Monthly Archive

In part one, I distinguished two views, M, and E. M is my own moral view. E is the view that accurately combines my confidence in various views being true with their implications, and so appropriately tells me what to do to maximise expected rightness of my acts. I previously finished with a question: How ought I live? By M, or by E?

In the personal case, I'm tempted by E. But let's put that aside. What about in the political case?

The fact that other people disgree with you should lower your confidence in your own view. If you were both equally good at truth-spotting, then there's no reason to think that they're mistaken rather than you. It's not the case that everyone is equally good at truth-spotting, but it still remains true that disagreement should make you double-check your views.

So perhaps, given enough knowledge of everyone's views, we'd all come to the same E. Call this E+. It's the view that accurately combines everyone's E given how good they are at truth-spotting, and puts all this together for a kind of civilisation's-expected-rightness for any action.

I want to say that E+ is politics. Democracy is a bizarre (but potentially least bad) way of getting at what E+ is. You can't weight votes according to ability to spot the truth, for numerous practical reasons. But weighting them equally gives us something, and it might be as close to the ideal of E+ as we can get.

This view amounts to the claim that politics is just ethics once we appropriately epistemically weigh our normative commitments. The collective nature of the political enterprise results from the fact that epistemic weighting is going to depend on what others think.

So that's the start of a view. This certainly needs a lot of work. But in it's favour is that:
a) It captures the fact that ethics should be relevant to politics.
b) It is parsimonious since it makes political reasons dependent on moral reasons.
c) It captures the collective nature of political decision making, and justifies democracy, without the need to posit any controversial values.
d) It explains why democratic decisions don't always come out correctly. When majorities vote badly, it's possible that the problem is that the majority are the worst at spotting the truth. Since that's what matters, majoritarianism does not rule (overrulling majority decisions might be permissible, depending very heavily on numerous pragmatic considerations).

That's the plus side. What's the negative?

Ah, Christmas-time. It's the time of year when I write on some vaguely "living together happily" type theme.

I'm fairly sympathetic to the war-on-moral-relativism that some corners of the philosophy blogosphere engage in. Some people do hold the view very naively, and some such naive views are both false and harmful.

But, I have two worries.

First, I wonder if the popularity of relativism is over-hyped. Having taught moral philosophy to first-years this semester, I noticed that very few were relativists. Most thought it was unsustainable (along with every other theory we've done, but that's beside the point). Perhaps we're attacking a view that's already dead?

Second is a broader, more interesting point. There are some perfectly respectable moral relativists. Gilbert Harman, for instance, is a clever guy, and he's a moral relativist. Now, you might say that popular naive moral relativism is very different to Harman's theory. But this isn't fair. People who don't study philosophy and who are naive moral realists are probably equally holding crap versions of potentially reasonable views. The same point applies to almost any view: Naive Christians certainly hold some bizarre views (e.g. the age of the earth), but so do naive atheists ("scientism"). Should we have a war on naivety instead? Why make a special effort in the case of moral relativism?

There's a broader issue at stake here: Should academics with an interest in popular opinion strive to influence the lowest of low believers, or the best of the best? Attacking only the most naive versions of the worst views seems like it's unfair to those who hold more sophisticated versions of those views. But attacking the most sophisticated version of a view can result in all sorts of other problems: You might merely attack one specific variant of a view rather than it as a whole, or fail to sway popular opinion at all on the issue. But weren't these things our very aim to begin with?

So it's a puzzle. On the one hand, some people hold very unjustifiable naive views. On the other hand, most of those views have sophisticated counter-parts, and it's not quite fair to attack only the (popular) straw-man.

So I have various normative beliefs about how society ought to be. Call my personal moral view M. Here's a deceiving question: Do I believe that M is likely to be true?

No. With an air of paradox, I'm well aware that all throughout time and space, people have held all sorts of competing moral views, and given that they've all been different, then very few of them, if any, have been correct. Since I have no reason to think that my ability to spot moral truth is vastly better than other people's have been, the likelihood is that my view is also false.

So what degree of confidence do I have in M? Let's imagine that I think that M has a 1% chance of being correct. This sounds low given that I do believe M. But this is ok so long as I also think that all other views that I'm aware of have a less than 1% chance of being correct (they may be so numerous that the aggregate confidence in them is 99%, or perhaps I'm mostly confident that the correct view is one I'm not aware of at all). Let's imagine that I give all other views precise confidence-values too. Note that my confidence in all these views will be raised or lowered by the opinions of others. "Why, if that clever guy Derek endorses alternative view C, then I must have been too little confident in that!".

After completing this process, we can, for each possible correct view, multiply the chance of it being correct by the rightness it assigns to an action, then sum these values. This gives us the expected moral rightness for any action. Call the set of these evaluations E. E is obviously not a true moral view: It's a mish mash of views given my confidence in each. (Indeed, E could possibly lead to apparently inconsistent decision making.)

Let's take a different simple example to illustrate E. Let's imagine that I afford each of 100 different views a 1% chance of being correct, and according to each of these views respectively, I ought to give 1, 2, 3, 4, and so on, percent of my income to Oxfam. So I am 1% confident that I ought to give 1% of my income to charity, 1% confident that I ought to give 2% to charity, and so on. After I consider this fact, I realise that I ought to give 50% of my income to charity, since this appopriately weights my actions with my confidence in various moral truths. Giving 50% of my income to charity maximises the expected moral rightness of my action.

Back to the original example. I now have two views: M, the view I have most confidence in, and E, the view that will not be correct itself, but which, if lived by, maximises the expected rightness of my actions. What should I live by? I'm tempted to say E, though I suspect that may be controversial (that nasty word integrity springs to mind, though I can never get a hold on what this might be).

Certainly it seems that living purely by M is a little over-confident. Let's imagine two further examples: In one I am 99% confidence that C is true, 1% confident that D is true. In the other, I am 51% confident that C is true, 49% confident that D is true. Is it really the case that I should equally live by C in either situation? And if we're not merely going to live by the view we're most confident in, it seems to me that the best non-arbitrary thing to do is to live by E.

What do others think?

Last year I had the wonderful opportunity of reading Bertrand Russell's History of Western Philosophy. It's a great (and entertaining) book, especially for filling in those areas and people you've never had the chance to study.

It's also interesting to get some perspective on how some philosophers are really people of their times. Amusingly, Russell spends a fair amount of time talking about how various philosophers led to the state of Germany at the time of writing (it was first published in 1946). I don't know enough history to know how compelling his statements on this are, but I found it deeply ironic that a book on history and how philosophers might be affected by (and affect) their social world should itself be so clearly affected by Russell's own social world. If you were to write a book on the history of philosophy now, German philosophy and nationalism might get a mention, but I don't think it would feature so heavily as it does in Russell's work.

Of course, now I being to think about how my own philosophical views might be affected by our current social climate..

There's some interesting discussion (with one or two very distinguished commentators) over at the Leiter reports in response to something Peter Singer said.

Personally, I agree with those who claim that Sidgwick is better than Mill. On the other hand, it's such a lengthy book that I think it'd be hard to know where to begin when using it as only part of a class on ethics generally.

Adam Smith's Theory of the Moral Sentiments, on the other hand, I've never read, and wasn't even aware it was highly regarded. Whether this is some idiosycratic position that I'm in, or if he really is so underrated that many young aspiring philosophers know little of him, I don't know.

I've just finished reading Bas van Frassen's The Empirical Stance.

Within, he tries to defend empiricism. One classic objection to empiricism goes like this:
However you state empiricism, it must amount to a statement itself not supportable by empirical evidence. So empiricism is self-refuting. So, for instance, if empiricism is the doctrine that empirical evidence is the only good justification there is, then this statement demonstrates itself to have no justification, for it cannot be justified by appeal to experience.

van Frassen avoids this by characterising Empiricism as a stance, which is a sort of combination of emotional and non-emotional attitudes, rather than a belief as such. Attitudes like this don't need justification in the same way that beliefs do, and so may not undermine themselves.

But he admits that stances may require beliefs - the empirical stance might just be a strongly favourable attitude towards the scientific method, but this attitude might also require the belief that this method is the most justified method there is. But if it requires that belief, then empiricism remains self-defeating.

So that's objection one: The self-refuting nature of empiricism only require that certain beliefs are necessary, not foundational. But van Frassen admits this much.

Another worry (perhaps incompatible with the first?) is that rational debate over stances is going to be impossible. If one's stance determines what counts as rational, justified, warranted, and so on, then debate between holders of differing stances will be impossible, because they hold differing standards of how to argue over their stances, and what counts as a merit in a stance. On the other hand, if there are some standards of rationality, justification, and so on, before holding stances, then empiricism is false: There are non-empirical standards of justification.

A friend of mine related the following example to me:
Imagine a moral theory that does not require any agent to give away all of their money to help those in poorer countries. So such a theory seems not to be too demanding. But now imagine that this theory also morally permits governments (or, I suppose, anyone) to force (tax) agents to give their money away to poorer countries.

One objection to such a theory might be the limits that it places on autonomy. But is there an additional objection to it: a demandingness objection?

Both answers are here tempting. It's tempting to say yes, because agents are possibly going to give all their money away, and this is what's apparently wrong with overly-demanding theories.

On the other hand, demandingness seems to be about what's required of agents, and in this example the moral theory generates no requirements, only permissions. How can the existence of permissions make a theory demanding?

Any thoughts?

One way the demandingness objection is sometimes cast is in terms of loss of welfare for the agent who must give up so much. But this looks to be a very odd way to cast the objection. For the demandingness objection ought to apply to utilitarians if it applies to anyone. Yet, the utilitarian has an apparently sound reply to this objection: They are well aware that such actions lower the welfare of agents in the west, and this is indeed to be regretted. But their entire point is that if welfare concerns us, the loss of welfare for western agents is less than the gain for agents elsewhere. Casting the demandingness objection in terms of welfare is going to be tricky (unless we just allow the demandingness objection to collapse into the objection that certain theories cannot account for partial concerns).

Here's another way to phrase the utilitarian reply to the demandingness objection: Sure, utilitarianism demands a lot of rich agents. But opposing theories demand far far more of poorer agents. It demands that they die. So the demandingness objection applies more strongly to theories that are not (say) act-consequentialist.

Can the demandingness objection be sucessfully characterised?