Methodologically, Aristotle starts off talking about injustice, then moves to justice.
He searches for definitions of justice first (and asks whether it’s a real thing), and then discusses its properties.
This is different to his usual methodology of starting with phainomena (appearances / intuitions), then aporiai (puzzles), before largely agreeing with the endoxa (common beliefs).
His taxonomy of justice has:
Specific vs general justice
Within specific justice, distributive vs corrective justice
Specific vs general justice
Aristotle starts off with the contraries, noting that unfairness seems to be a subset of unlawfulness. So fairness is a subset of lawfulness; the former is the domain of specific justice and the latter is general justice.
General justice is the whole of virtue with respect to relations with others.
He also characterises it in terms of lawfulness.
Idea: the law prescribes acts of every virtue (e.g. don’t desert your post = courage, don’t commit adultery = temperance).
So being law-abiding is exercising the whole of virtue towards others.
Specific justice is about fairness.
Aristotle starts with the endoxon that sometimes we blame someone even when there’s not an obvious vice that they’re displaying.
In these cases, the vice might be overreaching (pleonexia); a kind of special injustice where you allocate yourself more than your fair share.
Within specific justice, distributive justice is about how goods should be allocated; corrective justice is about rectification.
For distributive justice, goods should be allocated in “accord with some sort of worth” (V.3, 1131a26).
This is where Aristotle talks about geometric equality.
For corrective justice, everyone is treated as equal regardless of their worth.
“It does not matter if a decent person has taken from a base person, or a base person from a decent person”. (V.4. 1132a2-7)
So, this is about arithmetic equality.
Note that Aristotle doesn’t discuss punishment at all; the restoration here is merely compensatory.
Corrective justice is subdivided into voluntary (e.g. commercial transactions where you might be swindled by a merchant) and involuntary (for actual crime).
The latter is broken into forced [like robbery from the person] vs hidden [like theft].
There’s a distinction between acts of injustice and being unjust, analogous to virtuous actions vs being virtuous.
Aristotle gives a fourfold classification of harmful acts (V.8):
(1) Misfortune – injury contrary to reasonable expectation.
(2) Mistake – injury that could’ve been reasonably expected and where the agent was at fault, but due to some error.
(3) In knowledge but without deliberation – this is an act of injustice, and voluntary, but doesn’t make the doer an unjust person.
(4) Act done after deliberation – if the agent formed a prohairesis then did the harmful act, they’re unjust and vicious.
So you’re unjust only if you commit an act of injustice from choice.
Categories (1) and (2) involve ignorance of relevant particulars; there’s no injustice done at all, even though injury occurs.
Category (3) involves acting with knowledge, but not deliberate choice – e.g., someone acting unjustly out of spirit & passion.
This connects to blameworthiness, too.
Acting unjustly from ignorance is voluntary but pardonable (e.g. due to spirit & passion); acting merely in ignorance isn’t pardonable.
But it’s puzzling why anger diminishes responsibility yet in III.5 drunkenness increases it.
Aristotle thinks that acts of injustice done due to “natural passions” are pardonable, whereas “unnatural passions” aren’t. So this might explain why anger can partially excuse ignorance while drunkenness can’t – the latter is more voluntary.
Aristotle asserts that nobody can treat themselves unjustly (V.11).
When someone commits suicide, they are carrying out an act of injustice (since it is against the law). But Aristotle says the sufferer of injustice is the state, not the individual.
This explains why the state should punish suicide.
One justification for the view: it’s impossible to voluntarily be treated unjustly, yet the suicidal person voluntarily suffers [on Aristotle’s account of voluntariness].
He makes further arguments for the impossibility of treating oneself unjustly – e.g., it would involve the same thing being added to and subtracted from the same person.
One corollary of this: suffering injustice is not a vice, since it’s never voluntary.
This matters for discussion of justice as a mean; see below.
One doubt we might have about Aristotle’s account is that it seems like specific justice overlaps a lot with other vices.
Aristotle seems to think introducing the category of “overreaching” (pleonexia) helps to explain intuitions about how certain kinds of bad actions don’t accord with another one of the vices.
But it seems like these bad actions could in fact be subsumed under other vices! e.g., his example about doing adultery for profit rather than for appetite (V.II, 1130a24-28) – why should the vice be that of overreaching, rather than greed?
One way to put it: Aristotle has his conception of the vice of injustice wrong: it’s not driven by a characteristic motive like other vices, but comes about from the lack of good motivation – namely, fairness (Williams 1980).
So introducing pleonexia is getting things backwards.
Young (2006) is more generous, and thinks that Aristotle does view pleonexia as the absence of inhibition on the desire for gain, rather than desire for excessive gain.
Also, specific justice seems too narrow – it doesn’t designate as unjust many people who we’d intuitively consider to be so (Williams 1980).
He says specific injustice is “concerned with honour or wealth or safety and aims at the pleasure that results from making a profit” (V.2, 1130b2-4).
On Aristotle’s account of habituation, for someone to develop unjust character, they must be doing unjust actions with injustice as the goal.
But when we think of unjust people, they aren’t normally ones who actively seek out injustice!
Seems like it’s more commonly a side-effect of another vice, e.g. indifference / cowardice.
So if we really do restrict ourselves to only those acting with injustice in mind, that’s a very small group.
General justice is also puzzling, because the two definitions of it don’t seem like they cohere, given the treatment of decency (see below).
Specifically: if general justice is the whole of virtue with respect to others, then why is decency separated out? Surely complete virtue towards others already includes situation-specific practical judgement.
Although note that Aristotle says at V.10 1138a3 that decency is “a sort of justice, and not some state different from it”.
So his point is that it’s not a different virtue; it’s the genuinely just person correcting the letter of the law via phronesis.
It’s separated out as a concept, not as a state.
Decency is only really doing work if you’re using the lawfulness characterisation of general justice, since there are gaps in the law for particular cases.
(Although maybe this conflates lawful with legal. Still, it’s conceivable that the ideal laws would also have gaps.)
Justice, the law, politics
Aristotle identifies general justice with lawfulness, but it seems intuitively like being just doesn’t have a 1:1 relationship with following the law.
“Everything lawful is somehow just” (V.2)
But what about breaking laws in a way that doesn’t disadvantage or harm anyone – is that lawlessness and unjust? And can’t there be unjust laws?
e.g., you run a red light on an empty street. Then who is the victim of the injustice?
Irwin suggests there’s less of a contradiction here than it might seem, because there’s a distinction between the legal and the lawful and the fair. (V.1 footnote 9, V.2 footnote 6)
As noted earlier, fairness is a subset of lawfulness.
And when talking about the lawful Aristotle refers to the correct laws, not merely the actual laws – for those are fallible in their moral prescriptions.
So perhaps there are things which are illegal but not lawless, and therefore not unjust, like running that red light. (Similarly, things might be legal but unlawful)
The wise lawgiver, when laying down laws, can think about the fact that not all citizens are just/wise, and chooses their laws accordingly.
And this means that laws must be universal.
Cf the methodological point in I.3 that we should not expect too much precision from the political science.
Consequently, law-abidingness alone is not the fullest kind of justice possible; decency (epieikeia) goes beyond to adjust the law to particular circumstances.
Specifically, the equitable person “does what the lawgiver would have done had he been present”.
Aristotle says the equitable is “superior”, “better than one kind of justice”.
Aristotle picks out two components of political justice: natural justice and legal justice.
The former “has the same force everywhere”; the latter is laid down by humans (e.g. the price of a ransom).
This is a rebuttal to the conventionalist view that justice is entirely constituted by social arrangements.
Even though there’s variation in what communities call just, that doesn’t mean that there’s no natural fact of the matter.
Compare: people can be ambidextrous, but the right hand is naturally stronger.
Although the conception of worth used to distribute goods varies by political system, Aristotle thinks there is an objectively correct political arrangement.
“There is but one [constitution] which is everywhere by nature the best” (V.7)
In oligarchy worth is based on wealth; in democracy it’s for all citizens. (V.3, 1131a26-29). Aristotle says aristocracy is correct, for it distributes goods in proportion to excellence (see his Politics).
This connects to the objectivity of the virtues and the function argument: Aristotle’s view is that the virtues are natural excellences of the human soul.
So he doesn’t think that they can vary between communities, since the human ergon is fixed.
But then justice, as a virtue, mustn’t vary either.
Some relationships are not included within political justice proper, because justice can only exist among free people who bear certain relationships of equality with each other.
In particular, Aristotle excludes the master-slave, father-child, and husband-wife relationships from political justice; they only have justice by analogy (V.6).
So his supposedly natural justice is (of course) shaped by his own social context and assumptions.
Justice as a virtue
The biggest problem with justice as a virtue is that it doesn’t really seem to be a mean!
It looks like justice fails the location criterion for a mean: it’s not between two corresponding vices.
It can’t be a mean between vices of suffering injustice <> doing injustice.
Recall that Aristotle says suffering injustice involves no vice nor injustice.
It can’t be a mean between vices of taking too little <> taking too much.
Aristotle says that it’s excellent to take less than you might deserve: “the decent person tends to take less than his share”. (V.9, 1136b20-22)
Aristotle argues that justice is a mean because it’s about intermediate conditions, though this seems rather vacuous.
“Justice is a mean, not as the other virtues are, but because it is about an intermediate condition, whereas injustice is about the extremes.” (V.5, 1134a1)
For each kind of specific justice:
Distributive justice seeks a mean between the distributor taking too much and giving too much.
Corrective justice seeks the mean between profit and loss.
Even if we grant that justice is a mean about actions, it’s not really a mean about feelings, so the intermediacy thesis partly fails (as well as the location thesis).
This is because justice’s mean applies primarily to external distributions, not the agent’s feelings.
This helps illustrate some of the problems with the Doctrine of the Mean: ad-hoc-ness, triviality and non-action-guidingness.
Aristotle talks about “too much” and “too little” with respect to what is deserved – but obviously this trivialises his claim that the mean is what is best.
He’s simply defining it relative to two evaluatively suboptimal outcomes!
For all the mathematical analogies about geometric and arithmetic ratios, Aristotle doesn’t talk about how to determine the just amount (Winthrop 1978).
And indeed, the talk of ratios is at odds with calling the political science an imprecise one by its nature.
Justice is an unusual virtue in other ways too: it’s unclear how it benefits you, and it seems to admit of degrees in virtue. Aristotle does concede this himself (somewhat grudgingly).
Given that “justice is the only virtue that seems to be another’s person’s good” (V.1, 1129b25-27), it’s not obvious how it constitutes one’s own eudaimonia.
This is especially challenging for the decent person, who takes less than their fair share.
Aristotle’s solution seems to be arguing that they’re compensated for the smaller share of material goods with a greater allocation of honour or nobility/fineness (V.9).
Cf IX.8 discussion of how the virtuous person gives away money and honours to friends, and in doing so attains for himself more of the fine.
(This also raises familiar egoism concerns, as well as whether to kalon really is a genuine prudential good.)
But this runs into difficulty: how do you balance the incommensurable goods involved such that the decent person ends up with what is fair / the mean?
Aristotle asserts that the best person “exercises [virtue] in relation to another, since this is a difficult task” (V.1, 1130a3-5). But surely the mere difficulty does not explain why it is a valuable thing.
(maybe it’s for the fine / noble in some way, but it’s not really argued for)
How can decency (epieikeia) be a superior version of justice, if ethike arete involves phronesis already and the distinguishing feature of epieikeia is just that it adjusts actions to the particulars of a situation?
Either ordinary justice already involves phronesis and epieikeia is superfluous, or ordinary justice is more rule-bound than other virtues.
This also creates more problems for the idea of that justice is a mean: it doesn’t make any sense at all for there to be a mean and a better-mean along a single axis.
Aristotle can offer an OK resolution to this: having complete virtue towards others guarantees phronesis. So indeed, someone who has general justice will be decent.
And decency is better than the stickler-version of being just, where you just follow the letter of the laws.
But the residual problem is that the identification of general justice with lawfulness is in tension with the other characterisation of it as complete virtue towards others – decency seems to demonstrate that lawfulness actually isn’t the whole of general justice.