Does "a priori" have an implied substantive?

Question

Is a priori short for a longer phrase of the form a priori _____? If so, what is the elided substantive?

---

Background, or, How I Got Confused

I'm pretty sure that a fortiori is short for a fortiori ratione. I haven't found the origin of the phrase, but it seems clear from usage that there is an implied noun, and that that noun is either ratio or a synonym. Douglas Harper reports that a priori started c. 1300 to describe reasoning from a cause or first principle. That suggests to me that a priori might also have an implied noun, also obvious from usage and hence likely to be dropped—maybe a noun meaning "cause" or "first principle" or "antecedent".

Contemporary usage is strongly influenced by Kant, who used a priori as a sort of adjective for knowledge deriving from the nature of the mind rather than from the objects sensed or experienced. I had gotten the impression that the elided noun might mean something like experience, so the phrase would mean "from before experience". Today people often use the phrase to refer to things you can know before doing an experiment, either by deducing from already-established theory or just because you already found out; see e.g. this paper on designing neurophysiology experiments.

Now that I've been learning Latin, though, it appears that a priori is the start of a prepositional phrase that ends in a noun, like "from a previous _____" or "from the first _____", not capable of functioning as a sort of preposition itself, like "from before _____" or "from earlier than _____."

But I just read this in Guyer and Woods' translation notes on the Critique of Pure Reason (p. 74):

Our decision [to preserve ambiguous and obscure constructions in Kant's original text wherever possible] also means that where Kant's location of the adverbial phrase "a priori," which he always treats as a Latin bor­rowing rather than a naturalized latinate German term, is ambiguous between an adverbial modification of a verb and an adjectival modifica­tion of a noun, we have tried to leave it ambiguous, although we could not always do so.

So does a priori work like in concreto: as a prepositional phrase with no noun? On this hypothesis, as in concreto means "concretely", a priori would mean "beforely"—except that since it's a prepositional phrase, it can modify nouns as well as verbs, i.e. serve as adjective or adverb.

Answer 1

Here are some scattered observations, which may or may not constitute an answer.

I found¹ no classical examples of a priori, but several examples of a priore², most followed by an ablative noun. Pliny has an a priori without a following noun, but there is an obvious implicit defecto. (English version here.)

The Rule of Benedict has several examples of a priore. In this work prior is practically a noun, and this translation by Rev. Boniface Verheyen consistently translates it as "the Superior". This is the earliest example I found of substantivization of prior (including a priore), and it is consistent throughout the work. The meaning is very different from the philosophical or logical a priori.

It is hardly a surprise, but Gauss uses a priori in the modern sense in Latin.

Lastly, I am not sure if there is a real difference between a substantivized adjective and an adjective used with an implicit noun. There is an effect on gender (eg. res mala and malum), but all genders look the same for priore/priori. It is interesting to ask how the phrase was born — and I cannot give a convincing history on that — but after some thought I see little point in figuring out whether it is a substantivized adjective or has an implicit noun.

---

¹ The search was admittedly not exhaustive. I only looked at the texts of The Latin Library.

² In classical Latin the third declension the singular ablative ending is (almost) always -i for adjectives. In comparative (which is third declension for all adjectives) the ending is -e. And if prior was declined like a noun, it would get the -e. Therefore a priori with the -i looks like a later development. If it was abbreviated from a phrase used by a classical author, I would expect it in the form a priore instead.