The reasons against the converse construction are nearly, if not equally cogent. To have done anything at any previous time in order that a present effect may follow, is, ipso facto, to convert a past act into a present one, or, to speak in the language of the grammarian, to convert an aorist into a perfect. To say I did this that I may gain by it, is to make, by the very effect of the expression, either may equivalent to might, or did equivalent to have done.
I did this that I might gain.
I have done this that I may gain.
A clear perception of the logical necessity of the law of the succession of tenses, is necessary for understanding the nature of several anomalous passages in the classical writers. In the following, an aorist is followed not by an optative, but by a subjunctive.
Οὐκ ἀγαθὸν πολυκοιρανίη· ἑῖς κοίρανος ἔστω,
Ἑῖς βασιλεὺς, ᾧ ἔδωκε Κρόνου παῖς ἀγκυλομήτεω
Σκῆπτρόν τ' ἠδὲ θέμιστας, ἵνα σφίσιν ἐμβασιλεύῃ.
Here it is necessary to construe ἔδωκε, has given and continues to allow, which is to construe it like a perfect[[64]] tense. Upon similar passages Mathiæ writes, "but frequently the conjunctive is used, although the preceding word be in the time past, viz., when the verb which depends upon the conjunction shows an action continued to the present time." That means when the verb is really a perfect.
In Latin, where the same form is both aorist and perfect, the succession of tenses is a means of determining which of the two meanings it conveys. Veni ut videam=I have come that I may see. Veni ut viderem=I came that I might see.
Arnold states, from Krüger and Zumpt, that even where the præterite was clearly a perfect (i. e., =to have with the participle), the Roman ear was so accustomed to the imperfect subjunctive, that it preferred such an expression as diu dubitavi num melius esset to diu dubitavi num melius sit. The latter part of the statement is sure enough; but it is by no means so sure that dubitavi, and similar forms in similar constructions are perfects. There is no reason for considering this to be the case in the present instance. It seems to be so, because it is connected with diu; but an action may last a long time, and yet not last up to the time of speaking. Diu dubitavi probably expresses, I doubted a long time, and leaves it to be inferred that now I do not doubt.