Sometimes I hear people (well, philosophers) compare the goodness of one thing with the badness of another. They say, for example, the goodness of pleasure is less than the badness of pain; or pleasure is less good than pain is bad. Do such comparisons make sense?
Here’s a reason to doubt that they do. They seem similar to comparisons such as the following:
- Ben Wallace is less tall than Mugsy Bogues is short.
- Sydney is less hot than Glasgow is cold.
- Wood is less hard than feathers are soft.
But it’s not clear what these could mean.
Can anyone make sense of this?
[Update: I neglected to say before that my thinking about this issue was prompted by an interesting talk which Tom Hurka recently gave here at Edinburgh.]
How about: given a choice between having (i) a pain and a pleasure, or (ii) neither, people tend to choose (ii). I don’t think that’s necessarily true, but it would make some sense of the comparison, by suggesting that avoiding pain was more desirable than gaining pleasure. Hence, pain is more bad than pleasure is good.
The senseless examples are a bit unfair in that (with the possible exception of the cities) the property involved isn’t additive. Suppose good is additive. Then X is less bad than Y is good when X&Y is good.
With value, things can be value-neutral. But in your examples, I assume that no-one is height-neutral, no city is temperature-neutral, and so on. That is, value properties range, as it were, from zero downwards (badness), and zero upwards (goodness). But in your examples there is no “neutral zone” (or at least no non-arbitrary one). If this is right, then we can make sense of how X can be more good than Y is bad by analysing this in terms of how far it strays from the neutral zone.
So a better comparison would be with height above or below sea-level. One can legitimately ask whether, say, planes fly at a height greater than the depth to which submarines dive to.
i’m sorry, but how are we to compare what is good with pleasure, and what is bad with pain? are these things logically equivalent with one another?
and the examples seem to be relational values, not ethical. so do they really apply?
I don’t think the comparisons make sense in abstract. Pleasure and Pain per se do not have amounts of value and disvalue that could be compared. Instances of pains and pleasures do though – we can ask was the pain you experienced at the dentist worse than the pleasure you got from having shining teeth.
This shows that the pleasures and pains need to have intensities, durations, and so on before they can be compared.
This makes me think that what those philosophers have in mind is to take a pain and a pleasure that have the matching intensity and duration (is that possible?), and then ask is the instance of this pain less more bad than the instance of pleasure is good.
I think Alex is right in that this comparison requires that there is a level of zero value or indifference. I think the goodness of the instance of pleasure vs. badness of the instance of pain can then be tested as with betting behaviour. Would you be indifferent between getting the value neutral option certainly or 50% chances between the imagined pain and a pleasure (assuming that you are risk-neutral)? By changing the chances we could test whether we think that pleasures are better than the pains bad.
I think the correct analogy would be whether you have deposited more money in the bank than you owe money to the bank in debts.
Ben Wallace is less tall than Mugsy Bogues is short.
Something like what Jamie says seems right. But it might apply in these cases too, where the distance from Mugsy Bogues height to a height (or range) that is neither tall nor short is greater than the distance from Wallace’s height to a height (or range) that is neither tall nor short. Of course here the measures are much less precise, but they do make sense.
Surely it depends on the context, but the original post to me sounds like it isn’t about comparing discrete pleasures and pains. I think the idea is this: other things being equal, I’d rather avoid pain than incur pleasure. So, in accord with the Millian idea that if persons who know both systematically prefer one thing over another, it must be better, I might say that it is better to avoid (the badness of) pain than it is to have (the goodness of) pleasure. Another way of saying this: the goodness of pleasure is less than the badness of pain. I’m not saying this is right, I just think it makes sense from a certain point of view.
Good comments. A couple of quick replies.
To Jamie:
Perhaps height is additive. To get the height of Ben&Mugsy place Ben and Mugsy end to end (probably, best to put Mugsy on top). Then we might say: Ben is less tall than Mugsy is short iff Ben&Mugsy is short (that is, short for a concatenation of two people, not short for a single person).
To Alex:
I agree that the existence of a neutral zone (or neutral point) is important, and I like your plane/submarine example. As Mike points out, there might also be a neutral height: the height of someone who is neither short nor tall.
Suppose, however, nothing is neutral in value. If to be good is to be better than something neutral, then nothing is good. And if to be bad is to be worse than something neutral, then nothing is bad. But isn’t it possible for there to be good things and bad things without any neutral things?
But isn’t it possible for there to be good things and bad things without any neutral things?
You mean like, it’s possible for there to be things longer than a meter and things shorter than a meter even if there aren’t any things that are exactly a meter?
To be honest, I’m not sure whether this is possible. But why is this relevant? You’ll have to explain; I’m too slow. I get the feeling your question wasn’t meant to be friendly.
If I remember this right, there must be a natural point of being neutral in value, if it makes sense to say that one pleasure is twice as good as another. That is, without a neutral in value saying that would not make sense. So, saying that 20 degrees in Celsius is twice as warm as 10 degrees is Celsius does not make sense as 0 degrees in Celsius is not the neutral point in temperature. But, saying that 2m is twice as long as 1m makes sense because 0m is the zero point in length. So, if you want to argue that there is no such a thing as neutral in value you are committed to thinking that nothing is twice as good as something else.
Jussi,
Yes, that’s right, though we should distinguish two conditions:
It’s only the first condition which is necessary for us to compare ratios of value.
(I suspect this is related to Jamie’s last point about metres.)
Campell,
You wrote “Suppose, however, nothing is neutral in value. If to be good is to be better than something neutral, then nothing is good. And if to be bad is to be worse than something neutral, then nothing is bad. But isn’t it possible for there to be good things and bad things without any neutral things?”
I think the problem here is the accounts of good and bad. Why not just think that to be good is to be better than some possible neutral thing? (Ditto for bad)
This seems plausible and we can then accept that it possible for there to be [good things and bad things but no neutral things].
You might then switch your claim and say this: “But couldn’t it be the case that neutral things are impossible.”
But that seems implausible.
Brad,
Suppose I say ‘every possible thing is good’. This is implausible, perhaps, but not incoherent. It is, however, inconsistent with your suggestion, that to be good is to be better than some possible neutral thing (assuming nothing could be both good and neutral).
Something strange has happened. I replied, briefly, to Campbell’s 6:37 am and it showed up as posted when I looked, but now it’s vanished.
Anyway, I was just getting clarification. So the point is that for there to be things that are good, it isn’t necessary that there be anything that in fact is neutral in value. What’s necessary is that there be a neutral point on the value scale. (Just as, for there to be things longer than a meter it isn’t necessary for there to be anything that is exactly a meter long.)
This is probably just reiterating some of the previous comments. But I think in situations where there is a natural quantitative measure _and_ a mean value of said measure. Then one can easily talk about a deviation below the mean being greater or lesser than deviation above the mean.
Of course, agreed upon quantitative measures of pleasure or pain are hard to come by. But for the temperature example, I think people could easily say something like “Sitka, Alaska is colder than Nashville, Tennessee is hot,” and mean “Sitka has a mean temperature which is farther from the mean temperature of, say, US cities than Nashville does.”
The stuff about neutral points is interesting, but the same problem comes up when comparing only goods, right? If a pluralist about goodness says that virtue and pleasure are both good, but virtue is better, what could this mean? It would be like saying that gold is worth more than dollars.
Ben,
well – that depends. The term ‘goods’ seems to be used in two different meanings. Some people use it to mean things that are good, i.e., things that have the same evaluative property of intrinsic value or some such. On this view, saying that virtue is better than pleasure would be akin to saying that houses are more expensive than cars. I would understand the claim to mean that we have more reason to pursue virtue than pleasure.
There is another understanding of ‘goods’ which I don’t very much like. On this view, different goods have different evaluative properties that are incommensurable. Here it would be better to talk about different values than different goods. This would give you the odd result you refer to. But, I would like to hear an argument for why the goods of pleasure and virtue (i) have only different evaluative properties and (ii) why those properties are incommensurable.
Jamie,
Good, we agree. I think it could be possible to define a neutral point on our scale even in the absence of any neutral things. Suppose for example the following conditions hold: every thing is good (and therefore not neutral); there is no least good thing; and the goodness of things has a lower bound. Then we could define neutrality as the greatest lower bound. Does that work?
I think this mostly undermines my earlier objection to Alex’s suggestion.
Ben,
Excellent point. If anything, your case seems more problematic. At least with pleasure and pain we can make some sense of quantitative comparisons, in the way Jussi suggested earlier. We can understand what it would be for a given quantity of pleasure to be greater or less than a quantity of pain (in terms of duration, intensity etc.). But how do we make quantitative comparisons of pleasure and virtue? Jussi’s example of houses and cars seems different, because we can simply count the number of houses or cars.
Jussi,
I had the first thing in mind. I’m not getting your reply. What does it mean to say that we have more reason to pursue virtue than pleasure? “Houses are more expensive than cars” makes sense because we know what one house is, and what one car is, so we can compare their average prices. There’s no non-arbitrary unit of pleasure or virtue.
Campbell,
Even in the case of pleasure and pain, it’s not obvious how the intensities get compared. (Durations are easier.) Is there a unit of intensity that applies across types of mental states? If not, we have the same problem.
Campbell,
I fear I am being dense.
You say “Suppose I say ‘every possible thing is good’. This is implausible, perhaps, but not incoherent.”
Consider the horrific treatment of prisoners in concentration camps. That was very bad and not at all good. It actually happened. Therefore it is impossible for every possible thing to be good.
The fact that my suggestion is inconsistent with something that is impossible is no worry.
Is modal sloppiness causing problems?
Ben B,
Sorry for not being clear. I don’t think we need non-arbitrary unit of either pleasures of virtues. But, again, we cannot compare these in abstract. So, we have to take an instance of pleasure and an instance of virtue. So, say that our one option is to give one person a day at a spa. She is can describe in fair detail how pleasant that would be for her and we have fairly good idea of how much pleasure that would be. Another option for us would be to make a person who lies fairly often much more honest. Again, we would have reliable information about how many lies this would prevent and what kind of lies.
Now, you agree that both these options have amounts of single evaluative property – goodness. My thought that we can use reasons for choice to assess how much that property they have. Let us assume that by using the same amount of resources we are free to allocate to either person (and no personal ties or anything like that). The question is what do we have more reason to do – give one person a day at spa or another the given amount of more honesty. OF course, this example is not detailed enough for us to form an informed opinion about this but I see no principled reason for us to come a conclusion about which choice we have more reason to opt for. And, my idea was that this gives as the conclusion which one of these instances of pleasure or virtue is better, i.e., has more of the single evaluative quality we accepted these goods can have. None of this seems to require units of pleasure or virtue as such. The monetary value was just meant to illustrate the single scale or difference in amount of one property two options can have.
Each of the original three examples makes sense to me.
I haven’t read through the comments, and I suspect somebody’s probably already said this, but FWIW …
I understand them each basically the same way. I imagine a number line representing points along a scale measuring degrees of some attribute (height, average temperature, hardness), where ‘0’ picks out the baseline, ‘F’ is the predicate that names things on the positive side and ‘G’ is the predicate that names things on the negative side. To say that x is more F than y is G is to say that (x’s F-value + y’s G-value)>0. To say that x is less F than y is G is to say that (x’s F-value + Y’s G value)<0.
John,
You’re absolutely right: that has already been said, a few times. 😉
Campbell,
At least I got something right!
Wow, Ben, look:
http://investmentips4all.blogspot.com/
Gold is worth a lot more than dollars.
Jussi, right, if you’re comparing particular episodes, the claim makes sense.
I just discovered this blog and will definitely be coming back!
Would be be unfair to say that Barack Obama is less intelligent than George W. Bush was unintelligent?
I’ve been going about all morning and have just come from an interesting ethics debate at http://www.clever-bitch.blogspot.com
Adam Smith made sense of it in 1759, in much the same way as others have indicated here:
(OK, this isn’t about goodness and badness exactly, but it’s just a short step away)
By the way, I don’t think I agree with Jamie that good and bad are additive, or that this has anything to do with the possibility of comparisons of magnitude between them. If we accept:
Better to have loved and lost than never to have loved at all.
Can’t we consistently also hold the view that losing is more bad than loving is good?
No.
Brad–
My objection was that your suggestion makes (1) an analytic truth, which it isn’t.
(1) Possibly, something is not good.
Someone who denies (1) might have a false and implausible view of what is good, because, as you say, there seem to be obvious examples of things which are not good, e.g. concentration camps. But such a person might nonetheless understand perfectly well, or as well as anyone, the meaning of ‘good’.
Does that clarify things?
Jussi–
When the pluralist says virtue is better than pleasure, presumably they do not mean merely that some instance of virtue is better than some instance of pleasure. That’s too weak. Could they mean that the average goodness of instances of virtue is greater than that of instances of pleasure? I don’t think so; that’s too contingent.
I think they could mean normal instances of virtue and pleasure where normality is not a statistical notion. It could be something like the claims of the type: lions are more vicious than gorillas. This could be true even if all lions were tame ones that lived in zoos and gorillas still lived in the wild and were somewhat vicious. There are difficult issues of how to elucidate this sense of normality.
Simon,
1. I didn’t say that good and bad are additive.
2. Your example is supposed to show that additivity has nothing to do with the issue, but it seems to show the opposite. My suggestion works iff good is additive; in the ‘loved and lost’ example (as you gloss it), the suggestion doesn’t work.
3. Finally, my intuitions about the example actually support additivity. That is, when I imagine a case in which it is better to have loved and lost than never to have loved, I think I am imagining a case in which the goodness of loving is greater than the badness of losing. At least it seems to me that a conflicting view needs some kind of explanation.
Jamie,
I thought you implicated it. Why else would the examples be “a bit unfair in that … the property involved isn’t additive?”
I didn’t say that. I said additivity has nothing to do with the possibility of comparisons of magnitude.
OK, I’ll bite. Start with agnosticism about additivity. Now, we may say that losing is more bad than loving is good, since our existing relationships have intrinsic value that merely potential relationships cannot have for us (we might say: losing is to be avoided more than loving is to be pursued). But, consistent with this, we may say that someone who has enjoyed and since lost an intrinsically valuable relationship has had a better life than someone who has never had one.
Simon,
Here is how the dialectic looks to me.
Campbell asked what sense can be made out of a certain idea. I pointed out that if good is additive, then there is a natural suggestion for how to make sense of it. (I added that the analogous examples aren’t fair because their properties are not additive; Campbell thinks height might be additive, and I take that point.)
In response, you simply assert that the idea makes sense even without additivity, but you offer no suggestion for how to make sense out of it.
Is this how you understand the dialectic?
Jamie, apologies for misunderstanding the nuanced implicature of your first post.
Why do you think the example in my last post fails to make any sense out of the idea of good/bad comparisons without additivity?
Simon,
Well, I’m not sure how an example is supposed to make sense out of the idea.
Maybe it was the idea that “losing is to be avoided more than loving is to be pursued”? But how is that supposed to make sense? Here’s how I would try to make sense of it. If you are faced with the possibility of both losing and loving, where the alternative is neither, then choose the ‘neither’ alternative. But plainly this is not what you meant (since you recommend taking the ‘both’ option — as I understand your example). So what is the sense in “to be avoided more than”?
I’m doing my undergraduate thesis on *exactly* this topic for an independent major in Linguistics and Epistemology at Vassar.
I’m focusing on the types of sentences mentioned in OP – what Christopher Kennedy in “Projecting the Adjective” (1999) calls “cross-polar anomalies” and another type of sentence he thinks is responsible for the “incommensurability” problem, i.e.:
“My copy of ‘The Brothers Karamazov’ is heavier than your copy of ‘The Idiot’ is old.”
My intuitions tell me that these sentences do in fact have truth values, and that this reveals something about the semantics of gradable adjectives, but as others seem to have other intuitions, I decided to ask the English-speaking public to tell me what they think.
My thesis itself is a work in progress, but you can view/take the survey at:
https://www.surveymonkey.com/s.aspx?sm=zcBs1Nc9wQ2JSi8_2fk2HIYg_3d_3d
I would very much appreciate anyone taking the survey or forwarding the link to those colleagues or students who might be interested.
Anyone who would like to get in touch or would be interested in learning the outcome can use my Vassar email:
katracy@vassar.edu
Post suggestion from a lay reader:
Best Introductions to Ethical Theory
or
How to Catch Up with Contemporary Ethical Theory
Basically, if you could tell laypeople what to read in order to have the philosophical tools, background, terminology, and understanding to engage contemporary ethical debates, what would you point us to?
Jamie – sorry I’ve been away a little while. As to your question:
“How is losing to be avoided more than loving is to be pursued?”
The answer is that avoiding losing (what you love now) should be a higher priority for you than (initiating) loving. If you love A, and are faced with a possibility of either losing A and loving B, or neither, then you should indeed choose neither. The claim that losing is to be avoided more than loving is to be pursued, in this sense, is thus not in conflict with the claim that it is (or that it would be) better to have loved and lost than neither.
Kate – I answered your interesting survey and hope my possibly non-standard intuitions won’t mess up your results 😉
Kate–
Thanks for your comment. I wasn’t aware of the linguistics literature on this topic when I wrote the original post.
If others are interested, there’s a paper by Chris Kennedy available here:
http://home.uchicago.edu/~ck0/docs/salt97.html
(Hat tip: Billy Dunaway.)
Luke–
Thanks for your suggestion. Perhaps someone will take it up. In the meantime, you might check out this nice blog:
Ethics Bites
Simon,
Hm, I think you are comparing different prospects in the different comparisons. In each comparison, the goodness seems to be additive.
First, in the case in which you love A and you might stop loving A and start loving B, the idea is that the badness of losing A is greater than the goodness of loving B. These are additive, so you should choose Neither instead of Both.
The ‘better to have loved and lost’ case has a different comparison. There, the good of loving (A, let’s say) is greater than the bad of losing (A), and the value is again additive, so you should choose Both rather than Neither.
What if someone said, “No, the good of loving A is not as great as the bad of losing A, but still it is better to do both than to do neither”? This would be hard to understand, I think.
Hi Jamie,
Thanks, that’s helpful, I see how it would be hard to understand. But on the reading you suggest, don’t my hypotheses that:
(i)Losing is more bad than loving is good.
and
(ii) It is always better to have loved and lost than neither.
Entail:
(C) Successive loves and losses have diminishing value.
or, as Cat Stevens put it:
(C’)The first cut is the deepest.
Which is a surprising (if satisfying) result!
Seriously, I remain somewhat confused about how the kind of “status quo bias” represented by (i) should be understood, since I doubt that (C) should follow here. But I now see your point about the difficulty of denying additivity when making a comparison.
I never knew that was a Cat Stevens song. I’d always thought it was written by Rod Stewart (though the Sheryl Crow cover is obviously much better).
I think the status quo bias is an under-appreciated centeredness in value (Campbell has a view about this, appropriately enough). I have three children. The value of having these very three children is much higher from my actual perspective than it is from the perspective I would have occupied had I decided to delay having children (and ended up with three girls, say).
I believe John Broome eschews this move fairly explicitly in Weighing Lives, for very general methodological reasons, but I may be mixing up two different things, I’m not confident.
Campbell – I’d be much obliged if you could explain how to undo that last syllogism I wrote! (Or at least point me in the direction of the bit of writing of yours that Jamie just referred to, because I can’t find it.)
Jamie,
Please. Neither Cat Stevens nor Sheryl Crow can hold a candle to Rod (the ’70’s Rod’ that is.)
That’s weird, Robert. For some reason I’d always thought you liked music.
Hey, way off topic: Some PEA Souper should write the definitive reply to David Brooks NYT column on the end of philosophy.
It is difficult for some to understand just how good Rod once was given how bad almost all of his later work is.
I think Rod’s greatness was inexorably tied to Ron Wood. Without Wood, there was nothing to keep him from going Elvis.
The banality of Stewart’s lack of greatness as a song writer is exceeded only by the profundity of Wood’s lack of talent as a musician. Their product, however, is considerably smaller than the square of Faces’ unoriginality.
You’re just trying to bait van Roojen.
Hello everyone,
Thanks to all who took my survey or read my post.
I’ll post some summary results when all is said and done, but for now it’s actually looking like Kennedy is absolutely right, at least for the sentences and situations I’ve tested.
I haven’t accounted for standard variation or margins of error yet, so we’ll see.
Meanwhile, I’ve come across a paper that claims that not all cross-polars are anomalous. It doesn’t touch upon ethics, but it has to do with my thesis topic, and with the semantics of adjectives.
“Cross Polar Nomalies,” Daniel Büring, 2007
http://semanticsarchive.net/Archive/GNjMjljY/buring.SALT07.nomalies.pdf
-Kate
It’s like the old joke, “outside it is colder than tonight”
That’s funny, and I missed it entirely for more than a month!