Carnap Frameworks

The key distinction in the article is between "internal" and "external" questions.

Internal and external to what? To "frameworks." " Which kind of question are metaphysical questions? External.

What does that tell us about the answers to metaphysical questions? That how we choose to answer them depends on pragmatic considerations, not facts of the matter. He calls such questions "noncognitive," by which he means something like, they are not well-formed questions with objective answers.

Certain questions have answers in advance, answers we seek to discover. For example, Do unicorns exist? or Do Tasmanian wolves exist? Other questions do not have answers in advance, we answer them by stipulating what the answer is. Do you want Chinese for dinner? Perhaps more clearly, What do you want to name your child?

Carnap is saying that external questions are closer to being of the second type than of the first. We stipulate, or decide, or adopt the convention that, or something in that ballpark that, say, numbers exist, as an answer to the external question, "Do numbers exist?"

What about internal questions? What is the answer to the internal question, "Do numbers exist?" That depends on what framework the question is internal to. For the obvious framework, the one in which we use numbers, the answer is yes, and obviously so. Note the difference between the question, "Are there 100,000 twin primes?" and "Do numbers exist?," or, "Do Tasmanian wolves exist?" and "Do animals exist?" Consider a different framework, an "empiricist framework" in which only things accessible to the senses are said to exist. Do number exist in such a framework? No.

Some questions have both internal and external versions, others seem to have only internal versions. For the questions that have both versions, the answer to the internal question is always obvious (though it may be different in different frameworks).

Why? The questions that have obvious external versions are ones that are answered as part of setting up the framework. It is obvious that numbers exist in the mathematical framework because it is part of the point of the framework to incorporate numbers. It is obvious that numbers do not exist in the empirical framework because it is part of the point of the framework that abstract objects, like numbers, do not exist.

So, what is a framework? A framework is a coordinated system of assumptions. A framework serves as a method of interpreting the world. It predicates one or more kinds of objects and tells us what the criteria are for their existence. A framework is linguistic. It sets up a language for talking about certain kinds of objects and criteria for the truth of assertions in the language about the object. (Language is a funny word: he, and we, don't here mean something like English or French, but scientific language, sportswriter's language. Given that, it seems that more than one framework might be operative at a time. Carnap, it seems to me, trades on the ambiguity of the term "language" to avoid that issue.)

The article is intended for a "nontechnical audience": Carnap was a formalizer, and what he really has in mind quite often, is a technical framework specified in formal logic. In that context, the notion of a framework seems quite natural. He exports it to ordinary language without saying much more, but it is not clear that that is ok.

Can a framework have subframeworks? Carnap uses the word framework in such a way that the answer is no. That is clear because external questions are, for Carnap, outside any framework.

What is Carnap's definition of a framework? He never give one. He says some things about how they are formed and what they are used for. "A system of new ways of speaking subject to new rules." What he does do is give lots of examples of what he takes to be frameworks.

So, what is Carnap's answer to metaphysical questions? His answer has two parts:

1. If the question is internal to a framework, the answer is easy.
2. If the question is external then,
   1. It is noncognitive. That is, if you are in a disagreement about what the true answer is, you have made a mistake.
   2. We can adopt various frameworks for various purposes. Our reasons for doing so are pragmatic: convenience and compactness of expression for a particular purpose.

So, in the way in which people have thought they were asking "Do numbers exist?" Carnap's answer is that there is no answer, no fact of the matter. In the way in which Carnap thinks we should understand the question the answer is, "yes" on Tuesdays, "no" on Thursdays: We change frameworks, and with them the answer, as we see fit for various purposes.

To argue that numbers exist, Carnap might point out that they are useful for many purposes and hence appear in many frameworks.

-- ShaughanLavine - 23 Jan 2008