Analytic and synthetic

Kant was right to make a distinction between analytic and synthetic judgments. Contemporary criticisms of it seem all to miss the point, and try to make it about analytic and synthetic sentences or propositions, which obviously do not exist. Kant was speaking of mental judgments,[1] which propositions merely express. The same sentence, “all bodies are heavy”, expresses an analytic judgment when said by someone whose concept of body includes weight, and a synthetic one when said by someone whose concept of body only includes extension.

Kant did speak of some propositions as essentially analytic or synthetic. This should still not, however, be taken as saying something about the sentence, but about the truth of the thought that it typically represents. It is not that the meaning of the sentence, 7+5=12, is or is not the same on both sides of the equals sign. Like all simple affirmative propositions, it can be taken as an analytic or a synthetic judgment. Kant thought, however, that if taken as an analytic judgment, it is false, whereas if taken as a synthetic judgment it is true, given what he thought to be the usual concepts pertaining to five and seven and addition and twelve.

Clarifying convention

Since every sentence can be taken both ways, and this matters a lot to their truth, I find it useful to introduce a convention of explicitly marking propositions intended as analytic judgments with such a phrase as “by nature”, “by definition”, “essentially” or “necessarily”. Indeed, we often already do this, but not consistently. So, “all bodies are, by nature, heavy”, might be true to some people, but it is false to a Cartesian, who thinks that bodies are only extended by nature, and only acquire weight from the action of external forces.

This convention only makes everything clearer. Rephrased into this convention, take Kant’s claim that it would be false to say that “all triangles, by nature, have angles that add up to 180 degrees”. Some persons would find this claim strange, but they are the same persons who would find it strange to hear that the proposition, “all triangles have angles that add up to 180 degrees”, should express a synthetic rather than an analytic judgment.

Cognition markers

Calling them all by the general name of cognition markers, I give here some examples of sentences that would typically be used to mark a sentence as expressing an analytic or a synthetic judgment. I recommend that philosophers use them, especially the analytic judgment markers.

• analytic judgments: by nature; by definition; essentially; necessarily
• synthetic judgments: accidentally; not by nature; not by definition
• not sure if analytic or synthetic: for some reason; somehow
• judgments a priori: demonstrably; always and everywhere
• judgments a posteriori: empirically; by all experience; as far as I know

To the Kantian, all triangles have, not by nature but demonstrably, angles that add up to 180°.

Of course, since I only just thought about this, none of my posts written up to now have followed this convention consistently. I might update them later to be more in line with it.

Notes

[1] When stating the distinction in the first section of the Prolegomena, he was blindingly clear about this. There, he spoke about “judgment”, “concept”, “what has been thought”, and “cognition” – not “sentence”, “term”, “what has been said” and “communication”. In various places, however, he did speak about synthetic and analytic propositions, which was a mistake.