Incomplete

Valéry famously said, a poem is never finished, only abandoned. It is impossible to perfect a work, of course. Even if a line or stanza has been perfected by making all the best word choices, there is always the possibility of adding more lines to it, or the poem. And adding more alters the context, such that what was the best artistic choice may no longer be so. It is just like Gödel's incompleteness theorems, where extending a formal system to prove a previously unprovable theorem adds new unprovable theorems to the system. The trick then, is to find the right point at which to abandon a work.

Works must perforce be read by minds, even if it is only the author's own. In that case, it is the reader's mental resolution that counts. Above a certain threshold of polish, the difference becomes inessential or even indistinguishable. Each reader resolves works differently, but the author can use his own mind as a gauge, cognitive biases notwithstanding.

Quality of writing is one matter. Completeness of content is another. I am loathe to write anything incomplete, especially if it's non-fiction. After all, a single fact or concept can swing an argument. Thus, I have avoided many topics that I would love to expound on, such as thinking, ethics, and metaphysics, as I know that there are things that remain to be known about these issues before I can discuss them. After all, the more you know, the more you know what you don't know.

You can never rule out the possibility that you lack a decisive piece of knowledge. But you can stop actively looking for new facts and ideas once you can account for all the things you already know. And you can still write once you have pushed the things you don't know outside your area of focus.

Donald Rumsfeld spoke of known knowns, known unknowns, and unknown unknowns, for which he was pilloried for muddled English. But the expression is flawless; all that is lacking is the critics' comprehension of recursive sentences (Chomsky would be annoyed at their subhuman linguistic skills). Above all, it was an elegant yet profound explanation in simple English of arguably the key problem of applied epistemology. So, in those terms, how do we deal with the unknown? Known unknowns can be dealt with because you know how they could impact the big picture. However, worrying about unknown unknowns is too paralysing. The negative capability proposed by Keats is useful in (and only in) this regard, when having surveyed the unknowns, you can judge that the knowns will suffice for your purposes.

But some works need not be complete; they can be great because of what is left out, intentionally or obligatorily. For some works, their Shannon entropy is reducible by the reader, who can go from what is said to what is implied with degrees of certainty. For others, they are meant to provoke thought. Raising questions which the author himself might not be able to answer can lead the reader down interesting avenues, where knowledge is earned, even if not that which was set out for. This can reveal to the reader unknown unknowns, as well as unknown knowns. Incomplete works cohere (or not) with the reader's knowledge and values to form a more complete whole (as do pretty much complete works).

What of this piece of writing then? It can never be perfectly complete of course, but is it intended to be complete or incomplete? It seems fairly self-explanatory, but nonetheless is only a platform for readers to figure out what they know they don't know, what they don't know they don't know, and how that affects the way they see things. And this is the point at which I abandon the work.