There's no doubt that Mill et al had something specific in mind when they were thinking about "utility". But you wanted to associate Harris's philosophy with Utilitarianism as well, and it's obvious that Harris and Mill must at least slightly differ in their definition. So you must agree that there exists, in some sense, a generalized kind of utilitarianism where we can plug in different utility functions. Mill's specific ideas here don't seem relevant.
The usual definition I've encountered is that utilitarianism is universalist (it counts every morally important entity equally), maximalist (it commands you to maximise utility not merely to produce some), consequentialist (it is solely concerned with consequences, not act-types or virtues or rights) and it picks out one or more things which it defines as utility and would have you maximise.
Bentham defined utility as pleasure, Mill modified this by ranking "higher pleasures" (like art and philosophy) over "lower pleasures" (sex, drugs, rock and roll) and various other measures have indeed been plugged in by other people.
I probably need to watch the video again because it's been a while but I think I recall Harris talking about "flourishing" or some such thing, and I imagine that's going to cash out very much like "welfare utilitarianism" which defines welfare as the fulfilment of the preferences people would have if they were well-informed and rational.
That's probably close enough to what I'm getting at. How does one define utility besides "consistent value judgment"? It's tautological otherwise. What is not tautological is the ultimate definition of utility--Mill chose happiness and pleasure, which while themselves have a wide range of interpretation, are at least somewhat meaningful terms. Harris chose "human flourishing", which may overlap with happiness to some extent but is certainly not synonymous.
It's a specific kind of consistent value judgment, though. Kant would have you make consistent value judgments about which acts were right and which were wrong, and would have you do right acts and eschew wrong acts. That's consistent but he quite firmly rejects all consequentialist thinking - Kant is consistent, in fact, in holding that concern with outcomes is one of the things that we should not let distract us from what he saw as real morality.
Utilitarianism makes a different kind of consistent value judgment, weighing up the available outcomes and claiming that you should take whatever action maximises utility as you have defined it, even if that action is theft or murder and even if the resulting distribution of harms and benefits is unjust.
This is starting to get too deep into semantic issues, but I take a mathematical or computational view here. A function which does nothing but return zero, say, is still a function. You would agree, I think, that there is a large space of possible utility functions. Some of these are going to be simpler than others. To me, there is no obvious point to draw a line between trivial functions, functions that operate mostly on the self, and functions that operate globally.
I think if you set utility to always equal zero then your theory is no longer universalist, maximalist or consequentialist in any meaningful sense and hence it's no longer utilitarianism - or if it is, it's a kooky corner case that is utilitarianism only in a fairly debatable semantic sense.
That's close, but it still feels too limiting. I see no reason why the utilities cannot be interdependent. For instance, one might scale the value of diversity with another metric dependent on how the moral thought influenced overall outcomes. One may wish to encourage diversity but not to the point where you see highly negative results (say, authoritarian Communism). This requires a certain degree of interdependence.
I think any moral theory that needs a spreadsheet is definitely open to criticism on the grounds that it's of limited usefulness, and even relatively straightforward utilitarian systems (like the
QALYs used to make health care decisions) are already somewhat open to criticism on such grounds.
That said for things like government policy decisions that affect a wide number of people in different ways, such a theory might be a useful thing if you could sell enough people on it.
Anyway, I think we are coming to some sort of mutual understanding, so I put it to you: since you group Harris with utilitarians, you must have some idea for how similar a utility function must be to Mill's to qualify. You believe that highly atypical or degenerate functions like those of Nihilists do not deserve to be called such. So where do you draw the line?
I hope I answered this satisfactorily earlier, but if I didn't I'm happy to take another try at it.
