In doing so, I acknowledge that the actions of animals are not up for moral judgment; but their experiences are proper subjects of moral concern.
This framework of considering beings as moral objects whose experiences matter, but not agents whose choices "deserve" reward or punishment, is properly applied not just to animals, but to small children, the insane, and other "near persons" who lack the faculty of rationality.
This is nearly the opposite of Kantian "respect for persons" as I understand it, which accords the mysterious quality of "dignity" to all those with rational faculties. This "dignity" - this human-like rational function - is why we should respect the wishes of others, why what others want should matter to us.
I find it obvious from inspection that the pain of other experiencing beings should matter to us even if the others have no rational faculties at all. And I see the path from conscious-experiencing to conscious-choosing to be a continuum, rather than binary categories, with humans not even fully embodying the rational/choosing end of the spectrum.
I am concerned with suffering. Justice often concerns itself with suffering only so far as the suffering is "undeserved." I do not think any suffering is deserved. The notion of desert is entangled inside the context of a particular system.
In the bad old days, academics in criminology frequently wrote about victim-precipitated rape. Menachim Amir writes, in 1967:
We are accustomed to believe that forcible rape is an act which falls upon the victim without her aid or cooperation, but there often is "some reciprocal action between perpetrator and victim" in such cases.
Once the victim and the offender are drawn together, a process is set in motion whereby victim behavior and the situation which surrounds the encounter will determine the course of events leading to the crime. If the victim is not solely responsible for what becomes the unfortunate event, at least she is often a complementary partner.
"Victim Precipitated Forcible Rape," in The Journal of Criminal Law, Criminology and Police Science 58:4 (1967). Citations omitted; bolded emphasis mine.
I think victim-criminal interaction theory is a fruitful and interesting field, especially with an evolutionary psychology orientation. However, I also agree for once with mainstream academic feminists: nobody deserves to be raped.
The fact that a victim contributed somehow to cause a crime does not imply that the victim is a deserving victim. Rape is simply not an appropriate sanction for any behavior - even rape itself. Not even a rapist deserves to be raped.
But why should this be? To see this, we need to ask ourselves: what justification is there for saying a rapist deserves to be raped? Or that a thief deserves to have his hand cut off?
Considering the classical theories of punishment (justifications for a society imposing criminal sanctions), most of them - general deterrence, specific deterrence, incapacitation, and rehabilitation - justify punishment on the grounds that it is good for someone, or good for society as a whole. Punishment is justified in order to create the proper incentives, to maximize the happiness of everyone at the expense of the suffering of a few. Only the retributive theory of justice - vengeance, punishment for its own sake, spite - is compatible with a notion of desert on the part of the suspect. The rest justify punishment only on utilitarian grounds, and might equally justify punishment of the innocent!
The vengeance motive - the retributive theory of justice - is not explicitly utilitarian. It is the deeply-felt human idea that harm simply deserves harm - an eye for an eye. Under the framework of vengeance, it's a bad thing when a criminal dies before having the opportunity to be punished, even though everyone is made better off by his death. But there is a great deal of evidence that this spite function - the desire for revenge even when it doesn't make anybody better off - is an adaptation for realizing the most effective, versatile game strategy in social animals. A social animal that allows others to get away with defection unpunished encourages more defection, and meanwhile does not compete as well as a social animal programmed to follow tit-for-tat. In this light, we can see vengeance as evolution's tool to get a social organism to cooperate the optimum amount to maximize its fitness. I would argue that the justice of vengeance stands or falls with the justice of the utilitarian theories of punishment.
All the utilitarian justifications come down to this: we must punish people, make them suffer, so that overall, people in society suffer less. What this assumes is that we have a right to make people suffer against their will for the greater good. This assumption is wholly unsupported, and can never, in my view, be supported. How the unconsented suffering of some can be justified by the happiness of others is something I have never understood, and something that concerns me a great deal. I have argued that this is the same as the move in economics from "humane Pareto efficiency to ugly, realist Kaldor-Hicks efficiency."
Pareto efficiency is the idea that a transaction is just (and we should encourage it) if it helps someone and hurts no one. Any fully consensual transaction should have this characteristic, so a contractual exchange would be a Pareto improvement. (However, the justice of any transaction relies on the justice of the initial distribution, which is, in reality, totally unfair.)
Kaldor-Hicks efficiency comes from a recognition that consent is hard to do. With Kaldor-Hicks, we jump from requiring a transaction to help someone and not hurt anyone - that is, to be fully consensual - to allowing the transaction if the gains for some outweigh the costs to others, so that theoretically the losers could be compensated. (It doesn't matter if, in reality, the losers are compensated.) Many non-consensual transactions can be justified under Kaldor-Hicks; the good for some just has to outweigh the bad for others. For instance, rape is never a Pareto improvement, but if the rapist enjoys it more than the victim suffers from it, it could be a Kaldor-Hicks improvement. It is my contention that Pareto has a shot at being just, but Kaldor-Hicks is churched-up evil.
What is missing in any sort of justification for why it's okay to make some suffer so that most of us can be better off. And what's especially fascinating is that although in general in economics we do not compare utility functions of people, Kaldor-Hicks thinking essentially requires us to compare utility functions of different individuals. Why is it okay here and not okay in other places? There's no market here, by definition, so we're not using revealed preference as a guide.
But even a market based on actual consent is not grounded or justified in any way that should make us ethically comfortable. A market or social system may provide for individual choice in any given transaction, but a participant cannot decide whether to be part of a market economy. It's not consent all the way down, you might say.
Having a baby might be a Kaldor-Hicks improvement, but it cannot be a Pareto improvement. We need to pay more attention to this type of unconsented transaction, and our primary concern should be for its victims, rather than for the rights of agents making these harmful decisions for their own benefit.