This post is a bit less polished than ideal, but I wanted to get it out of my drafts
Introduction
One sometimes sees the following type of objection to consequentialism or utilitarianism:
P1. Consequentialist reasoning would justify X.
P2. X is not justifiable.
C. Therefore, consequentialism is wrong.
Similar argument structures can also be applied to any number of ideologies/philosophies:
P1. Libertarianism would justify X.
P2. X is not justifiable.
C. Therefore, Libertarianism is wrong.
Call this general argument format the Justifiability Argument. Some recent criticisms of longtermism arguably follow this structure. We can also call the ideology in P1 the "target ideology" for generic purposes.
However, note the potential ambiguity of "justify" as used in the Justifiability Argument. Depending on the sense in which it is used, P1 could mean either of the following:
A. Consequentialism, properly applied, implies X is justifiable.
B. A speaker can use consequentialist reasoning (e.g., consequentialist argumentation, consequentialist phraseology) to make a rhetorically convincing case for X.
A. is a normal premise in a modus tollens. A lot has been said on when and whether moral arguments of this form are compelling, especially where P2 is supported by ethical intuitions alone. I will not add to that discussion here.
B., however, is much more interesting. For the Justifiability Argument above to work under interpretation B., one needs an additional premises or argumentation. This post explores the Justifiability Argument under interpretation B, which I will call the Rhetorical Abusability Argument.
To focus solely on that question, for the sake of argument we can assume that P2. is true, though in reality we may also disagree with it.
Supplying the Missing Premises
Rhetorical Abusability
To make the Rhetorical Abusability Argument work, one could try to insert a premise like:
P3. It should not be possible for a speaker to make a rhetorically convincing case fora false conclusion using the reasoning of the true moral theory.
Call this the "Rhetorical Abusability Premise." I think most people who make the Justifiability Argument implicitly assume something like the Rhetorical Abusability Premise.
Unfortunately, the Rhetorical Abusability Premise seems incredibly weak to me. For one thing, what is rhetorically convincing to people is a product of moral psychology, not of moral truth. It would be very odd if first-order moral truth depended on the contingencies of how convincing different rhetoric is to different audiences. Different people will find different things rhetorically convincing, and different speakers are differently skilled at rhetoric. It should not be surprising if there are many false (by supposition) moral conclusions that a skilled rhetorician could make some number of people believe. Laypeople's moral beliefs are not especially coherent, and for most of history (and possibly still now) people have believed morally unjustifiable things. So, the mere fact that someone can rhetorically abuse some theory to convince some listener of a false moral conclusion cannot be a good reason to reject that theory.
There are other reasons to reject the Rhetorical Abusability Argument as formulated here. One advantage to the person using the Rhetorical Abusability Argument is that the Rhetorical Abusability Argument imposes no obligation to present the criticized moral theory correctly. What's important to the Rhetorical Abusability Argument is that the cited arguments or phraseology merely purport to apply the target ideology. This is beneficial to the person making the Justifiability Argument for a number of reasons. One is that it simply relieves them of the hard work of figuring out whether a bad consequence is actually entailed by the target ideology. As long as the person making the Rhetorical Abusability Argument can find some adherent of the target ideology attempting to use the target ideology to justify a bad outcome, they have a case against the entire target ideology.
Note some of the consequences this has. Ideologies with more adherents will naturally tend to have more people invoking their premises. Thus, all else equal, the odds that some of their adherents will engage in rhetorical abuse are higher. Thus, the Rhetorical Abusability Argument seemingly disproportionately targets more popular ideologies all else equal.
Even more perversely, this version of the Rhetorical Abusability Argument incentivizes people to cite the purported applications of the target ideology with poor reasoning or bad outcomes.
The Rhetorical Abusability Agrument thus often looks like a criticism of the ideology itself, but is in fact a criticism of specific purported applications of the ideology. This structure is often concealed in the Justifiability Argument, which makes it sound like "abusability" is an inherent property of an ideology, when in fact it is a contingent, social fact about how people are using and misusing it. The Rhetorical Abusability Argument as presented above does not, and in my view cannot, bridge the gap from showing that certain people misuse an ideology to showing that the ideology itself is wrong. Misusing any ideology is simply too easy, and consistently applying the Rhetorical Abusability Argument above would lead to nihilistic outcomes.
Heightened Rhetorical Abusability
We can modify the Rhetorical Abusability Argument to try to make it more convincing by claiming that the target ideology has heightened susceptibility to rhetorical abuse. Thus, the modified Justifiability Argument may look like this:
P1. Consequentialist reasoning is more susceptible to making a rhetorically convincing case for unjustifiable acts than other moral theories.
P2. The true moral theory would not be more susceptible to rhetorical abuse than other moral theories.
C. Therefore, consequentialism is wrong.
Call this the "Heightened Rhetorical Abusability Argument."
I still think this modified version of the Rhetorical Abusability Argument is wrong. For one thing, P1 still relies on psychological facts about what people find rhetorically convincing, so the foregoing counterarguments still apply, albeit to an aggregate of moral judgments rather than a single one. Furthermore, I have grave doubts about whether P1 is true empirically: I hear bad arguments couched in deontological and virtue ethics reasoning all the time, and it's very unclear to me that they are any less prevalent or rhetorically misleading than utilitarian arguments.
More importantly, though, an incorrect argument that relies on true premises will often (usually?) be more rhetorically convincing than an argument that relies on false premises. For example, conspiracy theories that weave together many true premises with false or unprovable ones are probably more convincing than conspiracy theories that have even less basis in reality. Consider the following two arguments for alternative explanations to the assassination of JFK:
P1. JFK fired CIA director Allen Dulles.
P2. JFK refused to provide support to the CIA-led Bay of Pigs operation.
P3. JFK proposed cutting the CIA's budget.
C1. Therefore, the CIA had a motive to remove Kennedy from office.
P4. Killing President Kennedy is one way to remove him from office.
C2. Therefore, the CIA had a motive to kill JFK.
P5. The CIA is expert at orchestrating assassinations.
C6. Therefore, the CIA also had the means and opportunity to kill JFK.
C7. Therefore, the CIA did kill JFK.
Argument 2
P1. Gremlins from the Earth's second moon dislike Irishmen.
P2. JFK was an Irishman.
C1. Therefore, gremlins from Earth's second moon had a motive to kill JFK.
P3. Gremlins from Earth's second moon can easily exercise mind control over humans.
P4. Lee Harvey Oswald was a human.
C2. Therefore, gremlins from Earth's second moon could exercise mind control over Lee Harvey Oswald.
C3. Therefore, gremlins from Earth's second moon had the means and opportunity to kill JFK.
C4. Therefore, gremlins from Earth's second moon killed JFK using mind control of Lee Harvey Oswald.
Both of these arguments lead to false conclusions. Furthermore, Argument 1 is much more rhetorically convincing than Argument 2! Should we therefore conclude the premises of Argument 1 are incorrect? If we had an analogous version of the Rhetorical Abusability Premise for empirical arguments, we would have reason to. However, we can see that an empirical version of the Heightened Rhetorical Abusability Premise is obviously bunk: facts can be abused just as much as falsities, and in any case their abusability is not a good reason to think that they're false.
Furthermore, we can actually see that rhetorical abusability may very well be positively correlated with truth! When an argument deploys true premises, it is more rhetorically convincing, even when its conclusions are false. This makes sense, since an argument with true premises will be less obviously-false than one with false premises. Invoking gremlins from Earth's second moon is a hard starting point for any argument. Thus, if something is more rhetorically abusable, it could in fact be evidence that it is true. A skilled rhetorician blends in just enough truth to make their claims—even their unjustifiable ones!—seem well-supported and indeed commonsense.
To recap, we have at least two good reasons to reject even the Unique Rhetorical Abusability Premise:
- It would obviously be false if applied analogously to empirical arguments, leading us to incorrectly reject empirical truths on the basis that someone was rhetorically abusing them.
- There may be a positive correlation between truth and rhetorical abusability.
Conclusion
It may be tempting to see someone purporting to follow an ideology to false conclusions and therefore conclude that the ideology is false. Indeed, it is painful to see one's own ideology applied to support repugnant conclusions.
However, we should realize that truths are intertwined. Rejecting a premise because it was use to support a bad argument may foreclose the opportunity to make good arguments that would rely on the same premise.
The lesson, then, is the same as it ever was: evaluate ideologies on their own merit, with attention to their actual implications and actual justification. If someone reaches a bad conclusion relying on a number of premises, carefully select which of their premises to reject, instead of permanently discounting all of them.
Thanks, I think this is a great post.
Another argument is something like the following:
This argument isn't directly about the truth of consequentialism. But if correct, it would mean that consequentialism is "self-effacing" - that people believing that consequentialism is true has bad consequences. Historically, some have argued that that's a reason to keep consequentialism, or utilitarianism specifically, secret. There is a debate within moral philosophy about to what extent being self-effacing is a problem for a moral theory.
One could construct an analogous argument concerning longtermism, or maybe specifically consequentialist versions of longtermism.
I'm not necessarily saying that this is what those who give the types of arguments that you cite have in mind, but in any event, it seems to me a type of argument worth being aware of. There's a large discussion on this topic - not the least in the literature on consequentialism and utilitarianism - that one can draw on.
(I may or may not return to the plausibility of this argument - I don't have time now.)
If this argument concluded that belief in consequentialism had bad consequences on net, it would be a more serious problem for consequentialism.
Another different and perhaps more relevant question is whether popularizing belief in consequentialism has net bad consequences on the margin.
(Just flagging that this is very related to the discussion in the first part of Reasons and Persons, and for the reasons presented therein I don't think it's a decisive argument against consequentialism as criterion of rightness.)
I wrote a more general post on arguments about consequences of misuse of consequentialism and longtermism.
Consider proposition P:
I think if it were the case that we received evidence for P, it would be reasonable to conclude that consequentialism is more likely to be wrong as a decision procedure[1] than if we received evidence for not-P.
Do you disagree? If not, we should examine the distinction between "(heightened) rhetorical abusability" and P. My best guess is something that I often tritely summarize as "anything is possible when you lie":
Notably, unlike your post, my argument is contingent upon a specific scaffolding of empirical facts. Strong and representative evidence that consequentialist beliefs are predictably harmful in history, or conceptual, empirically informed, arguments that consequentialist beliefs will lead people to predictably do harm in the future, will cause me to update against consequentialism as a decision procedure.
Though it was unclear to me from this post whether you were considering consequentialism as a criterion of rightness vs a decision procedure. If the former, I think the question is less interesting under moral antirealism or nihilism (which I suspect most critics of consequentialism subscribe to).