<

Should Harmful Speech Be Banned?

>

In this essay I explore whether causing harm is a sufficient condition to justify banning certain acts or types of speech. I argue that ‘harmful speech’ should be understood as ‘communicative actions causing undesirable consequences’, ultimately concluding that we should ban harmful speech, but that many information or argument sharing speech acts forward the interests of most people, and are not harmful. The question, “Should harmful speech be banned?” derives its philosophically quandarious nature from two presuppositions: Firstly, that the state is justified in banning harmful actions, and secondly that in a liberal society speech should be unrestricted. In analysing this question I take a Wittgensteinian approach, analysing the force and manner in which each of the concepts in the question are employed before offering a substantive answer to the question. As such, this essay is be separated into four sections.

1. Should

2. Harmful

3. Speech

4. Be Banned?

In the first section, I will consider whether we should be normatively sensitive to harms. I will argue that we have a axiomatic obligation to minimise harms, and furthermore, most often it is in our interests to. Secondly, I will consider what we mean by ‘harmful’ and a ‘harm’. I will consider physical harms, psychological harms, and objective, and subjectively recognised harms, ultimate arguing for a modified version of Joel Feinberg’s (1987, 34) conception of harms as a setbacks to interests, arguing that harms are subjectively recognised ‘undesirables’. In the third section, I consider what we mean by ‘speech’ as separate from action, as well as how this interacts with the previously put forward view of harm in the term ‘harmful speech’. I consider Stanley Fish’s (1994) argument for speech being considered a sub-category of action, given it is equally consequential and goal-directed, and therefore propose that ‘harmful speech’ should be considered as ‘communicative action that causes undesirable consequences’. Regarding the notion of ‘cause’ in that conception, I argue that the more easily one is able to demonstrate that a particular communicative action necessarily entails the harmful, undesirable consequences, (e.g. Yelling ‘fire’ in a burning building versus Putting forward an argument that convinces someone to engage in something harmful), the more weight the ascription of harm to the speech should be given. Lastly, having defined the constitutive elements in the question, I address whether or not banning is an appropriate response to speech with undesirable consequences. I argue that if a communicative action was entirely undesirable, and there were nobody whom benefited from it, we should ban it to maximise utility, understood as preference satisfaction. Most communicative action, however, I post, have both positive (desirable) aspects, and negative (undesirable) aspects, and in these situations, I argue that the state should engage in a careful balancing of these aspects in order to determine if the communicative action, on net, causes desirable or undesirable outcomes. Finally, I consider a Millian argument, using the analogy of the global brain, for why it is desirable in regards to most peoples interests to allow information and theories to be shared, ultimately concluding that we should ban undesirable communicative acts, but these might turn out to be fewer than we expect.

This concept-focussed structure is perhaps non-standard. What motivates this approach is that many of the concepts, (’harm’, ‘speech’) are host to a variety of different interpretations. By discussing and specifying which interpretation of the terms should be operative, two aims are hoped to be accomplished. Firstly, the mechanisms and theories behind the eventual verdict are laid clean on the table, like a watch’s parts on a watch-maker’s table, enabling one to more easily specify the defective part should they argument not working. Secondly, it hopefully avoids objections or dialectical ‘moves’ which seek to offer rebut the argument by relying on a suddenly different and ‘better’ interpretation of one of the key terms than the initial positive case was proposed in response to. By making the majority of dialectical moves prior to offering a substantive case, the conclusion should arrive naturally, with the reader already knowing where, if at any point, they disagree. Let’s see if it works.

1 . Should

Should our normative claims be responsive to harms? By this I mean, does the fact of one option being more harmful than another weigh on what decision we should make? To one who denies any normative sensitivity to harm, the question under examination would simply be translated as “Should we ban speech?”, the harmfulness would be entire irrelevant. This would then be evaluated according to whatever they do consider as having normative sensitivity. Potential alternative normatively-sensitive values include equality, autonomy, or liberty. Evaluating the question in regards to these values would consider whether banning certain kinds of speech maximises the preferred value. For the remainder of this essay, however, I will be continuing as though harms are normatively relevant. I offer two arguments for this, a more satisfying one and a less satisfying one, and allow the reader to decide which is which. 

The first is simply to double down that we have an axiomatic obligation to choose the least harmful option of two identical options, and the broader principle that we should minimise harm. Naturally, to one that is unconvinced as to the normative relevancy of harm, this assertion will simply be denied. However, my intuitive sense is that this is held almost ubiquitously as an axiomatic truth. Indeed, even asking “Why should we care about which option is the least harmful?”, is likely to garner you some odd looks in many circles. 

The second argument, which may be more likely to appeal to one unconvinced by axiomatic obligations, is that it is one’s rational self-interests to choose the option which is the least harmful, namely because it it likely to be in their best interests, both promoting a positive atmosphere for others to do the same should your positions be switched in the future, in Rawlsian ‘veil of ignorance’ fashion, and gets rid of the need for parties to attempt to sway your decision making through altering the payoffs of the decisions through things like threats, or revenge. One might rightly point out, however, that in this is not really an argument that one should be normatively sensitive to harm, but that one is normatively sensitive to their own interests, and that it is likely that choosing the least harmful option will have benefits that will serve these interests. In a situation in which choosing the least harmful option does not serve one’s interest, perhaps in circumstances in which nobody will ever find out what choice one made, and as such there will be no positive reinforcing atmospheric effects, nor any chance of retaliation, it would seem that harm would not be tracked as a normatively sensitive topic. These situations however, I put forth, are rare; most decisions are atmosphere-promoting or trackable, to some degree. For this reason, we should consider the second argument weakened slightly, but not without force, to the form that harms are likely but not guaranteed to be normatively relevant for decision making. This essay will therefore proceed as though harms are relevant for normative claims, endorsing the stronger position that we have an axiomatic obligation to minimise harm, whilst offering the argument that even to one unconvinced of the normative relevancy of harm, it is at least highly likely that minimising harm will serve one’s interests.

To whom are we levelling these claims? Who is the ‘we’ that ‘should ban harmful speech’? It is important to specify the agent whose behaviour we are proposing should change. One might hold different entities, federal states, local governments, university clubs, individual agents to different normative standards. The use of ‘banned’ in this question implies a legalistic or rule-making sense, and so in this essay I address the arguments to individuals of self-governing bodies, essentially arguing ‘We should use the law to ban harmful speech’, where ‘we’ are the self-governing populace, and not ‘We the government should ban harmful speech, regardless of what the citizenry think’, the position I am arguing for is an attitudinal one, and not an exact policy recommendation.

2 . Harmful

What do we mean by ‘harm’? Given all this discussion of obligations to minimise harm, how should we understand ‘harm’ such that banning some forms of speech might reduce it? One intuitive way of defining harm is in regards to physical harm. On this conception, X is harmful if and only if it results in physical damage to a person, where damage can be defined as diminished function. It is this conception that informs the famous adage “sticks and stone’s may break my bone’s, but words will never hurt me”. “Sticks and stone’s” are capable of physically damaging someone’s body, whereas mere “words” are unable to. One endorsing this view would require a story between any speech act to be banned and physical harm eventuating. Subsequently, they may agree to the banning of yelling “Fire!”, in a theatre and causing a stampede, or forms of provocation resulting in physical damage, but they are unlikely concede that slurs, insults, or lies are harmful unless some story can be told about how they cause physical harms.

But are physical harms the only kind of harm we recognise? Do we consider the statement “They were harmed, just not physically” to be nonsensical? If we do not, in what sense is the person the statement refers to harmed, if not physically? Nowadays, we commonly recognise the existence of psychological harms as well. Psychological damage, which equally may be defined as diminished function, involving pain, depression, or loss of focus or ability. One who accepts this broadened definitions of harms, such that X is harmful iff it causes physical or psychological damage, will require a causal story from speech to either physical or psychological harm to consider speech worth banning. One who accepts the broadened definition of harms will more amenable to including speech acts such as insults, slurs, bullying as appropriate for being banned. 

Is it the case that we have exhausted the appropriate meanings of harm, with physical and psychological damage? Could we ever sensibly say, “They were harmed, but not physically or psychologically damaged”? I think we would. Consider someone whose resume was thrown out on the sole basis that the reader did not like the person’s name. In what sense were they harmed? Joel Feinberg proposes a conception of harm as a “thwarting, setting back, or defeating of an interest”. An interest, he defines, is something which someone stands to “gain or lose depending on the nature or condition” (1987, 34) of it. Subsequently, X is a harm iff it thwarts, sets back or defeats an interest of someone’s, something they have a stake in (though Feinberg will go on to further refine this definition). In this sense, the person whose resume was thrown out was harmed, because their interests in acquiring a job were set back. Furthermore, it appears that this conception of harms being a setback to some goals, interests, or desires was already present in our conception of harm as physical or psychological damage through the definition of damage as ‘diminishment of function’. Any notion of diminishment of function of something must also have some idealised concept of the proper function, e.g. of a perfectly physically functioning leg, or psychologically healthy brain. 

This, however, raises a question. Who gets to decide what counts as interest? Are we endorsing objective interests, in that there are some definitional interests which all humans qua their humanness, necessarily have, or are they subjectively defined? There seems to be problems in both directions. The problem with asserting objective interests is that for any interest one chooses, there will always exist a counterexample of someone who denies that they have it. And it is odd to insist that someone is harmed by something, whilst they insist that they are not. For example, I find in myself the intuition to want to declare killing someone an objective harm, because it harms the interest that is universally held qua humanness; wanting to live. The issue being, that not everyone wants to live e.g. individuals desiring euthanasia. This is case I can see of the clash between the notion of objective interests, and subjective interests. For surely if we can find an example in which being kept alive is against one’s interests, then for whatever universal interest we choose, food, water, security, friends we will always be able to find a counter-example. Likewise, one intuitively runs into trouble defining interests, and consequently harms, as entirely subjectively defined. As it seemed strange to insist that someone is harmed, when they entirely insist that they are not, based on your perception of their objective interests being thwarted, so to does it seem strange for someone to insist that you are harming them, because they have chosen such minor and arbitrary interests, such as using the proper King’s English, or the example Feinberg uses, beating their team at baseball (1978, 42). These is especially relevant to the question of banning speech, due to the underlying presupposition that if any speech is to be banned it is certainly the harmful speech that should go first. Allowing people to subjectively define their interests, and therefore harms, seems to allow people to have an arbitrary basis for banning others speech based on their own choices. As such, it seems like defining harms as setbacks to interests captures many of our intuitions regarding harms, but we run into problems both when considering whether interests are objectively or subjectively defined.

Feinberg has his own solution to this problem. His solution, is to maintain that people subjectively choose their interests, but then to provide an objective translation of these into four categories, three of which we can safely ignore as ‘non-harmful’ to thwart, and incapable of motivating prohibitory laws under the harm principle. Under Feinberg’s picture ‘wants’ (interests) are sorted into four primary categories: 

  1. Passing wants which are flickering desires such as “the desire to eat an ice-cream cone” (1978, 60).  
  2. Instrumental wants which are wants in service of other interests, such as desiring to work late to fund other interests. Whilst instrumental wants might have harmful aspects such as the pain of having to work late, they equally have positive aspects, such as extra pay, that usually result in the event of them being pursued as “on balance” (1978, 40) not harmful. It seems then, that Feinberg separates harmful and non-harmful elements (e.g having to commute, missing time with one’s family, getting paid extra), from harmful and non-harmful events (e.g. working overtime’, ‘going on holiday’, ‘going to marriage counselling’) which are understood as the summation of their constitutive elements.
  3. Welfare interests which are requisite baseline interests that allow us to pursue other interests, such as physical and mental health, sufficient food, water, and security. 
  4. Finally ulterior interests are described by Feinberg as “ultimate goals and aspirations” (1978, 37) such as producing a great work of art, solving a scientific, achieving political office or building one’s dream house. Ulterior interests, Feinberg argues, fulfil the role of C. L. Stevenson’s ‘focal aims’, goals by which all other are in ultimate service to.

With this set-up, Feinberg solves the problem of people self-selecting arbitrarily minor interests, and therefore seemingly harmless things counting as harmful, by arguing that only setbacks to ulterior interests, or extensive damaging of one’s welfare interests such that one’s capacity to pursue their ulterior interests, count as harms. Therefore, one’s baseball team failing to win the pennant, though it might be annoying, does not count as a harm (Feinberg, 1987, 42). Equally, one may setback someone’s welfare interests momentarily, such as requiring an employee stay late one night, without harming them, as long as their ulterior interests are not damaged, which could be the case for example,  if the employee were required to stay back all night every night for a week, thereby forming an endemic setback to their ability to pursue their ulterior interest. Feinberg therefore solves the problem of subjectively defined harms by arguing that only setbacks to subjectively defined interests within an objectively defined category of interests, ulterior interests, count as harms.

I see two problems with Feinberg’s solution. The first problem is that ‘ulteriority’ or ‘importance’ of interests is still a subjective notion, and so it is still the case the people may ascribe high personal importance to seemingly arbitrary projects, and therefore still define their own version of harm. As such, it seems to merely to kick the question down the road, from ‘Who gets to define interests, and therefore what counts as harms?’, to ‘Who gets to define ulterior interests, and therefore what counts as harms?’. Any answer to what an ulterior interests is, will equally run into problems of universality; it seems to endorse one particular vision of the good life, which not everyone shares. It is equally possible that someone could insist, no matter what the interest is, that this is the thing that is their ultimate interest. Feinberg himself acknowledges this, writing that although it is unlikely, it may be possible that their might be “some Dodgers fans” (1978, 44) with the ability to create strong enough ultimate interests in outcomes such as their baseball teams winning or losing. But if this is the case, then it seems we are back with the problem of subjectively defined interests.

The second problem is that even maintaining that only setbacks to ulterior interests count as harms, Feinberg’s theory lacks a lower limit on how large that setback must be. It seems that all that is required to make something a harm is to come up with some causal story relating the event as setback the ulterior interest in some way. To this extent, the employee that is forced to stay late for even one night could call it harm, by saying that them staying late one night will throw off their circadian rhythm, which will disrupt their family life, which will make them more likely to be stressed etc. which will all setback their ulterior interest of retiring with their family to Switzerland. Essentially, to insist upon a ulterior interest with a large enough ‘world-description’ that it excludes the undesirable ‘harm’. It seems then that we may still have the problem of arbitrary things being designated as harms under Feinberg’s theory, because as long as someone can come up with a causal story of how something impacts their subjectively defined ulterior interest, they will be able to call it a harm.

The best solution, I argue, is to endorse the harms as setbacks to interests, bite the bullet on people subjectively defining interests and consequently harms, and to mitigate this, somewhat revise our understanding of the harm principle. There are two intuitive cases I feel that initially push back against this conception.

The first is that of the overly sensitive person, claiming that seemingly arbitrary, or obviously non-harmful things are harmful to them. The second is of someone we care about doing something harmful, but them insisting that it is not harmful to them. It seems like subjectively identified interests robs the ability to disagree about harms, if someone says it is then it simply is. However, I argue that there still remains the capacity to disagree. Firstly, one may argue that the person has incorrectly identified their own interests. This applies to the euthanasia case; we disagree the the person has correctly identified their own interests. Secondly, we may agree with their appraisal of their interests, but disagree about the setting-back or forwarding nature of their methods. If they still maintain the action is undesirable, after hearing out our arguments? I argue that we should accept that it harms them, that their preference provides a reason to avoid that action, but a defeasible one, that we might still perform it if we are aiming at some more desirable aspects. On the basis of these arguments, then, we should consider harms as undesirables to subjectively defined interests.

3 . Speech

What do we mean by speech? Intuitively we think of speaking. However, in regards to the subject manner of banning speech, we would hardly be happy in a world in which vocal utterances were unrestricted, yet published writing, texts, films, art or dances were restricted. In the beginning of the essay, I argued that the question derived its quandarious nature from two presuppositions: That we should ban harmful actions, and that in a liberal society speech should not be banned. Most of the force however, from this presupposition seems not to be derived from vocal utterances but rather from communication. Therefore, regarding this question, I’ll consider ‘speech’ more broadly as referring to communication.

A common way of alleviating the tension between the two presuppositions, is simply to argue that speech is separate to action, and our obligation to ban it does not apply, even if it is harmful. It seems the intuition is that speech is a sort of inconsequential, non goal-directed exploratory stage, in which ideas are shared and contemplated prior to any actual action taking place.

Stanley Fish, however, puts pressure on the speech/action divide, arguing instead that speech is a subset of action. He argues this by refuting the presupposition of speech as non-goal-directed and inconsequential, instead arguing that speech is goal-directed and consequential, that speech “is produced with the goal of trying to move the world in one direction rather than another” (Fish, 1994, 107) even in the simple act of discussion or convincing someone. If we accept this conception of speech, Fish argues, that people are doing things with words, that speech, then speech appears simply as a subset of actions, and now justified to be prohibited if necessary. Integrating this conception then, gives us a definition of speech as ‘communicative action’.

How, then, should we understand ‘harmful speech’? Combining the the proposed conceptions of harm and speech results in ‘harmful speech’ as ‘ communicative action causing undesirable consequences’. The reader may notice that addition of ‘causal’ linking the two conceptions, and I argue this is a key attribute in the debate. Utilising David Lewis’ ‘difference-making idea’ that “the difference [a cause] makes must be a difference from what happens without it” (Lewis, 1973, 557) in this case, for the speech to have caused the harms, for it to be considered harmful speech, it must be considered the harm would not have occurred, but for the speech. Subsequently, in banning the harmful speech, we do not expect the harms to be soothed or counteracted, but rather we expect them to vanish, because they will be uncaused. I argue that the degree to which a speech act is considered harmful is directly proportional to the degree that the undesirable consequences are able to be demonstrated to be a necessary entailment of the speech act. In fact, we can construct a spectrum of cases demonstrating this, from very clear to more diffused causal relations, and subsequently clear and borderline examples of ‘harmful speech’:

a). Saying the code-word near a bomb with a voice-activated trigger, causing it to explode
b). Saying “Shoot!”, to a firing squad, causing them to shoot
c). Yelling “Fire!” in a crowded theatre, causing a stampede
d). Yelling “There they are! That’s them!”, to an angry crowd, causing them to attack the person
e). Using emotionally laden or false arguments as a politician, causing people to act in a biased manner
f). Writing an academic paper, considering arguments for and against a position, causing people to believe the thesis of the paper

Progressing down the list of each of these examples, they different in how much the result is seen as necessarily following from the speech, the speech is that which causes the result to shift from non-actual to actual. In a) it is as though the speech flips a mechanical lever. In b) and c) it is as though the speech flips a ‘social lever’, there is almost no time between the speech and reaction. The more we progress down the spectrum, from the person provoking the crowd, to the politicians emotionally laden arguments though, there is increasing possibility for the result to be hearer-generated, we feel that it could have gone either way, the speech did not necessarily entail the results. Insults and slurs would probably be placed around c) or d), as speech acts that immediately entail an undesirable emotional response. Finally, in f), the deliberate consideration of arguments both for and against implies that the result will be very hearer-generated, they will have to make up their own mind, and therefore the casual link will be diffused. Therefore, I argue that in defining ‘harmful speech’ as ‘communicative action causing undesirable consequences’ the easier a causal link is established establishes the undesirable consequences as a necessary result of the speech, the more watertight the description of the speech as harmful will be.

4 . Be Banned?

We have now arrived at my substantive case. Having defined harm and speech, and how they interact in harmful speech, as well as defining speech as a subset of action, and therefore not axiomatically unprohibitable, we are in a position to consider: Is banning ’communicative action that has undesirable consequences’ the optimal move? In democratic states, it seems axiomatic that governments should minimise undesirables for their constituents. It is important, however, not to focus with tunnel vision on merely resolving the undesirable effects of particular speech acts, without considering the negative and undesirable effects that our proposed solution, such as bans, may haave. For example, whilst most people might agree that mild school-ground-insults have undesirable consequences, they also might agree that the legal restriction of a ban on all such insults is ultimately more undesirable.

What should be done in seemingly clear-cut cases, in which it is universally agreed by a populace that a speech act is undesirable, and that its banning brings, on-net, more desirable aspects than it prohibits? In these situations, it seems clear-cut that the populace is justified in implementing a ban, as it straightforwardly improves their lot. To this, one might respond that even in such clear-cut cases a ban would never be necessary, because given the universal agreement of the undesirability of the action, nobody would ever perform it. However, in these situations we might consider the state as solving a cooperation problem by adjusting the payoff structure of a particular strategy. For example: forms of public lying, such as false advertising, or by politicians on campaign. Whilst everyone may agree that such lies are harmful in the aggregate, individuals are tempted to use them to get an edge. A populace is therefore justified in banning an act which it rationally pre-emptively agrees it is likely to result in negative utility for the group, thereby manipulating the desirability for individuals in situations in which such acts would otherwise forward their interests. Totally harmful speech, that has not a single redemptive quality, then, we are justified in banning.

Universal agreement on the undesirability of a speech act, however, is practically never found. In most circumstances, it is in the interests of some citizens for a speech act to be prohibited and the interests of some others that it be permitted. In this situation, it seems that the state must engage in some variation of utilitarian calculus to maximise the forwarding of as many of its interests of its constituents as possible.

Intuitively however, this conclusion provokes unease. It seems we can easily imagine a situation in which the majority of a populace simply finds particular speech undesirable, whilst the purveyors of such speech feel it is of the utmost importance that they share their information. Is the knower of undesirable yet important information doomed to be silenced?

I argue that these intuitions can be rescued within the proposed framework, by arguing for the desirability of information-sharing speech acts, especially those which have a diffused causal relationship between the act and their consequences, by appealing to the analogy of the global brain (cf. Russell, 1983) and the arguments of John Stuart Mill. The global brain analogy refers to a thought experiment in which a group of humans is pictured as one large single organism, in which each human functions as a singular neuron. Humans function as a value-detector, sharing things in the environment that it finds valuable, and thereby allowing the global brain to coordinate its energy, directing more or less people to particular locations. In this sense, we might consider banning speech, ceasing the informational flow from a neuron or group of neurons, as equivalent to taking painkillers. On the one hand, the undesirable information flow has stopped. On the other hand, we are now blind to the that particular area. To take it to an extreme, if one desired to cease all undesirable information, one certainly could take massive amounts of painkillers, such that they could not perceive the outside world. This, however, would be dangerous; without information, one is unable to respond respond to things that require a response. This strategy is not viable for forwarding one’s interests long term. These arguments have the same logical form that John Stuart Mill puts forward in On Liberty, namely that it does forward a society’s interests to have free-flowing information, because it allows false beliefs to be rectified, and true belief to be re-affirmed and updated (2011, 9), thereby allowing more informed actions to be taken. The optimal move, in this sense, is not to do as Stanley Fish advocates, and attempt to garner the political power to silence your opponents (1994, 114), but, provided they are arguing in a non social-lever pulling method, to welcome their contributions. Using these arguments, then we may argue that information sharing speech acts, in which arguments are shared and discussed, are therefore not harmful for individuals, as they allow them to make more informed decisions, and therefore forward their interests, and therefore should not be banned. These will be characterised by a causally diffuse relationship between speech act and consequence, leaning towards the side that the consequences be hearer-generated, the central concern being that they make up their own mind. This is contrasted with speech acts detailed in section 3, such as insults, slurs, and inappropriate social lever pulling, which immediately cause undesirable and harmful effects. As such, by arguing for the interest-forwarding, non-harmful nature of information-sharing speech acts, we may maintain that harmful speech should be banned.

This essay therefore concludes, having argued for a specific framework of understanding the question “Should Harmful Speech Be Banned” culminating in the affirmative, that harmful speech should be banned, but broadly, that information sharing is largely non-harmful. First, it was argued that we have intuitive axiomatic reasons for being normatively sensitive to harm, and furthermore that it is often in ours interests to minimise it. Secondly, it was argued that ‘harm’ should be interpreted as a setback to a subjectively defined interest, after problems were demonstrated with Feinberg’s conception of ulterior interests. Thirdly, it was argued that speech should be interpreted as ‘communicative action’ and that we should reject the speech/action distinction, thereby placing speech as liable to be prohibited if harmful. Furthermore, it was demonstrated that the more easily one is able to demonstrate a causal relationship between the speech act and the undesirable consequent, the more force the ascription of harm to it will have. Lastly, having defined the constitutive terms, it was argued that banning is an appropriate response for harmful and undesirable speech acts, but that information and argument sharing communicative acts, given their ability to forward peoples interests, should not be considered harmful, and should not be banned.

Works Cited:

Feinberg, Joel. 1987. “Harms as Setbacks to Interest.” In The Moral Limits of the Criminal Law Volume 1: Harm to Others_, edited by Joel Feinberg, 0. Oxford University Press. https://doi.org/10.1093/0195046641.003.0002.

Fish, Stanley Eugene. 1994. There’s No Such Thing as Free Speech: And It’s a Good Thing, Too. Oxford University Press USA.

Lewis, David. 1973. “Causation.” Journal of Philosophy 70 (17): 556–67. https://doi.org/10.2307/2025310.

Mill, John Stuart. 2011. On Liberty. Luton, Bedfordshire, UNITED KINGDOM: Andrews UK Ltd. http://ebookcentral.proquest.com/lib/unimelb/detail.action?docID=770561.

Parekh, Bhikhu. 2012. “Is There a Case for Banning Hate Speech?” In The Content and Context of Hate Speech: Rethinking Regulation and Responses, edited by Michael Herz and Peter 

Molnar, 37–56. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9781139042871.006.

Russell, Peter. 1983. The Global Brain: Speculations on the Evolutionary Leap to Planetary Consciousness. 1st ed. Los Angeles : Boston: J.P. Tarcher ; Distributed by Houghton Mifflin.