What makes a material inference valid?

Jaakko Reinikainen

Tampere University

jaakko.reinikainen@tuni.fi

The problem

• Inference is a "normative" notion.
• This means that there is a sense in which one should believe in a valid inference just because it is valid: it is "worth believing", it has "authority", etc.
• Traditionally, validity is understood in terms of truth. The rational authority of inferences rests on truth.
• Is there an alternative?

​

An alternative solution by Brandom

• "If Today is Wednesday, then tomorrow will be Thursday" is a materially good inference.
• Material inferential goodness is independent both of truth and form.
• Yet it is distinct from inductive, abductive, and analytic "goodness", among other things (Kuukkanen 2023)
• So, what does material inferential goodness consist of?

• Material inferential goodness is determined by "norms implicitly instituted in practices".

Regulism about norms: "proprieties of practice are always and everywhere to be conceived as expressions of the bindingess of underlying principles" (MIE, 20)

Regularism about norms: the implicit norms are identical with regularities of behavior (MIE, 26-27)

• Since both of these fail, and since there are genuine norms, there must be a third alternative between them.
• The institution claim is meant to explain how taking the norms to be in force can lead to them really being in force, where "really" means that we could all be in error about the norms being in force.

The background of the Always Already argument

• (One major) culmination point of A Spirit of Trust (2019/ST).
• A new way to support the institution claim.
• The form of the argument follows a "pragmatic contradiction": a contradiction between what one says and a description of what one does when saying it.
• Involves a family of cases from Moore's paradox to ad hominem arguments.

• Brandom's AA argument draws from (at least) two complex premises.

The mutual recognition model of discursive practices. Implicit norms are instituted when recognition is de jure transitive and de facto symmetrical.

A historical model of determining conceptual contents via actual applications.

• The important point is that in order for recognition to be de facto symmetrical, its force has to be retroactive. The norms are instituted, in a way, ex post facto.

​

​

An example of (failed) institution: �the Queen's Shilling model

• The fundamental pragmatic move in the QS model is to attribute a commitment.
• What makes attributions of commitment correct and incorrect?
• Brandom is not explicit about this, but the only coherent answer he seems to have available is that the standing conventions of the practice determine the criteria by which attributions of commitments are entitled (Pippin 2005).
• Since not everyone in the community can be wrong about the criteria of entitling attributions of commitment (though they can be wrong about when they hold), the model is reductive in regards to rational normative force.

• In the mutual recognition model, the fundamental move is to attribute authority (i.e. an entitlement).
• What authorises attributions of authority are not the standing conventions but other attributions of authority (symmetry).
• The symmetry has to be diachronic since synchronic symmetry would lead to a regress, circularity, or reduction of some attitudes having primitive authority.
• The way how symmetry can be diachronic is explained by the "three judges model" of determining conceptual contents via actual applications.

The Always Already argument

• Two opposite meta-attitudes towards norms: Edelmütigkeit and Niederträchtigkeit.
• Edelmütigkeit holds that there are genuine norms that explain (discursive) behavior; Niederträchtigkeit denies this.
• “There is, however, a fourth way of understanding the status of these two stances. Its leading thought is that we have always already implicitly committed ourselves to adopting the edelmütig stance, to identifying with the unity that action and consciousness involve, to understanding ourselves as genuinely binding ourselves by conceptual norms that we apply in acting intentionally and making judgements. For we do judge and act, and we cannot avoid in practice taking or treating those judgements and actions as being determinately contentful—as materially incompatible with certain other judgements and actions, and as materially entailing still others.” (ST, 577)
• "If the determinate contentfulness of the thought and intentions even of the niederträchtig is in fact intelligible only from an edelmütig perspective, then anyone who in practice treats what he is doing as judging and acting is implicitly committed thereby to Edelmütigkeit.” (ST, 577)

​

A problem (?)

• The AA argument claims that the sceptic cannot rationally commit to reductive naturalism about norms, because in asserting such a commitment, the sceptic invests her authority in the truth of the view. Assertion as such understood is not possible from the sceptical standpoint; hence scepticism fails.
• But is this the only explanation for what the sceptic does in asserting a commitment to the reductive view?
• Compare: I can only make a self-identification explicit by using the first-person singular pronoun (e.g. by saying, "I am the shopper whose bag is leaking sugar".) But in order to explain what I do while expressing a self-identification, I need not use the pronoun "I". For that, I only need to use a description of the character of "I".
• Similarly, the sceptic can explain what she does in expressing commitment to reductive naturalism by telling some kind of reductive naturalist story of how she came to believe in the view. This is compatible with saying that she can only express the view by asserting it (where assertion is defined in genuinely normative terms). So while the use of normative vocabulary may be irreducible to naturalistic vocabulary, in the sense that there are no non-normative synonyms for normative vocabulary, the explanation of assertoric behaviour need not be irreducible.

Conclusions

• If the AA argument is incomplete, so what?
• Perhaps material goodness of inferences can be explained without it. Perhaps material goodness just is relative to forms of life like Kuukkanen (and Rorty?) has it.
• For me personally, that's too much relativism. But I must stop here.
• Thank you for listening!