The Last Artificial Virtue
Hume on Toleration and Its Lessons
Political Theory
Volume 37 Number 4
August 2009 511-538
© 2009 SAGE Publications
10.1177/0090591709335192
http://ptx.sagepub.com
hosted at
http://online.sagepub.com
Andrew Sabl
University of California–Los Angeles
David Hume’s position on religion is, broadly speaking, “politic”: instrumental and consequentialist. Religions should be tolerated or not according to
their effects on political peace and order. Such theories of toleration are often
rejected as immoral or unstable. The reading provided here responds by reading Hume’s position as one of radically indirect consequentialism. While
religious policy should serve consequentialist ends, making direct reference to
those ends merely gives free reign to religious-political bigotry and faction.
Toleration, like Hume’s other “artificial virtues” (justice, fidelity to promises,
allegiance to government), is a universally useful response to our universal
partiality—as Established uniformity, however tempting, is not. This implies
that toleration can progress through political learning, becoming broader and
more constitutionally established over time. A sophisticated Humean approach
thus shares the stability and normative attractiveness of respect- or rightsbased arguments while responding more acutely and flexibly to problems the
former often slights: antinomian religious extremism; underdefined political
agency; and internationalized, politicized religious movements.
Keywords: toleration; Hume; artificial virtues
T
oleration is widely recognized as essential to liberal democracy. Under
modern conditions of cultural pluralism and individual choice, progress
toward the ends of peace, security, political equality, economic opportunity,
and individual self-development requires that habits and norms of toleration prevail among both governing elites and ordinary citizens. Yet most
contemporary moral and political philosophy resists the idea of valuing
Author’s Note: I would like to thank for useful suggestions audiences at the American
Political Science Association’s 2006 Annual Meeting and at Queen’s University; more particularly Jeff Collins, Rainer Forst, Jock Gunn, Nadia Khalaf, Chris Laursen, Andrew Lister,
Alastair Macleod, and two anonymous reviewers for this journal; and LaMonica Andreoff for
research assistance.
511
512
Political Theory
toleration because it safeguards these ends. Instrumental justifications for
toleration are often seen as immoral and dangerous. They are allegedly
immoral because they are unprincipled, purely prudential. This seems “the
wrong kind of defence” for toleration; the right kind would appeal to the
intrinsic worth of individual autonomy. They are allegedly dangerous on
account of their fragility. The calculations of ends and interests that constitute even a durable modus vivendi can shift with political power—whereas
we want toleration to be stable, supported by principles that reproduce
themselves and grow stronger regardless of how powerful interests stand.1
In spite of these common criticisms, this essay will defend a sophisticated version of this allegedly immoral and unstable justification for toleration: one rooted in David Hume’s arguments regarding toleration,
establishment, and religious “policy.” Admittedly, articulating this sophisticated view will require drawing out the implications of Hume’s views
more systematically than he himself did: the account of toleration defended
here will be thoroughly “Humean” but not precisely Hume’s. That said, the
argument will follow Hume closely, not only drawing out and explicating
his political and historical arguments but also endorsing his approach to
political philosophy: instrumental, consequentialist, and determined to
redirect and channel the dangerous sentiments that we human beings persistently display and act on rather than (as Kantians tend to) trying to reason
as if they did not matter.
Hume’s consequentialist argument is strong for the same reason that it is
elusive and hard to articulate: it is radically indirect. The ultimate foundation
for toleration is prudential or “politic.” But a far-sighted or “enlarged” view of
both our own passions and others’ reveals the folly of trying to decide questions of toleration on prudential grounds, and suggests institutional and social
mechanisms that will let us act tolerantly while rarely thinking about why.
Given our natural tendencies to intolerance, thinking in politic or consequentialist terms will only replicate the biases that make toleration necessary.
Toleration works best when it seems (though it is not) natural, perhaps even
universal. Yet unless we understand why toleration is not natural, is in fact
difficult and hard to get used to, we may over time forget why toleration is
necessary, and we shall think about toleration the wrong way in key cases.
Toleration: Artificial, Not Arbitrary
For toleration to work maximally well it must become, through experience and social learning, second nature—what Hume calls an artificial
Sabl / The Last Artificial Virtue
513
virtue (one of only four or five such, along with justice, allegiance to government, fidelity to promises, and on some readings, chastity).2 Hume
famously argues that the rules of justice are artificial but not arbitrary. All
our experience and hard-won knowledge teach us that flaws of human
nature, like greed and partiality, make property, promising, and allegiance
to a settled government universally useful regardless of variation in human
societies and their circumstances. The laws of property, promising, and
allegiance are “Laws of Nature” in Hume’s special sense. They are far from
natural, but humans will always see fit to invent, inculcate, and enforce
them to guard against natural human inclinations that are both universal
and dangerous.3
Though Hume never says so explicitly, his argument for toleration can
be reconstructed along exactly the same lines. Seeing toleration as a learned
and artificial virtue fits perfectly with Hume’s claim that toleration is a “paradoxical principle and salutary practice.”4 Like a rule of property, worth
observing even when it means returning property to a vicious miser, toleration is both extremely useful—and, given the worthless or harmful nature
of some of the beliefs tolerated, often counterintuitive. The difference is
that while the idea of property comes to make great intuitive sense once we
are accustomed to it, toleration continues to baffle. This is because the
natural tendencies against which toleration guards us—superstition and
enthusiasm, and in their politicized form zeal and bigotry—are so strong as
to distract people in every generation from the likely consequences of
yielding to them.
Untutored prudence, on this view, is not as useful as it seems. The
primitive reason-of-state strategy of adopting either toleration or nontoleration as circumstances suggest may seem clever or politique but is in fact
short-sighted. It is predictably undermined by partiality and by the human
propensity to sympathize with the oppressed. The ends of peace and individual freedom that early modern “skeptical” and pragmatic thinkers
believed could be accomplished by either Hobbesian uniformity or by toleration (depending on circumstances)5 can in fact, and in the long run, be
furthered only by an equal and impartial policy of toleration. This is one of
the few policies whose predictable advantages neither disappear nor alter
over time. Like property, therefore, toleration may be called a law of nature.
If the interest-based argument for toleration in fact takes this shape, it may
allay not only the concern with stability but also a weak form of the “wrong
kind of defence” objection. Toleration so understood does not directly
express autonomy and mutual respect. Rather, it represents the instrumental
preconditions for promoting or safeguarding other ends that almost all of us
514
Political Theory
value: our lives; our Millian liberty to make our own choices provided they
do not harm others; and political peace and order. Institutions of property
protect all our goods (and the projects they enable) against our own and others’ partiality. Intuitions of toleration, similarly, protect life, liberty, and social
peace against our religious partiality or bias—against the human propensity
violently to prevent “threats” from groups different from ourselves, threats
that seem real but are in fact the figments of religious imagination.
We can distinguish three stages of what might be called politic toleration. First comes a crude stage, where persecution of religious innovators
seems (wrongly) the clearest means to the public good. This suggests the
need for an “enlarged” stage, where toleration is adopted by rulers and
citizens as the least bad alternative once religious diversity cannot be
avoided. Finally we can see (going beyond Hume) a contemporary stage,
where rulers and citizens, fearing their own propensities to misjudgment
and partisan passion, take matters of religious policy out of their own hands
and place them under the rule of constitutions and supermajorities. Once
we reach this third stage, we can for everyday purposes forget how and why
we reached it. We are then free to praise ourselves for having the wisdom
to ground our practice of toleration in deontological commitments to rights,
generosity, or practical reason.
But the difference between consequentialist and respect-based arguments for toleration does not completely disappear, nor should it. With religion, as with property, consequentialist arguments can help us work out
hard cases, by allowing explicit reference to what we know of the world.
Respect-based arguments can only draw legalistic distinctions. They cannot
effectively address questions of threat—a double threat, religious violence
on the one hand and bigoted overreaction to such violence on the other—
that the law acknowledges but whose particulars no law can judge. Here
too, Hume shows us how such analyses might work. In particular, changes
in the circumstances of religious violence and sectarian conflict have led in
the past, and should lead us now, to consider—with due self-questioning
and a constant bias against relaxing bans on persecution—changes in the
details of our toleration regime. This mitigated flexibility is a good thing.
Just as utilitarianism may be praised for not “churning out the same
defences of liberty and democracy whatever assumptions are made about
the state of the world or the preferences of its inhabitants,”6 broadly similar
arguments deserve credit for not spitting out inflexible extremes of toleration as if all the world were Canada.
Hume discusses toleration mostly in his History of England, where he
uses the term (and its opposite, persecution) to mean a bewildering variety
Sabl / The Last Artificial Virtue
515
of things. The apparent confusion partly stems from Hume’s use of “toleration” to mean both a sentiment and a policy.7 And it partly stems from
Hume’s project of showing that legal and moral contexts have changed over
time. Past actors must not be blamed for not practicing policies (toleration,
but also press freedom and habeas corpus) that seem dangerous in prospect
and prove viable only in retrospect.8
But even the more straightforward statements about toleration in Hume
are many and complex enough to arouse competing interpretations, some
stressing Hume’s support for toleration, some his apparent sympathy for an
Establishment that would limit religious competition. All these works tend
to see Hume as a simple consequentialist, who endorses as best that religious policy that efficiently promotes the political requirements of the time
and forestalls the most prevalent threats.9 But Hume’s furiously complex,
multifaceted treatment would be more usefully read as suggesting three
subtler points. First, the demand for consequentialist justification can help
us even when analysis of likely results yields no definitive conclusion. For
those who fail, as most do, to meet the demand at all thereby show their
true motive—bigotry. Recognizing this motive helps us guard against misconduct in the future and judge it in the past. Second, the combination of
Humean consequentialism with Hume’s own theory of parties, and still
more with the modern politics of public opinion, can motivate something
approaching the impartial toleration we are used to. Third, as suggested,
Humean toleration illuminates, as respect arguments do not, hard questions
that we may complacently think ourselves beyond: the possibility of extreme,
antinomian conscience; the question of agency, of who is supposed to be
implementing a régime of toleration and whether that person or institution
can be relied on; and finally, the proper reaction to religious and ideological
conspiracy on an international scale, where liberal institutions of domestic
authority cannot reach.
“Persecution”: A Warning Sign, Not a Standard
In response to Montesquieu’s suggestion that prevalent religions should
be tolerated but new ones may be extirpated, Hume appears to doubt
“whether persecution can in any case be justified”—only to hedge this
remark by appearing to give a very minimalist definition of persecution:
Laud’s expelling Puritan preachers from their (Anglican) pulpits did not
count as such. The particular case, and the quarrel with Montesquieu, are
too complex to discuss here,10 but the question of what counts as persecution
516
Political Theory
for Hume is crucial. Persecution, it turns out, is a matter of moral psychology;
the category’s boundaries track fanatical motives and the emotions that lead
us to flout our normal standards of dignity rather than the autonomy or
standing of the victims. Precisely for reasons of moral psychology, prudential reasons for or against toleration in their raw, immediate form will turn
out not to provide reliable guidance to what should be tolerated and why.
However, a demand for such reasons can provide crucial insight into moral
motives, resources for judging those who fail to articulate instrumental
reasons for their actions.
It is clear that Hume opposes what he calls persecution both for what it
does to the victims and their loved ones and for what it shows about the
persecutor’s moral and intellectual failings. Once Hume calls something an
act of persecution, he is bound to condemn it. Persecution is a synonym for
“absurd tyranny.”11 A determined policy of persecution amounts to “a long
train of violence and oppression”; “persecution and violence,” when
extreme, are said to make life “as insecure to [the persecuted] as a den of
robbers.”12 Persecution is closely associated, though not technically synonymous, with “ferocity,” “enormities,” “cruelty.” “Violent” or “furious” persecutions in particular involve, Hume suggests (though he does not quite
say) so much cruelty that they can never be justified—though here too
Hume’s reasons are usually instrumental, not intrinsic; violent persecution
all but rules out future peace.13
I submit that these positions fit together through a complex set of arguments. (1) Persecution is always regrettable, because leaving people alone
is always preferable when it can be done.14 (2) Persecution is sometimes
required for the public good when the sectarian doctrines being preached
are seditious and there is some chance of persecution’s working, though it
is not clear when exactly that is, except perhaps in the preprinting age and
in conditions where religious beachheads are (or were) very small.15 (3) When
the dissenting doctrines themselves are not explicitly seditious but call for
persecution, neither allowing these opinions to propagate nor suppressing
them is ideal, but on balance suppression seems slightly better because,
involving fewer numbers, it will require less violence.16 (4) The Protestant
religion as a whole, though in its early versions fanatical and seditious,
could not be suppressed once it had a mass following. Importantly, by the
early seventeenth century there was good evidence of this from French and
other examples—so the Montesquieu point is interesting but essentially
moot in a very large set of cases.17 (5) Torture and capital punishment in
particular tend to create martyrs as ordinary punishments (imprisonment,
test acts, or Hume’s favorite on fiscal grounds, monetary fines) do not.18
Sabl / The Last Artificial Virtue
517
(6) Private worship does not (ever, it seems) endanger the public good, but
attempting to suppress it endangers crucial procedural safeguards and
requires Inquisition-like measures involving guilt by association, secret
accusers, torture, and self-incrimination.19 A religious policy that accepts
these principles—all, except the last (ambiguously) being matters of policy,
not natural rights or equality or free conscience—does not, for Hume, count
as persecution in the negative sense. It is called either “severity” or perhaps
(in the case of [3]) justifiable persecution, not really worth the name.20 In
particular, a ban on publicly preaching seditious or politically divisive
forms of religion does not count as persecution.21
But a great deal of persecution in history does not observe these principles, and from that one can speak to both motives and likely consequences.
The motives of someone who persecutes without point—out of either religious malice or else culpable policy error, in the face of evidence that could
easily have shown the policy to be useless or harmful22—are likely to be
“bigotry” rather than policy: “revenge and cruelty, covered with the mantle
of religion.”23 And the results, given that instrumental reasons are not at
issue, are likely to be far more inhuman than a persecution based on policy
would be. For the latter will respect bonds of privacy, family, and criminal
procedure; the former delights in breaking these down in its zeal to end
moral pollution. That is why “furious” or “violent” persecutions are (consistently, it seems) those founded not on policy but on theology.24 “Human
nature appears not, on any occasion, so detestable, and at the same time so
absurd, as in these religious persecutions, which sink men below infernal
spirits in wickedness, and below the beasts in folly.”25
On this account, Hume’s argument is winding, social, and psychological, ultimately “consequentialist” but not in any direct sense. Motives matter but both their imputation and their predicted cause are indirect; offences
against normal standards of human decency matter partly in themselves but
partly because they go to motive. The test that asks whether someone proposing persecution has good instrumental reasons is therefore useful not
because it prevents persecution or even guarantees that persecution will in
fact achieve useful ends (one’s prudential judgment can always be wrong
or uncertain, and Hume gives many instances of that)26 but more indirectly,
as the miner’s canary. Someone who persecutes when it cannot possibly
have the desired effect must, the Humean may conclude, be motivated by
bigotry. Given his or her propensity to cause pain without adequate motive,
that person or party can be predicted to stop at nothing and to ignore—
without knowing it—both common humanity and social utility.27 In fact,
persecutors motivated by bigotry can be discerned partly by their penchant
518
Political Theory
for punishing religious deviance more harshly than they would “the greatest
ruffians, for [secular] crimes subversive of civil society.”28 Such (literally)
supernatural punishment is bad in itself and also provokes extreme enforcement mechanisms:
All persecution naturally, or rather necessarily, adopts the iniquities, as well
as rigours, of the inquisition. What a considerable part of the society consider
as their duty and honour, and even many of the opposite party are apt to
regard with compassion and indulgence, can by no other expedient be subjected to such severe penalties as the natural sentiments of mankind appropriate only to the greatest crimes.29
Someone prone to throw a newborn baby on the fire meant for its heretical
mother, or to tie accused heretics between the high tide and low tide lines to
drown them slowly, or as magistrate to take over interrogation from the professional torturer and draw a rack so tightly that the prisoner’s body is almost
torn in two,30 engages in acts that are not only horrific in themselves but signs
that he or she has revenge and hatred rather than public safety in mind. Queen
Elizabeth, along these lines, reasons (in Hume’s portrayal) after France’s
St. Bartholomew’s Day massacre that preventing the alleged crimes of
Protestants could not have been the real motive for the killing, since “the same
force, which, without resistance, had massacred so many defenceless men,
could easily have secured their persons, and have reserved them for a trial.”31
Compare the contemporary argument that torture is “not a reliable way
of gaining information.” The point is not normally to make a directly consequentialist argument—as if torture would be a good idea if it were reliable—
but to smoke out the conscious or unconscious motives of the torturers.
Since the torturers can be doing nothing for the public good, they must be
acting out of fear, conformity, careerism, hatred, or sadism. Similarly, Hume’s
argument against persecution provides a powerful skeptical and critical tool
against those who make vague arguments to justify vicious actions. By asking which severities precisely are necessary and appropriate to which ends,
one can—as Hume does—condemn the vast majority of persecutions as
rationalizations of something other than policy. This can be a strong arrow
in toleration’s quiver. Given some new factual circumstances (e.g., mass
communications techniques that make stamping out new religious and
political ideologies even more unlikely than Queen Mary’s stamping out
Protestantism), it can be even stronger. Attempts to suppress opinions by
force are even more suspect now than in Hume’s day, and the burden of
moral accusation they must bear all the greater.
Sabl / The Last Artificial Virtue
519
Making Lenity Stable: Establishment, Partiality,
and “Natural” Religious Equality
Hume generally supports free competition; some of his writings on trade
and monopoly influenced his friend Adam Smith. Religion is a famous
exception, and one that Smith noticed and criticized.32 In this sphere Hume
thinks that free competition would be disastrous. For if each minister competes for “customers” (Hume’s word), preachers will compete for the lowest common denominator, “no regard will be paid to truth, morals or
decency” in doctrine, and each religion will prosper by “practising on the
passions and credulity of the populace.” In a separate passage, Hume suggests that the possibility that hatred will arise from competition is the best
argument for an imposed religious uniformity, provided that this is feasible
(which Hume, as discussed below, strongly doubts).33 Rather that allowing
such competition, it is much better for the magistrate to
. . . bribe their [clergymen’s] indolence, by assigning stated salaries to their
profession, and rendering it superfluous for them to be farther active, than
merely to prevent their flock from straying in quest of new pastures. And in
this manner ecclesiastical establishments, though commonly they arose at
first from religious views, prove in the end advantageous to the political
interests of society.34
As Annette Baier notes, “[a] less religious justification for establishing
religion could scarcely be imagined.” Indeed, few doubt Hume’s scorn for
all religious positions prevalent in his time.35
So establishment in general seems a good thing, provided that it is run
by the magistrate. (Catholic establishments, in contrast, inevitably give
religious authorities interests contrary to civil ones.)36 In principle, Hume
opposes a thoroughgoing separation of church and state; at least in “civilized” governments, “the union of the civil and ecclesiastical power”
greatly promotes “peace and order.”37 Which establishment is best depends
on the civil ends to be served. In his speculative, self-proclaimedly impractical “Idea of a Perfect Commonwealth,” Hume proposes a Presbyterian
establishment, though with a civil check on religious trials.38 But in general,
and in the circumstances of Britain’s actual monarchical government, he
much prefers the Church of England. He indeed praises it more and more in
his later writings, abandoning (and revising in late editions of the History)
his early comments that its worship represents “superstition.”39 Anglicanism
fits Hume’s taste for social hierarchy, and has enough ceremonies to “allure,
520
Political Theory
and amuse, and engage the vulgar . . . without distracting men of more
refined apprehensions.”40 As the product of careful and incremental rather
than revolutionary religious reform, it has avoided for the most part the
latter’s doctrinal and moral extremes.41 And it is united to the State, encouraging (for reasons akin to Hobbes’s) a frugal church and a moderately
strong one—therefore not so jealous of its “power and riches” that it need
treat non-Established competitors with anything but “lenity and liberty.”42
Consistent with such arguments, Hume praises with little qualification
the reason and moderation with which the Anglican Church formulated its
rites under Elizabeth:
Rage and animosity against the catholic religion was as little indulged as could
be supposed in such a revolution . . . Many ceremonies, become venerable from
age and preceding use, were retained: The splendor of the Romish worship,
though removed, had at least given place to order and decency . . . No innovation was admitted merely from spite and opposition to former usage; And the
new religion, by mitigating the genius of the ancient superstition, and rendering
it more compatible with the peace and interests of society, had preserved itself
in that happy medium, which wise men have always sought. . . .43
Many commentators have taken this passage, with the quotation ending
there, to entail that Elizabeth’s moderate Anglicanism was Hume’s ideal.44
So it may have been in some absolute sense, stipulating that general
unbelief was not a live possibility. But Hume as usual does not think absolute ideals of much relevance to situations involving people with imperfect
and differing judgments, i.e., to politics. What men “have always sought”
is also, Hume notes at the end of the passage just cited, that “which the
people have been seldom able to maintain.”45 One problem with establishments is that rewards large enough to bribe indolence are also large enough
that powerful sects’ ministers will fight for the prize whenever the outcome
is in doubt—a potential argument against Establishment as such.46 Another
problem is reconciling establishment with freedom of religious inquiry;
Hume suggests that the Church of England was on shaky ground as soon as
Henry VIII allowed people to own and read English bibles.47
But the most fundamental problem is that sectarians will not be satisfied
with such placid, bribed ministers. Elizabeth’s middling Anglicanism was,
after all, a famous failure. Almost immediately after it was seemingly settled, sectarian pressures began to tear it apart.
Many of the English reformers, being men of more warm complexions and
more obstinate tempers, endeavoured to push matters to extremities against
Sabl / The Last Artificial Virtue
521
the church of Rome, and indulged themselves in the most violent contrariety
and antipathy to all former practices.48
The title of this section is, indeed, “Character of the Puritans.” It was the
Puritans, a sect and a party, who destroyed the possibility of moderate comprehension based on policy and a no-more-than-mild religious belief—
though later, High Church Anglicans under Charles II also abandoned
broad-Church “mitigated episcopy” in their zeal to attack Presbyterians.49
Though Hume does not spell it out, it is the constant danger of such
partisanship that renders dubious primitive, “politic” strategies for managing religion, and compels more impartial and durable schemes. The problem with what Hume calls party or faction (what we would now tend to call
ideological or even “movement” politics)50 is twofold. First, popular factions can overwhelm the elite politics of prudence: the politics of the few,
often of the one key minister (e.g., Richelieu, whom Hume mostly admires).
Already under James I and Charles I, pressure from parliaments and popular crowds were able to prevent monarchs from pursuing the religious
policies they most favored. By the time of the Restoration, the need for
kings to please popular religious opinion—and to shift along with
it—became palpable.51 Restoration England was no democracy but its king
still, like all governments, feared public opinion.52
A related but more fundamental problem is that factions think they are
being politic when in fact they are acting out of bigotry. “Biggotted zeal”
is “the most absurd of prejudices masqued with reason, the most criminal
of passions covered with the appearance of duty. . . .”53 When religious
enthusiasm takes expressive forms, heedless of consequences, it is at least
easy to recognize its character. But by the time of Charles II and ever since,
religious prejudice has seen the world through the lens of prudence.
Modern prejudice makes those in its grip see not angels or saints but imagined threats to their civil interests, threats that a dispassionate analysis
would prove imaginary.
This is Hume’s account of Charles II’s reaction to the Popish Plot.
Though Charles’s own tendencies were to indulge Catholicism—and later
to embrace it himself, in a moderate fashion—he was forced to adopt policies in response not to any real threat of rebellion posed by Catholics but to
the threats absurdly but sincerely perceived by parliamentarians and by city
mobs marinated in anti-Catholic bigotry.54 Unjust arrests and anti-Catholic
policies motivated by fear of the Popish Plot were seen by those calling for
them as rational responses to this-worldly threats, not against Protestant
souls but against their lives and liberty. That the plot’s alleged details were
522
Political Theory
ridiculous, and those who allegedly revealed it clearly lying, made no
difference. Bigotry prevented these facts from having influence. Belief in
the Plot remained common among English Whigs in Hume’s own day, this
false factual belief—a “bigoted delusion” as well as one “shameful” and
“barbarous”—being one of the “touchstones of partymen.”55 Nor is this
mechanism specific to anti-Catholicism. Lauderdale, in persecuting Scottish
Presbyterians, transformed his own unconscious religious bigotry into subjectively sincere but ungrounded fears of civil unrest—even when palpable
evidence disproved the existence of such threat.56
Hume’s analysis is all the more striking since he thinks that there was in
Charles II’s reign a genuine long-term threat from Catholicism. Though
there was a real threat, party feeling, “aided by the passions of hatred and
of terror” distorted beyond recognition its size and imminence:
Such zeal of proselytism actuates that sect [the Jesuits], that its missionaries
have penetrated into every nation of the globe; and, in one sense, there is a
popish plot perpetually carrying on against all states, protestant, pagan, and
mahometan. It is likewise very probable, that the conversion of the duke [the
future James II], and the favour of the king [Charles II] had inspired the
catholic priests with new hopes of recovering in these islands their lost dominion, and gave fresh vigour to that intemperate zeal, by which they are commonly actuated. Their first aim was to obtain a toleration . . . After they had
converted considerable numbers, they might be enabled, they hoped, to reinstate themselves in full authority, and entirely to suppress that heresy, with
which the kingdom had so long been infected. Though these dangers to the
protestant religion were distant, it was justly the object of great concern to
find, that the heir of the crown was so blinded with bigotry, and so deeply
engaged in foreign interests; and that the king himself had been prevailed on,
from low interests, to hearken to his dangerous insinuations. Very bad consequences might ensue. . . . But that the Roman pontiff could hope to assume
the sovereignty of these kingdoms . . . That a massacre could be attempted of
the protestants, who surpassed the catholics a hundred fold, and were invested
with the whole authority of the state: That the king himself was to be assassinated, and even the duke, the only support of their party: These were such
absurdities as no human testimony was sufficient to prove; much less the
evidence of one man, who was noted for infamy, and who could not keep
himself, every moment, from falling into the grossest inconsistencies.57
Whether or not Hume’s view of the long-term threat from Catholicism
is correct—and note that he puts it in the present tense, as something still
dangerous in his time—it is clear that there is a great distinction between
his attempt to soberly assess that threat and adopt measures addressed
specifically to it58 and the absurd prejudice that led the public to disregard
Sabl / The Last Artificial Virtue
523
all the rules of evidence in the Popish Plot. But the phenomenon of twisting
the evidence of threat to confirm one’s prejudices and bigotry is more general. We constantly fool ourselves and others, and Catholics and Protestants
are equally susceptible. The Catholic Philip II of Spain,
. . . actuated by an imperious bigotry . . . as he employed great reflection in
all his conduct . . . could easily palliate the gratification of his natural temper
under the colour of wisdom, and find in this system no less advantage to his
foreign than his domestic politics [employing “specious rules of prudence”
to rationalize his persecution policy].59
Hume’s account of artificial virtues generally is again useful. It explains
why toleration, to fulfill its ends, must develop into ever more systematic
and impartial forms. The purpose of such virtues is to guard against shortsightedness but also against a judgment biased by natural partiality. In the
case of property,
‘Twere better, no doubt, that every one were possess’d of what is most suitable to him, and proper for his use: But besides, that this relation of fitness
may be common to several at once, ’tis liable to so many controversies, and
men are so partial and passionate in judging of these controversies, that such
a loose and uncertain rule wou’d be absolutely incompatible with the peace
of human society.
Because of this, we need a system of rules of property “which must extend
to the whole society, and be inflexible either by spite or favour” (emphases
added).60 This is precisely why our rules of property do not distinguish the
miserly from the generous: our biases would prevent agreement on which
is which, and this would endanger social peace. But as Hume incessantly
reminds us, we are even more biased regarding religion. The Clarendon
Code was justified by the alleged “madness” of millenarians and the objective danger they were supposed to pose, but Hume’s analysis of the actual
danger (in hindsight) shows that these fears were based on “prejudices, narrow and bigoted.”61 “[F]actions being once excited, men can neither so
firmly regulate the tempers of others, nor their own, as to ensure themselves against all exorbitancies.”62 Perfectly instrumental rationality is, as it
were, more ardently to be wished than seriously to be expected; “the policy
of men is continually warped by their temper.”63
Because rationalization of bigotry regularly masquerades as good policy,
then, a superficially politic approach to religion—establishment, comprehension, bribing the indolence of some ministers, perhaps tolerating others
in a subordinate role—is itself impolitic, bordering on willfully obtuse. Just
524
Political Theory
as direct consequentialism in matters of property, i.e., allocating property
by fitness or merit, is less practical and useful than it seems, a superficially
“politic” attitude toward religion reliably replicates the very passions and
biases it is meant to control. Commentators have questioned how Hume’s
endorsement of establishment coheres, if at all, with his doctrines of toleration. To the extent that sectarian competition is dangerous, any regime
involving toleration would seem to leave room for such competition; and
efforts to defang enthusiasm through Establishment would seem beside the
point.64 If the above is right, the answer is clear. Hume thinks that Elizabethan
establishment and quasi-uniformity was a praiseworthy ideal but not practical, indeed destined to fail under factional and sectarian pressure. In the
contemporary age, we cannot be confident that ministers will be able to
make religious policy on prudence—or will even be able to know who is
prudent, including themselves. We who are judging the religious policy, in
fact, may be the ones exaggerating out of prejudice the civil risks posed by
people whose beliefs clash with ours; perhaps we all do so.
In the case of property, Hume calls for an “enlarged” form of instrumental reasoning, not one that calls for altruism but one that judges institutions
on the assumption that none who live under them will be altruistic or would
even want to be.65 “Enlarged sentiments” take account of “the civil interests
of society” or “general happiness of society,” but in a particular, deliberately
dispassionate mode. They are allied with the “cool reflection of a legislator,”
as opposed to “the intemperate zeal of a sectary.”66 To think in enlarged
fashion means, for Hume, treating religion instrumentally and politically,
not spending too much mental energy on the substance of theological disputes. Beyond this, however, it means taking a constitutional or polity-wide
perspective rather than restricting political concern to one’s party. “Affection
to a legal constitution” is more “enlarged” than “loyalty to [one’s] sovereign.”67
The antonym of “enlarged” is “narrow,” a word Hume regularly applies to
those who when judging tend to praise their sectarian or factional allies while
ignoring the predictable risk of conflict over such allegiances.68
Under the circumstances, universal toleration is the least bad alternative.
Reflecting on what Philip II might have thought had he not been motivated
by bigotry, and had he known what we know, Hume writes,
If any prince possessed such enlarged views as to foresee, that a mutual
toleration would in time abate the fury of religious prejudices, he yet met
with difficulties in reducing this principle to practice; and might deem the
malady too violent to await a remedy, which, though certain, must necessarily be slow in its operation.69
Sabl / The Last Artificial Virtue
525
This may imply that past princes should be forgiven for not trying toleration, given that its short-term effects were so hazardous.70 But it also
implies that toleration is in the long term the only policy worth trying. A
rough analogy can again be drawn to property. Rules of property do not
maximize utility in each case, but the whole system of property in the long
run is far better than the given the alternative of partiality and prejudiced
judgment and its attendant insecurity.71 Similarly, an institutionalized structure of universal religious freedom may, on a prudential account, lead to
particular impolitic results in the short term, allowing potentially dangerous
and antisocial beliefs to flourish. But in the long term and thinking ahead
to the likely effects of factional strife on political calculation, this policy
defeats the alternative of letting religious policy be governed by the swings
of partisan passions and the mistaken judgments that go with them.
This solution, is of course, open to its own problems or meta-problems,
since no standard for judging institutions can be wholly independent of the
biases that lead us to misjudge individual cases. As Hume dryly notes in the
Second Enquiry, when fighting an enemy we strive to assure ourselves that
our reflections are “enlarged” by making sure to attribute to the enemy bad
moral qualities that rationally justify hating and attacking him. Were we
less enlarged, we would have the barbaric honesty to hate him merely as an
enemy.72 Still, an institutional attempt at overcoming our predictable biases
is clearly better than nothing. Moreover, the psychological reasons for
likely institutional failure can be themselves be made the occasion for continuous and institutionalized reflection, discussion, and institutional reform,
especially since Hume’s “philosophical politics” in his History amounts to
a recurrent project of political and sentimental reflection, aimed at undermining our constant and natural belief that people of our own broad party
or ideological affiliation have been the ones to get religious and constitutional policy perfectly right. Through systematic irony and unmasking,
Hume casts doubt on every religious and ideological excuse for persecution—so as to make the reader suspicious of all religious factionalism,
including his or her own.73
Toleration so defined is not yet equal and universal. It implies that magistrates or those who determine their actions (this, in a democracy, includes
us all) should tie their own hands by laws of religious freedom and general
legal procedures to avoid the risks of sectarian competition and contagious
religious hysteria. But it does not follow from this that they—or we—owe
any respect or recognition to the substance of any religion or the opinions
of those who believe them. It is consistent with Hume’s theory of enlarged
toleration that all existing religious beliefs be ridiculous (many interpreters
526
Political Theory
hold that Hume himself thought so). Moreover, enlarged toleration would
also allow a policy of equally monitoring all congregations, and fining or
exiling ministers not for heresy but for any sign of sedition or the preaching
of violence. (Such monitoring, including of the dominant Reformed
churches, was practiced by Dutch magistrates in the seventeenth century, at
which time Hume does not hesitate to call them “tolerant.”)74 Finally,
political doctrines advocating direct and immediate violence, religious or
otherwise, could hardly claim protection under such a scheme, since preventing the social effects of such doctrines is the scheme’s whole point.
Sticking with the property analogy, “religious hate speech”75 or seditious
millenarianism would in this version of toleration be analogous not to
advocating theft but to theft itself: the natural infirmity that the impartial
religious order is meant to combat is extreme partiality that endangers civic
judgment. To cite another and happier Montesquieu doctrine, a pluralist
society must force each religion to tolerate the others, lest the victim religion in turn “attack the religion that oppressed it, not as a religion, but as a
tyrant.”76
With two centuries of experience beyond Hume’s we can adjust (and
largely have adjusted) these conclusions, moving toward a “Humean” system of religious policy more thoroughly and universally tolerant than Hume’s
own. Hume’s second-order politique strategy, which apparently entails a
widespread but not universal toleration and perhaps a mild Establishment
that is not expected to win universal favor, is not the last word. We can
achieve even more impartiality and immunity from faction. Thomas
Nagel has pointed out that while Hume’s rules of property are essentially
conservative—all get to keep what they currently have and transfer it by
consent—social learning can result in rules of property to take into
account other social good, yielding the ”social morality of the welfare
state.”77 In the same way, a development of Humean doctrine over time
might proceed in stages, First, minorities’ practice of religious observance narrowly understood might be rendered relatively immune from the
pressures of faction—perhaps by enshrining it in a law of quasiconstitutional status (as with the toleration of dissenters after the Glorious
Revolution). As circumstances change, the consequences of toleration
become more familiar, and political science develops, we might institutionalize more and more religious freedom than that for more and more
different religions (and eventually nonreligious beliefs). For instance, we
might allow the kind of technically “violent” religious doctrines that have
been shown to pose little threat in practice, or might enshrine religious
exemptions to noncrucial laws.
Sabl / The Last Artificial Virtue
527
But on a Humean account we can also have good reason to challenge
institutional safeguards for religious practices, in order to protect the
ends that toleration serves. For example, we might readopt surveillance of
potentially violent religious sects (though not indiscriminate persecution or
collective judgments of guilt) should this be justified by the level of observable, demonstrable, threats to citizens’ physical safety that their members
come to pose. Humean arguments would give reason in such cases to deliberately establish methods of checking our tendency to bigoted and partial
judgment and enforcement—these might include vigorous political debate,
and strong and independent administrative and judicial safeguards—and of
shining light on the remaining partiality with which these policies will
inevitably be pursued. Data on our own dangerous bigotry is equally relevant to institutional reform as data on others’ dangerous zealotry. The role
of toleration as policy, and its strength, is that it leaves the substance of
such decisions to political debate and social learning while persistently
warning us of the dangers to be guarded against. Toleration on this view is,
like all of Hume’s artificial virtues, an emergent convention rather than a
product of sovereign decision. “Magistrates,” or political actors generally,
can reinforce toleration as a social value through education, and if toleration is generally accepted they can coordinate a general social propensity to
punish occasional violators. But no clever executive or ruler can directly
control religious sentiments in the ways the politiques once hoped. The
practice of toleration works because it is no longer a deliberate policy.
Policy Arguments for Toleration Today
Besides accusations that they are unprincipled and unstable, the most
common criticism leveled against policy arguments for toleration is irrelevance. Even if the above analysis is true, one might argue, surely we no
longer need it. Religious persecution has few principled defenders in the
world (though many practitioners). Open and deliberate persecution seems a
problem that happens in other places, not a question whose merits pro and
con still concern us.
But this is too quick. Toleration as policy, in its sophisticated forms,
continues to address live questions, and a Humean approach continues to
help us solve them—as dignity-, respect-, or rights-based arguments cannot. For the latter yield uniform conclusions (respect everyone, give everyone equal rights) that fail to address complications of circumstance that
make it implausible or impossible to act on such conclusions.
528
Political Theory
Sects and the Single Worshipper
One of Hume’s central principles involves unbounded freedom of
conscience. He is against it.78 So are we. (This is why I have preferred
“toleration” to “religious liberty”: no one consistently, regardless of potential harms, favors the latter, regardless of potential harms.) Our failure to
realize this comes from assuming that individual conscience will be guided
by the doctrines of a group. Conscience in the sense we usually intend
means devotion to a long-established sect or belief system that has proved
compelling to a significant number of people, and not incompatible with a
peaceful and productive life, over time. And the idea of a religious person
evokes the idea of someone who belongs to a group with something institutionally to lose from violent or offensive behavior by its members.
Hume reminds us, however, not to judge religious inspiration generally
by the form that it takes after time and social interaction have socialized
and tamed it. When there were no such social restrictions on a newly freed
Protestant conscience, “the bands of society were every where loosened;
and the irregular passions of men were encouraged by speculative principles, still more unsocial and irregular.”79 Consciences quickly freed of traditional prejudices have proven in the past not rationally autonomous but
driven in their religious allegiances by “more powerful prejudices of
another kind”:
[t]he novelty itself of the doctrines; the pleasure of an imaginary triumph in
dispute; the fervent zeal of the reformed preachers, their patience, and even
alacrity, in suffering persecution, death, and torments; a disgust at the
restraints of the old religion; an indignation against the tyranny and interested spirit of the ecclesiastics. . . .80
Some enthusiastic Protestants before about the eighteenth century have
claimed inspired immunity from earthly magistrates.81 The Quakers went
further: they not only denied magistrates’ authority but broke into nonQuaker churches to call them false, and in some cases disdained clothes.82
To be sure, the dissenting Protestants of Hume’s day are nothing like this.
But Hume stresses that this represents a great change. Quakers, once fanatical, are now practically Deists—from Hume, a great compliment.83 More
generally, Protestants in Hume’s time pose no political danger because their
talk of free conscience is, fortunately, so much self-deception:
Though the liberty of private judgment be tendered to the disciples of the
reformation, it is not in reality accepted of; and men are generally contented
Sabl / The Last Artificial Virtue
529
to acquiesce implicitly in those establishments, however new, into which
their early education has thrown them.84
In practice, we allow full freedom of conscience, and accommodate certain
illegal actions motivated by sincere belief, primarily in the case of known
religious groups, not isolated individuals claiming an inspiration obvious
only to themselves. (The latter we call schizophrenics, and many selfproclaimed prophets taken seriously in England’s Civil War and other times
of strife would be so labeled today.) To be sure, believers can employ strategies like pacifism and conscientious objection to reassure governments
that making exceptions for even eccentric beliefs poses a limited danger.
But such strategies address a problem that still exists and that autonomy or
equality arguments for toleration tend to obscure: that of socializing a discourse of conscience that otherwise could serve to negate all the advantages
of law and government.
The Agent of Toleration
If the last point sounds harsh, this one will give us reasons to distrust our
own harshness. Treatments of toleration, and discussions of what sort of
things should be tolerated, do not typically specify the agent doing the
tolerating. The agent is left vague (“we”), or assumed to be an elite group
insulated from ordinary politics (“the state,” “the sovereign”), or left out
altogether for the explicit or implicit reason that ideal theorizing is supposed to posit a stable and universal consensus. In fact, toleration is one of
the last realms in which we continue to use nonchalantly terms like “state,”
“sovereign,” and “consensus” that we would otherwise question. When
thinking about politics generally we start with pluralist assumptions and
distinguish among the ends and values of different politicians, groups, parties, or ideological actors. But we do tend to think of toleration as something done collectively and impartially, equally by everyone, to everyone—or,
in the “state” case, by an insulated elite actor but in terms that all would
welcome.
Our wish to obscure the agent is completely understandable. Matters of
toleration are of fundamental importance to any society that wants to avoid
persecution, war, and insecurity. We want rules of toleration to be what
Rawls calls “constitutional essentials,” not subject to the political struggles
that drive normal policy matters. But the wish that toleration be general and
impartial should not be confused with the assumption that it automatically
is so. Hume reminds us that impartiality is very hard to achieve (perhaps
530
Political Theory
always unreachable). The normal case is for each group in society to project
its partiality, its prejudices, fears, and distorted views of harm and threat,
onto a group different from itself.
Liberal principles—Millian harm, Protestant (or modern Catholic or
other religions’) freedom of conscience, Kantian autonomy—are not so
much invalid as unhelpful. When some of us really distrust others, we will
perceive harm where none can reasonably be construed. Once we have
judged a particular group identity to be “incompatible with liberal commitments,” we will often hold its practices up to much higher standards of
autonomy or equality (or, we might add, whatever other values we consider
crucial) than we apply to majority groups considered less suspect.85 As a
result, we will persecute groups when we could have let them be, in a process that tends to snowball and produce the very threatening resistance,
insularity, hypertraditionalism, or uncooperativeness that we used to justify
our intolerance.86
Nothing in the Humean argument rules out the possibility that some
people who might claim toleration really do mean to harm us or really are
using “conscience” as a cover for projects that damage the public good
(however defined). What is ruled out is the likelihood that ordinary political
processes, and the public opinion that drive them, can be trusted to judge
correctly and impartially the harm or public bad involved. Rawls’ “burdens
of judgment” are applied by him optimistically. We are to tolerate others
because we know how hard it is to agree on what is good.87 But the obverse
of this is that we may be prone to be intolerant because we cannot easily
agree on what is harmful or whose actions, conscientious to themselves, are
rightly judged as deliberately wrong. There is no reason to suppose this will
be self-correcting. And our tendency to associate with those like ourselves
will make each of us think, rightly or wrongly, that “we” do have good
reasons for whatever fears we have.88
Realizing these things may lead us to be more skeptical of the rhetoric
of political agency in general. It is sometimes said that liberalism has
moved beyond toleration. But perhaps liberalism simply consists of the
argument for toleration, so thoroughly learned and generalized that its
original points and premises have been obscured. In its more realistic or
pragmatic versions, liberalism has learned from hard experience of biased
judgments of religious threat to become wary of any claim that social harms
in general can be systematically and permanently eliminated. Constitutional
checks on all government actions reflect this wariness. Their complete
absence normally corresponds with a real or imagined uniformity of belief
in the society involved, so that toleration does not arise.
Sabl / The Last Artificial Virtue
531
This may all sound obvious; indeed the idea that “who is to judge” is the
distinctive moral question in modern times is a liberal commonplace. But
it is a commonplace often forgotten in the specific literature on toleration,
which continues quite often to speak as if the problem were which other
people we (liberals?) should tolerate. To the extent that we think that judgments of cultural diversity and likely threats can be uttered with one voice,
independent of factional bias and collective egotism, we forget the origins
of our own assumptions. Hume helps us remember them.
The Limits of Leviathan
The history of the politique stance toward religion is often written as
footnotes to Hobbes. Surely (it is said) one solution to clashes in religious
judgment, and not the worst, is for a self-interested sovereign to impose a
uniform doctrine for the sake of social peace. Even those who now doubt
the utility of a uniform religion often still embrace the idea that a sovereign
must have the final word on matters of religion, ideology, and toleration for
the sake of the public good.89 If the agent is taken to be a society rather than
a sovereign, this fits in nicely with a mild cultural relativism that regards
societies or cultures as coherent units that can and should work things out
according to their own traditions and needs. The solution presupposes both
a “sovereign” who is somehow immune from public opinion and partisan
bias, and a commonwealth, state, or self-governing society that serves as a
self-sufficient unit so that an opinion banished from it can no longer
threaten people in it. Pluralistic democracy casts doubt, I have suggested,
on the former assumption. As Hume already realized, an increasingly global society casts doubt on the latter.
Elizabethan England would seem an Erastian’s dream. A sovereign, supported by a comprehensive spy system and in practice largely independent
of parliament and of common-law legal checks, enforced an official religion and prevented religious strife through command of laws, public symbols, and mandatory worship. Surely an island nation in such circumstances
could prevent religiously based political dissent. As Hume stresses, however,
this was far from the case. Religious factionalism and political threats were
by their nature international, and English Catholics merely set themselves up
in France:
The English seminary at Rheims had wrought themselves up to a high pitch of
rage and animosity against the queen [Elizabeth]. The recent persecutions,
532
Political Theory
from which they had escaped; the new rigours, which, they knew, awaited them
in the course of their missions; the liberty, which for the present they enjoyed,
of declaiming against that princess; and the contagion of that religious fury,
which every where surrounded them in France: All these causes had obliterated
with them every maxim of common sense, and every principle of morals or
humanity. . . . The assassination of heretical sovereigns, and of that princess in
particular, was represented as the most meritorious of all enterprizes; and they
taught, that, whoever perished in such pious attempts, enjoyed without dispute
the glorious and never-fading crown of martyrdom.90
Transfer the proper names from France to (say) Pakistan, from “Elizabeth”
to “Bush,” and this passage sounds quite contemporary.91
The Rheims seminary was not merely theoretical. It produced a would-be
assassin, John Savage, who was in fact sheltered and helped by extreme
Catholics in England including one Babington; this became known as
“Babington’s Conspiracy.” Elizabeth’s spymaster, Walsingham, foiled the
plot; the conspirators were, though Hume does not note it, hanged, drawn,
and quartered. Lest one draw bloody lessons, I would stress that Hume is the
first to recognize “root cause” arguments. Hume’s story of the relationship
between Elizabeth and Catholics is a story of action and counter-action in a
context of misunderstanding and tragic mistrust. The point is simply that
religious uniformity was not a sufficient policy. In the international context
(as theorists sometimes forget), it was no policy at all. It could not reach
Rheims. Elizabeth was safe from Catholic Fifth Columns when the Spanish
Armada invaded not to the extent she had repressed Catholics but, on the
contrary, to the extent that by (erratically) indulging them she had made most
unwilling to shelter such as Savage. Many English Catholics, in fact, encouraged enlistment in England’s army. Some even subscribed English ships.92
In our zeal to refine the categories and cases of toleration in countries
where religious wars are cases in philosophy books, we must not forget the
hard questions—of statecraft, faction, and maneuver as well as normative
theory—that attached to times when their peaceful resolution was in question.
Since their peaceful resolution, in international context, is still in question,
past answers to those questions can still help us. Whether we are the ones
in danger from potential religious enemies trained abroad, or the ones eager
to harbor, on grounds of toleration, religious or ideological dissenters who
will be judged as such enemies by other countries, or in fact such dissenters
and refugees ourselves, these are not matters that can be resolved within
one country or one political tradition. The problems of religious faction,
policy, and partiality are as they have always been: disrespectful of national
and cultural boundaries.
Sabl / The Last Artificial Virtue
533
Notes
1. See respectively Susan Mendus, “Introduction” to Justifying Toleration, ed. Susan
Mendus (Cambridge: Cambridge University Press, 1988), 10; John Rawls, “The Idea of an
Overlapping Consensus,” in Collected Papers, ed. Samuel Freeman (Cambridge, MA:
Harvard University Press, 1999), 421-48, quotation at 433. Rawls also suggests something like
Mendus’ argument in “The Idea of Public Reason Revisited,” Collected Papers, 589.
2. See David Hume, A Treatise of Human Nature, book III, part II, sections 1-6, ed. David
Fate Norton and Mary J. Norton (Oxford: Oxford University Press, 2000), or in 2nd ed., ed.
L. A. Selby-Bigge and P. H. Nidditch (Oxford: Clarendon Press, 1978), 477-534; henceforth
cited as T and cited by book, part, section, and paragraph, followed by “SBN” and the page
number in that edition. Also Hume, “Of the Original Contract,” in Essays: Moral, Political,
and Literary, ed. Eugene F. Miller (Indianapolis: Liberty Fund, 1987), 479-81. On chastity see
T 3.2.12, SBN, 570-73, and Annette Baier, “Good Men’s Women: Hume on Chastity and
Trust,” Hume Studies 5, no. 1 (1979): 1-19.
3. T 3.2.1, SBN, 484.
4. David Hume, The History of England from the Invasion of Julius Caesar to
the Revolution in 1688, 6 vols. (Indianapolis: Liberty Fund, 1983 [1778]), History, 5:130—
henceforth cited as 5:130. Compare History, 6:321: in the seventeenth century only the Dutch
had happened upon toleration as “the proper remedy” for “religious differences.”
5. See Richard Tuck, “Scepticism and Toleration in the Seventeenth Century,” in Mendus,
ed., Justifying Toleration, 21-35.
6. R. M. Hare, Moral Thinking (Oxford: Oxford University Press, Clarendon Press,
1981), II.9.9:167 (emphasis in original).
7. See, e.g., Hume, History, 2:356-57: Henry V is personally “tolerant” of Lollards, but
eventually sees no alternative to persecution as a matter of policy. Hume interprets Charles V’s
policy toward Protestants in just this way (notes Frederick Whelan, Hume and Machiavelli:
Political Realism and Liberal Thought [Lanham, MD: Lexington Books, 2004], 266-67, citing
Hume, History, 4:151-52).
8. For this “bifocal” theme in Hume, see Duncan Forbes, Hume’s Philosophical Politics
(Cambridge: Cambridge University Press, 1975), 284, 325; and more recently Andrew Sabl,
“When Bad Things Happen From Good People (and Vice-Versa): Hume’s Political Ethics of
Revolution.” Polity 35, no. 1 (Fall 2002): 73-92, both with extensive citations.
9. Richard H. Dees, “‘The Paradoxical Principle and Salutary Practice’: Hume on
Toleration,” Hume Studies 31 (April 2005): 145-64; Will R. Jordan, “Religion in the Public
Square: A Reconsideration of David Hume and Religious Establishment,” Review of Politics
64 (Autumn 2002): 687-713; Frederick G. Whelan, “Church Establishments, Liberty, and
Competition in Religion,” Polity 23 (Winter 1990): 155-85; John B. Stewart, Opinion and
Reform in Hume’s Political Philosophy (Princeton, NJ: Princeton University Press, 1992),
274-81; John B. Stewart, The Moral and Political Philosophy of David Hume (New York:
Columbia University Press, 1963), 256-87.
10. Charles de Secondat, Baron de Montesquieu, The Spirit of the Laws, ed. and trans.
Anne M. Cohler, Basia Carolyn Miller, and Harold Samuel Stone (Cambridge: Cambridge
University Press, 1989), bk. 25, chap. 9-10. Hume’s treatments of the question, not always in
his own voice, include History, 2:336, 355-56; 3:217, 433, 440-41, 478nQ; and 5:575-76nFF,
containing the text quotation. (I have altered Hume’s spelling “justifyed” to “justified.”) Tuck,
“Skepticism and Toleration,” 26-27, traces the thread of argument to Lipsius and Montaigne.
11. Hume, History, 6:418. Compare 6:330.
534
Political Theory
12. Ibid., 6:417. Hume’s near-Lockean language here is striking and very atypical.
13. Ibid., 6:226-27; on violent or furious persecutions, see ibid., 3:137; 4:205-06; 5:172
and esp. 339, 424; 6:65, 201, 330.
14. At History, 5:114, Hume famously endorses a proto-Millian definition of liberty
(though this was not original with either Hume or Mill).
15. Hume, History, 3:356-57.
16. This is my reading of the admittedly obscure History, 5:575-76, where Hume treats the
Montesquieu point, claiming that the Puritans could not complain of persecution when they
themselves would have spurned toleration as too indulgent of papist ceremony. Throughout
books 5 and 6 of the History Hume stresses the intolerance of all parties in the English Civil
War era (and before). Even the Independents, the most tolerant party, forbid Catholic and
Anglican worship (see 5:443 and 6:65); Cromwell’s greater indulgence toward Anglicanism
was unpopular (6:88). On prior ages see 3:390; 4:20, 24, 71.
17. Hume, History, 3:432-41; 4:18-19, 395-96nQ; and especially 6:322 (regarding
Scottish Presbyterians, persecuted under a “mistaken policy” by Charles II): “An unlimited
toleration, after sects have diffused themselves and are strongly rooted, is the only expedient, which can allay their fervour, and make the civil union acquire a superiority above
religious distinctions” (though “vulgar politicians” are too impatient to wait as long as this
takes).
18. Ibid., 3:432 (the emphasis on “capital punishments” is often missed); 3:439; 5:231,
244, 304; 6:500 (the last citing without comment William of Orange’s opinion that religious
tests for public office embodied not a “penalty” but merely “security provided for the established worship”). For praise of Charles I’s resort to “compositions” (nonenforcement of antiCatholic laws in return for money), see 5:229-30, 304.
19. Ibid., 3:364-67, 438ff.; 6:201, 227ff., 289, 466; and see the discussion below on conventicles. See also ibid., 3:372: a ruthless inquisition., since it leaves dissenters with nothing
to lose, gives them little reason not to rise in arms (on this point compare 4:147, 319).
20. “Severities” against “catholic preachers” [sic], dismissing them from pulpits and prosecuting public avowals of the Real Presence, seemed, says Hume, a “necessary policy” under
Elizabeth (Hume, History, 3:366-67). Compare 4:176-77: the fact that Elizabeth (when possible) “connived at” Catholic worship in “private houses” is said to entail that Catholics “could
not justly complain of violence or persecution”; also 4:265, 5:115, 130.
21. This was a common distinction at the time. Voltaire’s Treatise on Tolerance (in Treatise
on Tolerance and Other Writings, ed. Simon Harvey [Cambridge University Press, 2000],
chap. 5, 24-27) demands only freedom of private worship. I thank Chris Laursen for discussion on this.
22. The latter seems to be the case of Gardiner, a politique who argued for highly mistaken
civil reasons in favor of a policy of persecution that Queen Mary then adopted out of bigotry.
Gardiner’s proposal is called a “scheme of policy” (Hume, History, 3:431) but “surely never
enterprise was more unfortunate [note: not “morally heinous”] than that of founding persecution upon policy . . .” (Hume, History, 3:432). Here, as is his habit, Hume does not distinguish
sharply between moral and intellectual vices; stupidity is a vice like any other, and equally
blameable. See Hume, An Enquiry Concerning the Principles of Morals (henceforth EPM),
ed. Tom L. Beauchamp (Oxford: Clarendon Press, 1998), appendix 4, §3, 102, 314 in
Enquiries Concerning Human Understanding and Concerning the Principles of Morals, 3rd
ed., ed. L. A. Selby-Bigge and P. H. Nidditch (Oxford: Clarendon Press, 1975) (henceforth
“SBN” for this edition).
Sabl / The Last Artificial Virtue
535
23. Hume, History, 3:434-5 (re: Queen Mary). Compare 4:34 (Henry IV); 6:470-71 (on
Louis XIV’s revocation of the Edict of Nantes); 6:498-99 (on Louis’ persecutions as antistrategic, inspiring a League of terrified Protestant regimes against him: “no characters are more
incompatible than those of a conqueror and a persecutor”). To be sure, factional hatred,
founded on a passion for revenge, at least in some instances caused a level of tyrannical,
militarized persecution approaching that founded on religious bigotry—as with Charles II’s
persecution of Scottish “conventicles” (6:176, 226-28, 325-31, 416-18; see also 3:228).
24. Ibid., 5:339, 5:424, 6:65, 200-01.
25. Ibid., 3:437 (emphasis added). Here Hume’s argument is, I think, clearly indebted to
that of Pierre Bayle’s Philosophical Commentary.
26. Ibid., 4:280 As Whelan points out (Hume and Machiavelli, p. 144), Hume here labels
Henry III of France’s killing of the Duke of Guise a “cruel execution” but “a rational (if mistaken) calculation” of political necessity. In contrast, the friar who assassinated Henry III in
retaliation showed the “bloody spirit of bigotry” since he lacked similar instrumental and
public-spirited motives.
27. See ibid., 5:526: In 1648, only religious “bigotry” kept Charles I and Parliament from
compromising in the face of threats from Cromwell’s army.
28. Ibid., 3:439.
29. Ibid., 6:324.
30. See respectively: ibid., 3:438 (the baby was actually delivered by the pregnant woman
being burnt, then thrown on the flames with her); 6:418-19; 3:315.
31. Ibid., 4:164-65.
32. Adam Smith, The Wealth of Nations, ed. Edwin Cannan (Chicago: University of
Chicago Press, 1976), bk. 5, chap. 1, part 3, article 3, 313f. The reading here suggests that
Hume ultimately shares Smith’s suspicion that Church establishments reflect factional victories, not enlightened policy.
33. Hume, History, 3:434 (argument of Gardiner).
34. Ibid., 3:136 (emphasis added). Hume repeatedly portrays Catholic and Anglican priests
(though not so much sectarian Protestant ministers) as willing to barter theology for money:
1:270, 436; 5:413.
35. Annette Baier, “Hume on Different Species of Religion: The Case of Bishop Tonstal,”
paper delivered before the 33rd International Hume Conference, University Koblenz-Landau,
Koblenz, Germany, August 7-10, 2006. For an overview of Hume’s religious views, with citations to previous literature, see Shane Andre, “Was Hume an Atheist?” Hume Studies 19, no.
1 (April 1993): 141-66. For treatments of Hume’s account of religious faction, see faction, see
Jennifer Herdt, Religion and Faction in Hume's Moral Philosophy (Cambridge: Cambridge
Univ. Press, 1997), and Donald W. Livingston, Philosophical Melancholy and Delirium:
Hume's Pathology of Philosophy (Chicago: Univ. of Chicago Press, 1998).
36. Hume dissents from the fervent anti-Catholicism of his day, seeing certain benefits
even in a Catholic establishment (e.g., at History, 2:14, 3:137, 368-69; on this see Stewart,
Opinion and Reform, 278), but still insistent. See, e.g., Hume, History, 3:136. In particular,
indulgence-selling, which Hume compares to tax farming, gave Catholic priests an unfortunate incentive to make people more religious (ibid., 3:137, 207).
37. Ibid., 1:311; see also 1:163.
38. Hume, “Idea of a Perfect Commonwealth,” in Miller, ed., Essays, 520. (See also Hume,
History, 6:40.) I am among those who doubt Hume’s seriousness in this essay; by associating
commonwealths with Presbyterianism, which he elsewhere portrays as theologically silly and
politically dangerous (see esp. 4:45), Hume intends, I believe, a jibe at the commonwealth
536
Political Theory
form. Baier, in “Hume on Different Species,” 12, suggests an even more puckish possibility:
Hume endorsed Presbyterianism simply because he was writing in Edinburgh.
39. As noted by Stewart, Moral and Political Philosophy of David Hume, 283. Hume
expresses sympathy for “bells and smells” Anglican ceremony—though ruthlessly satirizing
its Laudian excesses (Hume, History, 3:354; 5:224-26).
40. Hume, History, 3:239, 339-40, 364; 4:121; 5:69, 260, 575; 6:166 (quotation at
4:122-23).
41. Ibid., 3:341 and compare passage in the text just below.
42. See, respectively, ibid., 3:207; 4:19.
43. Ibid., 4:119-20 (emphasis added). Compare 3:341, 354.
44. See especially Jordan, “Religion in the Public Square,” 694-95; also Stewart, Opinion
and Reform, 280, and Moral and Political Philosophy, 283; Baier, “Hume on Different
Species,” 10; Herdt, Religion and Faction, 230; Livingston, Philosophical Melancholy, 228;
Donald T. Siebert, The Moral Animus of David Hume (Newark: University of Delaware Press,
1990), 110f.
45. Hume, History, 4:120 (emphasis added).
46. Ibid., 6:170. On the other hand, toleration can lead to endless demands for accommodating ever increasing religious divergences (5:171-72). Hume, as often, finds these first-order
prudential arguments all plausible and therefore of limited use for policy.
47. Ibid, 3:289; compare 3:136.
48. Ibid, 4:120. Regarding the attempt of a committee after Henry VIII’s death to establish
a ceremonial yet vernacular liturgy very close to the later Elizabethan worship, Hume praises
its members’ moderation—but notes in hindsight that in hoping to please all Christians, “they
flattered themselves.” In the event, the liturgy could only be imposed through extreme persecution, including burnings; even this only produced superficial conformity (ibid., 3:364,
364-67).
49. Ibid., 6:165-66.
50. Hume’s analysis of parties preceded modern ideologies but anticipated them. For a
spirited if speculative treatment see Livingston, Philosophical Melancholy, esp. 256-89.
51. These opinions for a time were High Church and anti-Puritan, but later shifted to panProtestant and anti-Catholic. See Hume, History, 6:172, 176, 191 and especially 6:420.
52. Hume, “Of the First Principles of Government,” in Miller, ed., Essays, 32.
53. Hume, History, 5:31; emphases added.
54. Ibid., 6:332-95, esp. 345, 347-48, 353, where Hume implies that parliament’s fears
were irrational, immoderate, and sectarian—but sincere. Hume likewise judges that James I’s
and Charles I’s resistance to the most virulent forms of anti-Catholicism weakened them
politically: ibid., 5:32, 62, 115, 253.
55. Ibid., 5:348, 392, and 4:395nM.
56. Ibid., 6:328. (“Unconscious” is not Hume’s language.) Other examples of purely imagined threats by religious dissenters include 6:176, 191, 211.
57. Ibid., 6:340; italics in original. Note Hume’s casual mention that Catholics are a threat
to “pagan” and “mahometan” regimes: Hume is underlining his determination to shine light
on a purely political threat in terms that would gratify no Protestant sectarian.
58. Hume endorses exclusion of Catholics from the throne precisely on the grounds that
James II was not personally fanatical but still posed, because of the inevitable thrust of
Catholic intolerance in a country less than 1 percent Catholic, a threat to British liberty. Ibid.,
6:521, and see n18 above.
59. Ibid., 4:54. Note that “specious” at the time meant “plausible.”
Sabl / The Last Artificial Virtue
537
60. T 3.2.3.3-4, SBN, 502. Hume notes similarly at T.3.2.8, SBN, 489 that our “ideas of
morality, instead of providing a remedy for the partiality of our affections, do rather conform
themselves to that partiality, and give it an additional force and influence.” In Russell Hardin’s
nice gloss, “we think that a transgression of our partiality is immoral.” Russell Hardin, David
Hume: Moral and Political Theorist (Oxford; Oxford University Press, 2007), 138.
61. Hume, History, 6:166-67; quotation on 164.
62. Ibid., 5:330; emphasis added.
63. Ibid., 6:79.
64. Jordan, “Religion in the Public Square,” 713; Stewart, Moral and Political Philosophy,
284; and esp. Whelan, “Church Establishments,” 185.
65. A particularly clear statement is in the EPM, appendix 3.4f., SBN, 304-05; see the fine
summary in Hardin, David Hume, 137-39.
66. “Cool reflection”: Hume, History, 5:460; see also 1:359; “civil interests”: 5:483, 526;
“general happiness”: 5:458. See also 5:450, 453, 575nEE. To be a constitutional legislator in
general requires “large thought” and “comprehensive views” (6:39).
67. Ibid., 6:38.
68. Ibid., 3:69-73. Compare 5:264. Herdt, Religion and Faction, 80, notes that “enlarged”
views are associated with a process of modifying and correcting our sympathies rather than
taking them as is.
69. Hume, History, 4:54 (emphasis added).
70. As suggested by Jordan, “Religion in the Public Square,” 699-700.
71. T 3.2.6.9, SBN, 532; 3.2.7.3-4, SBN, 535.
72. Hume, EPM, 9n57, SBN, 274-75n.
73. On this anti-Whiggish aim, see Forbes, Hume’s Philosophical Politics; Livingston,
Philosophical Melancholy, chap. 10. Herdt’s analysis is particularly outstanding: Religion and
Faction, 191-213. To be sure, Herdt believes Hume’s antireligious program went too far.
74. See the essays by Willem Frijhoff and Samme Zijlstra in Hsia and Nierop, pp. 34, 114.
In discussing the Duke of Ormond’s rule in Ireland, Hume proposes intensive monitoring for
political threats as enabling toleration under conditions of intense and violent sectarian strife;
only because Catholics were “guarded with so careful an eye” were “the most timorous protestants” willing to abide them (Hume, History, 6:409-10).
75. See Yael Tamir, “Remember Amalek: Religious Hate Speech,” in Nancy L. Rosenblum,
ed., Obligations of Citizenship and Demands of Faith (Princeton, NJ: Princeton University
Press, 2000), 321-34, who draws first-order Humean conclusions: the use of religious doctrines to claim divine sanction for violence against a cursed group, however abstract and theoretically these claims are worded, should probably be suppressed.
76. Montesquieu, Spirit of the Laws, 487-88.
77. Thomas Nagel, “MacIntyre versus the Enlightenment,” in Other Minds (Oxford;
Oxford Univ. Press, 1995), 208; for a similar conclusion through a very different argument,
see David Gauthier, “David Hume, Contractarian,” Philosophical Review 88, no. 1 (January
1979): 35-36. Hume himself showed much more awareness that different institutions might be
appropriate for different societies in the case of religion than in that of property. I thank
Alastair Macleod for stressing this to me.
78. On the other hand, Hume seems to praise Charles II’s Restoration promise of “liberty
of conscience,” granted as a matter of indulgence and subject to parliamentary deliberation
(Hume, History, 6:139).
79. Ibid., 6:4.
80. Ibid., 3:211-12.
538
Political Theory
81. Ibid., 6:3f.; compare 3:136.
82. Ibid., 6:144-45.
83. Hume, “Of Superstition and Enthusiasm,” in Miller, ed., Essays, 78; Hume, unpublished Preface to second edition of the History of England, quoted in full in Ernest Campbell
Mossner, The Life of David Hume (Austin: University of Texas Press, 1954), 307.
84. Hume, History, 3:212 (emphasis added).
85. Melissa Williams, “Tolerable Liberalism,” in Avigail Eisenberg and Jeff SpinnnerHalev, eds., Minorities within Minorities: Equality, Rights, and Diversity (Cambridge
University Press, 2005), 36.
86. Hume, History, at 6:191 endorses the point, also made by Locke and Bayle, that “the
malignity of the sectaries” had as a chief cause “the restraint under which they laboured.”
87. Rawls, Political Liberalism, 54-63.
88. See Melissa Williams’s argument for grounding toleration in the attempt to forestall
“the force of our own fearfulness”: p. 23 (emphasis in original).
89. See, e.g., Alan Ryan, “A More Tolerant Hobbes?” in Mendus, ed., 37-59.
90. Hume, History, 4:223; see also 4:187-88.
91. Conspirators convinced that a heretical government did not deserve their allegiance
plotted to blow up the houses of parliament—in the Netherlands, in 2003. Ian Buruma, Murder
in Amsterdam (New York: Penguin Books, 2006), 211.
92. Hume, History, 4:265; see also 4:131. Elizabeth also at times practiced an unnecessarily “violent persecution” of Catholics. But this backs up Hume’s theoretical point: this persecution was impolitic and fostered sedition (4:205-09, 288; quotation at 206).
Andrew Sabl is associate professor of public policy and political science at UCLA. His
research is on problems of toleration and pluralism, democratic theory, political ethics, and the
thought of David Hume. He is the author of Ruling Passions: Political Offices and Democratic
Ethics (Princeton, 2002) and of articles in Political Theory, the Journal of Political
Philosophy, the American Journal of Political Science, NOMOS, Polity, the Journal of Moral
Philosophy, and other publications. He is currently writing a book for Princeton on the political theory of Hume’s History of England.