Mill’s Harm Principle

According to Mill’s harm principle, “the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number [through “compulsion and control”], is self-protection. That the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant” (Reference Mill and Robson1977b, 223, emphasis added). This implies that for x’s compulsion and control of y to be warranted, it is necessary but not sufficient that either (1) x’s purpose be self-defense against y (the defense clause); (2) x’s purpose be to protect some third party, z (the protection clause); or (3) x’s purpose be to promote the good of y, who is not a member of a civilized community (the civilization clause). In consequence, for Mill to endorse coercive assimilation without contradiction, he would have to do so via an appeal to at least one of these clauses.Footnote ¹

A key theme of On Liberty that would not seem to augur well for the view of Mill as a coercive assimilator is the danger that democratic reforms may pave the way for a “tyranny of the majority,” including coerced conformity in matters of religion and custom (Reference Mill and Robson1977b, 219, 272–273, 283–291). Nevertheless, in On Liberty Mill is not hostile to democracy as such, and in Considerations he argues that there is a moral imperative for civilized societies to create “[f]ree institutions,” which go beyond the mere rule of law to incorporate both the liberal rights enshrined by the harm principle, and the right to representation in a system of universal suffrage (Reference Mill and Robson1977a, 547; Reference Mill and Robson1977b, 217–218). Moreover, as explained in greater detail below, this defense is appended with two pages of controversial argument to the effect that such institutions “are next to impossible in a country made up of different nationalities” (Reference Mill and Robson1977a, 547). Supposing that “ought” implies “can,” this suggests that either (a) there may be civilized societies in which the moral imperative to create free institutions is inapplicable, namely, the multinational ones in which it cannot be applied; or (b) given that civilized societies ought to establish free institutions and multinational civilized societies can only do so by becoming mononational, they ought to make that transition.

One might think that Mill envisages scenario (a)—or what could be termed a civilized multinational despotism—in the case of Hungary. Of its population, “composed of Magyars, Slovacks, Croats, Serbs, Roumans [sic], and in some districts, Germans,” Mill observes that it is “so mixed up as to be incapable of local separation; and there is no course open to them but to make a virtue of necessity, and reconcile themselves to living together under equal rights and laws” (Reference Mill and Robson1977a, 549). In other words, given that Mill takes secession to be a nonviable option there, on the one hand, and speaks of “equal rights and laws” (emphasis added), on the other, one might think that he envisages a system that is benignly egalitarian insofar as it does not privilege one nationality over another but is nevertheless despotic in virtue of depriving the population of institutions that merit the appellation “free.” However, if this were his view, he would surely have said so, and thus it is more plausible to assume that when he speaks of “making a virtue of necessity” he means that they must somehow make free institutions viable. This leads to option (b), and the thought that, absent the option of secession, making free institutions viable in Hungary means making its multinational population mononational. But if this is Mill’s view, how might it be achieved?

Mill on Nationality and Free Institutions

To understand and evaluate Kymlicka’s answer to that interpretive question, it is necessary to examine in greater detail Mill’s claim that “Free institutions are next to impossible in a country made up of different nationalities” (Reference Mill and Robson1977a, 547–548). At first sight, Mill’s claim may seem surprising because, on the one hand, he takes Britain to have some of the freest institutions in the world (Reference Mill and Robson1977a, 551, 565; Reference Mill and Robson1984b, 121–123), and on the other, it is normally thought to be populated by at least three nationalities: the English, Scottish, and Welsh. However, once one attends to Mill’s definition of nationality, that puzzle disappears, for it becomes clearer why he does not consider the English, Scottish, and Welsh to be genuine nations.

Mill defines nationality thus: “A portion of mankind may be said to constitute a Nationality, if they are united among themselves by common sympathies, which do not exist between them and any others—which make them co-operate with each other more willingly than with other people, desire to be under the same government, and desire that it should be government by themselves or a portion of themselves, exclusively” (Reference Mill and Robson1977a, 546, emphasis added). In consequence, insofar as the English, Scottish, and Welsh are united by the salient common sympathies, and thus seek self-determination inclusively, which is to say, together in a single polity, they will not satisfy Mill’s conditions for distinct nationalities, and instead be subsumed within the broader “British nation” (Reference Mill and Robson1977a, 549, 551, 572). Nevertheless, each of the English, Scottish, and Welsh could be said to satisfy the rest of Mill’s definition, as each individual group could also be said to be tightly bound by its own common sympathies that induce desires for cooperation and communal life. In consequence, Mill’s definition still allows them to be said to constitute what will be referred to here as quasi-nations, which is to say, groups with a sufficient degree of the salient common sympathies to have the potential to seek self-determination, and thus (re)emerge as nations in and of themselves.

Of course, the preceding raises the question of what Mill takes to cause groups of people to be united by these nationality-constituting common sympathies, on the one hand, and why he takes the absence of such sympathies to be so problematic for free institutions, on the other. In the case of the first question, Mill’s answer is that the causes of nationality vary: “Sometimes it is the effect of identity of race and descent. Community of language, and community of religion, greatly contribute to it. Geographical limits are one of its causes. But the strongest of all is identity of political antecedents; the possession of a national history, and consequent community of recollections; collective pride and humiliation, pleasure and regret, connected with the same incidents in the past” (Reference Mill and Robson1977a, 546). However, he denies that any one of these most reliable of the causal factors is either necessary or sufficient. Instead, he insists only that, ceteris paribus, the more of these factors that are present, the more likely it is that nationality-constituting common sympathies will emerge, on the one hand, and the stronger these sympathies are likely to be, on the other.

In the case of the second question, Mill’s answer has two components: one conceptual and one empirical. The conceptual answer rests on the fact that Mill takes nations to seek exclusive self-determination as a matter of definition—recall that it is the willingness of the English, Scottish, and Welsh to seek self-determination inclusively, as a single communitarian body, that makes them one nation. This implies that in a multinational state national self-determination is a competitive good that one nation can only achieve at the expense of the others. Thus, the conceptual answer is that a multinational state is necessarily one in which at least some, if not all, of the component nations fail to achieve the form of communal freedom that is generally referred to as national self-determination.

Of course, this conceptual answer does not entail that a multinational state would necessarily lack free institutions in the sense of lacking a robust system of liberal-democratic rights. In consequence, this brings us to Mill’s second, empirical answer, according to which multinational populations tend to be causally inimical to the existence of such rights. Mill’s immediate concern here is that, where there is insufficient common sympathy between groups within a state for them to constitute one nationality, there is also likely to be insufficient common sympathy between them to cooperate to sustain these liberal-democratic rights against common threats, whether these threats stem from abroad or internally from a would-be native despotic government. Here one type of limiting case will be that in which the salient common sympathies are insufficient qua entirely absent because the respective populations are completely indifferent to each other. However, he also notes the tendency towards even worse scenarios in which the differences between the groups that preclude their uniting as a common nationality—whether these be of an ethnic, religious, linguistic, or historical nature—render them actively hostile, and thus desirous of depriving each other of liberal-democratic rights (Reference Mill and Robson1977a, 546–548; Reference Mill and Robson1985, 347–348).

One key point here is that, although Mill’s empirical analysis focuses on the difficulty of maintaining free institutions in a multinational state, the kinds of intrastate social fragmentation that concern him could suffice to preclude a common nationality without sufficing to produce multiple nations therein. After all, it is perfectly possible for a society to be divided thus without each of the mutually unsympathetic groups possessing a desire for exclusive self-determination, and thus satisfying his criteria for nationality. For example, in Considerations, Mill seems to interpret the Irish as unincorporated into the British nationality in virtue of the lack of the salient common sympathies,Footnote ² while treating them as a mere quasi-nation because most Irish people had not yet resolved to seek exclusive self-determination outside the United Kingdom (Reference Mill and Robson1977a, 550–551). In this respect, Mill’s conceptual analysis of the relation between nationality and free institutions implies that the British cannot be said to deprive the Irish of national self-determination if this entails frustrating an existent desire for that end.Footnote ³ Nevertheless, although Mill’s other writings show that he wavers on the question of how much liberal-democratic freedom the Irish can be said to enjoy within the United Kingdom’s institutions,Footnote ⁴ he makes clear that his empirical analysis still applies to the Irish as an unincorporated quasi-nation in a mononational (British) state. For example, in his commentary on Ireland for a radical political journal, the Monthly Repository, in the 1830s, he notes that historically the non-incorporation of the Irish into the British nationality facilitated the non-paternalistic despotism of the Irish “oligarchy” by allowing them to call upon the unsympathetic (“foreign”) British or English army whenever they needed to crush a revolt (Reference Mill and Robson1982c, 216).

There is some disagreement among scholars over whether Mill takes nationality—or the common sympathies upon which it is predicated—to have any noninstrumental value (Miller Reference Miller1997: 10, 193; Varouxakis Reference Varouxakis2002, 23). However, given the conceptual and empirical analyses just highlighted, Mill is clear that “Nationality is desirable, as a [defeasible] means to the attainment of liberty” (Reference Mill and Robson1985, 348). Moreover, he also doubts the viability of other potential routes to such liberty besides mononationality, free institutions being “next to impossible” without it (Reference Mill and Robson1977a, 547).

From Civilizing Despotism to Coercive National Assimilation?

In Considerations, while discussing the social and historical conditions in which representative government is suitable, Mill maintains that “like any other [form of] government” it “must be unsuitable in any case in which it cannot permanently subsist—i.e. in which it does not fulfil … three fundamental conditions … 1. That the people should be willing to receive it. 2. That they should be willing and able to do what is necessary for its preservation. 3. That they should be willing and able to fulfil the duties and discharge the functions which it imposes on them” (Reference Mill and Robson1977a, 413).

Mill’s conceptual analysis relates to condition 1 as follows. Given Mill’s definition of nationality, a multinational state will be one in which each nation seeks exclusive self-determination, and thus is “[un]willing to receive” representative institutions that include the others on an equal footing. This leaves the nations within the state with a choice between secession, the elimination or expulsion of the other nations from the state territory (that is, physical elimination), or ruling the other nations despotically (that is, segregation). The fact that Mill takes the absence of a common nationality or quasi-nationality to imply a deficit in the kind of common sympathies that produce a strong desire for cooperation and mutual protection is also salient here. After all, in this way, Mill’s empirical analysis also raises doubts about the likelihood of the satisfaction of conditions 2 and 3 in either a multinational state or a state that combines a nation with one or more unincorporated quasi-nations. In either case, the empirical analysis implies that there is likely to be too much indifference, if not outright hostility, between the various national or quasi-national groups, for one to be expected to actively defend the liberal-democratic rights of another.

However, such problems stemming from a lack of common nationality do not imply that if any of the given nations or unincorporated quasi-nations in question were considered individually, conditions 1–3 would remain unsatisfied. After all, a nation that rejects inclusive representative government in a multinational state might still embrace representative government upon secession (condition 1). Likewise, nations that are unwilling to do what it takes to make representative government secure and functional where this would benefit another nation or unincorporated quasi-nation with whom they share a state, might still be able and willing to do so in a state that was theirs alone (conditions 2–3). One reason this is important is that it raises the question of what kind of policy Mill favors in a situation in which a government presides over a population that is not united as a mononationality but it is the unwillingness, rather than the inability, of the various factions to sustain inclusive representative government that poses the major obstacle thereto. In this kind of case, does Mill think it can be legitimate to cultivate such willingness by using despotic methods to create a common nationality?

One reason for answering in the affirmative is that Mill makes it clear that despotic power can be legitimate when it is necessary to raise a people to the level of civilizational development in which they are able to sustain free institutions. For example, consider Mill’s view of those he takes to be least fitted to representative institutions, namely, those who live “in a state of savage independence”: “A representative assembly drawn from among themselves would simply reflect their own turbulent insubordination. It would refuse its authority to all proceedings which would impose … any improving restraint” (Reference Mill and Robson1977a, 415). Indeed, Mill thinks that the only form of government to which they are likely to submit voluntarily is that of a would-be military despot from within their ranks. With sufficient military prowess, perhaps supplemented with a degree of religious authority, such a leader might be able to instill a habit of obedience even if it is only to his own will (Reference Mill and Robson1977a, 394). However, Mill also makes clear that learning enough obedience to avoid anarchy is far from enough to live under free institutions. For example, Mill thinks that a “rude people,” that is more “alive to the benefits of civilized society” in virtue of being able to see that their interests can be served through moral and legal restraints (Reference Mill and Robson1977a, 377), may still be too barbaric to be able “to practise the forbearances which it demands: their passions may be too violent, or their personal pride too exacting, to forego private conflict, and leave to the laws the avenging of their real or supposed wrongs” (Reference Mill and Robson1977a, 377; Reference Mill and Robson1984b, 118–119). In consequence, he concludes, such a people will require a government “in a considerable degree despotic,” so that it can impose “a great amount of forcible restraint upon their actions” (Reference Mill and Robson1977a, 377).

Conversely, Mill also maintains that a people may be considered too barbaric for “representative government by the contrary fault to” those of “savage independence,” namely, “extreme passiveness, and ready submission to tyranny” (Reference Mill and Robson1977a, 416). In his view, this kind of fault is likely to be at its most severe among a people long enslaved, who have learnt thereby to obey the will of a master but lost their capacity for self-directed activity (Reference Mill and Robson1977a, 395). Mill’s rationale is that “[i]f a people thus prostrated by character and circumstances could obtain representative institutions,” they would be unable to sustain them because “they would inevitably choose their tyrants as their representatives” (Reference Mill and Robson1977a, 416). In consequence, he infers that such a people would also need a “parental despotism” that has enough of a master’s power to intimidate them into obedience whilst being progressive enough to uphold—and cultivate a respect for—the rule of law. Such a parental despotism will also design the laws to gradually increase people’s scope for individual decision making and personal initiative, increasing their independence from government direction. However, as before, Mill does not think that it is only the “extreme passiveness” (Reference Mill and Robson1977a, 416) of the “slave” (Reference Mill and Robson1977a, 395) that renders representative institutions unsuitable but also some lesser forms thereof. For example, Mill says that if a people are sufficiently submissive to criminals that they would sooner “pass by on the other side” when they observe a crime “because it is the business of the police to look to the matter,” the power of the “law and public authorities” may need to remain at a despotic level until the citizens are willing to play a more active role in sustaining the rule of law for themselves (Reference Mill and Robson1977a, 377).

Another obstacle Mill posits to a people’s ability to sustain free institutions is insufficient education and moral cultivation. One worry here is that people who combine extreme ignorance with extreme selfishness are likely to be sufficiently illiberal to be inclined to use their liberal-democratic rights to deprive others of the same (Reference Mill and Robson1977a, 327). Another is that, although a population’s capacity to read, write, and perform basic arithmetic is not enough to put democratic institutions to their optimal use, a population incapable even of basic literacy and numeracy would be highly vulnerable to manipulation by would-be despots, even if they could be moved by unselfish ideals. Indeed, this is Mill’s interpretation of Louis Napoleon’s path to becoming president, and subsequently emperor, of France upon the back of the “peasants,” “in opposition to nearly every educated person in the country” (Reference Mill and Robson1977a, 327). Each of these concerns about education are salient to the claim in On Liberty that free institutions are unsuitable until people “have become capable of being improved by free and equal discussion. Until then, there is nothing for them but implicit obedience to an Akbar or a Charlemagne, if they are so fortunate as to find one” (Reference Mill and Robson1977b, 224).

In this way, Mill maintains that there is a legitimate role for despotism in raising people to the level of civilization at which they are capable of sustaining and benefitting from representative government and the liberties secured by the harm principle. At this point, people will possess enough but not too much of the energetic independence of the “savage,” and enough but not too much of the obedience of the “slave”, while combining these traits with the minimal moral sympathies and educational attainments needed to form responsible judgments concerning their own good and that of society more broadly. As such, they will be capable of minimally well-motivated active citizenry, rational deference to the wise rather than the merely powerful or charismatic, and adherence to laws of their society’s democratic making (Reference Mill and Robson1977b, 269; Reference Mill and Robson1977a, 322–323).

The preceding also raises two possible logical routes from Mill’s harm principle to Kymlicka’s claim that Mill embraces coercive assimilation, which would depend upon an appeal to the civilization clause, on the one hand, and the defense or protection clauses, on the other. According to the first, Mill could be said to appeal to the civilization clause to justify the coercive assimilation as a necessary means to rendering certain groups civilized enough to be able to sustain free institutions. Given this approach, the underlying assumption would have to be that the inadequacies of the religion, language, or customs of the targeted population are such that their coercive assimilation into a nation of superior civilization would constitute a form of emancipation. In contrast, the second approach would make no such assumption, and hence apply where a people are civilized enough to be able to sustain free institutions but constitute a threat to others in virtue of their unwillingness to share free institutions with them. Given this approach, the goal of the coercive assimilation would be to generate enough commonalities to cause the common sympathies in virtue of which the groups would become co-nationals, and thus willing to share free institutions with each other. In what follows, it will be shown that although Kymlicka interprets Mill as taking the former route, one might also try to defend what could be referred to as a neo-Kymlickan interpretation of Mill, according to which Mill takes the second route instead.

Kymlicka’s Alignment of Mill with Engels

According to Kymlicka, Mill’s concerns about multinational states, on the one hand, and his conception of the causes of nationality, on the other, lead him to support majority nationalities in coercively assimilating recalcitrant minority nationalities within their shared borders (the intra-state thesis). At the same time, drawing on the role that Mill posits for despotism in civilizational development, Kymlicka claims that Mill does not justify coercive assimilation in the name of free institutions alone. Instead, he claims, Mill also conflates majority nationalities with those of greater civilization, thereby conflating such coercive national assimilation with a civilizing mission. Moreover, Kymlicka implies, having added such appeals to a supposed civilizing mission to the rhetorical arsenal of oppressive larger nations, Mill simultaneously licenses them to assimilate smaller nations outside their borders through imperial force (the inter-state thesis), thereby expanding the size of their nation-states and the scope of civilization in one fell swoop (Reference Kymlicka1995, 70–73).

Although Kymlicka offers some direct evidence for these claims from Mill’s texts, his immediate recourse is an appeal to the authority of Hobsbawm, who claims that Engels’s “essential stance” on national minorities was also embraced by Mill (Reference Hobsbawm1992, 34–35). Turning to Engels’s stance, Kymlicka claims that, although he “accepted the right of ‘the great national subdivisions of Europe’ to independence,” “smaller ‘nationalities’ were expected to assimilate to one of the ‘great nations’, without the benefit of any minority rights, whether it be language rights, or national autonomy.”Footnote ⁵ Moreover, Kymlicka maintains, this expectation was not merely a prediction but also a prescription predicated on the acknowledgment that it would necessitate force and coercion (Reference Kymlicka1995, 69–70).

Kymlicka’s principal evidence for this claim is Engels’s attitude to the outbreak of the First Schleswig War (1848–1851). This was fought between Denmark and some of the members of the German Confederation—most notably Prussia—over whether the then Danish Duchy of Schleswig, along with its largely German- and Danish-speaking population, should be incorporated into the German Confederation. Having claimed that the “Danish nation is in commercial, industrial, political and literary matters completely dependent on Germany,” and that the latter is more “revolutionary and progressive,” Engels backs the German nationalists thus: “By the same right under which France took Flanders, Lorraine and Alsace, and will sooner or later take Belgium—by that same right Germany takes over Schleswig; it is the right of civilization as against barbarism, of progress as against static stability … this right carries more weight than all the agreements, for it is the right of historical evolution” (Engels Reference Engels, Cohen and Benton1977a, 422–423).Footnote ⁶

Kymlicka’s interpretation of Engels as a coercive assimilator can also be reinforced with reference to the latter’s newspaper articles on the role of the “Slav barbarians” in the revolutions of 1848 (Engels Reference Engels, Cohen and Benton1977c, 228). For example, in one article on the attempted revolution in Hungary, Engels makes clear his hostility to the “Croats and Slovenes,” who he claims to have allied with the Hapsburg emperor against the Hungarian Revolutionary Army. This leads to a disquisition on the historical processes that have led eastern European peoples, aside from the exceptional Poles, Magyars, and Germans (along with the less significant “Rumanians [sic] and Transylvanian Saxons”), to acquire reactionary pro-feudal characteristics and aims. From this he concludes that they are incapable of supporting the bourgeois revolutions that he takes to be a precondition for communist emancipation, declaring them “destined to perish before long in the revolutionary world storm.” Thus, whereas Mill worries that free institutions are next to impossible in a multinational state, Engel worries that the capitalist preconditions of communist emancipation are impossible for a population weighed down by the presence of feudal-oriented “Southern Slavs” (Reference Engels, Cohen and Benton1977c, 230–236).

Turning to the agency that Engels thinks could lead these Southern Slavs to perish, he suggests that, sooner or later, Louis Napoleon will “conjure up” a “victorious uprising of the French proletariat,” and launch a war in which “the Austrian Germans and Magyars will be set free and wreak a bloody revenge” upon the national minorities who have tethered them to the feudal regime: “The general war which will then break out will smash this Slav Sonderbund and wipe out all these petty hidebound nations, down to their very names. The next world war will result in the disappearance from the face of the earth not only of reactionary classes and dynasties, but also of entire reactionary peoples. And that, too, is a step forward” (Reference Engels, Cohen and Benton1977c, 238, emphasis added).

Although some historians interpret this as a call for the “genocide” of the Slavs in question (Watson Reference Watson1998), Kymlicka categorizes Engels’s position in terms of coercive assimilation rather than physical elimination (Reference Kymlicka1995, 70, 211n24). Kymlicka’s rationale seems to be that, although it is difficult to see how the war and “bloody revenge” Engels envisages could not entail the death of many Slavs, he is not necessarily calling for the destruction or expulsion of every member of the group. After all, he also declares (paternalistically) that the coercion will be conducted in the Slavs’ own “material interests” (Reference Engels, Cohen and Benton1977b, 371). This is also supported by Engels’s attitude to the “Gaels” (of Scotland), “Bretons” (of France), and the “Basques” (of Spain), who, he maintains, will remain “fanatical standard-bearers of counter-revolution until their complete extirpation or loss of their national character” (Reference Engels, Cohen and Benton1977c, 234–235, emphasis added; see also Nimni Reference Nimni1989, 313). Thus, when Engels speaks of the elimination of the “Slav barbarians … down to their very names,” he seems to be calling for an initial military victory over them attended by acts of physical elimination, the threat of the repeat of which will subsequently coercively assimilate them to the point of eliminating their cultural and linguistic identities (Reference Engels, Cohen and Benton1977c, 238).Footnote ⁷

The Inter-State Thesis

In appealing to Hobsbawm’s authority to align Mill with Engels, one thing Kymlicka overlooks is that Hobsbawm’s rationale for positing this alignment runs contrary to his own. For Hobsbawm, the alignment shows not that both were coercive assimilators but rather that Engels is dealt unjustly by those who have “bitterly assailed” him “as a great-German chauvinist” for merely “predicting” as opposed to prescribing “the disappearance” of the Slavs in question. Indeed, for Hobsbawm, it “is sheer anachronism to criticize” Engels “for his essential stance,” as this did not “imply any hostility to the languages and culture of such collective victims to the laws of progress (as they would certainly have been called then)” (Reference Hobsbawm1992, 34–35).

Nevertheless, the preceding evidence indicates that Kymlicka has the stronger case with respect to Engel’s position, and that he is correct to dismiss (albeit implicitly) Hobsbawm’s suggestion that it is anachronistic to criticize it. However, a key reason for this that Kymlicka also overlooks is that Mill himself would have found Engels’s position deeply objectionable despite being his contemporary. In consequence, although Kymlicka’s appeal to Hobsbawm’s authority fails, his invocation of Engels is helpful qua providing an example of an authentic coercive assimilator with whom Mill can be compared. That said, when it comes to problematizing Kymlicka’s alignment of Mill and Engels with respect to the inter-state thesis, it will be helpful to start by examining an aspect of his interpretive package that misconstrues both thinkers.

As shown above, Kymlicka suggests that Engels takes the “carriers of historical development” to be “great nations, with their highly centralized political and economic structure” (Reference Kymlicka1995, 70). Thus, in suggesting that Engels takes this to include Russia, Kymlicka implies that Engels runs together the largest regional powers with the most civilized or historically progressive (Reference Kymlicka1995, 69–70) and thus attributes the same invasive “right of historical evolution” to Russia as to France and Germany (Engels Reference Engels, Cohen and Benton1977a, 422–423). However, contra Kymlicka, one of Engels’s key objections to pan-Slavism lies in its allegedly pro-Russian outlook. Engels reasons that, given its reactionary character, on the one hand, and the inability of the Slavs to defend themselves, on the other, its “direct aim” is “the creation of a Slav state under [feudal] Russian domination” (Reference Engels, Cohen and Benton1977c, 233). Indeed, it is for this reason that Engels replies to Mikhail Bakunin’s “sentimental phrases about brotherhood” in support of independent Slavic states, by declaring that “hatred” of Russians, along with the Czechs and Croats who allied with them, “is the primary revolutionary passion among Germans”: “only by the most determined use of terror against these Slav peoples can we, jointly with the Poles and Magyars, safeguard the revolution” (Engels Reference Engels, Cohen and Benton1977b, 378). Thus, although Engels might welcome the “Slav barbarians” being swallowed by Poland or Hungary, it does not follow that he would welcome either of the latter being swallowed by Russia because of its greater size and power.

Likewise, Mill makes it perfectly clear that civilization should not be conflated with mere size and power, insisting that when a small nation is absorbed by a less civilized nation of superior strength it is “a sheer mischief to the human race, and one which civilized humanity with one accord should rise in arms to prevent.” For example, “The absorption of Greece by Macedonia was one of the greatest misfortunes which ever happened to the world: that of any of the principal countries of Europe by Russia would be a similar one” (Mill Reference Mill and Robson1977a, 550; see also Varouxakis Reference Varouxakis2002, 10).Footnote ⁸

That said, there is a kernel of truth in Kymlicka’s intimation that Mill prefers larger nationalities. Ceteris paribus, Mill deems it a good idea for small nations that are geographically proximate to each other, and have enough in common to cultivate mutual sympathies, to unite under a federal union: “It has the same salutary effect as any other extension of the practice of co-operation, through which the weak, by uniting, can meet on equal terms with the strong. By diminishing the number of those petty states which are not equal to their own defence, it weakens the temptations to an aggressive policy, whether working directly by arms, or through the prestige of superior power” (Mill Reference Mill and Robson1977a, 559).Footnote ⁹

However, the unions that Mill has in mind here are supposed to deter aggression by larger powers rather than result from it. In this respect, Mill’s position is quite different from the position that Engels takes on the Slavs in the light of their historical inability to repel the “Turkish invasion of the fifteenth and sixteenth centuries” (Reference Engels, Cohen and Benton1977c, 232). Whereas Mill suggests that such aggression could potentially justify the civilized world in uniting to come to their defense, Engels infers that it justifies the most advanced countries conquering them instead. After all, Engels argues, since it was only the Germans and the Magyars who could save the Slavs from becoming “Turkish,” “indeed Mohammedan, as the Slavs of Bosnia still are today … this is a service which is not too dear even at the price of exchanging their nationality for German or Magyar” (Reference Engels, Cohen and Benton1977c, 232; see also Reference Engels, Cohen and Benton1977b, 370).

This brings us to Mill’s rejection of Engel’s “right of historical evolution” (Engels Reference Engels, Cohen and Benton1977a, 422–423). Because Mill frames the scope of the harm principle in terms of interference by members of “mankind” with “any of their number” (Mill Reference Mill and Robson1977b, 223), it remains salient at the international level. However, in A Few Words on Non-intervention (1859), in which Mill adds supplementary preconditions for legitimate foreign intervention, he makes clear that the harm principle leaves no scope for the aggressive inter-state violence in Europe that Engels’s “right of historical evolution” is supposed to entail. As Mill puts it: “among civilized peoples, members of an equal community of nations, like Christian Europe … [i]t would be an affront to the reader to discuss the immorality of wars of conquest, or of conquest even as the consequence of lawful war; the annexation of any civilized people to the dominion of another, unless by their own spontaneous election” (Reference Mill and Robson1984b, 120–121). Thus, contra Kymlicka’s alignment, Mill rejects Engels’s claim that France has a right to “take Belgium” (Reference Mill and Robson1977a, 423).Footnote ¹⁰ Moreover, in Mill’s contemporaneous references to the First Schleswig War in Vindication of the French Revolution of 1848 (1849)—a text that covers some of the same issues as A Few Words—he condemns that manifestation of German nationalism as a temporary reversion to “barbarous feelings” (Reference Mill and Robson1985, 347).

Of course, A Few Words restricts the scope of the preceding claim, that the “immorality of wars of conquest” should be taken for granted, to wars between “civilized peoples” (Reference Mill and Robson1984b, 120–121). In consequence, this claim is insufficient to refute the inter-state thesis as such, as it leaves room for wars of conquest against those considered to be uncivilized. In other words, it leaves open the possibility that Mill’s disagreement with Engels is less one of principle, and more one over whether the Slavs count as civilized, and thus immune from the kind of conquest that may be fitting for barbarians. Indeed, Kymlicka himself raises the possibility of such an interpretive move when he claims that many “nineteenth-century liberals, including John Stuart Mill, thought that liberal states were justified in colonizing foreign countries in order to teach them liberal principles” (Reference Kymlicka1995, 166, emphasis added). If this were true, would it not follow that Mill believes in a liberal variant of Engel’s right of historical evolution after all?

The first answer is that, as a matter of logic, it would not follow if the said right is one of coercive assimilation in addition to civilization as such. For example, although Mill makes clear that he takes British rule in India to be legitimate in his own day, and that he sees its legitimacy to depend upon the British engaging in a paternalistic civilizing mission, he does not see this mission as involving the coercive assimilation of Indians into the British nationality (Reference Mill and Robson1977a, 550). Indeed, when Mill compares British imperialism in India and Ireland in England and Ireland (1868), he says that one “advantage” in the Indian case was that:

This passage outlines two different approaches to colonial rule somewhat akin to Weberian ideal types. Although both are framed in terms of systems of rights rather than language and culture as such, the former pertains to the latter because it is an open question whether a colonizer will recognize a right on the part of the colonized people to retain them. In this respect, the first approach, which Mill associates with a colonizer without “conscience” and “sunk in … reckless savagery,” presupposes the colonizer’s right to obliterate the distinctive law, language, and culture of the colonized, and replace it with that of the colonizer (savage colonialism). In contrast, the second approach presupposes the colonizer’s right to obliterate only that which its “conscience” deems to be morally irredeemable—his examples in the case of British India include slavery, suttee, and infanticide (Mill Reference Mill, Robson, Moir and Moir1990, 121–125)Footnote ¹¹—along with a corresponding parental-despotic duty to use what remains of the distinctive law, language, and culture of the colonized as the basis for its future development (conscientious colonialism). This makes clear that even if Mill endorsed inter-state wars against the uncivilized for the sole reason of civilizing them, it would be inconsistent of him to take this to involve coercive assimilation, as this would amount to defending what he takes to be savagery in the name of civilization.Footnote ¹² Moreover, that Mill was consistent in this regard is demonstrated by his position on the Anglo-Orientalist debate over colonial education in India, in which he insisted that it would be wrong for the British East India Company to defund classical Indian higher education and replace it with a purely western curriculum (Tunick Reference Tunick2006, 603–608).

The second answer is that Kymlicka is mistaken to claim that Mill endorses inter-state wars against, or conquests of, supposed barbarians, initiated for no reason other than the paternalistic one of civilizing them to the point where they can sustain free institutions. Although Mill is willing to endorse a native civilizing despotism within any given uncivilized society, in On Liberty he states that “I am not aware that any community has a right to force another to be civilized,” a practice he refers to as “not a crusade, but a civilizade” (Reference Mill and Robson1977b, 291; emphasis added). In other words, although Mill makes clear that he takes the harm principle’s civilization clause to be necessary to demonstrate why British rule in India is justifiable, he denies that this clause serves as a sufficient condition to justify a civilizing conquest of one community by another.Footnote ¹³

The Intra-State Thesis

The preceding examination of Kymlicka’s alignment of Mill and Engels showed that Engels is willing to endorse inter-state wars of aggression in Europe in the name of a supposedly civilizing coercive assimilation. Since wars of aggression cannot be justified in the name of defense or protection, Mill cannot be charged with siding with Engels by leaving the door open to such wars via either of the harm principle’s first two clauses.Footnote ¹⁴ Moreover, since Mill rejects civilizades against supposed barbarians, on the one hand, and condemns coercive assimilation in the form of savage colonialism, on the other, he cannot be charged with leaving the door open to such wars via his civilization clause either. In consequence, the inter-state thesis can be dismissed entirely and the possibility of defending the intra-state thesis considered instead.

For his part, Kymlicka suggests that Mill endorses the intra-state thesis via an appeal to the civilization clause in the following passage of Considerations:

It would be hard to deny that Mill betrays a derogatory attitude to the members of the groups in question who insist upon exclusive self-determination. Moreover, some of the language employed is reminiscent of Engels’s description of the Gaels, Bretons, and Basques as “relics of a nation” (Reference Engels, Cohen and Benton1977c, 233). But does this indicate that Mill is pointing to the actual or potential justification of, say, the intra-state coercive assimilation of the “originally … inferior and more backward” Welsh by the English via the non-civilization clause?

The first point to be made at the level of general principle is that in On Liberty, Mill makes clear that the period in which the civilizational threshold that immunizes peoples from legitimate paternalistic coercion has been “long since reached in all nations with whom we need concern ourselves,” and hence, when it comes to these nations, compulsion is “justifiable only for the security of others” (Reference Mill and Robson1977b, 224).Footnote ¹⁵ In consequence, at the level of application, Mill cannot be interpreted as claiming that it would be possible to justify the paternalistic intra-state coercive assimilation of the Welsh (or any other European people) in his day. After all, Mill says that the Welsh were “originally … inferior” (emphasis added), not that they remain so, or that the cultural and political losses of participation in the United Kingdom (or the British Empire) would be so terrible for the Welsh that they would lose their ability to sustain free institutions.

The second point pertains to the historical dimension of Mill’s claim. There is no denying that Mill intimates that Wales benefited from English rule at some point in history, on the one hand, and benefited from some aspects of the Anglicizing assimilation that resulted from the intermixture of the peoples, on the other. However, it does not follow that Mill should be read as justifying, say, the Edwardian Conquest of Wales (1277–1283), and the subsequent incorporation of Wales into the Kingdom of England through the Laws of Wales Acts (1535–1542), by claiming that whatever coercive assimilation that entailed was justified on paternalistic grounds.Footnote ¹⁶ Firstly, although Mill does not discuss the medieval English conquest of Wales, he discusses that of Ireland, and rather than justifying it he describes it as mere “usurpation” (Reference Mill, Robson and Robson1986, 929). Secondly, as shown above, Mill’s position is that insofar as despotism (or colonialism) includes coercive assimilation, it can only be considered paternalistic (or conscientious) where the aspects of a culture that are targeted for coercive elimination are irredeemably immoral. In the case of the Laws of Wales Acts, the Welsh people were incorporated into the English legal system, whilst Welsh was denied the status of an official language of state, with the goal being to “extirpe all and singular the sinister usages and customs differing from the laws of this [English] Realm” (Raithby Reference Raithby1811, 243). However, Mill says nothing whatsoever to imply that retaining the use of Welsh as one of the official languages of state would have been morally analogous to retaining suttee or slavery. In consequence, he is committed to viewing the first medieval steps towards the Anglicization of Wales to be just as much a case of savage colonialism as those taken in Ireland.

Of course, given the difficulty of defending the intra-state thesis with reference to the civilization clause, the question is raised as to whether Kymlicka would have been able to make a more plausible case if he had framed it in terms of the defense or protection clause instead. Here Ireland might seem to provide a promising case for such a neo-Kymlickan interpretation because Mill discusses the dangers its independence could pose to Britain. One of his concerns in England and Ireland is that the “mere geographical situation of the two countries” renders “them far more fit to exist as one nation than as two. Not only are they more powerful for defence against a foreign enemy combined than separate, but, if separate, they would be a standing menace to one another” (Reference Mill and Robson1982a, 521).

A second is that Ireland’s historical grievances—and the tensions and conflicts that could be created through the process of separation itself, such as those of a “a civil war between the Protestant and Catholic Irish, or between Ulster and the other provinces” (Reference Mill and Robson1982a, 522–523)—would make it easy for future “trifles” to “become causes of quarrel” (Reference Mill and Robson1982a, 522–523). A third—but not the last that Mill offers—is that such a standoff could draw the pair into a security dilemma. For example, fearing reinvasion by Britain, Ireland might seek to shield itself by forming a continental alliance, which would in turn threaten Britain with the prospect of Ireland serving as a stepping-stone for an invasion force. In this way, Ireland’s defensive measures could threaten Britain to the point of inducing the invasion that the continental alliance was designed to prevent.

Given such worries, it seems that one should expect a Mill who believes that coercive assimilation is justifiable in self-defense to urge Britain to block Irish secession and proceed accordingly. However, Mill makes clear that if a majority in Ireland come to favor independence, Britain should accept this verdict. In consequence, advocates of the intra-state thesis need to be able to explain away Mill’s opposition to coercive assimilation in this case. One option would be to point to Mill’s doubts about the feasibility, as opposed to injustice, of holding Ireland by force. For example, Mill suggests that, although Russia might be able to hold down a “people in desperation” through “military violence” in virtue of being “almost inaccessible to a foreign enemy,” and thus immune to retaliation by others, “the attempt could not long succeed with a country so vulnerable as England, having territories to defend in every part of the globe, and half her population dependent on foreign commerce” (Reference Mill and Robson1982a, 520). Similarly, although Mill acknowledges that some Britons might support crushing Irish secession with “fire and sword,” he deems it unfeasible because “the mass of the British people” would not “permit the attempt.” However, this focus on considerations of feasibility overlooks the fact that Mill shares what he takes to be the sentiments of the British majority on this question, deeming them “those who are not yet corrupted by power,” and guided by “the sense of right” (Reference Mill and Robson1982a, 520). In short, Mill’s principal objection to using force or coercion to block Irish independence is moral, and he makes it despite believing it could pose a danger to Britain (Reference Mill and Robson1982a, 520; Reference Mill and Robson1977a, 551; cf. Martins Reference Martins2012, 99).

In the British-Irish case, Mill deems Irish secession to be a viable option because the two communities are (more or less) “geographically separate” (Reference Mill and Robson1977a, 551). In consequence, an alternative approach to justifying the intra-state thesis via the defense or protection clauses could be to argue that the cases in which Mill embraces coercive assimilation are those in which he does not take secession to be a viable option. As indicated above, Mill judges Hungary to be one such case due to the intermixture of the ethnic minorities with the Magyar majority. In doing so, he endorses implicitly the use of force or coercion against any minority group that attempts to secede by seizing territory on a unilateral basis. Moreover, given the intermixture of the populations, there is a potential justification for this via the defense or protection clauses, namely, that it would entail either segregation or physical elimination for other groups in that territory. After all, given his definition of nationality, national self-determination for such a group would have to be exclusive, and thus predicated on denying inclusive self-determination to other groups in that territory, by either killing them, forcing them out, or denying them a right to political participation should they opt to remain.Footnote ¹⁷ Could this, then, show that Hungary provides a case in which Mill endorses coercive assimilation via the defense or protection clause in the name of avoiding the greater evils of segregation or physical elimination?

Once again, the answer is negative, as even if Mill endorses defensive force or coercion in response to aggressive manifestations of nationality, it does not follow that he thinks this should take the form of coercive assimilation. It is true that Mill implies that measures should be taken to form a Hungarian nationality when he says that “there is no course open to them but to make a virtue of necessity, and reconcile themselves to living together under equal rights and laws” (Reference Mill and Robson1977a, 549). However, he does not say that this should be achieved through merging the minorities into the Magyar language and culture (as advocated by Engels). On the contrary, his point is that, although it is easier to build “equal rights and laws” upon the foundation of the common sympathies of a mononationality, in this case those common sympathies will have to be forged through equal rights and laws. As shown above, Mill thinks such a path is perilous and success far from guaranteed, but that is why he defends it on grounds of necessity rather than optimality.

At the same time, Mill is optimistic enough to believe that a just political framework can constitute a defeasible cause of nationality-constituting common sympathies when other factors are present that favor fragmentation. For instance, he maintains that historical grievances notwithstanding, “consciousness of being at last treated not only with equal justice but with equal consideration” is persuading some Irish that it is better to be “fellow-citizens rather than foreigners” to the British (Reference Mill and Robson1977a, 551). However, in suggesting that it may be possible for the Irish to become British, by acquiring the sympathies to embrace inclusive self-determination, he is not suggesting that this would entail the disappearance of the Irish quasi-nationality, any more than he thinks this has occurred in the case of the English, Welsh, and Scottish. In consequence, there is good reason to think that he would take the same to be true in the case of the incorporated quasi-nationalities he envisages forming the Hungarian nationality, even if the cultural and linguistic boundaries between them softened over time (Reference Mill and Robson1977a, 549–550).

Diagnosing Kymlicka’s Error

If the intra-state thesis is indefensible with reference to the civilization clause or the defense clause, it must be dismissed along with the inter-state thesis. However, to dismiss Kymlicka’s interpretation thus necessitates a deeper diagnosis of the source of his error. Here once again the Hungarian case is enlightening. By highlighting how Mill can envisage a Hungarian mononationality composed of so many quasi-nationalities, the case also serves to highlight the peculiarity of Mill’s definition of nationality, especially the way in which it is framed in terms of common sympathies rather than the cultural or linguistic characteristics that he takes to be the most reliable causes thereof. For his part, Kymlicka’s definition of nation differs by providing language and culture with a constitutive rather than a causal role in its foundation. As he puts it, “‘nation’ means a historical community, more or less institutionally complete, occupying a given territory or homeland, sharing a distinct language and culture” (Reference Kymlicka1995, 11). Thus, whereas Mill takes the common sympathies that unite the Swiss Germans, Swiss French, and Swiss Italians in a desire for inclusive self-determination as a sign of their Swiss mononationality (Reference Mill and Robson1977a, 546), Kymlicka takes their cultural and linguistic differences to show that they are multiple nations with a shared patriotic loyalty to the same state (Reference Kymlicka1995, 13, 187).

The preceding suggests that Kymlicka misinterprets Mill by taking what he says in Considerations to mean by nationality what he, Kymlicka, means by nation, and thus that their deeper underlying nonnormative disagreement is over the nature of nationality itself. After all, once one acknowledges that Kymlicka’s nations may only count as Millian quasi-nations, it becomes easier to see why Mill is much less troubled by, and thus less prone to insist upon, the coercive assimilation of Kymlicka’s nations than Kymlicka himself suggests. Indeed, Mill’s awareness of Ireland’s historical resentments would have awoken him to the fact that such coercion is a very crude tool for creating nationality-generating sympathy as opposed to nationality-fragmenting hatred (Reference Mill and Robson1977a, 550–551).

The plausibility of this hypothesis can also be illustrated with one final example. In his discussion of Mill’s prescriptions following the (largely French Canadian) Lower Canada Rebellion against British rule in 1837–1838, Kymlicka claims that Mill follows Lord Durham—the general and high commissioner for British North America—in advocating “the more or less forcible assimilation of the French, so as to create a homogeneous English nation-state” (Reference Kymlicka1995, 55). Although Kymlicka offers little evidence to support the claim (Varouxakis Reference Varouxakis2002, 15–18), in Mill’s political analyses of Canada in 1838 (Reference Mill and Robson1982b), shortly prior to Durham’s infamous Report on the appropriate British response to the uprising, it is possible to find phrases that make Kymlicka’s point intelligible.Footnote ¹⁸ For example, whereas Durham speaks of finding a way of “coercing the present disaffection, and hereafter obliterating the nationality of the French Canadians” (Reference Durham1839, 221), Mill speaks of finding the “legitimate means of destroying the so-much-talked-of nationality of the French Canadians” (Reference Mill and Robson1982b, 459–460). If Mill’s “nationality” is lent the same meaning as Kymlicka’s “nation,” the phrase makes Mill sound like Engels. However, when Mill’s phrase is understood in terms of his own definition, it signifies only that he is seeking legitimate means whereby British and French Canadians can come to form enough common sympathy to embrace inclusive self-determination.

The key conceptual point here is that, whereas Mill’s conception of nationality entails that only one nation can enjoy self-determination in a single state, Kymlicka’s conception allows for multiple nations to enjoy self-determination within a federal state over which none of them maintains complete control, provided they have the appropriate group rights. This is crucial for eliminating the appearance of normative disagreement over coercive assimilation in the case at hand. After all, when Mill speaks of the “only legitimate means of destroying the so-much-talked-of nationality of the French Canadians,” the means he has in mind is a “federal body” (Reference Mill and Robson1982b, 459)—ultimately rejected by Durham (Reference Durham1839, 225–227)—that will unite the British and French Canadians in a single state, while allowing the two peoples to have separate regional legislatures. Of the French Canadians, Mill says that such a federal body would “compel them,” but not by “bringing into their house and home, into their social and domestic relations, the customs of another people (which, whether practised on all of them or on a part, would be one of the last excesses of despotism)” (Reference Mill and Robson1982b, 459). The compulsion would instead be “to consider themselves, not as a separate family, but an integral portion of a larger body,” thereby merging “their nationality of race in a nationality of country” (Reference Mill and Robson1982b, 459; see also Varouxakis Reference Varouxakis2002, 18).Footnote ¹⁹ In this way, a phrase that could be misconstrued as evidence of Mill’s support for coercive assimilation actually shows that it would even be hard to charge Mill with neglect in this particular case.

Nevertheless, the preceding still suggests that at the normative level Kymlicka would have been able to build a stronger case if he had charged Mill with a general tendency to favor policies of neglect instead. Mill’s defense of a federalist response to the French-Canadian uprising shows that he was more open to external protections than Kymlicka acknowledges. However, one might speculate that Mill only proposes that solution because the status quo ante failed to produce the kind of linguistic assimilation that he would have deemed most conducive to the mononationality he deemed necessary for free institutions. One reason for this is that Mill highlights the difficulty of generating a sense of common democratic purpose when linguistic barriers mean that the “same books, newspapers, pamphlets, speeches, do not reach” the entire population (Reference Mill and Robson1977a, 547).Footnote ²⁰ A second reason is that, in the case of the United Kingdom, Mill defends the creation of “municipal and provincial representations” (Reference Mill and Robson1977a, 535) so “that those who have any interest in common, which they do not share with the general body of their countrymen, may manage that joint interest themselves” (Reference Mill and Robson1977a, 537), but falls short of calling for a federation. Although this leaves scope for quasi-national minorities who form regional majorities, such as the Welsh and Scottish, to demand a federal body or alternative external protections, Mill’s concerns about the national fragmentation that could result therefrom mean that he is not the kind of thinker to initiate such a campaign. Indeed, although he encourages individuality ardently, he actively discourages the resurrection of “obsolete customs” and “declining languages” where this could generate national differences between groups (Reference Mill and Robson1977a, 551). In this respect, he seems to favor neglect, even though his commitment to liberal democracy implies that if it leads to popular demand for greater quasi-national devolution or linguistic revivalism, there is no legitimacy in a despotic veto.

Of course, some may reject Kymlicka’s distinction between neglect and coercive assimilation, arguing that the former should be conceived as a subcategory of the latter instead, especially where it is preceded by conquest or other forms of historical injustice (Blake Reference Blake2003, 222–223).Footnote ²¹ For example, one might argue that the efficacy of neglect in promoting assimilation in the present often depends upon the lingering effects of antecedent historical injustice that serve (1) to place the (quasi-)national minority in question into diminished or disadvantaged circumstances; and thereby (2) to convert subsequent government-generated incentives to encourage their assimilation into coercive offers that it is punitively costly for them to reject. Nevertheless, even if it could be shown that Mill’s apparent preference for policies of neglect is malign rather than benign (see Kymlicka Reference Kymlicka1995, 110), and that this means that Mill and Engels should be aligned as coercive assimilators after all (relative to some non-Kymlickan taxonomy), this would not render Kymlicka’s ethical alignment of the two thinkers any less misleading. After all, there would still remain a world of moral difference between turning a blind eye to unfair and pressurizing forms of disadvantage, and embracing what Engels referred to as “the most determined use of terror” (Reference Engels, Cohen and Benton1977b, 378).