Thursday, December 11, 2008

Dogs and a Sense of Fairness

There is a study out that reports to show that dogs have a "sense of fairness". (See, Associated Press:Studies Show Dogs Have Sense Of Fairness)

Since I deny that a sense of fairness exists, I naturally have some objections to the claim that a particular study shows that dogs have that which does not exist.

I object to moral sense theories in general. Moral sense theories are theories that say that moral properties exist "out there" in the universe, and that we have a moral sense organ that allows us to detect these properties. In much the same way that our eyes see trees and cars, our "moral sense" sees fairness and unfairness – a theory that says that fairness and unfairness exists "out there" in the same way that trees and cars do.

The problem with moral sense theories is that we do not sense moral properties. We only sense our own likes and dislikes. Of course, it is tempting for us to view our likes and dislikes to be "moral properties" that exists "out there".

That interpretation then gives us a justification for forcing others to act in way that fulfill those particular desires. "I am not forcing you to do X merely because I like some result. I am forcing you to do X because it realizes something that (I sense) to hold a moral property that makes it worthy of being promoted, even if I did not want to promote it."

It’s like the argument, "I'm not condemning you for your homosexuality. God is. I am simply following God’s will." In fact, God has no will. God has no morality other than that which the theist assigns to God. So it is not "God's will" the theist is enforcing – it is his own prejudices.

So, a study that claims that dogs have a "sense of fairness" is as objectionable as saying that dogs have the capacity to sense God’s presence. In fact, why not use this explanation? What the dogs are sensing is not "fairness" but "God's will." Recognizing God's disapproval, and having an innate disposition (provided by God) not to do things that God disapproves of, they avoid that which we call "unfair". That is why the researchers got the results they did.

Ask the researchers why it is that the "sense of fairness" theory is better than the "God's approval" theory. The only difference between the two theories is the assumption that "fairness" exists as an entity to be sensed and that we (including dogs) have an organ capable of sensing it, versus the assumption that God's approval exists as an entity to be sensed and that we (including dogs) have a capacity for sensing it.

Both options represent equally good "science".

I would like to add that I am a moral realist – I think that moral properties do exist. I simply deny that that both divine command theories and moral sense theories are mistaken about what moral properties are. The entities those theories refer to do not exist. Clearly, other alternatives are available.


CrypticLife said...

So, a study that claims that dogs have a "sense of fairness" is as objectionable as saying that dogs have the capacity to sense God’s presence. In fact, why not use this explanation?

I don't think "sense of fairness" is being offered as an explanation here, but as a description of behavior. The reason you wouldn't appeal to a deity to explain it is because you might be able to explain it in other ways.

Friar Zero said...

I recently stumbled upon a blog by a political science professor that argues that his concept of darwinism fills in the blanks of aristotles virtue theory. You can find it here.

I thought this might be germane to the discussion as he seems to be advocating exactly what your addressing here and in the last post.

David Parker said...

Just stumbled across you blog a few weeks ago, and have enjoyed reading.

The entities those theories refer to do not exist. Clearly, other alternatives are available.

What alternative(s) would you suggest?

Anonymous said...

The problem with moral sense theories is that we do not sense moral properties. We only sense our own likes and dislikes.

Sorry, I have to strongly disagree with this.

The same argument can be used to declare that there is no such thing as a sense of humor, because the only thing capable of distinguishing between something that is funny and something that isn't is inside a human brain. Obviously, we don't sense humor, we only sense our own likes and dislikes, so there is no such thing as a sense of humor!

If we understood humor well enough, we could write a computer program that reads jokes and tells us whether or not they were funny.

"Fairness" is similar to humor. If men get paid more for doing the same work than women, that's not fair, regardless of whether one uses a fairness detector, human or not, on it - or whether or not one cares about the unfairness of the situation.

Fairness is a vaguely defined property, but it does exist, in that situations can be classified by people as "fair" or "not fair", and the classifications are consistent and appear to be meaningful.

However, being able to detect properties and caring about them are entirely different things. "Fairness is intrinsically valuable, regardless of whether or not people care about it" is a proposition of questionable validity at best, although I would say that people generally have reason to promote a desire for fairness.

Martin Freedman said...

Hi Doug S.

Some very interesting points although I "strongly disagree" with your position :-)

You actually make IMHO two separate arguments so I offer two responses.

A sense of humour
Here I think the issue is that a "sense of humour" is another metaphorical extension of "sense" but in a different direction to "moral sense". That is it is better to regard humour as an emotional response more akin to others such as, say, anger.

A person can act to make someone angry, as a person can act to make someone laugh. Someone can become angry at the unintended (to cause anger) actions of a person, as someone can find funny the unintended (to cause humour) actions of a person. Some people have anger problems, whereas others do not. Some people have a sense of humour whereas others do not. And so on and so forth.

A sense of fairness
Now when it comes to a "sense of fairness" this is actually a different point, since either fairness is a token or sub-species of a moral sense or, maybe, one might consider fairness more basic, as a driver or proximate antecedent of a moral sense perhaps?

Of course people attempt to provide an argument as to why a situation is unfair, as they would to why a situation is immoral. Both can make reference to facts but there is no (un)fairness property intrinsic in the world as there is no goodness/badness property either. So your fairness argument IMHO begs the question.

Anonymous said...

Is "intelligence" a property that exists "out there" in the universe? It's at least as difficult to measure as "fairness", and similarly difficult to define. Yet we can say that a human is more intelligent than a chimpanzee, and a chimpanzee is more intelligent than a rock. We can even compare the intelligence of two different humans. It would be silly to say that intelligence does not exist.

What I claim is that fairness exists, but it is not a moral property in the sense that it has mysterious intrinsic value. It is a property, because, like "intelligence", some configurations of atoms have fairness and some configurations of atoms do not. People "sense" fairness in a way similar to the way they can "sense" kindness, truthfulness, or logical consistency. We don't literally sense fairness with a sense organ the way we sense light or sound. We infer the presence or absence of the rather abstract property of fairness based on the computations our brains perform.

However, fairness does not have any value independent of desires, because desires are the only things that give value to states of the universe. Fairness is relevant to D.U. morality because 1) the desire for fairness is a malleable desire, and 2) the desire for fairness is one that people generally have reason to promote.

(Incidentally, would a device that could scan a brain, list its desires, and compute whether those desires are good or bad in D.U. terms be like a moral sense organ in the way that a camera is like a visual sense organ? Wouldn't any organ that could distinguish between good desires and bad desires be a moral sense organ?)

Anonymous said...

In other words: if ethical naturalism is true, then it should be possible to build a moral sense organ, in the sense that a computer with a microphone and voice recognition software has a speech sense organ. Desire utilitarianism implies ethical naturalism, so...

(Or am I misunderstanding what is meant by a moral sense organ?)

Martin Freedman said...

Hi Doug S.

What you are now saying more or less supports Alonzo's position.That is fairness is a relational property, that is for person A with certain desires finds state of affairs S fair, where as person B with a different set of desires finds S unfair. We can do this because we have evolved cognitive and connative capacities to do this. Can we fully explicate this or do we, most of the time, just "sense" it - that is use cognitive and connative heuristics and so on?

Taking "sense" literally there is no distinct "sense of fairness" but taking it metaphorically we can talk about a "sense of fairness". However, when challenged, one can hopefully explain why a state of affairs is (un)fair which would be an explanation containing (tacitly or explicitly) our desires and the state of affairs under debate. In doing so, we would not be using a sense of fairness organ but would utilising whatever the relevant cognitive and connative capacities we have to explicate the (un)fairness of the situation.

Dogs cannot do this - rationally state why a situation is unfair - but they can certainly act in ways congruent to us in suitably designed situations. The mistake made by researchers is to infer that there is a "sense of fairness" in dogs, by confusing our metaphorical usage with respect to us.

As for a "moral sense organ", the standard usage in ethics is that such an organ detects intrinsic values. Since you agree that these do not exist, should one talk about designing a "moral sense" organ in a machine, such as a BDI agent - but this would be analysing extrinsic relations - between desires and states of affairs? I hold that we could design such a machine (a different topic), and we could choose to use metaphorical extensions of "moral sense" to talk about such a module. However one would have to be careful as this is not what is meant by "moral sense" in ethics literature nor "sense" in the biological (ahem) sense.

Anonymous said...

I think we're getting closer to agreement here.

Dogs don't have a literal sense of fairness in the way they have a sense of smell. However, they have an aversion to being treated in a way that we humans would describe as "unfair", so it seems reasonable to conclude that their brains are making similar inferences to the ones made by human brains and monkey brains.

Apart from the evidence found in this experiment, it seems reasonable to assume that dogs would have some kind of aversion to unfair treatment; dogs are descended from, and are very similar to, wolves. Wolves are pack hunters which routinely face the problem of how to distribute the food from a successful kill among several pack members - and demanding a "fair" share, but no more is (presumably) an evolutionarily stable strategy.

Martin Freedman said...

Doug S.

I think we are getting close enough too. It is always nice having a constructive debate.

I certainly grant that our notions of fairness have pre-human antecedents and am not surprised to find higher mammals exhibiting capacities such as an aversion to unfairness, albeit of a far more constrained and here and now variety. This supports my thesis that we have evolved cognitive and connative capacities - for beliefs, desires and dispositions - it just that humans can reflect upon and abstract from those and then reflect upon those abstractions and end up operating with a far more expanded temporal and spatial scope than other animals are capable of.

Still the point that Alonzo made and I agree with is, as I put it, that researchers can sometimes confuse some of our abstractions - metaphorical extensions - with literal usage and this is sort of what Alonzo was complaining about here.

And yes you have given an empirically plausible explanation (descent from wolves as social animals) as to why dogs have a (limited) aversion to unfairness. I think Alonzo was correctly concerned with avoiding the mistake often made to selectively chose only those animal features that support some evolutionary based theory of morality, as Alonzo highlighted in his recent post on altruism (albeit he was focused there on how that did not avoid Euthyphro). Eliminating such mistakes as these researchers made prevents their conclusions being recruited to support such arguments.

Looking forward to future constructive debates with you.

Anonymous said...

I don't think that the original article was suggesting there was a "fairness" faculty in any creature, and that we would all seemingly agree therefore with Alonso's original assertion that there are no moral properties "out there" in the universe.

I also agree that our likes and dislikes (actually all our emotions play into this not just these) play an important role. However, I don't think we can go directly from "like" to "think fair". The first is mainly emotional, the second ism mainly rational. Indeed in practise, we might often think things "fair" (like paying for damgage we cause to our neighbour's property)even if we don't particularly like it. Similarly, we might like a situation even if we accept it being unfair (eg the ref calls time a few minutes early just before the other side gets their shoot for goal to equalise the match).

In the case of dogs, I don't think that they think in terms of "fair" or "not fair" at all - that is far too abstract. Rather, as we observe in ourselves, every emotion has some end behind it - something I am attracted to or repelled by. Deprive Fido of an expectation of food (or show him Rex getting it instead) and you will observe the natural emotion of anger.

In the case of humans, we would also display anger, but this would be compounded by our "sense" (don't take this literally) of injustice, which is our reason examining the sense data and saying that the course of action taken by the person dishing out the food is not only depriving me of food but is also unreasonable, ie against reason.

Alonzo Fyfe said...

Thanks to all participants for an interesting discussion.

I am a moral realist - I take moral properties to be real (though I take them to be relational properties and not intrinsic properties). So, "fairness" is real and, as such, is something that can - at least theoretically - be sensed.

However, the claim that an entity has a sense of fairness is a much stronger claim than the claim that fairness can, theoretically, be sensed.

The article cannot be separated from a social context in which moral sense theories are very much alive - theories that make claims like "let your conscience be your guide" where, if you "sense" something to be fair or unfair, then this should be taken as a reliable indicator that it is fair or unfair.

Yet, moral sense theories such as "let your conscience be your guide" run up against real-world observations that show people having no problem whatsoever doing attrocious things. Jihads, crusades, slavery, genocide, all sorts of abuse and crimes committed by one person against another, a large number of whom can reliably be said to have "let their conscience be their guide".

One of the observations in the article is that, while dogs will accept a reward when given, monkeys will refuse a reward from somebody showing favoritism. Not only will the monkey not given the reward go on strike, so will the monkey being rewarded.

Now, do dogs have the capacity to sense fairness better than monkeys? Do monkeys have a better sene of fairness than dogs?

Making the claim that one group is more human-like than another simply begs the question as to whether humans themselves have a well-developed or poor-developed moral sense. Perhaps, as with smell, dogs sense fairness better than humans.

These types of questions cannot be answered unless one first has a theory of fairness - a theory that then allows one to measure the ability of entities to pick out what is fair and unfair reliably. (You cannot measure reliably without a theory that gives you the right answer.)

Here I assert, but I would ultimately argue, that a theory of fairness will show it to be an entity that is not "sensed" at all - even by humans. The only thing we "sense" are our own (socially molded) desires and aversions - our own likes and dislikes. Calling this a "sense of fairness" gives our likes and dislikes a moral weight they do not deserve, and invites the agent to accept that moral weight without question - as a "sense" of something morally real.

Tshepang Lekhonkhobe said...

That link has moved. Maybe the story is here: ?