Covers M 3.1–3.21
Why there is justice in society
Justice is not a permanent feature of the moral universe. It exists because a particular combination of human nature and material conditions makes it useful, and Hume proves this by changing the conditions one at a time and watching justice disappear.
- That justice is useful is too obvious to argue. The bold thesis is that public utility is the sole origin of justice and the sole foundation of its merit, which makes the claim testable: find one case where justice is useless yet binding, and the thesis fails. M 3.1
- The method is a controlled experiment. Change one variable, hold human nature fixed, and see whether justice survives. The more unusual the angles tried, the more robust the result. M 3.2–3M 3.17
Abundance
Moderate scarcity
Extreme necessity
Perfect benevolence
Golden ageThe poets remove storms from nature and from the heart, and leave out "mine and thine". M 3.14
Universal loveNo need for contracts or landmarks; families and marriage come close. M 3.6–7
Limited generosity
Effortless plentyTaking costs nobody anything, so property is "an idle ceremonial". M 3.2–5
Our worldPartial to ourselves but able to learn; few goods from nature, many from labour. Justice is useful here. M 3.12–13
Shipwreck, siegeSociety is perishing anyway, so the rules give way to self-preservation. M 3.8
Perfect malice
Ruffians, warJustice kept alone protects no one; it lapses where the other side keeps no rules. M 3.9–11
State of natureThe philosophers' war of all against all: no law, no property. M 3.15–16
- Under abundance, taking my apple costs me nothing because I pick an identical one. Property protects against injury, and where no injury is possible there is nothing to protect. M 3.2–3
- Under universal benevolence, every person is "a second self" to every other; contracts and boundaries exist only because interests diverge. Communal experiments in moments of zeal returned to separate property as selfishness returned, on Hume's account. M 3.6–7
- Under extreme necessity, "no greater evil can be dreaded from violence and injustice", so seizing a plank in a shipwreck is no crime, and magistrates may open private granaries even in lesser emergencies. M 3.8
- Justice is reciprocal. A virtuous man among ruffians must arm himself with whatever sword is at hand, and the laws of war themselves lapse against an enemy who keeps none. M 3.9M 3.11
- Our actual condition is the medium: limited generosity plus scarce goods that "art, labour, and industry" can multiply. Hence property, hence the usefulness of justice, "and hence alone arises its merit and moral obligation." M 3.12–13
- Hume doubts the state of nature ever existed, since people are always born into families with some rules, but if it did, justice would be suspended there. M 3.16
The circle of justice widens with interdependence
For discussion
- Unique goods (a particular view, a gift from someone you love) have no perfect substitute. Would property survive in the abundant world, shrunk to whatever remains scarce? M 3.3
- The kibbutzim kept common property for decades, then moved towards private ownership. Does that count for or against Hume's "ardour" that cools? M 3.7
- Rawls borrows Hume's "circumstances of justice" to say when questions of justice arise. Could questions of fair treatment arise even in the golden age? M 3.12–13
- Did trade ever contract the circle, as in the slave trade? M 3.21
Possible answers
- Unique goods. Property would survive but shrink to whatever stays scarce: unique objects and places, and also time, attention and status. That supports Hume rather than refuting him, since justice tracks scarcity wherever it appears. M 3.5 already shows rules attaching to particular scarce goods (wells) and not to plentiful ones (land).
- The kibbutz. It counts both ways. Decades of common property show that strong shared commitment, small size and mutual monitoring can sustain sharing far longer than an "ardour" that cools. The later move to differential pay and private ownership fits Hume's mechanism of returning self-interest. The fairest verdict is that Hume overstated the speed, not the direction. Also, the kibbutz was never rule-free: it ran on its own internal rules.
- Rawls. Hume's claim is functional: justice is useful only under moderate scarcity and limited generosity. Rawls uses the same conditions to say when questions of justice arise. In a golden age, property questions vanish, but questions of fair treatment (respect, turns, standing, recognition) might remain. Hume's result is easier to reach because his "justice" is so narrow.
- The slave trade. Commerce with outsiders coexisted with treating some of them as property. Interdependence extended justice to trading partners, not to those who were traded. That is the logic of M 3.18–19: justice reaches those whose cooperation one needs. The case shows the limit of Hume's optimism. Trade can widen the circle, and an existing circle makes trade possible, but neither guarantees inclusion of the powerless.
Suggested lines of argument, not settled answers.
Covers M 3.4–5, M 3.22–3.34, with M App3.10 note
Why property rights
Part I explained why possessions must be separated at all. Part II asks a second, separate question: which rules should separate them? The answer has the same source. "The good of mankind is the only object of all these laws." M 3.22
Property follows scarcity, not importance
Air and waterMost necessary, yet plentifulLeft in common; no lavish use is unjust. M 3.4
Land in empty, fertile countriesPlentifulTreated "on the same footing" as air. M 3.4
Wells in Abraham's landScarce, while land was notProperty in water but none in land: the reverse of Hume's Britain. M 3.5
The seaFree while inexhaustibleHume predicts claims of exclusive dominion once its advantages become exhaustible. M 3.4
- Merit-based allocation fails. A designer ignorant of human nature would give the most to the most virtuous. But merit is obscure and everyone overrates his own, so "no determinate rule of conduct" results. A rule best "in speculation" can be "totally pernicious" in practice. M 3.23
- Equality is weighed, then rejected. Hume states the case for equality at its strongest before answering it.
What Hume concedes
- Nature is generous enough that equal shares, improved by industry, would give everyone the necessaries and most comforts. M 3.25
- Departing from equality, "we rob the poor of more satisfaction than we add to the rich": diminishing marginal utility, a century early. M 3.25
- Sparta and the agrarian laws of Rome and Greece show equality was partly practised. M 3.25
Why he rejects perfect equality
- Instability. Different degrees of "art, care, and industry" break any equal division at once. M 3.26
- Incentives. Checking those virtues makes want "unavoidable to the whole community". M 3.26
- Tyranny. A "rigorous inquisition" would be needed, yet levelling property also levels the authority required to run it. M 3.26
Four rules common sense finds useful
What you make or improve is yoursto encourage useful habits of industry
Property passes to children and kinfor the same useful purpose
Property can be transferred by consentto make commerce possible
Contracts and promises must be keptto secure mutual trust
- Even natural-law writers, whatever principles they start from, end by appealing to "the convenience and necessities of mankind", and a concession made against one's own system carries extra weight. M 3.29
- When several rules are equally useful, "the slightest analogies" pick one, because having no rule would breed endless dispute. First possession is the standard example, a focal point in Schelling's later sense. M 3.31
- When utility and analogy both fail, civil law decides: reason cannot say how many years of possession should give title, so legislators set different periods for different assets. M 3.33
- In extraordinary cases property yields to "the safety of the people", because it was only ever respected because public interest "commonly" requires it. M 3.32
- Civil laws adapt the rules to each society's government, manners, climate, religion and commerce (Montesquieu), but their only aim remains the interest of society. M 3.34
- Hume separates the need for some stable division, which is "obvious, strong, and invincible", from the rules assigning particular things to particular people, which often rest on slight considerations of utility, humanity or imagination. M App3.10 note
For discussion
- Is inheritance the weakest of the four rules? Hume himself later says it rests largely on "connexions of the imagination". M 3.28M App3.10 note
- Is Hume's real position on equality a rejection or a tradeoff? His essay "Of Commerce" (1752) argues that great inequality weakens a state. M 3.25–26
- Where do modern merit-based schemes (performance pay, grants, admissions) run into the two problems of M 3.23?
- Property is inflexible in ordinary cases yet overridable in emergencies. Is the line between a sympathetic exception and suspending the whole system stable? M 3.32
Possible answers
- Inheritance. It is the weakest of the four on incentive grounds. Whether bequest motives drive effort is contested, inheritance may blunt heirs' incentives, and Hume himself rests it partly on imagination and notes that it exists where the incentive argument cannot apply. Its defence is security of family provision and the danger of the alternative: confiscation at death would need the enforcement apparatus M 3.26 warns against. So the utility question is about degree (estate taxation), not existence.
- Rejection or tradeoff. Tradeoff. All three objections in M 3.26 bite at perfect equality, while the diminishing-marginal-utility concession of M 3.25 stands, and "Of Commerce" favours broad enjoyment of the fruits of labour. A utility-based theory implies an interior optimum, something like Okun's leaky bucket. What Hume adds is a caution: because stable titles matter so much, redistribution should work through general rules such as taxes, not reassignment of particular holdings.
- Merit-based schemes. All three meet both problems. On information: performance pay rewards noisy, gameable measures (multitasking, teaching to the test); grant review shows low reviewer agreement and a Matthew effect; admissions rely on proxies that track preparation and money. On self-conceit: most people think themselves above average, so honest rankings read as unfair and invite influence activities, appeals, and disputes over what counts as merit. They survive by doing what Hume recommends. They allocate on verifiable proxies under fixed rules, keep stakes bounded within a stable property system, and secure finality, sometimes by lottery among fundable proposals, which is M 3.31 in action.
- Inflexible yet overridable. The distinction holds if the trigger is itself rule-governed. An exception for a sympathetic individual invites every claimant to argue he is special (the Cyrus problem). An emergency suspension responds to a public condition, applies generally and is temporary. It becomes unstable when whoever benefits also declares the emergency. Hence constitutional limits and compensation for takings, which preserve expectations even when property is overridden. Hume himself offers little protection here.
Suggested lines of argument, not settled answers.
Covers M 3.9–3.11, M 3.18–3.19
Justice and power
The most uncomfortable consequence of the theory. If justice exists because mutual restraint benefits both sides, it reaches only those who can make restraint worth the other side's while.
JusticeEach can harm the other, so each gains from restraint. "Society" in Hume's sense "supposes a degree of equality". M 3.18
Rights won by leverageWomen, reduced to near slavery in many nations, usually "break the confederacy" through their own influence. M 3.19
Humanity onlyRational but powerless beings are owed "gentle usage", but no justice and no property; the relation is command and obedience. M 3.18
- Imagine rational creatures too weak ever to make us feel their resentment. Restraint brings the stronger party no benefit, so justice has no function; only the separate duty of humanity remains. M 3.18
- This is "plainly" our relation to animals. Europeans' sense of superiority "tempted us to imagine" the native peoples of the Americas in the same position, and so they threw off justice "and even of humanity", which Hume presents as a moral failure. M 3.19
- Read that sentence carefully but not generously: Hume also wrote a notorious racist footnote in "Of National Characters". The passage is not evidence of a general anti-racist stance. M 3.19
- His account of women is condescending in tone but consistent in logic: rights expand as bargaining power expands. M 3.19
- Reciprocity sets the same limit between enemies. Justice lapses among ruffians, and even the laws of war, which are themselves rules of mutual advantage, lapse against a side that keeps none. M 3.9M 3.11
- Inside society, punishment applies the logic to the state's own power: for a criminal the ordinary rules are "suspended for a moment", for the benefit of society. M 3.10
- Later writers call this view "justice as mutual advantage" (Gauthier, 1986), contrasting it with "justice as impartiality" (Barry, 1989).
For discussion
- Nussbaum (2006) argues mutual-advantage theories fail exactly here: people with severe disabilities, animals, distant foreigners. Is Hume's duty of "humanity" enough, or does the weaker party need rights? M 3.18
- Rights have historically expanded with leverage (the franchise, married women's property, labour rights). Does that confirm Hume as history? Does it make him right as a norm? M 3.19
- The Geneva Conventions bind "in all circumstances" and forbid reprisals. Rejection of Hume's reciprocity, or a firmer rule that cannot unravel? M 3.11
Possible answers
- Humanity or rights. Humanity alone is fragile. It depends on the stronger party's sympathy, which Hume admits weakens with distance (M 5.41) and yields to interest. Rights give standing that does not wait on anyone's feelings. A Humean reply is that rights for the weak can be useful to everyone: anyone may become old, ill or disabled, and the general point of view can extend approval beyond those who can retaliate. Still, pure mutual advantage leaves a gap that many think only an impartiality principle fills.
- History versus norm. As history, there is real support: rights often expanded with organisation, votes and economic weight. As a norm, no, since how rights spread does not show how they ought to. Two complications: Hume himself calls the abandonment of humanity a failure (M 3.19), and many expansions were driven by people who had nothing to gain, such as abolitionists. That is sympathy at work, which Section V would predict.
- The Geneva Conventions. They are best read as a more sophisticated application of Hume. A reciprocity-contingent rule unravels, because each side claims the other violated first and reprisals escalate, which is the vault losing a stone. An unconditional rule is more useful than a contingent one, like the inflexible rules of App3.6. But this does go beyond the letter of M 3.11, which explicitly allows dropping the rules against an enemy who keeps none.
Suggested lines of argument, not settled answers.
Covers M 3.1, M 3.40–3.48, M App3.7–8
The main source of justice
Hume poses a dilemma. Our sense of justice either comes from reflecting on its usefulness, or it is a simple instinct "like hunger, thirst, and other appetites". He eliminates the instinct and leaves utility standing. M 3.40
- The thesis has two parts. Origin: the institution exists because it serves the public good. Merit: we approve of just conduct because we reflect on those consequences. M 3.1M 3.48
- Property is not one relation but many: occupation, labour, prescription, inheritance, contract. An instinct for it would need "ten thousand different instincts" for ideas that a thousand volumes of commentators have failed to define. M 3.41–42
- Kings, senates and judges change property, and we accept even mistaken verdicts for the sake of peace and finality. Nobody has "innate ideas of prætors and chancellors and juries". M 3.43
Instinct: all birds of a species build the same nest, in every age and country. M 3.44
Reason and custom: houses differ across times and places, yet all have roofs, walls and windows because they serve the same needs. So do property laws. M 3.44–45
- Property rules are full of "finer turns and connexions of the imagination" that no instinct could produce. M 3.46
- The one real difficulty: through education we condemn injustice automatically, without feeling any calculation. Habit hides the origin, yet in daily life we keep asking "What must become of the world, if such practices prevail?" M 3.47
- Justice arises from convention, but not in the sense of a promise, since we cannot be bound to keep promises by having promised to. It is a sense of common interest, felt by each, observed in others, and carrying everyone into a general plan. M App3.7
- Since utility shows such force in the case of justice, Newton's rule ("to the same effects, the same causes") lets Hume extend it: it is the sole source of approval for justice and fidelity and a large part of the merit of benevolence. M 3.48
For discussion
- Maynard Smith's "bourgeois strategy" and the endowment effect suggest a simple instinct favouring possession may exist. Does Hume's argument survive in a narrower form? M 3.41
- "What if everyone did that?" sounds Kantian, but Hume asks about consequences. Which reading fits how people actually use the question? M 3.47
- Is extending utility from justice to benevolence a legitimate use of Newton's rule, or an overextension? M 3.48
Possible answers
- A possession instinct. Hume's argument survives in a narrower form. A natural bias favouring the possessor may explain the starting points (present and first possession), which Hume himself says the imagination favours (App3.10 note). It cannot explain prescription periods, inheritance, contract or the authority of courts. The modern picture of conventions built on a natural focal point, in Sugden and Skyrms, is Humean in spirit.
- "What if everyone did that?" Most everyday uses are consequentialist: people point to the disorder that would follow. Some uses are about fairness: "you are making an exception of yourself", said even when one person's act has negligible effects, which is closer to Kant. Hume's App3.3 bridges the two: utility belongs to the scheme, so the free-rider is blamed for exempting himself from a useful system.
- Newton's rule. Partly legitimate. Approval of justice and approval of benevolence are similar effects, so assigning them a common cause is reasonable. Hume claims sole for justice but only a considerable part for benevolence, which shows he sees the risk. The overextension would be to deny that benevolence also pleases immediately, as App3.2 and his later sections on agreeable qualities allow.
Suggested lines of argument, not settled answers.
Covers M 3.30, M 3.35–3.39 with notes
Social constructs and superstition
Hume sets a trap for his own theory. The reasoning that exposes superstition seems to expose property just as well: in neither case does anything in the object explain the strength of our feeling.
- "Uninstructed nature, surely, never made any such distinction" as mine and yours. Objects are physically separate from us; only "the general interests of society" connect them to an owner. M 3.30
- What is property? Whatever it is lawful for one person alone to use, as fixed by statutes, customs, precedents and analogies. Ignore their common aim and the laws of property look "whimsical, unnatural, and even superstitious". M 3.35
| Superstition M 3.36 | Justice M 3.37 |
|---|---|
| A Syrian would starve rather than eat pigeon; an Egyptian would not approach bacon. | I may eat from this tree, but the fruit of an identical tree ten paces off is criminal to touch. |
| Fowl is lawful on Thursday and abominable on Friday; eggs are allowed in this diocese and damnable a hundred paces away. | An hour ago, wearing this coat deserved punishment; "a few magical syllables" (a sale) have made it mine. |
| A building profane yesterday is holy today, after certain words are muttered over it. | The same house is lawful to live in on this side of the river, but not across it. |
| Examine the objects by sight, smell, taste, chemistry or medicine: no difference is found in either column. M 3.36–37 | |
| Frivolous, useless, burdensome. | "Absolutely requisite to the well-being of mankind and existence of society." M 3.38 |
- The debunking points at superstition are "too obvious to have any influence". Superstitions persist not through ignorance but through "education, prejudice, and passion". M 3.36
- Strip away usefulness and a sale is as unintelligible as a liturgy that makes bricks sacred. Usefulness is the whole difference. M 3.38
- Promises show this most clearly. Will alone binds no one; words must express it. Once spoken, words bind even against a secret contrary intention, but not when used without understanding or in evident jest. These rules look inconsistent if promising is about inner intention and make sense if it exists to secure reliable expectations. M 3.38 note
- The casuists' "secret direction of the intention" outraged people because society could not survive it. Notably, nobody applies the doctrine of the priest's secret intention to civil contracts, which shows that even dogmatic superstition never wins full practical belief. M 3.38 note
- Debunking strengthens rather than weakens justice: "what stronger foundation can be desired" than that society, even human nature, could not subsist without it? M 3.39
The point in one sentence
Property is a social construct in the same sense as a sacred building (an "institutional fact", in Searle's later term), but a construct that sustains society is not a superstition, and seeing why it exists gives the strongest reason to respect it. M 3.38–39
For discussion
- Once people see property as a convention, might some feel freer to break it when unobserved, as Hume's "sensible knave" in the Enquiry's Conclusion would? M 3.39
- Are there features of contract law that protect something other than reliance, such as the promisor's autonomy or fairness? M 3.38 note
- Money and marriage are also constructs. Where would Hume place a contested convention whose usefulness is disputed?
Possible answers
- The sensible knave. Some might. Seeing the function of a system gives reasons to support the system, not to keep every rule when unobserved; the knave can still free-ride. Understanding strengthens commitment in those who already care about society through sympathy, but does little for those who do not. Hence the role of enforcement in the Coda, and Hume's reply in the Conclusion that knaves usually get caught and lose the peace of mind integrity brings.
- Beyond reliance. Some doctrines seem to protect fairness or autonomy: refusing to enforce unconscionable bargains or agreements made under duress, and the freedom not to contract at all. Many can be recast in Humean terms as useful general rules. Bargains extracted by pressure, for instance, undermine the trust that promising exists to secure. Whether every such doctrine reduces to utility is the open question.
- A contested convention. Hume's test is usefulness on the whole, which is an empirical question and may vary by society (M 3.34). A convention that is "frivolous, useless, and burdensome" deserves the treatment of superstition. But stability is itself valuable, so the case for reform must outweigh the cost of unsettling expectations. Changes in marriage law show conventions being reassessed against what they actually do for people.
Suggested lines of argument, not settled answers.
Covers M 5.1–5.47
Why utility pleases
Section III showed that we approve of justice because it is useful. Section V asks why usefulness should please us at all, especially when the use goes to other people. Hume's answer is sympathy.
- Usefulness pleases everywhere, even in objects: a ship with a swelling prow looks better to a sailor than a geometrically regular one, and square doors would "hurt the eye" because they do not fit the human figure. M 5.1
- Yet a useful person excites affection and esteem, while a useful object does not; we call herbs "virtuous" only by "the caprice of language". M 5.1 note
- Morality is not an invention of politicians. Education can strengthen or redirect moral feelings, but without a natural basis words like "honourable" and "shameful" could never have been understood. M 5.3–4
- So the end that the social virtues promote must please us, either through self-interest or "more generous motives". M 5.4
- The selfish theory was "excusable": private and public interest usually move together, so everyday cases cannot tell the two explanations apart. M 5.5–6M 5.16
The crucial experiment
| Case | Self-interest predicts | What we observe |
|---|---|---|
| Virtue in distant ages and countries M 5.7M 5.10–11 | Indifference | Approval, without even asking where or when |
| A noble deed by an adversary M 5.8 | Disapproval | Approval, as the exiled Athenian admired his generous enemy |
| Virtue that also benefits us M 5.9M 5.17 | Approval | Approval plus a distinct extra feeling: gratitude (or, when harmed, revenge) |
- The reply that we imagine ourselves as contemporaries fails: a real feeling cannot arise from an interest known to be imaginary. Fear of heights fades with habit, but moral discrimination grows sharper with practice. M 5.13–14
- Conclusion: "the interests of society are not, even on their own account, entirely indifferent to us", since nothing pleases as a means when its end leaves us cold. Why we have fellow-feeling at all is a first principle Hume declines to explain further. M 5.15M 5.17 note
- Sympathy is emotional contagion: "the human countenance … borrows smiles or tears from the human countenance" (Horace). Moved by misery, we cannot be indifferent to its causes. M 5.18
Four scenes, rising feeling
A warm apartmentpleasure in ideas of ease M 5.19
A happy family"a pleasing sympathy" M 5.20
An inheritance threatenedimmediate indignation M 5.21
Provinces enslavedhorror and antipathy M 5.22
- Theatre, pastoral and history show sympathy where self-interest is plainly absent. Crowds amplify it; indifferent characters on stage damp it; fishermen's toil spoils Sannazaro's seaside pastoral. M 5.24–29
- Intensity depends on scale (the Athenian army at Syracuse versus Pisa's petty wars) and on vividness (Tacitus versus Suetonius on the same tyrants). M 5.33–34
- Sympathy has a dark side: "popular sedition, party zeal, a devoted obedience to factious leaders" come from the same source. M 5.35
- Benevolence need not be strong, only present. Where self-interest is silent, even a weak concern for others decides: nobody treads as willingly on a stranger's gouty toes as on the pavement. M 5.39
- Pure disinterested malice probably exists in no human heart; a creature that had it would invert every moral sentiment. M 5.40
The general point of view
- We judge by tendencies, not accidents: a peach tree whose fruit the snails destroyed is still a good tree, and good intentions defeated by fortune earn the same general praise. M 5.41 note
- Why correct our feelings at all? Because language is "formed for general use". People in different positions could never converse about character unless they adopted "some general unalterable standard". The heart does not fully follow it, but it serves "in company, in the pulpit, on the theatre, and in the schools". M 5.42
- Virtue far away is "like a fixed star": as bright as the sun to reason, but giving neither light nor heat until acquaintance or eloquence brings it close. M 5.43
- Two routes converge. A priori, creatures like us cannot be indifferent to others' welfare; a posteriori, utility is the observed source of praise, and the sole one for justice, fidelity and allegiance. Each confirms the other. M 5.43–46
For discussion
- Do dictator and ultimatum games, or inequity-aversion models (Fehr and Schmidt, 1999), settle the question Hume raised? They measure something narrower than "humanity". M 5.17
- Slovic's "psychic numbing" suggests concern can fall as victims multiply. Does that contradict Hume's point about scale, or fit his stress on vividness? M 5.31–34
- If sympathy also drives party zeal, what makes moral sympathy different from crowd feeling? M 5.35M 5.41–42
- Smith's impartial spectator, Harsanyi's impartial observer, Rawls's original position: what changes when the spectator becomes a formal decision-maker? M 5.41
- Hume judges a statesman by the good done to his own country, yet in "Of the Jealousy of Trade" he prayed for France's commerce. Are these consistent? M 5.38 note
Possible answers
- Experiments. They support Hume but do not settle the question. Positive giving to anonymous strangers in dictator games shows other-regarding preferences exist; giving falls when anonymity is tightened, which suggests image concerns matter too. Hume's claim is about a spectator's approval, which third-party punishment experiments track more closely: people pay to punish unfairness done to others. His claim is modest, that concern for others has some weight, and the evidence confirms at least that.
- Psychic numbing. It fits Hume. His scale examples work through vivid narrative; Syracuse is a story, not a statistic. Statistics are faint ideas and an identified victim is a lively one, so the identifiable-victim effect is Hume's vividness. The general point of view is then the correction: judgment says harm scales with numbers even when feeling does not, which is a reason for institutions that act on numbers.
- Moral sympathy versus crowd feeling. The mechanism is the same; the difference is correction. Party zeal is uncorrected sympathy with one's own group, amplified by contagion. Moral sympathy is corrected for position and partiality, attends to the tendencies of actions for all affected, and must be stable enough to share with people in other positions (M 5.42).
- Formal spectators. Smith's spectator is an imagined person whose sentiments we try to match; it still judges characters, as Hume does. Harsanyi's observer, with an equal chance of being anyone, maximises average utility; Rawls's parties behind the veil choose principles for institutions. Formalisation shifts the task from judging characters to choosing institutions, and from sentiment to rational choice. Conclusions become determinate, but now depend on assumptions such as attitudes to risk.
- The statesman and the jealousy of trade. The two are consistent wherever trade is positive-sum: a minister who promotes trade serves his country and its partners alike. The footnote concerns cases where "a competition arises", and the essay argues these are rarer than mercantilists think. The tension returns where gains really are at a partner's expense, as with a large country's terms-of-trade-improving tariff. The footnote would praise that minister; the spirit of the essay would not.
Suggested lines of argument, not settled answers.
Covers M App3.1–3.11
What Appendix III contributes
The appendix answers the puzzle left open at M 3.1. If justice rests on utility, why must we return a fortune to a miser, or let a bad heir inherit? Because utility belongs to the system of rules, not to each act. M App3.1–3
Benevolence is a wall
Each stone adds height in proportion to each worker's care. Take one away and the wall is one stone lower. M App3.5
Justice is a vault
Each stone alone would fall; the whole stands only by "the mutual assistance and combination of its corresponding parts". M App3.5
Both structures are standing.
- Benevolence acts directly on one person; each act is "compleat and entire", and a lone benefactor is admired all the more for his rarity. M App3.2
- Justice benefits society only as "the whole scheme or system, concurred in by the whole, or the greater part". A single just act can be harmful: riches inherited by a bad man are "the instrument of mischief". M App3.3
- Cyrus and the coats. The young Cyrus let a big boy swap his small coat for a small boy's big one, since each now fitted. His teacher corrected him with "general, inflexible rules". The swap improved the fit for both, but a voluntary trade could have achieved it without letting the strong reassign by force. M App3.4
- Property laws are general and blind to persons: they will take goods acquired without good title from a generous man and give them to a selfish miser. It is enough that the system's good "much preponderate" over its evil. M App3.6
- Convention means a shared sense of common interest, not a promise. Two men row a boat "without any promise or contract"; gold and silver become money and words become language in the same way. Whatever benefits all only if all perform "can arise from no other principle". Hume says Grotius's account of property is "in the main" the same. M App3.7–8
- Is justice natural? Not if "natural" means instinctive rather than artificial; yes if it means usual or non-miraculous. For "so sagacious an animal", what necessarily follows from its intellect may fairly be called natural. The dispute is "merely verbal". M App3.9
- Courts seek general rules free of favour. Precedent binds even when the first decision had no good reason, and analogy fills the gaps, so that in many fine questions there is no truth or falsehood on either side, only a need for some stable rule. M App3.10
- The assignment rules follow the imagination: small rivers belong to the owners of their banks, the Rhine and Danube to nations of "suitable bulk", slowly deposited land to the adjacent field, and land torn off suddenly only once roots have grown into both. M App3.10 note
- Once rules exist, every injustice is both a public wrong and a private harm, and the private harm powerfully reinforces our blame. Wrongs with no identifiable victim are "more lightly thought of". M App3.11
Three contributions of the appendix
- Utility at the level of rules. It resolves the tension of M 3.1: the scheme is useful even when single acts are not, which is why the rules must be inflexible (rule consequentialism; rules over discretion). M App3.3–6
- A theory of convention. Justice is a self-sustaining pattern of mutual expectations, like rowing, money and language, not a contract. M App3.7–8
- Two levels of property. The need for stable separation is invincible; who gets what is often settled by slight utility, humanity, precedent and imagination. M App3.10
For discussion
- Is a lone just person, in a society where nobody else respects property, admirable or foolish? Compare the lone benefactor of App3.2 with the ruffians of M 3.9.
- Rules or standards (Kaplow, 1992)? Are there areas of property law where flexible standards work better than Hume's inflexible rules? M App3.6
- Rowing is an assurance game, but respecting property looks more like a prisoner's dilemma. When does repeated interaction turn one into the other, and when does it fail? M App3.8
- Hume says the initial assignment is "pretty indifferent". Compare Coase (1960): indifferent for efficiency only with zero transaction costs, and never for distribution. M App3.10 note
Possible answers
- The lone just person. In Hume's extreme case it is foolish: among ruffians, justice is "no longer of use" (M 3.9). Under partial compliance it can be admirable, because a visible example can help establish a convention, and reputation still pays. The asymmetry is the point of the wall and the vault: benevolence gains value from rarity; justice depends on concurrence.
- Rules or standards. On Kaplow's logic, standards suit heterogeneous, infrequent situations: nuisance ("reasonable use"), negligence, good faith in contract, unconscionability. Rules suit recurring, high-stakes questions where predictability matters, such as title, transfer and registration. Hume's insistence on inflexible rules fits the core of property better than its margins.
- Assurance or dilemma. In a one-shot interaction, respecting property is a prisoner's dilemma. Repetition with patient players who can observe each other makes cooperation an equilibrium through trigger strategies (the folk theorem), and the game behaves like an assurance game. It fails with large numbers, anonymity, present bias (a low discount factor, as in the Coda), or a known last round. That is exactly where Hume brings in government.
- Hume and Coase. Hume anticipates half of Coase: for production, what matters is a clear and stable assignment, not who gets what. He has no account of transaction costs, under which the assignment matters for efficiency too. He also says nothing about distribution, where the assignment always matters. His "pretty indifferent" is a claim about society's need for stability, not about welfare.
Suggested lines of argument, not settled answers.
Covers T 3.2.7.8 (optional reading)
Coda: two neighbours and a meadow
The Enquiry shows how conventions of justice arise. This passage from the Treatise shows where they stop working, and why government is needed.
Two neighbours may agree to drain a meadow, which they possess in common … But 'tis very difficult, and indeed impossible, that a thousand persons shou'd agree in any such action. T 3.2.7.8
Two: each knows the other's mind, and each sees that if he fails, the whole project is abandoned.
A thousand: hard to plan, harder to execute, "while each seeks a pretext to free himself of the trouble and expence, and wou'd lay the whole burden on others."
- Government first gives security in the execution and decision of justice, protecting us against others' weakness and passion, and against our own. T 3.2.7.8
- The root flaw is present bias: no quality causes "more fatal errors" than preferring "whatever is present to the distant and remote", judging objects by their nearness rather than their value. The ruffians of M 3.9 showed the same "blindness to future consequences". T 3.2.7.8
- Government does more than protect existing conventions: it obliges people to make them and to pursue common ends. T 3.2.7.8
- Magistrates solve both problems. Their own interest is bound to that of many subjects, they need consult no one to form a plan, and any failure in execution harms their interest, so they prevent it. Hence bridges, harbours, ramparts, canals, fleets and armies. T 3.2.7.8
- Government, "compos'd of men subject to all human infirmities", becomes by "one of the finest and most subtle inventions imaginable" a composition partly exempt from them. T 3.2.7.8
The main idea
Conventions of justice can arise spontaneously where groups are small and each member is pivotal. Scale and present bias break them. Government is the further invention that sustains conventions, and secures collective goods, in groups too large to manage them alone. T 3.2.7.8M 3.20–21
For discussion
- Public choice (Buchanan and Tullock, 1962) says rulers pursue their own interests too, and Hume himself wrote that "every man ought to be supposed a knave" in designing a constitution. Is the "subtle invention" precisely institutional design that makes knaves serve the public?
- Ostrom (1990) documents large communities managing commons with their own rules, monitoring and graduated sanctions. What lets them escape the thousand-person problem?
- How does this limit change your reading of Section III's optimistic expanding circle? M 3.20–21
Possible answers
- Knaves and design. Yes. The two views reconcile if the "subtle invention" is institutional design: assume knaves, then arrange things so that even knaves serve the public, which is what Hume means when magistrates "find an immediate interest" in their subjects' interest. Public choice shows the alignment fails under concentrated interests and weak accountability, so the reconciliation depends on institutions Hume only gestures at.
- Ostrom's commons. Her design principles recreate Hume's small-group conditions at larger scale: clear boundaries, rules fitted to local conditions, participation in making them, accountable monitoring, graduated sanctions, cheap conflict resolution, and recognition of the right to organise. Defection becomes visible and each member's compliance matters again. This does not refute Hume: it shows the "magistrate" can be a community institution rather than the state.
- The expanding circle revisited. The circle extends only as far as conventions can be sustained. Beyond small groups that requires institutions, and between nations there is no common magistrate. International cooperation, trade agreements included, must therefore be self-enforcing through repeated interaction. Section III's optimism holds where interdependence is strong and observable, and needs institutional support everywhere else.
Suggested lines of argument, not settled answers.
Questions
- Hume claims utility is the sole origin of justice. Which thought experiment is most convincing, and which is weakest? M 3.1–3.17
- If justice exists only where people can harm each other, what do we owe to those who cannot? M 3.18
- Hume concedes equality would raise total happiness, then rejects it as impracticable and dangerous. Rejection or tradeoff? M 3.25–26
- Is moral approval grounded in sympathy a reliable foundation, given that sympathy also fuels faction? M 5.35
- How does the general point of view compare with Smith's impartial spectator and Rawls's original position? M 5.41–42
- Two neighbours can drain a meadow; a thousand cannot. What does this imply about the limits of justice as convention? T 3.2.7.8