Before I move on to more current things, I want to go back and respond to Vera Flocke’s response to my post about her 2024 paper “Carnap is not against metaphysics” in the Interpreting Carnap volume. That was over two years ago, and I rudely left her without a reply. At the risk of responding to things she no longer believes (or has any interest in), I will now finally do her the rudimentary courtesy of responding to her objections to my post. I apologize for the tardiness of this follow-up.

She makes two main points: First, that — contrary to my critique — she did in fact (implicitly) make the principle of tolerance a criterion for distinguishing between the earlier and later Carnap, and second, that Carnap was not against metaphysics but against certain meta-metaphysical connotations of metaphysical positions. (I’m less sure about this second one.)

Regarding the first point, she argues that Carnap’s well-known statement of the principle of tolerance in §17 of the Syntax can be interpreted to mean “that there are ‘frameworks’, some statements are internal and some are external to a framework”. But there is no mention of statements internal or external to a language (the expression “framework” dates to 1950) in Syntax §17, and in fact, as I pointed out in my original post, Carnap already employed frameworks (under a different name) in the Aufbau, as the criterion for distinguishing between “scientific” and “metaphysical” realism. He did not use ESO language there, but the idea is the same. To claim something as “real” is an example of scientific realism, he says in §§175-8, if it is constitutable within the constitution system outlined in the first part of the book, but if it isn’t so constitutable, then this claim is an example of metaphysical realism (and therefore vacuous); so existence is framework-relative (constitution-system-relative) just as in ESO. Since Carnap is explicitly pluralist about constitution systems (and presents the Aufbau as merely an example of a constitution system, not as the definitively correct one), this criterion can be assumed to apply to any constitution system. Carnap is already “tolerant” regarding constitution systems, then, in the Aufbau (as Michael Friedman liked to point out). He is also tolerant about different philosophical ways of articulating the world; he explicitly regards e.g. realism and phenomenalism as merely different façons de parler. But what he is not yet tolerant about is different systems of logic, and different ways of constructing logical and mathematical systems; the principle of tolerance in the Syntax (and in the 1932 paper replying to Neurath about the form of Protokollsätze) specifically extended tolerance to these most basic considerations, and opposed the previous “absolutism” of thinking there was a “correct” way of articulating them.

So “tolerance” in the sense of the Syntax principle of tolerance is not equivalent to the idea “that there are ‘frameworks’, some statements are internal and some are external to a framework”. Carnap had the second idea long before the first. Nor do they imply each other in any way. Take as an example the three competing doctrines about the foundations of mathematics during the Vienna Circle period: logicism, intuitionism, and formalism (the discussion of these is the context of that famous quotation from the Syntax enunciating the principle of tolerance). There is no contradiction between the idea “that there are ‘frameworks’, some statements are internal and some are external to a framework” and the idea that one of those three doctrines about the foundations of mathematics is the correct framework, and the other two mistaken. That is what most people thought, and indeed still think, about these matters — along the lines of the proto-intuitionist Kronecker, who thought “die ganzen Zahlen hat der liebe Gott gemacht, alles andere ist Menschenwerk.” Carnap’s point was that it’s ALL Menschenwerk; there’s no criterion of correctness anywhere out there in the world to tell us what logic to use or how to conceive of mathematics. There’s no more empirical evidence for deciding between logicism and intuitionism than there is for deciding between realism and phenomenalism. It’s up to us.

So not only does she not mention the principle of tolerance explicitly, as one of her three criteria to distinguish the earlier from the later Carnap, but the principle of tolerance is not even implicit in any of those three criteria, including her third one, in which she claims, in her response, the principle of tolerance is implicit. It is not.

Flocke’s second point concerns the distinction between metaphysics and meta-metaphysics. She says, “I argue that there’s a difference between a metaphysics, e.g. ‘Necessarily, everything is necessarily something’ (necessitism), and a meta-metaphysics, e.g. ‘Necessitism describes the fundamental structure of reality’. Carnap did not want to liberate us from metaphysics (e.g. necessitism), he said to accept a metaphysics is a matter of accepting a framework, which is a fine thing to do. He wanted to liberate us from meta-metaphysical mistakes, such as believing that accepting necessitism is to have a correct view of the fundamental structure of reality.”

I’m not sure how to respond to this. To start with, I guess I would ask, where did Carnap ever say anything remotely like “to accept a metaphysics is a matter of accepting a framework”? Presumably she is aware that he never said any such thing, but relies rather on her interpretation of the Syntax §17 passage she quotes, according to which: “to accept a logic is a matter of accepting a ‘framework’. The types of logic Carnap had in mind when writing the Logical Syntax of Language involved metaphysical claims, regarding the existence and nature of numbers. So, in that regard, to accept a metaphysics is a matter of accepting a ‘framework’.”

But how could the “frameworks” (Carnap did not use this term in the Syntax; there he just called them “languages”) involve metaphysical claims, when they don’t even have interpretations (that came several years later)? They are just bare syntactic marks on a page or a blackboard, with no semantics, just as Hilbert distinguished purely mathematical geometry from physical geometry (a distinction that Einstein claimed was crucial to his own discovery of relativity), and Carnap accordingly called his syntactic approach a “pure geometry of language”. There aren’t even any physics involved, let alone metaphysics!

And then I would want to ask how the assertion of necessitism (using her statement of it) differs from the claim that necessitism (in this same sense) “describes the fundamental structure of reality”. This is essentially the same redundancy as the one we encounter between “The grass is green” and “‘The grass is green’ is true.” — where the first statement is in the object language and the second in a metalanguage. They say the same thing. So I’m unsure what distinction is being drawn here.

Finally I would insist on a certain principle of interpretive charity, or perhaps interpretive integrity or something, according to which the author’s own use of words is to be deferred to when interpreting that author’s work. (Perhaps this is just a principle of sound hermeneutics — in the old, Schleiermacher sense rather than the newfangled Heideggerian, Gadamer sense.) If I advance the startling claim that Kant rejected synthetic a priori statements, it’s off limits — according to this principle — to then say, “oh well, I’m using the words ‘synthetic a priori statements’ to mean ‘Swedenborgianism’, which Kant clearly rejected”. So I really don’t see the point of attributing to Carnap the statement (which he obviously would have rejected outright) that “to accept a metaphysics is a matter of accepting a framework”. I am open to the idea that Carnap became more willing to recognize a grey area between meaningful and meaningless statements in his later years, and in that sense became more tolerant of things some of us might count as metaphysical; I’ve more or less endorsed this view myself. Also I’m open to the idea that some of what Carnap wrote in the Aufbau (notably the sections §§155-7, where he talks about eliminating the basic relation) is, as David Lewis later said, metaphysical by Carnap’s own standards at that time — Fraser Macbride has meticulously documented this. So it’s not like I’m allergic to associating Carnap with metaphysics in any way; but let’s play by the rules!