Carnap vs. Whitehead on Demonstration vs. Description

A feature of Carnap's system discussed in the episode was his his attempt to objectivize our talk of objects by removing any demonstrative or ostensive elements. Though the "elementary experiences" as I examine them are of course mine, and not analyzable in themselves according to Carnap's account, the only way they become useful to science is through their connection with other experiences, which enables quasi-analysis that picks out qualities which we can then talk about in common.

So even though what I call green and what you call green might not even be phenomenally the same thing (the inverted spectrum problem in epistemology and philosophy of mind), the fact that we call all the same things green makes this a functional concept. Ultimately (and we didn't get into this on the episode), at the end of the book, Carnap opines that an ideal base for his system wouldn't even be these elementary experiences and the recollected similarity relation, but just the relations/structural properties themselves, as in the train line example we discussed, where the whole definition of a node in the train network is its relation to other nodes.

I've started Whitehead's The Concept of Nature to research that for a future (not yet scheduled) episode, and was surprised to see that Whitehead reverses this relation between demonstrative and structurual (descriptive) properties in pursuit of the same goal.

Whitehead states (p. 9): "...Thought places before itself bare objectives, entities as we call them, which the thinking clothes by expressing their mutual relations." What science does, according to Whitehead, is describe the "bare objective" of a demonstrative: that thing. In order to say anything about that thing, we have to use a common concept word, which is to compare it to other things, and thus bring in a lot of connotations that aren't actually relevant to that bare objectivity we want to talk about. So for Carnap, the ideal is to talk only about the structural properties, while for Whitehead, the ideal is to shore away all the structural, comparative, descriptive properties and talk about the unspeakable: the immediate object of experience.

Clearly, there's a lot of overlapping ground here, though, in that they both recognize the same relationship between bare demonstratives and the objects of our speech (and moreover, they both refer positively to Russell's theory of definite descriptions). However, they'd describe this relationship differently: For Carnap, the concept is built out comparison of the bare demonstratives (elementary experiences), and all talk is actually about sets of these elementary experiences, whereas for Whitehead (who I'm still working to figure out), talk by its nature is impure, losing the content of an individual experience in having to use a general concept term. For both of them, the immediate experience is the ultimate ground for objectivity insofar as that's possible, which is ironic because the immediate experience is of course the essence of the subjective; it's mine and mine only, incommunicable in some sense.

-Mark Linsenmayer


  1. Kallan Greybe says

    Huh, small world, Mark. I haven’t had a chance to listen to the Quine ‘cast and the Carnap’s passed me long by, but they’re high on my list now. Basically the goal of my dissertation is to argue that something like a Positivist understanding of the verification principle could survive the various problems of underdetermination by appealing to externalist arguments in semantics (particularly Kripke and Evans but I think that any of them could work.) You would lose stuff in the process, particular logics and possible the naive empiricism, but the verification principle can largely stand seperately of those. The idea is that the way we normally save evidence from underdetermination theories, by explicitly accepting that evidence is theory laden and then by focusing on the way evidence decides between relevant alternatives, also happens in linguistic contexts.

    The way this works is that meanings are the usual way we traverse across linguistic divides and externalists all argue that the way the world actually is plays a significant role in setting those meanings. A good analogy is to think about navigation. If I want to get from here to there it’s a good idea for me to either keep some landmarks in view or orient myself and the direction I’m heading cardinally. If I don’t keep some features of my picture of the world fixed then I’m going to get lost. Kripke’s arguments in Naming and Necessity show that whatever these features are however, they can’t be internal to a language, which is what descriptivism amounts to, and instead have to track some real feature, in his case a causal relation to a dubbing ceremony.

    The idea then is that evidence on the one hand is what allows us to actually decide between relevant competing alternatives, under at least some post-Positivist theories of evidence, and on the other hand is the bit that stands fixed and allows us to finish the work of bridging the semantic gap between what is meant by the alternatives on offer.

    Okay, hopefully you get the gist from that.

  2. Victor Mansella says

    I’m thrilled to hear that you guys will eventually be covering Whitehead! The Concept of Nature is an excellent introduction to his process metaphysics but I would suggest, perhaps for another episode, another shorter reading that concerns the function of reason. It is called the Function of Reason. Whatever you choose to cover, his ideas are distinct enough from most of the other metaphysicians that y’all have covered that I think he deserves some attention from the PEL.
    Looking forward to it!

Leave a Reply

Your email address will not be published. Required fields are marked *