Discussion about this post

User's avatar
Silas Abrahamsen's avatar

Well, I'm honored that my article is deserving of a response!

I'm not quite sure what purpose the Cheesecaketarianism is playing here. My argument wasn't that weird implications can't be reasons to doubt a theory. Rather the argument was that you can't infer from the fact that, given some moral principle plus evidence about the way the world actually is, the moral valence of actual events are different than you expected prior to getting that evidence, to that moral principle or empirical finding being wrong.

The reason I reject cheesecaketarianism is that it gets obviously wrong results in hypothetical cases where we stipulate the facts (as well as the theory being highly arbitrary, etc.). For example, it claims that given the choice of saving the lives of 541709734 happy non-cheesecake producing people, and saving a single gram of delicious cheesecake, I should save the cheesecake. The problem is that the theory is a function from possible situations to moral judgements, and the fact that it sometimes turns some possible input into the wrong output is a sign that the function is wrong. I have no problem with this inference!

But suppose that I have imagined all these strange hypotheticals, and come to terms with them and cheesecaketarianism. Now I find myself *actually* in the scenario just described above. I already know that *if* I were in this scenario, *then* I should save the cheesecake given cheesecaketarianism. Simply finding out that the scenario is actual gives me no reason for adjusting my belief in the theory of cheesecaketarianism. I already knew that given this input, the function would spit out that output. Seeing this input actually being put in the function and that output be spat out should in no way affect whether I think the function is right--I already knew it would happen.

It also shouldn't change how confident I am that the input actually happened. Why? Because the output is always inferred from the function plus the input. That means that given the function, my credence in any output should simply be my credence in the disjunction of the inputs that would lead to that output. Thus the output should be no more surprising than the inputs in question, and so however likely you think the given input is, prior to considering what output it would lead to, should never be changed by finding out that it leads to that output.

But it is this last type of inference you would need to infer from social progress being good to animal welfarism or shrimp consciousness being false. After all, you already knew that *if* shrimp were conscious, *then* animal welfarism would tell you that social progress is bad. Finding out that shrimp are actually conscious then gives you no new evidence against animal welfarism, and you also get no reason to doubt whether shrimp are conscious.

Now, perhaps you antecedently thought that it is false that in the hypothetical world where shrimp feel pain, it's wrong to cause them lots of pain for minor personal pleasure. That's another matter, but you certainly shouldn't think that animal welfarism has some probability, get evidence that the *actual* world is in such and such a state, and then lower your credence in animal welfarism or the world being in such and such a state.

(P.S. just to clarify, I'm not necessarily claiming that social progress has been bad. I think it is pretty probable, but it's also a difficult and complex issue. My problem is simply with the inference made.)

Expand full comment
Ari Shtein's avatar

> Of course, we are not psychologically capable of loving cheesecake, just as we are not psychologically capable of loving shrimp. But that makes it even more virtuous to maximize the amount of cheesecake

I agree with this bit—Bentham's piece on the virtue ethics of shrimp welfare was silly, and I think I even brought up Cheesecaketarianism in the comments. Weirdness is not virtuous in itself.

But your broader argument is still, uh, dumb.

> You might object that Cheesecaketarianism doesn’t line up with our intuitions, but our moral intuitions evolved to track what’s beneficial for survival, not what’s objectively good!

That's not why I object to Cheesecaketarianism! I object because caring about the amount of cheesecake in the world is much more arbitrary than caring about the amount of pain in the world.

I, and everyone else around me, have regular qualitative experience of pain being unpleasant and wanting less of it, which gives us good reason to think that morality commands us to decrease the amount of pain in the world. Further, there's not really an obvious candidate to reduce pain any further—no deeper experience that we can say is *actually* being had when pain is felt.

When it comes to cheesecake, on the other hand, sure, I might say that cheesecake gives me a regular qualitative experience of pleasantness. But I can reduce it into constituent parts like satiety and nutrition, which a moral theory can take account of without specifically caring about cheesecake.

> There is an extremely strong moral intuition that social and economic progress over the past few centuries has been more important than the suffering of invertebrates.

Is there? What does "intuition" mean to you? Is it similarly intuitive that, say, 19th-century English literature is higher quality than the medieval Persian canon?

I don't think so! Your intuition says that Dickens is good because you've read Dickens and enjoyed it—if I presented you with a compelling argument for why medieval Persian writing was really great, and gave you some excerpts that you enjoyed, then you'd have good reason to reevaluate your stance!

Expand full comment
17 more comments...

No posts