Neither realist nor anti-realist

In a previous post, I attacked metaphysical realism. The problem is, now that I've delved deeper into anti-representationalism and the views in its cluster, I think the issue is more complex than that.

The thing is, in the anti-representationalist cluster, the realism/anti-realism opposition, whether it concerns morality, mathematics, logic, etc., becomes inarticulable and/or a person in the anti-representationalist cluster can be simultaneously a realist in one context and an anti-realist in another without contradiction.

For the anti-representationalist cluster, a word is inseparable from its language game and the norms of use it has by virtue of belonging to that game. And in everyday contexts, the norms of use are such that it is justified, even required, to endorse statements like “there are objective facts”.

Just as easy ontology (a view of the anti-representationalist cluster) insists that the word “property” is inseparable from rules of use, such as “if a car is red, then it has the property of redness”, which eliminate (in ordinary language games) property antirealism, so we must insist that the word “fact” is also inseparable from rules of use, such as “if p is true, then p is a fact”, which eliminate (in ordinary language games) fact antirealism. This is also the case for the word “true”, the word “objective”, the word “reality”, etc.

For the views of the anti-representationalist cluster, the meaning of a word is inseparable from its norms and rules of use, so that (in ordinary language games) the realism/anti-realism debate is trivially resolved by realism because of the very meaning of the words, anti-realism being trivially incoherent.

Realism/antirealism debates can only be articulated non-trivially in another language game, a language game where the norms are different.

But here we encounter against a problem, or at least a subtlety. The thing is, this other language in which we can articulate a non-trivial realism/antirealism debate is just that; another language. This implies (1) that it uses words (such as “fact”, “property”, etc.) in a different sense, (2) that the conclusions reached in this language have no impact on the conclusions reached in ordinary languages, and that (3) there is no “superior conclusion” between that of this language and that of ordinary language.

That is, we can conclude in ordinary language “objective reality exists”, and simultaneously conclude in this other language “objective reality does not exist”, without contradiction. The reason is that these two statements belong to different languages and don't have the same meaning. They don't have the same meaning simply because, in order to allow the debate to be non-trivial, we had to change the rules of use of words (create a new language).

Above all, we must insist that neither “objective reality exists” nor “objective reality does not exist” has a “higher status”.

In fact, it's not that neither has a higher status, it's that this very question is meaningless and inarticulable. To ask which language is “the right one”, or which language makes “the right conclusions”, we need to place ourselves in a language, and the answer to that question will be “inextricably connected” to that language.

For example, if we have a language without compound objects and a language with compound objects, we can take a step back and ask which is better (for example, which is more useful). But to ask and answer this question, we'll be in a language, and the answer to the question will therefore be inextricably connected to that language, so that another language might give a different answer.

If one sees “realism” as seeking some kind of answer detached from any language about which languages are the best and which languages give the right conclusions. If a realist knocks on the table and says that his questions about whether we “really really believe” in “facts”, in the “strong sense” and not the “weak sense”, whether we believe in them “independently of any language, including the one in which we are speaking”, are not answered. The anti-representationalist cluster will just consider that the realist is deeply confused, that what he's saying makes no sense, and that he's in love with an inarticulable chimera.

I'm a realist in some contexts and not in others, because the meaning of words and their standards of use change from one context to another. If I'm asked whether I'm a realist or an anti-realist “independently of any context and any language”, either I'll consider that the question is being asked in a very special language, and I'll try to understand the norms of use that govern it, or I'll consider that the question doesn't make sense.
For example, if I'm asked in an ordinary context whether I believe there are “moral facts”, I'll say yes. But if I'm asked in a more philosophical context, I might answer no. And I don't consider that there is any contradiction or that one answer is more valuable than the other. These are two games with different rules and meanings.

And you have to realize how radical and paradoxical this is. The answer to “does truth depend on the framework” itself depends on the framework. Read the sentence carefully, and realize how paradoxical it is, and how it completely breaks the realism/antirealism framework.

But doesn't this lead to a kind of resurrection of the analytical/synthetic dichotomy?
No, we continue to consider that any connection between two nodes is separable. To say that “there are objective facts” is trivially true in ordinary language is not the same as saying that there are inseparable nodes. What we're saying is that we need to realize the distinction between debating which cluster to connect to which cluster, and debating which clusters to have or what are our clusters. Even if the distinction is vague, it's important, because otherwise, we may not understand what our interlocutor is saying, we may not understand the gap that separates us, and so on.