wrote what he wrote. His books are not filled with references to the “N-word.” So did Malcolm X, or Hitler, or Che, or anyone else.
    c) we may not–in fact, we almost certainly do not–use the same definitions as you do, for a variety of things including but not limited to “racist” and “___phobic” and, most-hated-of-all, the dreaded and never-really-defined “dehumanizing.”

    Again, I don’t run around quoting Mark Twain to people, that would be obnoxious and racist. But

    And why do we decry the take-down of past authors as racists and sexists?
    Well, among other reasons:
    a) we think many of the past authors were products of their time and we question the hubris of judging ancients by modern standards;
    b) we think that there is an unfortunate and growing inability to distinguish the morality of the speaker from the truth of the speech, and therefore we think people will (improperly) ignore important or relevant views because the “wrong” thoughts of the speaker;
    c) we are generally under the impressions, that “take-downs,” overall, act collectively to diminish and depress the emergence and analysis of ideas, which we view as a net good;
    d) we think that, to the degree such people arr sexist or racist, that this is usually pretty self-evident from their speech/writings and we think it is also important for people to draw their own judgments without our input or supervision.

    Under the label of free speech, of all things?


    1. Thanks for clarifying! I’m not sure we disagree all that much. If I check your list and compare it to what I say in the last paragraph, I don’t see that much deviation, once I take into account all the qualifications (about educational contexts etc.). Rather, there might be differences of emphasis.
      That said, there is one issue that we might disagree about strongly. You write. “we think that, to the degree such people are sexist or racist, that this is usually pretty self-evident from their speech/writings.” – I doubt that self-evidence. Evidential status and sensibilities are subject to (historical) change. When I grew up, for instance, I didn’t recognise the n-word as racist. I had to be educated about this. Clinging to what I happen(ed) to find self-evident would be a very shaky ground. (That’s why I generally think such questions should be addressed in on-going dialogues. If you tell me that I’m being offensive I have to deal with that and find perhaps a compromise.)

      Just three further points:
      – You asked about “inventing reasons”: Imagine we’re having a conversation and I ask you to stop using a particular word because it offends me. If you respond by saying that you’ll have to continue, say, because you want to make an important point, that’s fine. Looking at your response, I’d say you’ve made a choice, perhaps a legitimate one, but one that comes with a moral cost (namely the cost of offending your interlocutor, at least for the moment). You could bite the bullet and acknowledge that it comes with a cost that you think is outweighed by the benefits. And that would be ok. What would not seem ok (to me) is if you were to start giving reasons why I have, say, no right or grounds to be offended. – This latter response is what I’d call “inventing reasons”.

      – You say: “Asking these questions, I’m assuming that none of you would be inclined to endorse these items. That’s not the point.
      Well, if you’re operating from such a strange assumption it is very much a relevant point, don’t you think?” — That might be a misunderstanding: My assumption is simply that you do not endorse the practices of ISIS or Breivik. So I’m simply assuming that you’re not sympathising with terrorists. I thought that assumption isn’t so strange. Or am I missing something?

      – You say “we question the hubris of judging ancients by modern standards”. While I doubt that modern standards should be the only standards, I think it is literally impossible to completely switch off one’s modern standards. One can pretend that they are not there, but that’s just pretence. NB: My response would not be to “cancel” Aristotle, but to try and recognise how far I am myself part of the tradition that I condemn. The point is not cutting off oneself from the past (as if one were morally superior), but to trace involvement in traditions.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s