On Philosophy

August 10, 2007


Filed under: Ethics,Metaphilosophy — Peter @ 12:00 am

One persistent problem for philosophy is the tendency of some philosophers to attempt, through reason alone, to uncover the ultimate or hidden structure of reality. Of course few would describe what they do in those words, but as soon as we have things like substances and universals being incorporated into the theory as more than descriptions then we are in trouble. But I have covered this in detail already; there is no need to say the same thing over again. A second problem for philosophy, which is almost as troublesome as the first, is the idea that what we should do, what is normative, can be uncovered through pure reason, and/or that the normative is the sole and central domain of philosophy (the two claims really go hand in hand, because if what is normative involves more than pure reason then questions concerning normativity may fall under other disciplines on occasion, with philosophy simply handling the more abstract cases).

The problem with philosophy and normativity is that the problems surrounding it have been created by philosophers, in other words, they are pseudo-problems, and so attempting to answer them is likely to be a waste of effort. Normativity, as it is found “in the wild” is a relatively simple thing. But when philosophers get hold of it normativity becomes something altogether different. Questions about normativity take on an almost supernatural character; normativity is often treated like something outside of the natural world, but nevertheless something that we should be guided by, and that which reason alone (or at least primarily reason) can reveal.

But instead of tearing down the existing conceptions of normativity, of which there are many, it is more productive just to take a look at normative questions with as few preconceptions as possible. “What should I do?” is the essential normative question, as I see it. Saying what is normative, what we should do, is essentially to form an answer to that question. And this question is basically asked for a single reason, because the person asking it lacks information. In the simplest cases the information missing is simple factual information about the situation. If I am presented with two cups, one of which contains poison, and I ask myself “which should I drink?” I am wondering which one contains the poison. Other times the information missing is about the best strategy for action. If I am playing chess and I ask “what move should I make?” I want to know which move in this situation is likely to lead me closer to victory. And of course some situations contain an element of both, in poker for example asking “how should I bet?” contains curiosity about both the best betting strategy and the cards the other people have, and can be answered in both ways (which leads to answers such as: “given what you knew you should have raised, but given the cards he already had you should have folded”).

At this point some will object, saying that I have avoided discussing the areas in which normativity is dealt with most often in the context of philosophy, namely ethics (and where, presumably, it is different in nature). But I do not think that normativity in ethics is really any different from normativity in the situations discussed so far; it is thinking that it is different which leads to problems. Consider, for example, someone who asks “what should I do?” when choosing between keeping a promise and helping someone in need. I claim that this is like “what should I do?” when asked when playing chess, the person wants to know the course of action that will allow them to best fulfill their joint desires to keep their promises and help those in need. Now philosophy might be able to help them in this situation, for example by getting them to think of why they want to keep their promises and why they want to help those in need, and then evaluate, in that light, which is more important. We can even save some of the idea that ethics is universal. Ethics, I would maintain, is like a successful team strategy, team members should follow the team strategy because they want to win as a team, even if that means sacrificing the spotlight sometimes. Ethics is like a team strategy for society, by playing ethically we help society win (be a better society to live in), and that is what we desire. So telling someone that they should act ethically is, in this sense, to explain to someone how a different pattern of action is best, just as we might explain an alternate poker or chess strategy to someone. And of course there is plenty of room for philosophers to work at developing the best strategy.

But ethics, so described, does not have the kind of super-normativity that some philosophers seem to expect of it. There may be people who, when you explain you chess strategy to them, reply that they are trying to loose. This is, of course, unusual, but there is no point in stamping our foot and insisting that they must follow our chess strategy, and that if they don’t they are playing contrary to reason / nature / the divine will. Similarly if we find someone who doesn’t care about the state of society then there is no point in insisting that they should be ethical, it just doesn’t make sense. The proper reaction is to lock them up or exile them, not to try to explain to them that they really should be ethical. Or so I claim.

The opposing view is that what we should do can be determined by reason, that from certain axioms follow what we should and shouldn’t do (this is how universality can be recovered). But how are we to know that these axioms are the correct ones, because there is no consensus about them (so obviously there is no special access to them)? Well, what we do is apply those axioms to various situations, and if they tell us to do something that we clearly think we shouldn’t do we reject them as flawed. And that is absurd, when you think about it. Rather than the axioms guiding how we are to act, how we choose to act in fact guides the selection of the axioms. One problems with that is that there will never be an agreement as to what the right axioms are, because people simply make different choices in certain situations (such as abortion), and feel strongly about those choices, and so this process will result in not a single axioms set, but rather a different set of axioms for each person. A more serious problem is that if this method of evaluating the axioms is taken seriously then, over time, the axioms will become more and more complex until they agree perfectly with how we think we should act. But then the axioms tell us nothing, we already have opinions about how to act, opinions about which things are important (keeping promises, saving lives, and so on) and their relative importance. Formalizing those innate desires with an axiom system doesn’t add anything to them (except complexity, which is good for getting tenure, but bad in general).

There is, of course, more that could be said about normativity, for example, about how questions such as “what should I want?” or “what should my goals be?” still can be answered on the basis of existing desires/goals, or about how normative language such as “you should be nice to strangers” is occasionally used as a way to change people’s behavior, and not really in a normative sense at all (in fact I think that’s where the philosophical confusions about normativity get started, with the idea that one person can tell another that they should do something without the possibility of them being wrong, an idea entrenched by those in positions of authority who would like to boss people around in this way). But I don’t think I have to say anymore to drive home my central point, that normativity is much simpler than is supposed by many philosophers, and that it is by treating it as more than it is that problems arise. The ultimate test of normativity is this: if someone should do something then, if you explain to them the facts of the situation and the likely consequences of their acting that way (such as explaining why a strategy is likely to be successful), they will follow that advice, or at least admit that it influences their choice (because we must allow for various competing desires). But if they are unaffected then it is not in fact normative, at least for them. To claim otherwise is to deprive the word normative of its force.


  1. [I tried posting this yesterday, but it doesn’t seem to be showing up…]

    I take the core point of your post to be that most of the time our ultimate ends are not in question, so we’re merely interested in the instrumental question of how best to achieve them. But the more philosophically interesting question is whether we can assess ultimate ends — which would explain why some philosophers focus more on it!

    The opposing view is that what we should do can be determined by reason, that from certain axioms follow what we should and shouldn’t do (this is how universality can be recovered).

    Another form of rationalism would be more ‘coherentist’ in nature, and so not appeal to axioms at all, but simply recognize that some desire sets are more internally unified and coherent than others.

    in fact I think that’s where the philosophical confusions about normativity get started, with the idea that one person can tell another that they should do something without the possibility of them being wrong

    It would be a strange form of objectivism that did not admit of human error! (That’s more the mark of relativism: whatever an individual prefers is thereby “right for them”.)

    The ultimate test of normativity… if they are unaffected then it is not in fact normative, at least for them.

    Of course, that only works if they are assumed to be rational. Some idiot could always return a blank stare when faced with an instance of modus ponens, but that’s no reflection on the normative status of the inference!

    Comment by Richard Chappell — August 11, 2007 @ 12:16 am

  2. But the ultimate ends question must be adressed from the standpoint of our current ends, as I mentioned parenthetically. To approach the question otherwise is to blow normativity up into something it is not, and hence create unnecessary problems.

    I know you love coherentism in general, but it just doesn’t carry any weight with me, perhaps because it is incoherent with my other beliefs (does that make it self-defeating?). Suffice to say you certainly don’t recover objectivity; different people end up with different belief sets coherent in different ways. In fact I have made the idea of modifying your goals to be more coherent to be a central idea in my approach to the “good life” (as detailed on this blog), but it just doesn’t get us back to a “supernatural” conception of normativity that reason alone has access to, normativity remains mundane.

    As for irrarionality: it depends – if it is because they can’t follow the reasoning involved then I agree, but if it is because they have “irrational” desires (like the desire to lose at chess) I disagree, there is no priviliaged position to decide which of their desires are irrational. Secondly it is a confusion to bring the status of deductive inference into the realm of normativity. The fact that a->b, a, ∴ b is not a normative fact, it is a simple fact about truth preservation. The inference is not normative, it is true or reliable or whatever you want to call that sort of things. Normativity only enters the picture when we use the inference to argue for something.

    Comment by Peter — August 11, 2007 @ 12:35 am

  3. On the last point, I had in mind a person with inconsistent beliefs, i.e. accepting that ‘a->b’ and ‘a’, but rejecting ‘b’. Any rational agent will be moved by the normative force of coherence to revise their beliefs. Some humans, on the other hand, might either fail to see the incoherence, or else admit it but simply insist that they don’t care, and don’t see anything wrong with having inconsistent beliefs, etc. Does that mean that epistemic normativity doesn’t apply to them? No, it just means that they’re irrational. Same for someone who doesn’t respond to practical incoherence (i.e. an incoherence in their desires).

    I’m with you on rejecting ‘privileged positions’ — again, that sounds like a holdover from foundationalism. My point is just that one’s desires are not automatically immune from criticism. They can be criticized on grounds of internal incoherence, just as beliefs can — and this is ultimately a rational matter. (Nothing “supernatural” about it, of course!)

    As for whether coherentism entails pluralism or just one universal ideal, that’s an open question — as discussed in comments here.

    I’d be interested to hear more about what beliefs of yours you take to conflict with coherentism. Perhaps a topic for another post, some day?

    Comment by Richard Chappell — August 11, 2007 @ 1:14 am

  4. Here’s a question for you: what’s so bad about incoherent beliefs/desires? For some sets of incoherent beliefs/desires (as long as they don’t include a desire to be right) there is nothing within them that leads to a rejection of incoherence. Thus to say that they are bad or undesirable is to be making a judgement from a priviladged position. I’m against incoherent desires myself, but only because usually people want to avoid unhappiness/unfufilled desires, I don’t see anything “wrong” with them by themselves. (Obviously incoherent beliefs aren’t conductive to truth, but that only matters from our point of view, as people who care about the truth.)

    I think it is pretty obvious that coherentism implies pluralism. Consider a person with only a single desire: to kill people. Consider another person with only a single desire: to save people. Neither is inconsistent, both are different. Thus the possibility of pluralism exists. Of course you could argue that real people will all converge, but given that pluralism is possible, I doubt it. Which is why I don’t approach ethics from the standpoint of trying to satisfy people’s desires, because pluralism is a kind of defeater for ethics (or at least is easily turnined into one).

    I’m not opposed to coherentism when it is restricted to justification, but I am opposed to it when it stands by itself, when the only standard by which we hold our beliefs to is coherence. Because I fundamentally believe that there is, for many questions, a right answer, independant of our beliefs. We could all be crazy and develop belief systems in which the sun going arround the earth was coherent, but that wouldn’t make it so. Coherence by itself is insufficient to move us from incorrect ideas to correct ones, we need, in addition, are external standards by which we try to measure up our ideas to. In science that would be predictions. I guess this is a kind of coherence, but its not the kind that the coherentists push for (which is coherence amoung existing beliefs).

    Comment by Peter — August 11, 2007 @ 1:45 am

  5. I think Richard Chappell is right. That’s the conclusion I reached myself.

    Essentially, one’s ethical outlook it based on one’s understanding of objective reality. If you are a racist who thinks that there is a sharp division between races, that their behaviour is markedly different and that races are engaged in a Darwinian struggle for survival, don’t be surprised to find your political views are similar to that of the Nazis. If you believe that race differences are inconsequential, that there is a loving God and that God commands us to love each other and live in peace, your political views are going to be radically different from that of the Nazis and you are going to be committed to international peace and reconciliation.

    As Richard points out you could have a position where someone adopts an inconsistent and contradictory approach. For example, someone who has racist/social darwinist views may nevertheless press for an end to racial discrimination and for international brotherhood. They are free to do so, but it makes no more sense than someone saying “I want to travel by the quickest route from Paris to Berlin” and then going via Johannesburg.

    Of course one can have a range of views which might match
    a set of beliefs about objective reality. For instance, a belief in God is compatible with a whole range of political beliefs – but not I would suggest with a programme of atheistic persecution and mockery of religion as pursued by the Bolsheviks. Furthermore, within the range of “matches” against a particular set of beliefs in objective reality, there will of course be differences of strategy and tactics, which may lead people to oppose one another. For instance, whilst the Bolsheviks were essentially agreed in their analysis of objective reality, there were factions based around different strategic approaches (e.g. socialism in one country versus world revolution). The situation gets confusing because of course in such circumstances opponents often present the disagreement as one about objective reality – whereas in fact the difference is far more to do with strategy and tactics.

    Also, we have to consider “flying under false colours”. Many communists came to think that Stalin wasn’t actually a communist , but was a megalomaniac motivated by personal aggrandizement. Whatever the truth of that (I personally think he was genuinely committed to the communist belief system), it is certainly the case that many people carry around these secret belief systems i.e. although they may be publicly committed to a certain ideology, their real view of objective reality is that humanity is made up of sharply differentiated individuals of varying abilities who have opposing interests and are engaged in a kind of Darwinian struggle. One often finds such people attracted to the world of business where such views are tolerated. But in other fields – religion, politics, charities, the arts – such people have to disguise their motivation to a greater or lesser degree.

    Comment by field — August 11, 2007 @ 3:17 pm

  6. I have no idea how your comment relates to anything that anyone has been talking about, feel free to explain.

    Comment by Peter — August 11, 2007 @ 4:01 pm

  7. I think many societies would reach their aims more easily by practicing capitalism. So, I consider capitalism normative for their aims. On the other hand, I don’t consider it unethical to not practice capitalism.

    My question is does “ethics is normativity” capture the difference between what I wish people would do, and what I think they “must” do?

    Comment by Carl — August 11, 2007 @ 5:33 pm

  8. Carl –

    Is your use of “normative” to do with how things should or ought to be? I am assuming it is.

    It is unclear what your statement “I think many societies would reach their aims more easily by practicing capitalism.” means. Does it means that you personally think Nazis in a Nazi state should be encouraged to adopt capitalism so as to more easily achieve their aims? Or are you simply saying that for each society, capitalism offers them the best chance of reaching its aims. That’s not normative.

    It seems to me if you don’t mind whether or not a society practises capitalism that probably suggests you have a very open view of humanity, as I do myself – based on a broad understanding of history, anthroplogy, economics etc., so there is no reason for you to adopt a strongly normative position in relation to such matters, since you are not in conflict with your world view.

    In terms of ethics, I don’t recognise any real distinction between should and must. It’s really a sliding scale. At one end I don’t really like people sneezing in public without covering their nose and mouth. It’s not illegal – there is no lawful punishment for the action (although I believe there have been laws enacted during epidemics). However, even for this minor act there will be some sanctions – looks of disapproval from fellow passengers for instance. We can moe to the middle of the scale – it’s not illegal to criticise your boss’s tie, but if a young person was starting out in business, someone accustomed to offering his opinion on people’s dress sense, I would probably advise him that he “must not” criticise the boss’s tie. Moving up the scale, I might advise my children that they “mustn’t” play their music so loud as to annoy the neighbours. Technically it is probably an offence punishable by fine or even prison for repeat offences, but the likelihood of police intervention is small – it’s really an extension of my personal morality. Then we get on to the big offences: eating people is wrong territory i.e. you most definitely must not cannibalise your fellow human beings and if you do you must be derpived of your liberty or your life. At the far end we have the ideologues who want to murder millions of people: and I believe we mustn’t allow people to commit mass murder.

    How one views the scale is a reflection of your view of objective reality. For the Nazis an accidental blood transfusion between a Jew and an Aryan was a highly serious matter that called for severe punishment – something we wouldn’t even regard as an offence. For the Nazis there was no “should” about people getting married before raising a family, with regard to the SS. All these views can be traced by to the objective reality description of Nazism – with its well defined races, Aryan supremacism, social darwinism, analysis of religions (as corruptive of the race) etc. , absence of any clear deity.

    Comment by field — August 12, 2007 @ 9:34 am

  9. I’m using the word “normative” the way Peter is — something the doing of which helps you attain your goals.

    Comment by Carl — August 13, 2007 @ 12:05 am

  10. If that is the case I don’t see how “ethics” can be “normativity”. Normativity must be what helps you attain your ethical goals.

    [However doesn’t appear to be using the word in the sense you claim when he says

    ““What should I do?” is the essential normative question, as I see it. Saying what is normative, what we should do, is essentially to form an answer to that question”

    That passage implies Peter is using it as not just an efficiency term but as an advisory term. The two are obviously not the same. It might be more efficient for prisoners to eat other than be fed at public expense, but I wouldn’t advise we do that.]

    Normativity, according to your definition, sounds to me just like a fancy word for strategy or tactics.

    For me the sequence is:

    1. One develops a view of objective reality – what it consists of and how it is structured.

    2. One develops personal, group and community goals on the basis of of that view of objective reality. (The goals emerge because life doesn’t let you be. Things happen – you have to act or react – and in doing so, because of the sort of beings we are, we look to the future and set goals. It is at this point that goal setting is usually found to be consistent with the view of objective reality.)

    3. One develops a strategy or tactics for achieving those goals.

    For instance, Tony Blair ex PM clearly saw objective reality as including a God and included the fact that God viewed people as being of equal worth in His sight. In terms of personal goals, he saw it as important to develop a meaningful relationship with that God. In terms of group goals, he helped define the Labour Party’s goals in line with his view of objective reality. He then tried to win power so that he could shape the nation’s goals in a similar direction.

    His strategy for achieving his goals took different forms. In his personal life he drew close to the Catholic Church. With the Labour Party he organised it so that the Marxist left were completely marginalised. In terms of pursuing the national goals, he made full use of law and policy instruments – taxation etc.

    All seems v. straightforward to me. Where’s the problem?

    Comment by field — August 13, 2007 @ 12:17 pm

  11. I think it is pretty obvious that coherentism implies pluralism. Consider a person with only a single desire…

    But a single desire by itself is ad hoc. A mutually-supporting web of desires is thus more coherent. (In general, a web may be made more coherent by adding more supporting strands, and removing isolated or ad hoc strands.) So I actually think it’s fairly plausible to deny pluralism, and think that there is only one maximally coherent desire set. Of course there are a plurality of more-or-less coherent desire sets — ones that at least avoid outright inconsistency. But they are not all maximally coherent.

    We could all be crazy and develop belief systems in which the sun going around the earth was coherent

    No we couldn’t. Maybe we could add enough ad hoc epicycles to make it yield identical predictions to a Copernican theory, but we reject this because we recognize that the simpler scientific theory exhibits greater coherence. Or perhaps you mean that we could dismiss all the empirical data and just brutely assert a simple Geocentrism, but that move too would be flawed for a similar reason.

    Comment by Richard — August 13, 2007 @ 11:04 pm

  12. I admit it is ad hoc and unrealistic, but given that is it possible I think it is safe to assume that reality contains more than one consistent network of beleifs beleived by someone. Certainly, for example, it would be consistent to believe in a different kind of logic, nothing anyone can say about that, so long as you introduce some rules that make it agree with classical logic when it comes to ordinary things.

    No, you could hold the Earth’s position constant, let the sun revolve arround it, and let everything else go arround the sun. Perfect agreement with everything observable. Needs a different theory of gravity though, but that’s no biggie (could be easily done with two flavors of gravity). I can also create perfectly coherent theories of relativity where space-time is euclidean (Poincare demonstrated this was possible). We choose relativity and non-euclidean space time for reasons other than coherence.

    Why does simpler imply more coherence? That’s not what coherence means as I understand it. Two different axioms systems (say two geometries) can both be perfectly coherent so long as they don’t contradict themselves, and one can be much simpler than the other, without making it more coherent.

    Comment by Peter — August 13, 2007 @ 11:16 pm

  13. Let me put things another way. Suppose I have a set of beliefs K, and beliefs a, b in K conflict. There are two ways to maximise the consistency of K, discard a (and associates if necessary, i.e. beliefs that entail a, such that if a conjunction of c d and e entail a one of them must be discarded, and it’s associates, etc) or discard b (and associates if necessary). Since both can be discarded (dicarding either of them resolves the inconsistency, so seeking maximal consistancy does not favor discarding one or the other) even if all people started with the same set of beliefs K and pruned their beliefs to be maximally consistent (by iterating this process until no inconsistencies remain) they would not converge to the same set of beliefs, J, without other principles besides consistency at work in the revision of beliefs.

    Comment by Peter — August 13, 2007 @ 11:30 pm

  14. I think it is safe to assume that reality contains more than one consistent network of beleifs beleived by someone.

    Of course, that’s why I’m talking about *coherence*, not *consistency*. The latter doesn’t even come in degrees, so it would make no sense to talk of a set of beliefs being “maximally consistent”. It’s either consistent or it’s not.

    So I absolutely agree that we need “other principles besides consistency at work in the revision of beliefs.” Once inconsistencies have been ironed out, we need to ask the further question of which belief sets are more or less coherent.

    Comment by Richard — August 14, 2007 @ 1:03 am

  15. Ok, I’m pulling your leg a bit, I was hoping to get your version in order to address your specific opinions, but since I don’t want to leave you hanging I will suppose that you will add simplicity, and since you want to avoid the reductio that believing nothing is the simplest set of beliefs you will add in the requirement that it explain everything observed, but not include more than is necessary to do that (Occam’s razor).

    The problem with this coherentism is that it is simply another way of phrasing the belief that there are external standards (reality) that we hold our beliefs up to, + a kind of minimal naturalism. Here’s why. Suppose your beliefs contain the existence of something non-natural. Let’s say a non-physical mind. Now I have a naturalistic explanation of mind, but you respond that you have certain beliefs which entail that the mind cannot be physical. But, according to this version of coherentism, it is simpler (more coherent) for you to reject those beliefs and be a materialist. Suppose I am wrong about this; then coherentism implies pluralism, since certainly the materialist is not incoherent, they just lack certain beliefs that exclude materialism (about the existence of non-reducible phenomenal charachter or modal reasoning or whatever). On the other hand if my assesment is right then dualism and everything else non-physical is ruled out. Which leaves us with a narrow brand of naturalism / empiricism, in which the only things that are true are those which explain our observations.

    You could avoid all this of course by leaving out Occam’s razor, but then you would allow, say, different religious belief about distinct, exclusive, and non-interfering, but non-identical, divine beings to be coherent. And that, again, implies pluralism, since two people can believe in two different such gods, and a third person in neither, but obviously no one can believe in both.

    Surely this is not what the coherentist wants (right?), they wanted an internal criterion of truth. And in any case if your criterion for being true is the simplest explanation for experiences we might as well just say that. Of course I wouldn’t buy into something that minimal myself (I doubt many would), since it seems to exclude philosophy itself.

    Comment by Peter — August 14, 2007 @ 2:04 am

RSS feed for comments on this post.

Create a free website or blog at WordPress.com.

%d bloggers like this: