Finally the name of accident is commonly accepted in three ways principally. One for a thing which by its own nature inheres in a subject, such as color and taste of milk, and similar. Another for any attribute not pertaining to the essence of the thing of which it is said; in which way not only all contingent attributes accrue to things, as to iron to be hot, and to man to be white or a grammarian, but also very many necessary ones; nor by any necessity alone, but drawn from the very nature of the subjects, as to man to be capable of laughter or of disciplines. By which signification we used a little before in the verb Accidit, where we explained the fourfold division of proper. Third for an attribute which neither pertains to the essence of the thing, nor flows from it as it were by necessary emanation: in which way by another reason it accrues to man to be black, another to be capable of disciplines, because this in the very nature of man insofar as he is man has whence to arise, that from another source comes forth. And this is the proper and peculiar reason of accident, which makes it the fifth, extreme species of universals; provided accident is taken not as this or that in number, but as it is common to many. For although the name of accident carries nothing universal before itself, it can be accepted for singulars equally as for universals: from the very use of philosophers, however, it is accommodated to signify this extreme species of universals, of which the rest remains for us to speak.
It is usually defined accident taken in this way, that it is present and absent without corruption of the subject. Likewise, that it chances to be in and not to be in. And finally that it is neither genus, nor species, nor difference, nor proper, but inheres in the thing: the sense of which explanations is the same. In which definitions the place of genus is understood universal, so that the first definition may have this: Accident is a universal which can be present and absent, be affirmed and denied of subjects without their overthrow. And in the same way the second definition can be understood. The third truly thus can be exposed: That which indeed can agree with many by the same reason, but not as genus, nor species, nor difference, nor proper. So that in all these definitions the contingent is put in the last place as the constitutive difference of accident. More explicitly, however, accident will be defined if it is said: universal, or that which is predicated of many in the question what kind it is, accidentally and contingently; in apt order: first place individuals are excluded; then genera and species; then truly differences; lastly properties.
This genus of universals comprehends not only real accidents, but also attributes of reason and negations which are said contingently of many, and even substances themselves circumposed after the manner of accidents, as to be a predicate, a subject, blind, dark, clothed, adorned, gilded, fortified. For all these can contingently agree with many and be affirmed of things not destroyed of which they are said.
Accident is further divided into that which is truly separable, and that which is inseparable. For although every accident of this species of predicables can by thought, and even by negative thought, be separated with the subject preserved, since in its nature considered in itself nothing repugns it to lack it; not all however can be separated in reality, but some can, some not. For to sit is separable both by thought and in reality; to be snub-nosed, blind, and similar are separable by thought, not however in reality. In which thing a broad distinction is seen between accident and proper, which neither in reality nor by negative thought can be separated with the subject preserved. For if, for example, you deny man to be capable of disciplines, by the very negation you are forced to confess him not even to be a man, since from the very nature of man necessarily follows aptitude and propensity for undertaking disciplines.
Accident can also be divided into general and special, just as difference and proper; as general indeed is said of species differing, as white, to watch, to sleep; special truly of those differing in number alone, as to laugh, to be a grammarian.
Now if accidents are compared with properties, and the discourse is of true and real accidents and properties, that is usually observed how properties first agree with universals, then through universals are communicated to singulars; so accidents first agree with singulars, then through singulars are attributed to universals.
But if accident is compared with all four superior genera of universals, this distinction, besides others already said, is seen between them, that those four always have the act of being in many conjoined with aptitude, whether they themselves or their particulars exist or not (for it was always true that Socrates is man and animal and participant of reason and capable of disciplines even before he and such predicables existed in the nature of things; true it is now after he closed his extreme day; perpetually true it will be) accident truly although from all eternity it had aptitude to be in many; it does not have the act, however, unless both it itself exists and the particulars in which it exists; as heat existing in water and wood existing.