1
$\begingroup$

The distinction between sets and proper classes has interesting ontological consequences.

One way to define a proper class, in the context of set theory, is to state that it is not an element, in a set. As a consequence, if c is a proper class, $\neg\exists x(x=\{c\})$. If we take an object to exist just if it is an element, it follows that proper classes do not exist.

Nevertheless, we may quantify over classes. If e.g. R is Russell's class $\{x|x\notin x\}$, we have that $\exists x(x=R)$; but $\neg\exists x(x=\{R\})$. One may say that Russell's class subsists, but does not exist.

Do "empty names" behave similarly, so that e.g. Sherlock Holmes subsists, but does not exist?

Does the element-strategy have noteworthy advantages? Has it been explored in the literature?

$\endgroup$
2
  • 5
    $\begingroup$ What you describe sounds formally just like a two-sorted set/class theory (e.g. Gödel–Bernays and similar) but with different terminology, saying “$a$ exists” for what would conventionally be read as “$a$ is a set”. Is this what you have in mind, or if not, what difference besides terminological are you thinking of? Systems with terms denoting not-necessarily-existing things have certainly been defined (e.g. Fourman/Scott’s partial logic), but in all such systems I know, “$a$ exists” is equivalent to “$\exists x\ (x=a)$”, i.e. what you’re calling “$x$ subsists”. $\endgroup$ Aug 30, 2021 at 18:46
  • $\begingroup$ @PeterLeFanuLumsdaine It indeed has such a provenance. The terminology I opted for has bearings upon how we think about some important problems. Yes, there are many free logics, with non-denoting terms. The approach I suggest has advantages, it seems. To wit: (1) The mere subsistence, and thus also non-existence, of proper classes, combines well with the idea that classes merely abbreviate formulas. (2) It provides a nice semantics for fictional characters, by including them in the domain of discourse, while making statements as SH is F and SH is not-F false (if we have choice-negation). $\endgroup$ Aug 31, 2021 at 0:38

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.