Tag Archives: Philosophy of language

The Harm of Racial Slurs

by Tim Harding, B.Sc. B.A. (philosophy)

In this paper I would like to analyse the derogatory nature of racial slurs.  In particular, I shall try to answer the following questions: ‘What harm do racial slurs do?’ and ‘Do racial slurs deserve protection under the principle of free speech?’.  In answering these questions, I enlist the methodology of speech act theory.  I argue that racial slurs are perlocutionary speech acts that result in harmful consequences to target groups and individuals. They are not just offensive insults or taboo words.  There are widely-accepted exceptions to the right of free speech for words that cause harm, and I argue that racial slurs should be included amongst these exceptions.

By the term ‘racial slur’, I mean derogatory words targeting the race of a person or group.  Examples include ‘n**g*r’, ‘k*k*’, ‘ch*nk’ and ‘w*g’.  I do not include derogatory words targeting a person’s religious, political or other beliefs.  This distinction is important, because the ability to criticise beliefs and ideas is an essential component of free speech. Also, people can change their beliefs but not their race. So, on this basis, I would not count a derogatory word like ‘jihadist’ as a racial slur.

Another point of clarification is that in this paper, I intend to discuss the philosophical principles of racial slurs and free speech, rather than the policy or legal aspects of these concepts.  The question of whether racial slurs should be permitted, regulated or prohibited by law is one that lies more within the scope of political philosophy or jurisprudence than the philosophy of language.

I would like to illustrate the conflict between racial slurs and free speech by quoting some examples of actual slurs gathered by Stanford law professor Charles Lawrence III (1990: 431).  Each example is followed by a typical response, appearing in italics, that Lawrence has personally heard many times over against taking any action against such racial slurs.

  • Dartmouth College:

Black professor called ‘a cross between a welfare queen and a bathroom attendant’.

Yes, speech is sometimes painful Sometimes it is abusive. That is one of the prices of a free society.

  • Purdue University:

Counselor finds ‘Death N**g*r’ scratched on her door.

More speech, not less, is the proper cure for offensive speech.

  • Smith College:

African student finds message slipped under her door that reads, ‘African N**g*r do you want some bananas? Go back to the Jungle.’

Speech cannot be banned simply because it is offensive.

The common point being made in these responses is that such racial slurs are merely offensive insults, and that the principle of free speech outweighs taking any action against them.  That is a point that I would like to challenge in this paper.

To assist my analysis, let me contrast the impacts of these racial slurs with some non-racial slurs.  Whilst all slurs are by definition insulting to their targets, there are at least some non-racial slurs that probably cause little or no harm.  For instance, slurs like ‘ignorant’, ‘stupid’ and ‘lazy’ are often heard in heated debates; and whilst they may cause offence, the targets of such slurs rarely if ever claim to be harmed by them.  So, what are the harms caused by racial slurs?

I will commence my answer to this question by referring to speech act theory.  In linguistics and the philosophy of language, a ‘speech act’ is an utterance that has a performative function.  In colloquial terms, a speech act is ‘doing something’ as well as ‘saying something’.  Since the middle of the twentieth century, recognition of the significance of speech acts has demonstrated the ability of language to do things other than to describe reality or states of affairs (Green 2017:1).

In his seminal 1962 work ‘How To Do Things with Words’, John Austin developed his theory of performance utterances.  In a distinct break from the logical positivist view of statements, Austin (1962: 4-5) argues that there are meaningful sentences without truth values.  He claims that utterances can be found such that:

  1. they do not ‘describe’ or ‘report’ or constate anything at all, are not ‘true or false’; and

  2. the uttering of the sentence is, or is a part of, the doing of an action, which again would not normally be described as saying something.

Examples that Austin gives of such utterances include saying ‘I do’ or ‘I will’ during a marriage ceremony, the naming of a ship at its launch and the offering of a wager.  He calls such statements ‘performance sentences’ or ‘performance utterances’ (Austin 1962: 6).  The action that is performed when a ‘performative utterance’ is said belongs to what Austin calls a ‘speech act’ (Austin 1962: 40).  Austin distinguished between three types of speech acts as follows:

  • a locutionary act is the uttering of a sentence with a certain sense and reference, that is, meaning;

  • an illocutionary act has a certain performative force, such as informing, ordering, warning or promising;

  • a perlocutionary act results in consequences, such as persuading, deterring, threatening, and as I discuss below, subordinating (Austin 1962: 109).

Verrochi (2015:1) enlists Austin’s theory to provide us with a methodology for actively addressing the harm that is done by racial slurs.  She argues that attempts to describe racial slurs as merely insulting or offensive, or to locate the harm in the intention of the speaker (that is, racist attitudes) are inherently problematic (Verrochi 2015:20).  In terms of speech act theory, these could be categorised as locutionary or illocutionary speech acts.  Instead, Verrochi advocates a perlocutionary speech act categorization of racial slurs:

What is needed and warranted is a system that locates the harm of [racial] hate speech not in the feelings, emotions, or thoughts of the audience, nor in the heart, intention, or thoughts of the speaker, but in the force of certain speech acts — when uttered in the ‘right’ context by people with the ‘right’ authority — to do as much as they say (Verrochi 2015:20).

Verrochi (2015: 15-20) argues that racial slurs that target historically subordinated groups (such as African-American slaves and their descendants) have the effect of intimidating these groups and re-establishing a harmful social hierarchy based on race.  In her opinion, these perlocutionary speech acts result in racism (by which I assume she means racial stigmatisation and discrimination) and promote racial subordination.

Similarly, Tirrell (1999: 43) states that the perlocutionary effect of racial slurs is clear: ‘they are angry put-downs that attempt to reduce the person to one real or imaginary feature of who they are’.  She argues that it is not simply because such a particular speaker has a particular (racist) attitude that a racial slur is harmful (Tirrell 1999: 44).  In other words, racist attitudes are not in themselves harmful – it is the perlocutionary speech acts that do the harm.  She says that:

…the heart of the expression is its designating of the person as subordinate…To call someone a ‘n**g*r’ today is at minimum to attribute a second-class status to him or her, usually on the basis of race and, arguably, to take that lower status to be deserved (Tirrell 1999: 45).

So rather than racism being an attitude or mental state of a speaker, Tirrell (1999: 45-46) regards racism to be ‘a structure of social practices that supports and enforces the subordination of the well-being of some races to the well-being of members of other races’.  (For instance, slavery in the United States had not only a negative effect on the well-being of the slaves, but it also had a positive effect on the well-being of the slave-owners and their descendants for all the free labour from which they profited).  In terms of speech act theory, Tirrell argues that ‘the social, psychological and economic practices of treating dark-skinned African-Americans as less valuable that light-skinned European-Americans give content and force to the term n**g*r’ (Tirrell 1999:49).

Tirrell (1999: 58) examines a theory that the derogation of the term n**g*r is a pragmatic effect, not a semantic aspect of this term.  Because if the derogation were a semantic aspect of the term, there could be no non-derogatory use of it, and yet there is.  Some African-Americans use this term amongst themselves as a strategy for the reclamation and blunting of the racial slur, or even as a term of endearment.  According to this theory, such pragmatic factors are the means by which the derogatory force is detached (Tirrell 1999:58).

Tirrell adds a corollary that when people who are not African-Americans use the term, it is impossible for the term not to carry derogatory force (Tirrell 1999:58), as argued by Whiting and further discussed below.  In speech act terms, Tirrell says that ‘individual speakers cannot escape the socially established meaning of their utterances, except occasionally by the grace of the communities in which they live and speak’ (Tirrell 1999: 61).  In this way, Tirrell is implying a perlocutionary force of racial slurs.

While Verrochi and Tirrell have focused on the harm caused by racial slurs to target groups, Delgado (1982: 135-149) has identified various sources of harm to individuals.  He argues that ‘such language injures the dignity and self-regard of the person to whom it is addressed, communicating the message that distinctions of race are distinctions of merit, dignity, status, and personhood’ (Delgado 1982: 135-136).

The psychological harms causes by racial stigmatisation are often much more severe than other insults, because membership of a racial minority is neither self-induced nor alterable (Delgado 1982: 135-136).  They not only impair the victim’s capacity to form close interracial relationships, but affect even their relationships with their own group.  Such psychological harms can result in mental and psychosomatic illnesses (Delgado 1982: 137).

Racial stigmatisation can also damage a victim’s pecuniary interests, by limiting his or career prospects and social mobility.  In this way, it can be seen as a force used by the majority to preserve an economically advantageous position for themselves (Delgado 1982: 139-142).

Finally, Delgado (1982: 142-143) argues that racial slurs have an even greater impact on children than adults.  Empirical studies have shown that the effects of racial slurs are discernable early in life, with minority children exhibiting distress or even self-hatred because of their colour; and majority children associating dark skin with undesirability and ugliness (Delgado 1982: 142).

In contrast to Verrochi and Tirrell, Hom (2008: 416-440) argues that the semantic strategy fares better than the pragmatic strategy for explaining how racial slurs or epithets work.  The difference is that according to the semantic strategy, their derogatory content is fundamentally part of their literal meaning; whereas according the pragmatic strategy, their derogatory content is derived from how they are used (Hom 2008: 416).

A problem for the semantic strategy is that it fails to explain the non-derogatory uses of racial slurs referred to above.  On Hom’s view the derogatory content of a racial slur (he calls them epithets) are causally determined in part by factors external to, and sometimes unknown by, the speaker (Hom 2008: 430).  Hom calls this view combinatorial externalism (CE), where the meanings of racial slurs are supported and semantically determined by their corresponding racist institutions (Hom 2008: 431).  In this way, racial slurs ‘express derogatory semantic content in every context, but they do not actually derogate their targets in every context’.  In other words, racial slurs are words with derogatory content; speakers derogate by using words with such contents (Hom 2008: 432).

To provide an example of what Hom probably means here, let is consider two sentences provided by Elizabeth Camp (2013: 330):

(1) Isaiah is a k*k*.

(2) Isaiah is not a k*k*.

On Hom’s view, sentence (1) is derogatory towards Isaiah, but sentence (2) is not, because Isaiah is not being described by a racial slur.  On the other hand, if these sentences are view pragmatically rather than semantically, they could be interpreted as meaning the following:

(3) Isaiah is Jewish.  And by the way: boo to Jews!

(4) Isaiah is not Jewish.  And by the way: boo to Jews!

So, by this interpretation, the use of the word ‘k*k*’ in sentence (2) is still derogatory towards Jews in general, even though it is not derogatory towards Isaiah in particular.  Camp calls this view subjectivist expressionism (Camp 2013: 331-332), although it is not a view that she necessarily subscribes to herself.

Whiting (2013: 366–368) supports the second interpretation above, and argues that it is typically no less derogatory to make negative claims using racial slurs than it is to make positive claims using them, whereas Hom’s combinatorial externalism suggest otherwise.  Whiting presses his point by providing a further example:

  • A. The US President is a n**g*r.

  • B. A said that the US President is a n**g*r.

Whiting (2013: 368) argues that intuitively, B’s utterance is racially derogatory, but combinatorial externalism does not seem able to explain why this so.  He says that on Hom’s account, B’s report in factually true and non-derogatory.  Whiting (2013: 368) reports Hom as later acknowledging this problem, and trying to explain that a sentence like (B) as not derogating but causing offence.  Whiting (2013: 368) responds by arguing that if a sentence like (B) was uttered to an audience of white racists, none of them might take offence but the utterance is still a derogatory racial slur.

Anderson and Lepore (2013: 350–363) reject the assumption that we need to understand the use of a racial slur as expressive of derogatory content.  Instead, they propose what they call ‘prohibitionism’ which is the view that racial slurs are prohibited or taboo words, and so a violation of this taboo might provoke offence.  They argue that this taboo is ubiquitous, that is, embedding a racial slur inside a sentence does not immunise its users from transgression (Anderson and Lepore 2013: 353).  Presumably, this taboo includes the reclamatory or affectionate use of the word ‘n**g*r among African-Americans.

In contrast, Whiting (2013: 368-369) argues that prohibitionism seems false, because it is possible for there to be racial slurs in the absence of a taboo or prohibition.  He provides the following example:

Imagine a deeply racist society in which the use of ‘n**g*r’ is not prohibited but nonetheless is expressive of racist thoughts or attitudes concerning those to whom its neutral counterpart applies. Members of the society derogate in using ‘n**g*r’, even though they do not violate any prohibition. The word, as used in that society, is surely a slur.

In my view, the above scenarios provided by Camp and Whiting neatly separate the derogatory force of racial slurs from merely causing offence or uttering taboo words.  The significance of this distinction is that, as we saw in the examples provided by Charles Lawrence earlier in this paper, the argument that racial slurs are merely offensive insults is used as a defence against taking any action regarding the use of such slurs.  Put simply, this argument is that insult and offence are common components of public debate, and to restrict their use would constitute an unjustified restriction of free speech.

On the other hand, it is commonly accepted (including in law) that free speech is not absolute.  There are well-known exceptions to the right of free speech, for instance, in cases of public safety (shouting ‘Fire!’ in a crowded theatre), causing riots, incitement to crime and defamation.  The rationale behind these exceptions is that such statements cause harm.  But as I have argued in this paper, racial slurs also cause harm, possibly to a lesser extent than threatening public safety, causing riots or inciting crime, but harm nonetheless.  The harm caused by racial slurs undermines the argument that racial slurs should be protected as free speech, on the grounds that they merely cause insult and offence in a similar manner to non-racial slurs.  (To be clear, I am not generalising that all harmful speech is impermissible. I have given reasons above why I think that racial slurs cause sufficient harm to be impermissible).

In conclusion, speech act theory can provide us with a methodology for showing how racial slurs cause harm to target groups and individuals, rather than just insult or offence.  I have argued that racial slurs are perlocutionary speech acts, meaning that they result in consequences (in this case harmful consequences) rather than just express racist attitudes or emotions.  The harm lies in these adverse consequences to their intended targets, rather than in the racist attitudes or emotions of the speaker.  Because there are widely-accepted exceptions to the right of free speech where sufficient harm is caused, I have argued that racial slurs should be included in such exceptions.


Anderson, L. and Lepore, E. 2013 ‘What Did You Call Me? Slurs as Prohibited Words’. Analytic Philosophy 54: 350–363.

Austin, John L. How To Do Things with Words. Oxford: Oxford University Press, 1962.

Camp, Elisabeth. 2013 ‘Slurring Perspectives’. Analytic Philosophy Vol. 54 No. 3 September 2013 pp. 330–349

Delgado, Richard. 1982 ‘Words That Wound: A Tort Action For Racial Insults, Epithets, and Name-Calling’. Harvard Civil Rights-Civil Liberties Law Review, Vol. 17 (1982).

Green, Mitchell, ‘Speech Acts’. The Stanford Encyclopedia of Philosophy (Winter 2017 Edition), Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/win2017/entries/speech-acts/&gt;.

Hom, Christopher. 2008. ‘The Semantics of Racial Epithets. The Journal of Philosophy, Vol. 105, No. 8 (Aug., 2008), pp. 416-440

Lawrence, Charles R. 1990 ‘If He Hollers Let Him Go: Regulating Racist Speech on Campus’. Duke Law Journal, June 1990 pp. 431-483.

Tirrell, Lynne 1999. ‘Derogatory Terms: Racism, Sexism and the Inferential Role Theory of Meaning’ in Oliver, Kelly and Hendricks, Christina (eds.), Language and Liberation: Feminism, Philosophy and Language, SUNY Press, 1999. pp 41-79.

Verrochi, Meredith. 2015.  ‘Uncooperative Engagement: An Active Response To Hate Speech’. Ph.D Dissertation, submitted to Michigan State University. <https://d.lib.msu.edu/etd/3808>  Viewed 25 May 2018.

Whiting, Daniel 2013. ‘It’s Not What You Said, It’s The Way You Said It: Slurs And Conventional Implicatures’. University of Southampton Analytic Philosophy Vol. 54 No. 3 September 2013 pp. 364–377.

Copyright notice: © All rights reserved. Except for personal use or as permitted under the Australian Copyright Act, no part of this website may be reproduced, stored in a retrieval system, communicated or transmitted in any form or by any means without prior written permission. All inquiries should be made to the copyright owner, Tim Harding at tim.harding@yandoo.com, or as attributed on individual blog posts.

If you find the information on this blog useful, you might like to consider making a donation.

Make a Donation Button


Filed under Essays and talks

The Gricean Maxims

by Tim Harding

There are certain social conventions and assumptions that are normally made by people engaged in meaningful conversations.   Vocabulary and the rules of grammar combine with knowledge of the situational context to fill in what’s missing and resolve ambiguities.  For example, when we ask at the dinner table whether somebody can pass the salt, we are not literally enquiring as to their physical ability to lift and move the salt container.[1][2]

Listeners and speakers need to cooperate with each other and to mutually accept one another to be understood in a particular way.   In sociolinguistics, this is known as the  Cooperative Principle.  As phrased by British philosopher Paul Grice, who introduced it, the principle states,

‘Make your contribution such as it is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.[3]

Though phrased as a prescriptive command, the principle is intended as a description of how people normally behave in conversation, to ensure that what they say in a conversation furthers the purpose of that conversation.  The principle describes the assumptions listeners normally make about the way cooperative speakers will talk.

Thus the cooperative principle works both ways: speakers (generally) observe the cooperative principle, and listeners (generally) assume that speakers are observing it.  This allows for the possibility of implicatures, which are meanings that are not explicitly conveyed in what is said, but that can nonetheless be inferred.  For example, if Alice points out that Bill is not present, and Carol replies that Bill has a cold, then there is an implicature that the cold is the reason, or at least a possible reason, for Bill’s absence.  This is because Carol’s comment is not cooperative — does not contribute to the conversation — unless her point is that Bill’s cold is or might be the reason for his absence (see the Maxim of Relevance below).  If Bill’s cold had nothing to do with his absence, then Carol’s comment would be irrelevant, misleading and thus uncooperative to the conversation.


The cooperative principle can be divided into four maxims, called the Gricean Maxims, describing specific rational subordinate principles observed by people who adhere to the overarching cooperative principle.  Grice proposed four conversational maxims that arise from the pragmatics of natural language.  The Gricean Maxims are a way to explain the link between utterances and what is understood from them.

Maxim of Quality

  1. Do not say what you believe to be false.
  2. Do not say that for which you lack adequate evidence.

Maxim of Quantity

  1. Make your contribution as informative as is required (for the current purposes of the exchange).
  2. Do not make your contribution more informative than is required.

Maxim of Relation

  1. Be relevant.

Maxim of Manner

  1. Avoid obscurity of expression.
  2. Avoid ambiguity.
  3. Be brief (avoid unnecessary prolixity).
  4. Be orderly.

Without cooperation, human interaction would be far more difficult and counterproductive. Therefore, the Cooperative Principle and the Gricean Maxims are not specific to conversation but to verbal interactions in general. For example, it would not make sense to reply to a question about the weather with an answer about groceries because it would violate the Maxim of Relevance. Likewise, responding to a simple yes/no question with a long monologue would violate the Maxim of Quantity.

However, it is possible to flout a maxim intentionally or unconsciously and thereby convey a different meaning than what is literally spoken. Many times in conversation, this flouting is manipulated by a speaker to produce a negative pragmatic effect, as with sarcasm or irony, or to convey a meaning by what is not said in the situational context. For example, a student named Luisa Casati has asked her tutor Jeremy Hirst to write a letter of recommendation.  The letter reads as follows;

‘Dear Colleague,

Ms. Luisa Casati has asked me to write a letter on her behalf. Let me say that Ms. Casati is unfailingly polite, is neatly dressed at all times, and is always on time for her classes.

Sincerely yours,

Jeremy Hirst’

Jeremy has violated the Maxim of Quantity by providing insufficient information as to Luisa’s suitability for further study or employment.  He has also violated the Maximum of Relevance by discussing some of her positive personal qualities, but which are not centrally relevant to her abilities as a student or employee.

Jeremy might have deliberately violated these two maxims in an attempt to be truthful whilst not hurting Luisa’s feelings (so as not to violate the maxims of quality or manner).  He may thus be conveying a subtle negative message to the reader by the nature of what he has left out of the text rather than what he has included.

Assuming that Jeremy’s letter is rational and purposeful, then in my view it does not disprove the maxims in question.  Indeed, by deliberately violating these maxims, the letter may well be conveying a subtle negative meaning that might not be conveyed if these maxims did not exist.  That is, if the Gricean maxims did not exist, and the letter was read literally and simply, then the letter may convey only a positive message that Jeremy might not really intend.

Speakers who deliberately flout the maxims usually intend for their listener to understand their underlying implicature. Conversationalists can assume that when speakers intentionally flout a maxim, they still do so with the aim of expressing some thought.  Thus, the Gricean Maxims serve a purpose both when they are followed and when they are flouted.

References and notes

[1] Fromkin, V., Rodman, R., Hyams, N., Collins, P., and Amberber, M., (2009) An  Introduction to Language (6th edition) South Melbourne: Cengage Learning. pp. 196-7.

[2] Indeed, it is thought that one of the main limitations to artificial intelligence is that machines are likely to interpret language too literally, unless they have been programmed with all the knowledge of situational context that humans accumulate over a lifetime.

[3] Grice, Paul (1975). ‘Logic and conversation’. In Cole, P.; Morgan, J. Syntax and semantics. 3: Speech acts. New York: Academic Press. pp. 41–58.

If you find the information on this blog useful, you might like to consider supporting us.

Make a Donation Button

Leave a comment

Filed under Essays and talks