The Nicomachean Ethics — trans. F. H. Peters (1881)

For in all such cases, roughly speaking, though the terms are not always quite appropriate, we talk of the doer's "gain" (e.g. the striker's) and the sufferer's " loss ; " but when the suffering has once been assessed by the court, what the doer gets is strictly called "loss" or penalty, and what the sufferer gets is called "gain."

Now, what is fair or equal is a mean between more or too much and less or too little ; but gain and loss are both more or too much and less or too little in opposite ways, i.e. gain is more or too much good and less or too little evil, and loss the opposite of this.

And in the mean between them, as we found, lies that which is equal or fair, which we say is just

That which is just in the way of redress, then, is the mean between loss and gain.

When disputes arise, therefore, men appeal to the judge :* and an appeal to the judge is an appeal to that which is just ; for the judge is intended to be as it were a living embodiment of that which is just; and men require of a judge that he shall be moderate [or observe the mean], and sometimes even call judges " mediators " (jitaitiovg), signifying thatp. 150 if they get the mean they will get that which is just.

* The SiKcuTToi at Athens combined the functions of judge and jnrv.

That which is just, then, must be a sort of mean, 8 if the judge be a " mediator."

But the judge restores equality ; it is as if he found a line divided into two unequal parts, and were to cut off from the greater that by which it exceeds the half, and to add this to the less.

But when the whole is equally divided, the parties are said to have their own, each now receiving an equal or fair amount.

But the equal or fair amount is here the arith- 9 Tnetic mean between the more or too much and the less or too little. And so it is called SiKaiov (just) because there is equal division (Sixa) ; SUaiov being in fact equivalent to Stxatov, and SiKarrTtjg (judge) to ci^a(TTr]g.

If you cut off a part from one of two equal lines 10 and add it to the other, the second is now greater than the first by two such parts (for if you had only cut off the part from the first without adding it to the second, the second would have been greater by only one such part) ; the second exceeds the mean by one such part, and the mean also exceeds the first by one.

Thus we can tell how much to take away from ii him who has more or too much, and how much to add to him who has less or too little : to the latter's portion must be added that by which it falls short of the mean, and from the former's portion must be taken away that by which it exceeds the mean.

4, 8-13.] JUSTICE. 151

12 To illustrate this, let A A', BB', CC be three equal lines : —

A E A'

I

B B'

C

From A A' let AE be cut off; and let CD [equal to AE] be added to C C; then the whole D CC exceeds EA' by CD and CZ [equal to AE or CD], and exceeds BB' by CD.

And this * holds good not only in geometry, but in the arts also ; they could not exist unless that which is worked upon received an impression corresponding in kind and quantity and quality to the exertions of the artist. 13 These terms, "loss" and "gain," are borrowed from voluntary exchange. For in voluntary exchange having more than your own is called gaining, and having less than you started with is called losing — in buying and selling, I mean, and in the other trans-

* The point to be illastrated is, that in these private transactions what one man gains is equal to what the other loses, so that the penalty that will restore the balance can be exactly measured. Of this principle (on which the possibUity of justice does in fact depend) Aristotle first gives a simple geometrical illustration, and then says that the same law holds in all that man does : what is suffered by the patient (whether person, as in medicine, or thing, as in sculpture or agriculture) is the same as what is done by the agent. This paragraph occurs again in the next chapter (5, 9) : but it can hardly have come into this place by accident; we rather see the author's thought growing as he writes. I follow Trendelenburg

(who omits the passage here) in inserting 2 before «ro«e«, but not in omitting rh before rdaxor.

p. 152