It can be tempting to think that science is the pursuit of universal truths. A hypothetical fact would then only be scientific when it describes some (nearly) universal pattern. And indeed some of the best examples of good science are of this kind; the paradigmatic one is of course Newton’s laws that unified the falling of apples and the circling of planets. Universal indeed! (If you read Dutch: I found Floris Cohen’s Isaac Newton en het ware weten worth my while.)
But one should not give in to this temptation. Universals are scarce, while defeasibles are ubiquitous.
Science is the pursuit of interesting truths. And interesting truths can be very specific. The discovery of the coelacanth is one example; another is the building of a computer that can beat the chess world champion, or more recently, also by IBM, a program that can play the strange game of Jeopardy well. In such cases it seems to even matter less how well the result is described in a peer reviewed highly ranked journal. Some results are just interesting in themselves.
Of course greater universality can make a truth more interesting. But more often it is the defeasibles that are the most interesting: defeasibles have the power to describe patterns as they exist in their relevant context, including a specification of how they depend on circumstances.
The here-and-now character of many interesting truths is (in my interpretation of his work) one of the themes of Stephen Toulmin; e.g., in his book Return to reason.