Justin E. H. Smith in his own blog:
There is a familiar distinction in philosophy between contingent and necessary truths. Truths of the latter sort are those the negation of which implies a contradiction, or those that are true simply in virtue of the meaning of the words involved. For example, “A triangle has three sides” is true simply in virtue of the meanings of the words 'triangle', 'three' and 'side'. If you encounter a figure with four sides, then necessarily you have not encountered a very unusual triangle, but rather a non-triangle.
Contingent truths are those the negation of which implies no contradiction, or, to put this somewhat differently, those that could have been false (whatever that might mean!). Some contingently true statements involve particular cases, e.g., “This swan is white.” A special class of contingent truths are those expressed by empirical claims about how one expects all entities or phenomena of a certain kind to be. These are the sort of truths established by inductive reasoning, and it is characteristic of them that they can always turn out to be falsified by any given case. Thus, “All swans are white” was held to be true for a long time, as the instances of observed swans grew and grew, and in each case, each swan observed turned out, in fact, to be white. This contingent truth however, turned out to be false, as European travelers to Australia, home of the Cygnus atratus, realized toward the end of the 18th century.
Now, any member of the genus Cygnus is a swan, and there was a prior fact of the matter, prior that is to Captain Cook's expedition, about the color-independent features of an entity that determine whether it is a member of this genus or not. This is what makes “All swans are white” a mere empirical claim rather than an analytic truth, or a truth that can be established simply in virtue of the analysis of a proposition into the meanings of its component parts.