The origin of the medieval Liar Paradox has often been attributed to Aristotle's Sophistic Refutations, which became available in 1130, spawning a medieval tradition of commentary. In Chapter 25 of the Sophistic Refutations, Aristotle writes:
(Grab your Dramamine and your barf bags)
All arguments of the following kind have this feature: 'Is it possible for what is-not to be? No. But, you see, it is something, despite its not being. Likewise, also, Being will not be; for it will not be some particular form of being. 'Is it possible for the same man at the same time to be a keeper and a breaker of his oath?' 'Can the same man at the same time both obey and disobey the same man?' Or isn't it the case that being something in particular and Being are not the same? On the other hand, Not-being, even if it be something, need not also have an absolute 'being' as well. Nor if a man keeps his oath in this particular instance or in this particular respect, is he bound also to be a keeper of oaths absolutely, but he who swears that he will break his oath, and then breaks it, keeps this particular oath only; he is not a keeper of his oath... The argument is similar, also, as regards the problem whether the same man can at the same time say what is both false and true...
The first half of this passage is typically referred to as the Perjurer, and the second half of this passage as the Liar. Aristotle seems to imply that they can both be solved in the same way, but there are difficulties in doing so.
The Perjurer passage from the Sophistic Refutations amounts to saying something like this: if a man swears that he will break his oath, then later swears that he will go to Athens, but then breaks his promise and refuses to go to Athens, then it could be said that he has both kept and broken his oath, which would be a contradiction. However, Aristotle says that it is not a contradiction (i.e., not an insolubilia or paradox), but a fallacy - specifically, a fallacy secundum quid et simpliciter, because it is false simpliciter (absolutely and taken on the whole), but true secundum quid (in a certain respect).
The application of secundum quid et simpliciter seems to work in the case of the Perjurer: he is, on the whole, an oath-breaker, although with respect to his particular vow to break his oath he is an oath-keeper.
Aristotle's suggestion that "the argument is similar for the problem whether the same man can at the same time say what is false and what is true" is more problematic. Many medieval authors assumed that this "problem" referred to the Liar Paradox, and then attempted to solve the Liar in the same way Aristotle solved the Perjurer, secundum quid et simpliciter. They then either discovered that such a solution was prevented by an antinomy that needed to be worked around, or ignored the problem altogether.
The antinomy is this: to solve the Liar through the attribution of this fallacy, it must be argued that the Liar is false simpliciter and true secundum quid. To do so, the Liar must be considered as having two parts, as in Aristotle's Perjurer. However, we see from the example of just such an endeavor given in Paul Spade's "The Origins of Mediaeval Insolubilia" that reconstructing the Liar Paradox as a formal parallel to the Perjurer doesn't create the two instances required to support the secundum quid et simpliciter solution:
"Supposing that Socrates promises that he will say only falsehoods to me, afterwards when he comes to me, he says that I am a stone. Then he speaks the truth with respect to discharging his promise; therefore he speaks the truth. Likewise he lies, because he speaks falsely. Therefore the same man lies and speaks the truth."
The Liar considered thus, in the same mode as the Perjurer, can be solved without considering it as a logical fallacy. One observes that the Liar (in this case) both fulfills a promise and speaks falsely in a single act, which is not a paradox. The only way to yield the paradox (and therefore, the applicability of the solution via the application of the secundum quid et simpliciter fallacy) is to consider the incompatibility of two distinct acts, two utterances which generate confusion between keeping or breaking one's oath simpliciter and doing so secundum quid. The solution to the Liar as constructed in parallel with the Perjurer is simply to observe that the liar speaks truly and falsely at different times.
Ultimately, the difficulty revealed by attempting to solve the Liar and the Perjurer in parallel via attribution of the fallacy secundum quid et simpliciter is that even though the two can be set up in exactly the same way, their solutions will still be different due to the modal difference between the Perjurer (who makes two oaths) and the Liar (who makes one).
Note: To get around the antinomical problems posed by Aristotle's locution on the Perjurer/Liar solution, some authors (Thomas Aquinas, Adam of Balsham) have the Liar (and/or the Perjurer) speak in the present tense, which is not clearly the case in Aristotle. Others adapted the secundum quid et simpliciter argument terminologically, to restrict the way that terms in an insoluble supposit (e.g., that they supposit "in a certain respect" rather than "absolutely", which amounted to a restriction on self-reference, or through restringentes. Ultimately, the use of the secundum quid et simpliciter (and efforts to address the difficulties in making the Liar parallel to the Perjurer) constituted the core of early commentary tradition on the Liar Paradox./
Source:
Paul Vincent Spade, "The Origins of Mediaeval Insolubilia Literature", in Lies, Language and Logic in the Middle Ages, London: Varorium Reprints, 1988.