Two notions of possibility
For Aristotle the term ‘possibility’ is homonymous (Pr. An. I, 3, 25a37–40): on some occasions the possible and the impossible are contradictories (e.g., De int. 12, 22a11–13; 13, 22a32–38), while on others possibility is incompatible not only with impossibility but also with necessity (e.g., An. pr. I, 13, 32a18–21). Whenever the distinction is relevant, I shall call possibility in the first sense ‘possibility proper’ and possibility in the second sense ‘contingency’.
In the Latin translation of De interpretatione by Marius Victorinus which was used by Boethius Aristotle's two terms for ‘possible’ were translated by the Latin terms ‘ possible’ and ‘ contingens’, which Boethius understood to be synonyms. This was the usual view in early medieval logic, and it can still be found in the squares of opposition for modalities presented by William of Sherwood and Peter of Spain in the middle of the thirteenth century. Already in the twelfth century, however, there were attempts to give separate meanings to the two words. For instance, John of Salisbury criticised those who used the terms as synonyms; according to usage in his time a mere absence of impossibility did not warrant calling something contingent. Even in those works in which the terms are used as synonyms there often is a remark referring to a related distinction, according to which ‘contingens’ is opposed to ‘necessarium’ in the sense that some possible sentences are necessary and others contingent. This became the dominant use of the words in later medieval logic.