Print This Post Print This Post

In her ‘Two Distinctions in Goodness’, Christine Korsgaard drew attention to an overlooked distinction between two distinctions about value. One is the distinction between final and instrumental value. The other is that between intrinsic and extrinsic value. Something has instrumental value only if we desire it for the sake of some further end; something has final value if we aim at it for its own sake, not as a means to some other end. And something has intrinsic value if it’s valuable only in virtue of its intrinsic properties; something has extrinsic value it it’s valuable also in virtue of its extrinsic/relational properties.

Korsgaard rightly claimed that discussion of value from G. E. Moore onwards has mistakenly conflated these two distinct distinctions . Before her paper came out philosophers would often contrast intrinsic with instrumental value, thereby overlooking the possibility of value that is neither intrinsic nor instrumental—the possibility of extrinsic final value. Korsgaard’s paper had a larger agenda. She wanted to persuade us to adopt a peculiar Kantian/subjectivist theory of value. On this theory, things have value only because we value them. The fact that we value them confers value upon them because we ourselves are intrinsically valuable. So most of the things we value only have extrinsic value, value derived from the fact that we value them. I think it’s safe to say that Korsgaard hasn’t been successful in persuading philosophers that this is an attractive theory of value. Her theory, after all, can itself be accused of conflating a conceptual distinction between types of value and a meta-ethical distinction between two accounts of the source of value—the distinction between subjective and objective value.

Although most philosophers weren’t persuaded by Korsgaard’s theory of value, some were certainly persuaded by her meta-distinction, which is what I’d like to consider here. What might it mean to deny that there is such a thing as extrinsic final value? In one sense, Korsgaard is surely right that there is a distinction to be drawn. The distinction is between the attitudes called for by something’s value (final or instrumental) and between what explains that thing’s value (intrinsic or extrinsic properties). Suppose someone dearly wants to own a certain pen. We can ask him why he wants that. He may answer that he needs it in order to write something. If so, then his desire is instrumental. But he can tell us that he wants to own this pen for its own sake. That’s simply something he wants. If we now again ask him why he wants this, we’ll be asking a different question. We already know that the man values owning this pen—values it as an end. Now we are asking him to explain the source of this value. If he cites only its intrinsic properties, he takes it to have intrinsic value. If he also cites its relational properties, he takes it to have extrinsic value. So if someone denies that there is such a thing as extrinsic value, he can’t be denying that there is this distinction. So what would he be denying? He could be denying one of two things. He might simply be denying that anything in fact has extrinsic value. This would be a substantive claim about value. Or he might be denying that our concept of value allows these two admittedly distinct things to come apart. That there could be such a distinction doesn’t yet show there is such a distinction. After all, the man can answer the second question by citing facts that are neither intrinsic nor (genuine) extrinsic properties of the pen. He may say ‘Because 5+7=12’ or ‘Because the French Revolution started in 1789’. The fact that someone could cite such properties to explain why he wants something for its own sake hardly establishes that there is an overlooked type of value that is neither intrinsic nor extrinsic. Someone who seriously cited such facts is obviously failing to understand what ‘value’ means.

My aim here, however, isn’t to directly deny the existence of extrinsic final value. When philosophers were persuaded by Korsgaard’s meta-distinction, they were especially persuaded by some of her examples. It’s fairly easy, when we think of it, to come up with examples of things that seem to have extrinsic value. We value a wedding ring, or Abraham Lincoln’s pen, only because of their relations to other entities and facts—because this particular ring was used in our wedding ceremony, because Lincoln wrote using this particular pen. We value them because of their relational properties.

In this post I want to suggest an alternative account of these and similar examples. Because I will not consider the full range of putative examples of extrinsic final value discussed in the literature, my suggestion wouldn’t be an out-and-out attack on this very notion. But if my suggestion is successful, it will undermine one main source of motivation for thinking there is such a thing. And I think it has independent interest because it may identify a new distinction in value.

Consider the distinction between instrumental and final value. Pleasure is a final value. It’s something we should want for its own sake. And there are many things we value only because they tend to cause us pleasure. We value these things only instrumentally. I have reason to go on the Ferris Wheel only because I have reason to have some fun.

Pleasure is the kind of thing we ought to promote. We have at least some reason to try to get more of it, if we can. This is how we ought to respond to its value. But as many have pointed out, not all values call for this kind of response. For example, it would be a mistake to try to promote friendship in this way. It would be a mistake for us to aim to have as many friends as we can, or to give up one friend simply because this will have the effect that we gain three new friends. This is just one example of something whose value calls not for promotion but for something like respect or honor. And there are other examples.

The suggestion I now want to float is simple. Certain final values we ought to promote. Things that are merely means for promoting those value are only instrumentally valuable. Certain final values we ought to respect rather than promote. But what is the parallel of instrumental value for value of this kind? I call it expressive value. (This may not be the best label, but it’s better than the more familiar but narrower and seriously misleading ‘sentimental value’. Do you have a better suggestion?) You can say that things have expressive value if they are vehicles through which we express our respect towards the things we ought to honor. Or perhaps better, that through their connections to entities we ought to respect, some things also call for special attitudes, attitudes ultimately directed at the What I’m suggesting is that Lincoln’s pen and wedding rings have only expressive value. That is, they have extrinsic but not final value, although they aren’t, strictly speaking, instrumentally valuable.

To be sure, if we define ‘final value’ as ‘non-instrumental value’, then expressive value would trivially have final value. If someone cherishes the photo of a long lost relative, she is not doing this as a way of bringing about some further end. Of course, you might say she’s doing this in order to express her respect for the long lost relative. But that’s a bit misleading. It’s not as if she needs to think to herself: ‘I must find some way of expressing my respect,’ and then looks around and finds that old photo, her means of achieving this goal. She might simply find the photo, quite by accident. She might have all but forgotten the dead relative. But in seeing this photo, she may recognise something she ought to preserve and treat in a special way.

In their ‘A Distinction in Value: Intrinsic and For Its Own Sake’, Wlodek Rabinowicz and Toni Ronnow-Rasmussen at one point seem to be considering and rebutting something like my suggestion. They are discussing a supposed example of extrinsic final value, Princess Diana’s dress. They write

The dress is valuable just because it has belonged to Diana. This is what we value it for. But, one might object, is it really a case of a non-instrumental value? Diana’s dress is perhaps valuable merely as a means: merely because it allows us to establish an indirect connection to a person we admire or find important in one way or another. Having such a connection may be something that we set a final value on. Couldn’t this be what is going on here? Not necessarily. Even if the desire to establish such an ‘affiliation’ with Diana may well be a part of the causal explanation of our evaluative attitude towards the dress, this does not imply that the evaluative attitude itself is of the instrumental kind: if we idolise Diana, we do not simply find the dress useful for some purpose; we ascribe an independent value to it.

They only come close to considering my proposal. Look at the last sentence. From the fact that we don’t find the dress useful for some purpose, it doesn’t follow it has independent value. And if something is not a final end, it doesn’t follow that it must therefore be an instrumental end. The dress doesn’t serves any purpose, is not a means to an end. Not even the end of valuing or honoring the dead princess. That gets things the wrong way round. Given the fact that it stands in a certain close relation to Diana, some of the value ascribed to Diana is ‘transferred’ to the dress. We treat this dress in certain ways we wouldn’t treat some other dress.

It’s a strange feature of this type of supposed example of extrinsic final value that the object’s extrinsic properties always seem to involve something of prior and independent value—something of intrinsic final value. Why is that? There is nothing in the mere notion of extrinsic value that calls for this. This feature should already make us suspicious of the standard account. It means, for one thing, that the value of items such as Lincoln’s pen is not really ‘independent’. It is derivative in an obvious way. As our valuation of Lincoln changes, so would the value we ascribe to the pen. And it seems very doubtful that the pen would in itself make a world better simply by existing—say, by surviving hidden in some cellar. Yet Lincoln pen would possess value not only because it actually causes certain responses in people.

There is of course more work to be done. There are potentially troublesome examples such as, say, the value someone may ascribe to preserving Hitler’s boots. More importantly, we need a better way of distinguishing between final and non-final value than the distinction between what’s wanted for its own sake and what’s wanted merely for the sake of something else. But I think I’ve said enough to get the idea of expressive value off the ground.

(After writing this I saw that some related points are made in Anthony Hatzimoysis’s nice PQ paper on ‘Sentimental Value’.)


  1. 1. Posted by Jeff Huggins | August 8, 2007 4:38 pm

    Guy, thanks for the great and thought-provoking post.

    Although I’m not a value theorist in any formal sense, I’ll provide my ‘two cents’ (coming mainly from a scientific and logical standpoint and from my study of morality in general).

    First, I’d like to comment on the more general issue (i.e., in your preamble and context-setting). Then, I’ll offer a couple brief thoughts that may serve as helpful considerations (I hope) as you continue to consider your concept of ‘expressive value.’

    Regarding the more general discussion (and Korsgaard’s point, and responses to it), I’ll offer a case to consider. (I don’t personally know if this specific case has been defined before, although I can’t imagine that is hasn’t been.)

    Imagine that there are only five humans remaining on Earth, and you are one of them. Imagine that you are beyond your reproductive years, i.e., that you can’t have any more children. Imagine that the other four people consist of two older adults, who also can’t have more children, and two children, one boy and one girl. So, the two children are the last children on Earth and the last humans who are capable of having children and continuing the species. Now, the two other adults are ‘bad guys’ (to use a shorthand), have captured the two children, and are threatening to kill them. You are a ‘good guy’, of course, and would like to protect the children (for whatever hopefully good reasons you may have). In other words, you value the children, and they do have value. The question: From your standpoint, or in general, what type(s) of value do the children have, using the terminology ‘final/instrumental’ and ‘intrinsic/extrinsic’? And why (i.e., what is your reasoning?)?

    (I’m not really requesting an answer. I’m posing the case as a hypothetical, for consideration.)

    My other two or three thoughts are these:

    You mention that “Pleasure is a final value”, that it’s “something we should want for its own sake.” I (of course) like to feel happy, and I like many pleasures, but from a scientific standpoint, I think it’s roughly accurate to say that our human quests for pleasure and for the feeling of happiness came about (and are within us) as rough imprecise mechanisms (means) to prompt us to do the things necessary for survival, mating, and reproduction. Pleasures and happiness can be considered to be ‘icing on the cake’ of life (when we have them). They can be considered two of the most powerful motivating ‘carrots’ that guide (very imprecisely, as Daniel Gilbert points out in his great book, Stumbling on Happiness) our daily decisions. Thus, from one important standpoint, it is fine and reasonable that we value them. They are enjoyable, and they are there for us to value. From that standpoint, yes, one can consider pleasure to be a ‘final value.’ But, from another standpoint, and relatively speaking, the quest for pleasure and happiness are means to a more foundational ‘end’, I believe.

    Finally, I would suggest that, when it comes to Lincoln’s pen or Diana’s dress, people ultimately value owning something like those either because doing so makes them more happy (or less sad) or because doing so will in some way, they believe, help them with the more basic elements of success in life, or both. For example, if I owned Lincoln’s pen, I could simply enjoy it, I could use it (consciously or subconsciously) as a status thing, or to connect with like-minded colleagues, or I could sell it to put food on the table or to buy a house.

    I don’t know if these several considerations support the idea of ‘expressive value’, challenge it, or don’t move the argument one way or the other. But, I hope they are helpful in some way.

  2. 2. Posted by Guy Kahane | August 8, 2007 9:33 pm

    Dear Jeff

    Thanks for these remarks. Let me give a quick reply.

    (1) Things (and people) can be valuable in more than one way. The children in your example have value simply because they are human beings (or persons) and deserve to be protected simply for this reason. This would be their intrinsic value. But they may have a further value, the value of being the only ones who can continue the species. This would be their instrumental value — the value of the good they could bring about (though in this example this value might be very great). Finally, being among the last humans to still exist might give them yet a further form of value. This would indeed be extrinsic value. Being among the last humans isn’t anybody’s intrinsic property.

    (2) Pleasure as a final good: I think you may want to consider distinguishing, on the one hand, causal explanations of why we value this or that, and on the other, the question of what has value and of what kind. There may be an evolutionary explanation for what we find pleasure in various things, but it would not follow that the value of pleasure is purely instrumental. We’d have reason to seek pleasure even if, in this case, this would not benefit us in any other way, and even if our evolutionary history was utterly different. (Pleasure does of course ALSO has much instrumental value.)

    (3) You are right that it’s possible to try to give accounts of the value of items such as Lincoln’s pen on which they turn out to have merely instrumental value — say that of the pleasure they give. On these accounts, they have NEITHER extrinsic final value nor what I called ‘expressive value’. So this is a separate debate. I myself am not persuaded by such a view. We could ask, for example, WHY people enjoy possessing such items, or why they might confer on us higher status. It might be that the best answer is that we enjoy them BECAUSE we appreciate the prior value they possess.

  3. 3. Posted by Jeff Huggins | August 9, 2007 12:47 am

    Dear Guy,

    Thanks very much for your responses. I’ll offer a few thoughts in response, at least one of which could also apply to discussions in the Kamm group, perhaps.

    Using the numbering system from your recent comments:

    (1) If we classify each type of value using two words (for now), i.e., one from each of the dimensions ‘final/instrumental’ and ‘intrinsic/extrinsic’, then the value(s) of the last two children on Earth would or could be (according to your recent comments):

    Their ‘final intrinsic’ value as human beings; and,
    The ‘instrumental extrinsic’ value associated with the fact that they are the last two people with reproductive prospects on Earth who could continue the species; and,
    Perhaps another form of extrinsic value that you mention (final? instrumental?) associated with the fact that they are among the last people.

    As you mention, the second value in the above list might (indeed, would) be very great.

    To make the case more interesting and revealing, and to consider it (hypothetically) in the style of the Kamm Reading Group analyses, what if there were five (rather than two) ‘bad guy’ adults who had captured the last two children, and what if they were intent on killing the children because of their view that the human species should be brought to its end? You are the only ‘good guy’ adult. You face a choice: You could kill the five adults or you could let them kill the two children. Would it be permissible to kill the five adults in order to protect the two children and thus the future of the species? Would the answer change if you had to kill 500 adults (without reproductive prospects) in order to protect the last two children? And, are these answers any different than they would be if the two children were not the last two children on Earth? If so, why? Finally, what if, of the 500 adults that had to be killed to save the two children, 400 of them were completely innocent bystanders, i.e., not involved in the plot to kill the children in any way?

    (I realize that these are hard-to-imagine cases, and somewhat intricate, but they are interesting to consider.)

    (2) You mention that “There may be an evolutionary explanation for what [why?] we find pleasure in various things, but it would not follow that the value of pleasure is purely instrumental. We’d have reason to seek pleasure even if, in this case, this would not benefit us in any other way, and even if our evolutionary history was utterly different.” I agree of course that, from our human perspective, it is very reasonable to conclude that pleasure has some substantial value in itself, that is, aside from the degree to which it motivates us to do things that contribute to other aspects of our well-being and species survival. We humans evolved to feel that way, and we do feel that way, and pleasure feels … well … pleasurable. So, from that standpoint, pleasure can be both ‘instrumental’ and ‘final.’ That said, relatively speaking, if a situation requires a direct choice between a substantial increase in someone’s ‘pleasure-for-pleasure’s-sake’ (as a final value) and another person’s basic survival, all else equal, the nod ought to go in favor of preserving the second person’s life, of course.

    As to the interesting question of what would be the case “if our evolutionary history was utterly different”, the answer to that, in my view, depends on the meaning of “utterly different.” If our evolutionary history was utterly different to the degree that we humans did not evolve to seek and enjoy pleasure, then we would have no reason to value it for its own sake, i.e., as a final value. For example, our evolutionary history was such that we did not evolve to like the feeling of being in freezing water, so we generally don’t value that feeling.

    (3) You write, “We could ask, for example, WHY people enjoy possessing such items, or why they might confer on us higher status. It might be that the best answer is that we enjoy them BECAUSE we appreciate the prior value they possess.” I’m not sure that I understand your point, and I might be missing it. When we ask “WHY people enjoy possessing” something, we go even deeper into ‘human nature’ (using the term as an admittedly ambiguous shorthand). If someone argues that the reason does not have to do with pleasure or happiness (or enhancing them) and does not have to do with instrumental value (e.g., to sell Lincoln’s pen to put food on the table), then what other reason might there be that is consistent with our understanding of human nature but that is not based largely on semantics? When you say that the best answer might be that “we enjoy them BECAUSE we appreciate the prior value they possess”, I’m not sure I understand your point: This phrase includes the words ‘enjoy’ and ‘appreciate’ within itself. And, I’m not sure what you mean by “prior value they possess”. For example, in the case of Lincoln’s pen, are you talking about the original cost of the pen that Lincoln paid, or the instrumental value of the pen associated with its use in writing (as any other pen would have), or some other ‘prior value’ that exists other than that which results from the ability of Lincoln’s pen to give pleasure to someone or provide instrumental value to them? (Or are we talking here about the very real differences between pleasure and happiness?)

    One more thought on point (3): Although I may be missing your point, if we did a hypothetical scientific poll with people to ask them why they would value owning Lincoln’s pen (or Diana’s dress, or etc.), they would probably give a range of reasons that could be categorized into “it would make me happy for X reason” and “I’d sell it so I could use the money to do Z.” If we then took the first category of answers (“would make me happy”) and pushed those people for deeper reasons, we would (probably) soon get to the point where we would either have to accept the expected/usual answers or find ourselves pushing for answers that somehow don’t lead us back to pleasure or happiness. As we know, people often can only go down into themselves so far in trying to explain why they feel a certain way, or why they would or wouldn’t do something. If you go (in research) beyond that point, other dynamics start to substantially influence answers.

    That said, I may be missing your point. If we pick a specific example, such as Lincoln’s pen, what type of value do you think people might see in owning Lincoln’s pen other than value associated with Lincoln’s pen’s ability to help them experience more pleasure/happiness and other than its instrumental value toward other ends (such as selling it to put food on the table)?

    I hope some of these thoughts are helpful or at least interesting. Sorry if I’m confused or missing the point.

  4. 4. Posted by Richard Chappell | August 9, 2007 2:26 am

    But what is the parallel of instrumental value for value of this kind? I call it expressive value. (This may not be the best label, but it’s better than the more familiar but narrower and seriously misleading ‘sentimental value’. Do you have a better suggestion?)

    Would “derivative value” capture the idea you have in mind?

  5. 5. Posted by Justin Blank | August 9, 2007 3:26 am

    Hopefully this question isn’t too obvious, but why not think that friendship is a final value which we ought to promote? If so, your examples would just show that we go wrong in conceiving of friendship as a value which is promoted by maximizing the number of people I can call “friend.” But often, it seems like we rightly take actions to promote (or at least maintain) our friendships, and that this is a reasonable response to the value of friendship.

  6. 6. Posted by Guy Kahane | August 9, 2007 2:56 pm

    Richard, thanks for the suggestion. Yes, what I called ‘expressive value’ would be a form of derivative value. But this label wouldn’t single it out. Instrumental value is also derivative. And I think that some of those who defend the idea of extrinsic final value would be happy to admit that it’s derivative — that, e.g., the value of Lincoln’s pen is derivative from the value we ascribe to Lincoln. They would however claim that this value can be BOTH derivative and final.

    Justin, I made only a brief reference to a view that is very much under debate. One place where the claim that friendship is not a value we should aim to maximize is in thomas Scanlon’s book ‘What We Owe To Each Other’. Thomas Hurka critically discusses Scanlon’s views in ‘Value and Friendship’. In any case, and without going too deeply into the debate, Scanlon’s view would be compatible with the points you make. Let me just note that to raise the possibility of expressive value, I just need that some things, with friendship perhaps one of these, are such that they call for certain attitudes. It is compatible with thinking that they also ought to be promoted (we could instrumentally value something that helps us make more friends!)

    Jeff, you raise very many interesting points, though some would lead to discussion of new and different topics. For example, the fascinating question about the continuation of the human race partly revolves about the moral status of merely potential people as opposed to existing ones, not so much about the notion of final & extrinsic value. About your last point, we may have a different understanding of what kinds of value might exist. Again, let me point out that we can make claims about the distinctive value some things have without denying they also possess others, less controversial forms of value. Some people would like to possess a Picasso painting only because they want to make money. But one main reason why it is worth much money is that many ascribe to it a high and prior aesthetic value. Similarly, rare historical objects in a museum have value not just because it’s pleasant to see them. You may disagree and you would not be alone. Hedonists, for example, think that only pleasure is an intrinsic (and final) good and other things only good because they cause pleasure. With you, they may deny these claims about further value. But in that case, they would be disagreeing both with me and with the proponents of extrinsic final value. So this would be a further debate.

  7. 7. Posted by Nick Shackel | August 9, 2007 3:44 pm

    I think it’s true that some things have a kind of value that is well termed ‘expressive value’ and that their value is related to what should be respected. I also think you’re right in identifying something analogous to instrumental value with respect to values to be respected. I’m not sure if these two are the same, and so I have some terminological and other worries.

    First of all, I think what you are calling expressive value just is a special kind of instrumental value, but well worth distinguishing for all that. You might say that it is not a means in the simple causal sense that, say, the causal power of a corkscrew is a means to wine. But some means are constitutive means, and your expressive value might be of that kind.

    Secondly, what I think of as having expressive value has final value because of it’s expressive significance. In this case I have in mind especially acts rather than things.

    Thirdly, let’s map the logical space. This seems one way to do it, but perhaps not in line with your thought:
    {value to be promoted, value to be respected}x{extrinsic, intrinsic}x{instrumental, final}

    your expressive value is {value to be respected, extrinsic, instrumental}
    whereas I’m suggesting the expressive value might also include {value to be respected, extrinsic, final}

    Another way, perhaps more in line with your thought, would be

    {value to be promoted, value to be respected}x{extrinsic, intrinsic}x{instrumental, expressive, final}

    when your expressive value is {value to be respected, extrinsic, expressive}

  8. 8. Posted by Guy Kahane | August 9, 2007 4:07 pm

    Nick, I need to think of this further but just two quick comments:

    (1) I’m not sure I want to deny that expressive value is a form of instrumental value. Or at least, the development of the parallel with instrumental value should give us a single type of value with two forms, ‘promotional’ and ‘respectful’.

    (2) About some acts having both expressive and final value — but I wonder there these wouldn’t also have INTRINSIC value. They get they value in part from their intentional content, not from their (extrinsic) causal relation to something of value. (The test case of course would be cases of expressive acts that are blamelessly directed at a mistaken object.)

  9. 9. Posted by Jeff Huggins | August 9, 2007 4:56 pm

    Guy, thanks for your helpful comments.

    I agree that the discussion mentioned in our posts involving the continuation of the human race, as well as the associated cases and a corresponding discussion of values, would be different from that intended in this thread. That said, I encourage the site (i.e., Ethics-etc.) to begin a thread on that subject, as I believe that the subject is rather foundational to a complete framework of morality/ethics, can lead to many fruitful explorations, and can shed some light on many of the more specific and detailed dialogues in some of the other threads. Perhaps a philosopher who is allowed by the site to begin a thread (by making a post) could begin the discussion, or perhaps the site itself could simply begin the discussion by posting the case (and its variations) from my recent two posts (without the other clutter in those posts, of course)?

    Regarding the discussion of ‘expressive value’ and your recent comments. I’d like to clarify one thing and then add a couple related thoughts that might be helpful in some way, hopefully.

    To clarify, I do see all sorts of value(s) in all sorts of things. I’m not a hedonist (although I do appreciate some pleasures and the times when I’m reasonably happy). I would very much enjoy owning Lincoln’s pen because I am an admirer of Lincoln and respect and value what he was about and what he did. I do many things for their educational value (to me and to others, hopefully). So, in my earlier post, I was not trying to suggest that only ‘pleasure’ matters. Instead, what I was trying to convey was something more like this: If you combine the notions of happiness, pleasure, fulfillment, and interrelated things (words that represent feeling ‘good’ in various ways) into a category; and if you set aside into another category (for this discussion) the instrumental value that any object or action might have toward our human well-being in other senses (including toward our basic survival and reproduction); then what other general type of value is there that cannot be linked, even indirectly, to these two categories (within the context of the discussion we’re having, that is)?

    Here is an example that might prove helpful and that perhaps demonstrates one of the loosest of linkages (to these other values): Regarding an object such as Lincoln’s pen, there is probably a value associated with someone, or some institution, owning it relative to the alternative that it was lost in the trash or burned in a fire. In other words, no particular individual needs to own it (the pen), and this aspect of its value anyhow need not be to a specific individual. There is value that the human race still has access to Lincoln’s pen rather than no access to it (i.e., if it had been destroyed). That value, of course, is sort of an inspirational, educational, reminding, historic value to all people who have access to the pen (in a museum perhaps) or are at least aware of it through books and History Channel documentaries. One could argue that this type of inspirational, educational, and historic value is very far removed from personal ‘pleasure’, especially if no person owns the pen (e.g., if the pen is owned by a museum). One could also argue that this type of value is fairly far removed from direct ‘instrumental’ value if that term is viewed in a narrow sense. On the other hand, it seems to me that the very type of person who might respect and value Lincoln’s pen most would value it for reasons that do ultimately tie back to one of these two broad categories, even if such person did not own it himself and had to see it in a museum. In seeing such a pen, the ‘respect’ and ‘hope’ that such a person might feel on the occasion would lead to some mixture of fulfillment/happiness/comfort and/or to a respect and appreciation for the fact that the pen could (and should) inspire others to help make the world a better place as Lincoln did. (Indeed, the pen could serve as a reminder to people of the ‘power’ that a pen can have (versus the sword).) So, in these ways at least, the value that an admirer of Lincoln’s pen might see in it would tie back to one or the other of these types of value, or probably some mixture of both.

    In any case, I hope these thoughts are helpful in some way.

  10. 10. Posted by Jussi Suikkanen | August 9, 2007 8:52 pm

    I thought a typical example of an object that has extrinsic final value would be a rare stamp. I’m not sure you can deal with this case with the expressive account. It’s not immediately obvious what the further more basic value would be the respect of which you could express by valuing the stamp. This is not to say that things do not have expressive value – clearly many things do. But, I’m slightly suspicious that this category would fill the logical space of extrinsic, final value.

    Also, I’m not very happy with the subjectivist reading of Korsgaard – it sounds a lot like Berkeleyan reading of Kant but that’s another story.

  11. 11. Posted by Guy Kahane | August 9, 2007 10:41 pm

    Jussi, thanks. Rarity was one thing I had in mind when I qualified the aims of this post. I have some thoughts on this but they’re still half-baked and would require a bit of space. Let me just point out that it isn’t the property of being rare that by itself confers value on an entity. Pretty much everything comes out rare under some description.

    Korsgaard’s subjectivism about non-moral value departs from Kant; as I read him, he was a hedonist about such value. But although Korsgaard is hard to pin down on many things, she is pretty explicit about her subjectivism in a number of papers, and certainly in the one I refer to above. Oddly enough, at that stage she held a hybrid view one which some things have only subjective value and others — rational beings — have objective value. In later work she’s given up this view and seems to have adopted a more general subjectivism. Although this is tangential to my post, I’d be interested to hear why you think that this is a misreading of her views.

  12. 12. Posted by Jussi Suikkanen | August 10, 2007 2:34 pm

    I agree that we need to spell out more about rarity. Rarity in stamps makes the object in that class valuable whereas rarity in snow-flakes does not. But, the same conditionality seems to for any good-making property. Even with pleasure, you might think that it is good-making only in so far as is not sadistic pleasure.

    But, this isn’t to say that these properties are not the final good-makers but rather that they too have enablers and disablers and are thus conditional in a way. As Kant already noticed, conditionality and finalness are compatible and often go together. That there are conditions under which rarity makes an object good does not imply that rarity itself does not make the object good. The crucial question is, are there any good-making properties that are relational but do not fall in the category of expressive values? The answer seems to be yes for me. Rarity seems to be one good example and I’m sure there are others (like being a winner). I’m not sure what the motivation would be to argue that there are none.

    I guess I’m not comfortable with the word subjectivism whereas constructivism or procedural realism (as she often puts it) would sound better. I take it that subjectivism is the view that a value claim is true if it accurately reports the de facto valuing attitudes of the agent towards the object. I don’t think this is Korsgaardian or Kantian view – even though I don’t want to contest your Korsgaard reading. In any case, I mean that you write:

    “On this theory, things have value only because we value them. The fact that we value them confers value upon them because we ourselves are intrinsically valuable. So most of the things we value only have extrinsic value, value derived from the fact that we value them”

    In a sense this is right, but in many other ways quite misleading. I’d like to put it by saying something like that the value of objects is on the Kantian view conditional on whether the objectes can be play a part in what a rational/good will wills or whether the object can be a possible object of good willing. If they can, then value is conferred to the object. Notice that this makes value objective in a way – independent of what the attitudes of actual agents happen to be. Even if there were no rational agents, things could be valuable if they could be rationally willed. And, what can be rationally willed does not depend on the content of the agents willling but rather on its formal, universal characteristics. In this way, value, via being conditional on objectively characterised willing, becomes objective too. This seems to be far away from the subjectivist views.

  13. 13. Posted by Guy Kahane | August 10, 2007 3:56 pm

    Jussi: In philosophy it’s hard to pin down labels on people. There are more people who are described as subjectivists than who describe them as subjectivists. In the end this doesn’t matter much as long as we are clear what is claimed and what isn’t. You describe ‘the Kantian view’ as neither properly subjectivist and neither fully realistic; that’s the work procedures and constructions are supposed to do. I agree with that but think we are talking about different things. Kantians are not properly subjectivist about morality. Granted. But we are talking about non-moral value. Do Kantians have a non-subjectivist account of that? They do perhaps go beyond subjectivism in denying value to what violates the moral law. Perhaps this rules out malicious pleasures. But it leaves us with a whole world of things that could be rationally willed: pains, pleasures, cakes of mud, blades of grass… It would be extremely interesting if Kantians could also ‘construct’ non-moral value in some plausible way. But no Kantian I know has tried, or even thinks they ought to try.

    In your comments, you use some notions I’m somewhat suspicious of: enables and disablers, conditions on value… I’d point out first if these notions can do work in value theory, they add further distinctions to those we’ve discussed so far. That is, if the goodness of pleasure is conditional, it can still be intrinsic. (Why should conditionality imply either ‘conferred’ or extrinsicness?)

    Even if we accept that all the above play a role in determining whether something has value, how are we to distinguish between them? That is, how do we tell whether it’s being rare that in itself makes a stamp a final value, whereas other factors only play an enabling role? No doubt they might, but this doesn’t in itself answer the doubt I was hinting at, the doubt that being rare doesn’t matter in itself. (The world doesn’t become a worse place if we discover that a stamp isn’t really rare!)

    It seems to me that if we reflect on specific examples where rarity seems to make a difference, we discover pretty complicated structures (think in this context also of the value of monogamy and ‘exclusivity’ or even about Walter Benjamin’s ‘The Work of Art in the Age of Mechanical Reproduction’). I don’t myself have any good account of this value.

    But the general point that I think was raised by my post is that we don’t yet possess a good criterion of ‘finality’. The ‘for it’s own sake’ vs. ‘for the sake of something else’ isn’t a good criterion and without one it’s a bit hard to resolve disputes about putative examples of extrinsic final value.

  14. 14. Posted by Jeff Huggins | August 10, 2007 6:19 pm

    Guy, you raise a great point when you write:

    “But the general point that I think was raised by my post is that we don’t yet possess a good criterion of ‘finality’. The ‘for it’s own sake’ vs. ‘for the sake of something else’ isn’t a good criterion and without one it’s a bit hard to resolve disputes about putative examples of extrinsic final value.”

    I’m not sure whether the following is the source of the same ambiguity or lack of ‘good criterion’ that you refer to, but it seems to me that the word ‘final’ itself, taken in its common meaning, points to different things depending on what perspective is taken, i.e., who is the ‘speaking voice’, as follows:

    • ‘Final’ takes on one meaning if we are viewing things from the internal perspective of a living human and referring to an individual human’s scope of experience. In that sense, the death of the individual is ‘final.’ And, because we humans enjoy and seek the feeling of happiness, even for its own sake, happiness can be seen as a ‘final’ value from this perspective. From this perspective, humans value a number of things in this ‘final’ sense, and the number probably depends on how finely we want to parse these various things. The ultimate number and identity of ‘final’ values in this sense, if one wants to go deep, may well depend on language, that is, unless someone wants to tie it to something universally measurable, such as the brain region that lights up (using some super-fine future measuring device) for each different ‘final’ value, allowing a value to be seen as a separate final value only if it lights up the brain in a substantially different way than do other final values. (I’m not suggesting this approach, at least not at this point in our understanding.) If we try to parse the ‘final’ values in this sense too finely, into many different ones, then unless we are willing and able to tie each individual ‘final’ value to something universally measurable (presumably of a biological nature), and especially if people of differing languages and cultures view these finely-parsed values differently, then the only way to resolve the issue would be through voluntary agreement to the definitions and parsing, i.e., as in an ‘incompletely theorized agreement.’

    • ‘Final’ takes on a moderately different meaning (with a different set of criteria) if we use a broader understanding of ‘life’ and ‘morality’, allowing ourselves to (and trying to) view things from a somewhat external standpoint (at least to the degree that our understanding helps us do so). From this perspective, many things that might be considered as ‘final’ values may not be (perhaps depending on other considerations) but, instead, are really instrumental values, along the lines of comments in my previous posts. Or, perhaps they would still be considered ‘final’ values in one sense (but less foundational, relatively speaking, than other ‘final’ values) while also serving as instrumental values toward the more foundational final values.

    • Of course, in the broadest sense, we don’t quite know what ‘final’ really is. From the standpoint of the human species, it could apply to the death of our species (i.e., so the ongoing life of the species would be a, or the, central ‘final’ value). From the standpoint of the universe, the only (or main) ‘final’ value may be the ongoing life (as in existence) of the universe.

    The case (and its variations) I posed in earlier posts calls for us to explore the differences between these perspectives (especially the first and the second) and understand the implications. Depending on what you meant by your comment (repeated above), this line of thinking may help sharpen the criteria of ‘finality’ to which you refer. (If not, sorry for the digression!)

  15. 15. Posted by Jussi Suikkanen | August 10, 2007 6:56 pm

    hmh. I think I’m thinking of Kantians as more Kantian than you. Yes. Kant did have a theory of non-moral value in Groundwork. It’s surprising how often his value theories are ignored. After all, he did start the book with the claim that only *a good will* has unconditional *worth*. By worth, he did mean special kind of value that was essentially a moral notion. Granted it does confer moral worth also to actions (if they result from a good will) and persons (who have the good will).

    But, he did think that other things can have conditional value – like happiness and health. And, such conditional values were non-moral in nature because moral value is always unconditional. Kant did have a value theory that was supposed to account for this value. As I argued above, this was a non-subjectivist account in the sense that the value of these objects did not depend on actual attitudes of willing. Here is Kant himself (Groundwork 4:428):

    ‘The ends that a rational being proposes at his discretion as effects of his actions (material ends) are all only relative; for only their mere relation to a specially constituted faculty of desire on the part of the subject gives the their worth’.

    This, as usual, is not the clearest passage. But, as I read it, the domain of value-conferring beings are the *rational beings* and their *specially constituted faculty*. So, any willing or desiring will not do in the sense it does for the subjectivists.

    You are right that this leaves a whole lot of things that can be valuable by being rationally willed. But, I don’t think it leaves the things many of the things on your list. Keep in mind that with ‘rational’ here we must mean what Kant meant by rational. Pleasure surely goes in and that’s fine – but some of the others it is difficult to see how they could be willed without a contradiction in will. Kant, after all, thought that there are necessary ends rational beings have like humanity (i.e., rationality) and in the case of humans also happiness. So, willing pain would create a contradiction (and eating mud I presume). The trivial activities like counting grass are a more interesting case. I’m sure there are ways for Kantians to try to construct contradictions out of them.

    You are right about conditionality. It’s a Kantian notion and he thought that many intrinsic values are conditional. I don’t think I said anything in opposition to that.

    The distinction between good-makers and enablers and so on wasn’t doing any real work in what I said. It’s just a nice way to present things. If we take the whole set of the good-makers and enablers in the rare stamp (without making the said distinction) and are left with any relational element at all that affects the value while thinking that the stamp has final value, we have an example of an extrinsic final value. There are others ways to explain why rarity in stamps is good but not in the shape of snow-flakes. One could be that they are different properties. Another would be that goodmakers are instantiations of properties and abstract universals.

    For what it’s worth, I think there is a working test for finality of value. One is to use the relevant counter-factuals. I think there is an even better one. We can look at the considerations that are the reasons for valuing the valuable thing. It seems plausible that these considerations correspond to whatever considerations make the object good. Now, my view is that we can trace the final value of an object to the basic or fundamental reasons for valuing objects. The distinction between basic and derivative reasons seem to be in some workable order. It seems that we have some grasp of why we value things.

    So, a basic reason for valuing a rare stamp seems to be its rarity. This would make it finally valuable. In the case of Lincoln’s pen, it seems that the reason to value the pen is based on the fact that the pen has a certain relation to Lincoln. But, it looks plausible that you are right about the case – the fact that the pen is in a relation to Lincoln seems to get its status as a reason derivatively from the reason we have for valuing Lincoln. So, that too seems to fit your idea.

  16. 16. Posted by Guy Kahane | August 11, 2007 12:14 pm

    Jussi, this is very helpful. One disagreement between us now turns out to be about Kant exegesis and thus won’t be resolved without a lot of textual detail. But let me indicate where my reading of Kant is different. A number of commentators have made much of Kant’s remarks on unconditionality and rejected the Rawlsian ‘priority of right to good’ reading. This is in line with your suggestion. Our disagreement however is about non-moral value. Since I’d prefer to leave out the Critique of Judgement, let’s focus on prudential value. As I understand Kant, he holds the following, (1) Prudential value = one’s happiness; (2) Happiness = hedonic states, more or less as understood by classical empiricists; (3) We have some moral duties with respect to our own happiness, but these are straightforwardly instrumental — things we need to do to sustain rational nature; (4) In themselves our ‘woe and weal’ are not even properly described as good or bad; (5) Our reasons to seek things that promote our happiness are merely hypothetical imperatives — if you want, are desire-based reasons; (6) As it happens, we all necessarily desires our own happiness, so we always have these reasons. Kant never explains in what sense this end is necessary.

    To the extent that Kant was a hedonic about prudential value, he of course wasn’t strictly speaking a subjectivist. As I see it, many contemporary Kantians see his hedonism as an embarrassment, and think that replacing it with a desire-satisfaction view is a step forward. (Actually, Andrews Reath has worked hard, unconvincingly to my mind, to try to show that Kant already held a desire-satisfaction view.)

    Also, while rational willing is no doubt different from mere desire, I’d say that for matters of meta-ethical classification I’d count a view on which rational willing confers value on things as subjectivist, so long that ‘rational’ here means no more than procedural rationality. (Most subjectivist views place some constraints on what attitudes ‘confer’ value.)

    I think you’re too quick to grant me the Lincoln pen case. I think advocates of extrinsic final value claim both that the value of such objects is derivative AND that it is final. On this view, X can have final value because of its relation to Y, which itself, has say, intrinsic value. So neither the counterfactual test nor the cited reasons test settles the dispute. I think a better test here might be whether in failing to respect the value of X one would also be failing to respect the value of Y. I think my view easily wins this test in examples such as wedding rings. You can’t (1) value your marriage, (2) value your wedding ring, but (3) use the ring in gambling or to flirt with stranger, without also showing lack of respect to your marriage or spouse. So the value of ring seems dependent on that of the marriage in a stronger way than the extrinsic final view can allow.

    I’ve been a bit evasive about the rare stamp example because I’d like to think about it further. I rather doubt we are all that clear about our basic reasons. Citing ‘because it’s rare’ as a reason and expecting the conversation to stop here doesn’t yet show that there is no further reason in the background (e.g. ‘because I’ll have something nobody else does’).

    Perhaps more important is the question your raised earlier, why would anyone even want to deny that there are extrinsic final values. That’s the fundamental question, and it’s not easy to answer. As others have noted, intrinsic isn’t the same as essential or necessary, and it’s final value that seems to do the real normative work. On the other hand, some of our central examples of final value are also of intrinsic value — persons, sentient beings, pleasure and pain… Rare stamps are not on this list…

    Jeff: let me just say that I think my question roughly falls in your first category of what ‘final’ might mean. For some of the reasons I mentioned above, we answer the empirical question of why people ‘finally’ value without answering the conceptual question of how to distinguish this particular valuing attitudes from others ones.

  17. 17. Posted by Guy Kahane | August 12, 2007 3:48 pm

    Chris Grau has kindly directed my attention to his very interesting and relevant ‘Irreplaceability and Unique Value’. Grau argues that unique value is a distinctive form of extrinsic final value where the valuable object is irreplaceable. At some point he considers the example of

    a crude drawing given by a child: it seems quite plausible to say that the drawing might possess little or no intrinsic value, and yet I might nonetheless value it in a noninstrumental manner (as an end or a “final good” if you don’t like the term “end” referring to objects) and take it to be irreplaceable.

    I think this is a good example of what I called expressive value, a form of non-final value. Objects with expressive value can certainly be irreplaceable in Grau’s sense.

    Perhaps more important, however, and relevant to my exchange with Jussi, is that some paradigmatic instances of unique value don’t seem to be instances of expressive value. The value one places on one’s loved ones would be an example. These would also be strong candidates for being extrinsic final values (uniqueness and rarity, while obviously distinct, are probably connected). And many examples of expressive value — such as wedding rings — seems to derive their value from their connection to things or persons with ‘unique’ value. If this is correct, then my claims about expressive value would not in themselves make a strong case against the very notion of extrinsic final value.

  18. 18. Posted by Jussi Suikkanen | August 13, 2007 8:26 am

    Yes. I’m sure there are differences in our Kants. I like Kant exegesis – I wish I was a Kant scholar sometimes… For instance, I don’t think Kant held (2). This is what he says about happiness in Groundwork:

    “All people have already, of themselves, the strongnest and deepest inclination to happiness because it is just in this idea that all inclinations unite in one sum. … Yet one can form no determinate and sure concept of the sum of satisfaction of all inclinations under the name of happiness”.

    To me this sounds more like the view that happiness is the state in which all our inclinations are satisfied than any sort of hedonism. So, I’m with Reath here. This fits the idea why he thought reason is so bad in helping us to pursue happiness – it’s quite blind to what the sum of our inclinations would be as satisfied (there’s a great passage about this in 4:418). Nature is much better in this. But, if happiness is the satisfaction of all our desires then it is no wonder that it is something we desire. So, I don’t think this is a replacement.

    And, I would think that he thought that happiness and the satisfaction of the ends of our inclinations are conditionally good on the condition of good willing.

    I’m not sure about that definition of subjectivism. Some people (Kant, perhaps Smith and some others) think that given enough formal constraints on rationality all rational agents converge in their willing (or advice or whatever). I don’t think the result is subjectivism.

    I don’t think I understand how a value could be final and derivative. These are supposed to be exclusive categories. An object can of course have both final and derivative value. I think these people hold that the final value of the pen is grounded on the relational property to Lincoln who was valuable. But, they would deny that the final value of the relational property is derived from Lincoln’s value – it is just a condition on the relational property.

    I don’t know assessing the examples is different. There’s so many of them – wilderness, fascinating persons, chart hits all kinds. And, some of them seem pretty plausible to me. I think there is an answer to the ‘fundamental question’. The usual motivation is to think that sole value bearers are propositional entities – facts, states of affairs, and so on. These are then thought to have independent existence and thus final value is intrinsic to them. But, if you think that concrete objects can bear final value, then this motivation is gone.

  19. 19. Posted by Jeff Huggins | August 13, 2007 2:55 pm

    Guy and Jussi, I’m not an expert on Kant, but I thought I’d briefly make some comments (from my standpoint) regarding happiness in light of what has recently been said about Kant in your recent posts.

    Guy writes: “As I understand Kant, he holds the following, (1) Prudential value = one’s happiness; (2) Happiness = hedonic states, more or less as understood by classical empiricists; (3) We have some moral duties with respect to our own happiness, but these are straightforwardly instrumental — things we need to do to sustain rational nature; (4) In themselves our ‘woe and weal’ are not even properly described as good or bad; (5) Our reasons to seek things that promote our happiness are merely hypothetical imperatives — if you want, are desire-based reasons; (6) As it happens, we all necessarily desires our own happiness, so we always have these reasons. Kant never explains in what sense this end is necessary.”

    The last sentence in particular—“Kant never explains in what sense this end is necessary”—is particularly interesting.

    Jussi writes: “For instance, I don’t think Kant held (2). This is what he says about happiness in Groundwork: ‘All people have already, of themselves, the strongest and deepest inclination to happiness because it is just in this idea that all inclinations unite in one sum. … Yet one can form no determinate and sure concept of the sum of satisfaction of all inclinations under the name of happiness’. To me this sounds more like the view that happiness is the state in which all our inclinations are satisfied than any sort of hedonism.”

    Jussi adds: “But, if happiness is the satisfaction of all our desires then it is no wonder that it is something we desire.”

    In my view, roughly and loosely, ‘Happiness’ is the summarizing name we give to the broad (and often multi-component) feeling we humans get and have (for a while)—and that we have evolved to have, that is, as a product of our evolutionary development—when we are (or at least think we are) being roughly successful at doing the things necessary for ongoing human survival, that is, for our own survival and for our successful creation of a ‘next generation’ that can continue the process. The quest for happiness is a very rough, imprecise, directional motivator. In the overall scheme of nature, it (happiness) is a big carrot. This notion (while not new, of course) has many implications relevant to an overall understanding of morality. It should not (and is not meant to) demean happiness. It just helps put happiness into an overall, broader, and more grounded perspective relative to other aspects and qualities of life such as ongoing survival itself, family, reproduction, human social-moral dynamics (and morality), death, and etc.

    It is correct, of course, that we humans seek feelings of happiness. But the reasons why we seek happiness should not be a mystery. One (that we all know very well) is that it feels good. But the other broader and more grounded reason—the reason that we can better understand by trying to take a perspective outside of our selves (as much as we humans can anyhow) based on what science can tell us—is as mentioned above, I believe. This second reason is the more foundational reason, i.e., the underlying reason and the one that explains the first reason.

    This also supports what Jussi is getting at (if I understand those particular comments). Put another way, in a sense, ‘happiness’ is what we humans like (and seek) by definition, because it is the name we have given to the (sometimes complex) feeling that we do, by nature, like and seek.

    Why do we naturally (and normally) like and seek this feeling that we have called ‘happiness’? Because doing so has worked from an evolutionary standpoint. Doing so has helped us survive and reproduce from generation to generation. Doing so reflects how we have evolved. Indeed, the feeling we call happiness (and its subsidiary feelings or components) almost certainly co-evolved with our other basic survival mechanisms.

Post a comment

Name: (required)

Email Address: (required) (will not be published)