What I wrote about in the previous post is actually only a side issue in Tobia’s paper. His main point is that intuitions are used as evidence for and against philosophical theories, if “intuitions” are understood as “special kinds of philosophical assumptions, ones to which we are invited to assent that are suitable for argument (for example, providing evidence without further justification) and that are not purely inferentially formed.” He gives five examples of supposedly using intuitions this way:
- Judith Jarvis Thomson makes two judgments: that it’s morally permissible for the tram driver to turn his tram onto the side track and kill one person instead of killing five other people on the main track, and that it’s not morally permissible for the surgeon to kill the healthy patient in order to save five sick patients. She then uses these judgments as evidence for her principles that “killing five is worse than killing one” and “killing one is worse than letting five die”, respectively.
- Larry Temkin puts forward judgments like “discomfort of level 100 for 2 days is better than discomfort of level 80 for 4 days”, “discomfort of level 80 for 4 days is better than discomfort of level 64 for 8 days”, and so on, up until “discomfort of level 13.4 for 1,024 days is better than discomfort of level 11 for 2,048 days” on the one hand and the judgment that for someone he loves he’d choose 2,048 days of discomfort of level 11 rather than 2 days of discomfort of level 100, on the other. All those judgments are used as evidence against the transitivity of “all-things-considered better than” relation.
- Henry Kyburg uses the judgment that we’re not sure about certain things to the extent that we would be, had we conformed to the probability calculus, against a certain interpretation of the subjectivist theory of probability.
- Timothy Williamson uses the judgment that “it is silly to say there is a first second at which Rembrandt was old” to reject the principle of bivalence for vague languages.
- Harry Frankfurt uses the judgment that “we are unmoved by inequalities between the well-to-do and the rich” as evidence against economic egalitarianism.
All those judgments, argues Tobia, have one thing in common: they are “not purely inferentially formed”. This is why they can be called “intuitions”. Is he right about that? I don’t think so.
First of all, there are some inaccuracies in how Tobia portrays these arguments. For example, Williamson doesn’t endorse the above mentioned argument, he merely mentions it. After all, the whole point of his book is that there was a first second when Rembrandt was old, no matter how weird it seems – he explicitly dismisses the “intuition” that Tobia says he relies on. If Frankfurt uses a judgment that can be suspected of being “non-inferentially formed” in some sense, it must be “there’s nothing wrong with inequalities between the well-to-do and the rich”, not “we are unmoved by inequalities between the well-to-do and the rich”. What Tobia calls “Thomson’s principles” are actually Foot’s principles that Thomson doesn’t accept. In addition to that, some other elements of Tobia’s interpretation are perhaps not outright inaccurate, but certainly controversial. For example, Tobia thinks that Thomson’s judgments are meant to support “Thomson’s principles” while Cappelen, to whose analysis of Thomson he’s replying, thinks it’s the other way around. And neither of them seems to offer an awful lot of evidence in favour of their respective views.
I’m not going to try to settle the dispute here, I’ll just assume that Tobia’s take on all the arguments is roughly accurate. It doesn’t really matter that much as there still looms a more basic problem: the expression he uses all the time: “not purely inferentially formed” is ambiguous and its two meanings seem mixed up. A proposition can be non-inferential in a textual sense, which means it’s just out there in the text, not inferred from or supported by other propositions. It can also be non-inferential in a psychological sense, which means someone comes to believe it (or perhaps it starts seeming true to someone) not as an outcome of this person’s conscious inferential process. On a generous interpretation, Tobia shows that in his five examples the judgments in question are non-inferential, or not fully inferential, in a textual sense. But this clearly doesn’t mean we can call them intuitions and say they’re used as evidence.
It is hardly surprising that philosophical arguments, or arguments in general, rest on unsupported premises. It simply follows from the fact that texts cannot be infinitely long. If you try to support your unsupported premises, you’re only going to end up with another set of unsupported premises. Instead of solving the problem, you’ve just made your argument longer. It should be obvious that all arguments must have unsupported premises, however this doesn’t mean that all arguments are based on intuitions.
For example, I can start an argument with a textually non-inferential claim that the Earth is not flat. Does it mean I’m using the intuition that the Earth is not flat as evidence? Of course not. Most likely I’m simply assuming it’s something my readers already know, so I don’t need to waste their time explaining why this is the case. I can start an argument with a textually non-inferential claim that getting from Lancaster to Manchester is generally cheaper than getting from Lancaster to London. Does it mean I’m using the intuition that getting to Manchester is cheaper as evidence? Of course not. Most likely I’m assuming my readers either already know it, or, should they not know it, that the claim’s truth and evidence in favour of it is quite uncontroversial and easy to look up, I can therefore expect the readers to take my word for it. Whether anybody finds my claims intuitive – in any possible sense of “intuitive” – is completely irrelevant.
In short: even if it can be shown that all the judgments from Tobia’s list are textually non-inferential, proving textual non-inferentiality doesn’t do Tobia any good. If he wants to convince us that intuitions are used as evidence in his examples, he should have something to say about psychological non-inferentiality. Does he? Well, not much, at best. Perhaps he hints at psychological non-inferentiality of propositions here and there, but he never really proves any of them to be psychologically non-inferential to anybody in particular. Worse still, even if Tobia did prove it, he’d still not accomplish his goal of showing that philosophers use intuitions as evidence in his sense.
Why? Take Temkin’s proposition: “discomfort of level 100 for two days is better than discomfort of level 80 for four days”. Perhaps it is psychologically non-inferential for Temkin. Perhaps it is psychologically non-inferential for many people. Perhaps all people. Does it mean that Temkin is using an intuition as evidence? In a sense yes, but this is not the sense that Tobia is interested in.
Temkin’s proposition has infinitely many properties. For example, it is expressed with a sentence that begins with the letter “D”. But this is clearly not why Temkin picked it to start his argument with. The fact that it starts with the letter “D” is irrelevant as far as the evidential weight is concerned. And I think that the fact it’s psychologically non-inferential (if it actually is) is similarly irrelevant. Even if everyone in the world came to believe it in the course of a fully conscious inferential process, it wouldn’t change anything methodologically substantial.
If Tobia disagrees, he should demonstrate that non-inferentiality of the relevant kind plays an evidentiary role in his examples. Alternatively, that some other potential symptom of intuitiveness plays such a role. Sadly, he doesn’t even begin to do any of that. And so we’re left with no good reasons to believe that philosophers use intuitions as evidence.
The Metaphilosophy Blog 