An important question came up in class today. "Why don't poets just say what they mean."
If an absinthe maddened verse-jockey somewhere wants to tell me that he's upset over his most recent lover's decision to spend the rent money on a collection of needlepoint replicas of scenes from the patriotic body-art documentary "Tattoos of our Founding Fathers," he could just say it. This is often a source of frustration for science students attempting to navigate literature classes. The science student, as an ostensibly logical being, wants information that is direct and free of distortion. In fact, this is an issue for any person with a growing expertise in one field. After working so hard to compile a library of information and skills, we might not be terribly interested in having to adapt it to another, purposefully complex, system.
The poet is more likely to wonder (or probably opine) about the apparent sterility of mathematics and the engineer laments the theorist's ignorance of utility. What I want you guys to recognize is that underlying all these fields is a meaningful foundation: There is information in this system and you can get it.
Subscribe to:
Post Comments (Atom)
Posts
3 comments:
The thing of it is that sure, he could say it, but there is intrinsic information in the way that it's said. The poet's use of a specific image is meant to affect the bias in the representation used to view the text. This is a major problem in NLP, because while we can extract meaningful information from the text based upon a number of it's properties, how it's meant to be interpreted is dependent upon the way the text is structured.
While, as our expertise in different fields increases we have a tendency to want to stick to those, rather than having to learn a whole new system. I myself find that an interesting counterpoint to adapting my expert system (In my case physics and Math) to the new one (Also in my case Machine Learning and NLP) is finding how I can view the new system in my old terms. By being able to work on both, I'm able to construct an abstraction of both systems which allow me a better understanding of each. A great test that I used to have for this was going to parties and engaging art students in conversations about their views about life, the universe and everything. Though there would be common terminology involved, inevitably what they would mean by the sentence "The universe is made of energy" would be completely different from what I might mean if I were to make such a statement. As such, the trick then was to figure out what they meant in my terms, come up with a response, and then respond to them in theirs.
There's also the matter in the scientist viewing the poet as obscure and the poet viewing the mathematician as sterile of both missing a key point. As scientists we have many different ways of representing a given system (this is entirely the subject of modelling). It is on us to recognize that the poet simply has his own model, figure it out, and learn what the useful and unique information it provides is. Likewise, it is on us as poets to recognize the beauty in the abstraction as well. To treat the letters and symbols as, at the very least, dadist ramblings, and at best to learn what this other language means, understand what the mathematician is trying to say, and find the beauty in it.
Personally I find the mixing of these things to be essential to the development of new research.
Part 2 of this post deals with the inherently encoded way that any information gets transmitted, including the notion that poets write in code just like computer scientists, general relativists (the physics equivalent of haiku addicts) and interpretive dancers.
Where's part 2 then? I'm itchin' for the read man... Jonesin' even...
Post a Comment