Bernard Williams on "The analogy of city and soul in the Republic"

At 434e, Plato himself says that in the end, the analogy may not work to flush out justice, but he does seem confident that it can be made to do so:

• "But if something different is found in the individual, then we must go back and test that on the city. And if we do this, and compare them side by side, we might well make justice light up as if we were rubbing fire-sticks together."

The force of the analogy must be built upon a claim that justice is one and the same thing, whether in a city or in an individual:

• 435b: "Then a just man won't differ at all from a just city in respect to the form of justice: rather he'll be like the city"

At 435e, Plato says:

• Well, then, we are surely compelled to agree that each of us has within himself the same parts and characteristics as the city? Where else would they come from? It would be ridiculous for anyone to think that spiritedness didn't come to be in cities from such individuals as the Thracians, Scythians, and others who live to the north of us who are held to possess spirit, or that the same isn't true of the love of learning, which is mostly associated with our part of the world, or the love of money, which one might say is conspicuously displayed by the Phoenicians and Egyptians.

So, we call a city 'spirited' because individuals in it are spirited.

That is one model of how a composite thing comes to be F: it's constituent parts, or most of them, are F. Williams calls that the 'whole-part' theory.

That cannot be the full story, however, because then there would be no point in the analogy: if we already know how justice comes to be in a city (namely, because the people in it, or most of them, are just), then the analogy will not be informative.

Think of it this way as well: if you have a crowd of large sailors, do you have a large crowd of sailors? NO! But if you have a crowd of angry sailors, do you have an angry crowd of sailors?
The question is whether justice is like anger or like largeness. Maybe having a city of just people does not mean that you have a just city.

Go back to 420a:

• ". . . in establishing our city, we aren't aiming to make any one group outstandingly happy but to make the whole city so, as far as possible. . .
• . . .  we take ourselves, then, to be fashioning the happy city, not picking out a few happy people and putting them in it, but making the whole city happy."

So Plato is aware of the fact that the happiness of the city does not depend on putting happy people in it or making certain people happy in it: he should also be aware that justice in the city is not created in that way either. So a large crowd of sailors is not a crowd of large sailors. (But an angry crowd of sailors IS a crowd of angry sailors.)

So there are some predicates that work one way (the parts impart a character to the whole) and others do not (the whole has a character which no part need have).

Williams says that Plato wants at least the following claims:

• a) A city is F if and only if its humans are F (the 'whole-part' theory: more generally stated, such a theory holds that "X is F if and only if its parts y are F")
• b)The explanation of a city's being F is the same as that of a man's being F (the same Form of F-ness applies to both) (this is not the same as the 'whole-part' theory, but is not incompatible with it).

Plato wants to explain justice as:

• c) Each of the elements (the rational, the spirited, and the appetitive parts) does its job.
  which requires or implies:
• d) The rational part rules.

• e) A city is just if and only if its humans are just.
  and
• f) An segment of the city is rational, spirited, or appetitive if and only if that segment's humans are.

It looks as if Plato is committed to a city that has some humans who are rational, others who are appetitive, and others who are spirited.

A problem with that is that appetitive humans surely are not ruled by their rational part, and so surely are not just (per d 'The rational part rules').
Since the majority of the humans in the city are appetitive, that means that the majority of humans in the city are unjust, and that means that the city is not just (per e).

Thus that version of a whole-part explanation of the justice of a city will not work!

Williams tries another possible tack:
Accept e) A city is just if and only if its humans are just.
Maybe that means that to have a just city, the majority of the humans must be just.
Then d) must be true of every or almost every human, and so the rational part will rule in most humans. We now need to explain how that is true in spite of the potential problem raised above: that the appetitive people are not just.

At 433c, Plato identifies justice in the city as "the FACT that every child, woman, slave, freeman, craftsman, ruler, and ruled each does his own work and doesn't meddle with what is other people's."
That FACT is not in the mass of the people: there must be something in them that brings the fact about.
We know that the mass of the people are the appetitive part of the city.
Being the appetitive part means that they are not the rational part of the city. Are they to have a rational part of them that rules?

We can approach that question by asking about the parts of a person from the masses. It seems that each person has a basic appetitive part, a basic thumetic part, and a basic rational part. It might be the case that the basic appetitive part of the individual has its own subordinate rational part which exists in addition to the basic rational part of the individual.
That would create an infinite regress, however, for if each part has little versions of the other parts, then you just get triads ad infinitum: the subordinate rational part will have its own sub-subordinate parts.
That would be absurd, says Williams (I am not so sure, but most people think it would be absurd).

It must be the case that ideally the individual has an appetitive part that is weakened/transformed and kept in place by the rational part through the help of the spirited part. 

Now that we have established that it would be absurd for each individual person to have a triad of parts each of which has a further subordinate triad of partseach of which has a further sub-subordinate triad of parts etc., we can try to apply that conclusion to the city. In the city, we find a ruling class, a spirited class, and an appetitive class of people. We can conclude that  the ruling class must be totally rational, the spirited class must be totally spirited, and the appetitive class must be totally appetitive (if the explanation of justice is to be the same for the individual as for the state). If that were not so, what would mark the appetitive people off from the rational people? It now seems that maybe the appetitive class can have NO rational part.

So Williams asks, does Plato want the humans in the just city to be just?
433-4 implies that he does, but he never says so.
At 431e-432a, he says that sophrosune (soundmindedness) is shared by all the humans, and consists in the mutual "agreement between the naturally worse and the naturally better which of the two is to rule both in the city and in each one."
A problem with this is that it's not clear what role justice is left with, because sophrosyne seems to cover the ground which c) and d) claim for justice:

• c) Each of the elements (the rational, the spirited, and the appetitive parts) does its job.
  which requires or implies:
• d) The rational part rules. 

And yet even these soundminded folk are not clearly 'in agreement,' for he says at 431d:

• 'Then, don't you see that in your city, too, the desires of the inferior many are controlled by the wisdom and desires of the superior few?'

If the superior CONTROL the inferior's desires, that means the inferiors are not themselves soundminded, for they have those desires and would act on them were it not for the superior people's rule. (we might answer this for Plato by saying that part of the superior people's control is actually a transformation of desires: Williams considers that later).

The result of all this discussion is still that the analogy is supposed to justify the rule of the rational element in the city, which is identified with a class of persons. It does so via the clear superiority of the soul of those whose rational part controls them.

So maybe we have to say that the whole-part theory is wrong: most of the people in the city will not be just. So maybe we need a new theory: 'the predominant section' theory.

• g) A city is F if and only if the leading, most influential, or predominant citizens are F.

g) comes out clearly in Plato especially in book VIII, where Plato talks about the cities that have degenerated (hence they are called 'degenerate cities'). That discussion starts at 544d.

In degenerate cities, it is clear that not all citizens share the character of the city. For example, the tyrannical city is not entirely composed of tyrannical sorts. See 577c, 567a, 568a.
There may be a few tyrannical humans in a law-abiding city (575a), but as long as they are few, they have little influence. If there are many, they can produce a tyrant (575c).
At 564d, the 'drones' are said to be present in an oligarchy, but if they become the leading element, it changes to democracy.

Democratic cities are composed of all manner of people (557c): like a multi-colored garment, the democratic city has every sort in it.
Democratic individuals, however, are always shifting, without expertise in anything, indulgent of their desires (561d). The common thread is that both have many elements calling for attention, with no rational principle to govern the whole.

To form a democracy, it certainly would not be sufficient to have a whole bunch of democratic individuals as the predominant part.
The problem is that the democracy then will have no particular character, rather than a democratic character. A democratic state has all sorts of individuals, not just democratic ones.
So Plato has confused a state that has various characters amongst its people and a state that has democratic people.

How are we to picture these democratic individuals when they are quiescent (controlled)?
Maybe what we have been doing is to talk about the desires as though they are the same whether they are controlled or not (I said that Williams would return to this question above). If the desires of some people are controlled by other people, perhaps that control somehow transforms the desires of the people who are controlled.
Williams says that thinking of the desires as either controlled or not controlled will not help, because of the following.

One might think of having one's desires controlled in two ways:

• 1) they are still there, every bit as powerful, but controlled. This is internal civil war with the rational part winning.
• 2) Inner peace: the appetites have been transformed and are only mildly and harmoniously active (we might say they are complicit in their own oppression).

What if that can be solved and the appetitive part of the city corresponds to the workers, the craftspersons, etc. How does that work? Are cobblers naturally appetitive? It seems ridiculous to suppose so (although aristocrats often do have exactly that sort of prejudice).

Plato needs to say that expertise is what qualifies one to rule, and that has been a source of criticism of Plato. Williams thinks he needs more than just expertise. He needs the following triad:

• 1. Being capable of being a ruler (i.e. having an extra-ordinary rational part) goes with having rational motivations.
• 2. Being capable of being a good soldier (i.e. having an extra-ordinary thumetic part) goes with having thumetic motivations.
• 3. Being capable of being a good worker (i.e. having an extra-ordinary appetitive part?) goes with having appetitive motivations.

Let's just give Plato 1. What about 3? There is no reason to think that Plato thought that being a good worker entailed or was entailed by having strong desires. The most Williams will attribute to Plato is the claim that the two go together, but not in a strong entailment way. Plato seems to have thought the workers as a class are people of strong unruly passions AND goodhearted loyal workers.

The thumetic part has particular problems. It is ambivalent: it seems to help the rational part only if it is well raised. Children and animals can display it, but it is also the part responsible for noble indignation. It's corresponding form of life is the military and competitive life.

Conclusion: the city-soul analogy, the tripartite soul, and Plato's explanation of justice in city and soul have grave problems that are unresolved.