Category: language


Semantics.

January 6th, 2013 — 6:53pm

John Searle, the philosopher who does not believe that computers will ever manifest artificial intelligence, is constantly stating that textual computation, or syntax, isn’t semantic, or meaningful. It seems disingenuous to me, since syntactical interactions are the dynamic structures within which semantics occurs, and semantics can be entered into any system through the use of a feedback loop that relates to a larger environment. But his arguments are always floating around in one way or another if the possibility of artificial intelligence is being discussed, and so the supposed lack of semantics in digital content naturally comes to mind while studying up on CSS3, or the most recent development in Cascading Style Sheets for HTML, which is the way that web designers are able to separate the structure of a website, it’s HTML, from the content and the presentation of that content.

By using the design aspects of CSS, the HTML that remains is left to better describe the structural semantics of the content by labeling it as something like Headline, or Aside, or Emphasized Text, etc. Which means that as the web develops, textual content will more and more manifest itself in semantic terms, a feature that seems to emerge naturally as an outgrowth of the human desire for efficiency and clarity. Not that this is a new idea or anything, but the concept of the semantic web as a naturally occurring emergent refutation of Searle’s arguments against the possibility of computational intelligence is interesting. It’s like the universe itself is telling the guy he’s wrong.

el-placer-del-poetaBut since CSS is being used as a presentational layer over the structural semantics of HTML, I was thinking that it is also evolving into a separate emotional overlay of the content. The designer is not only making things pretty by using CSS, but also helping to better define the emotional semantics of the content in the same way that our brains overlay emotional semantics over everyday situations. When you look at a DiChirico painting there is a weird, lonely, nostalgic quality that comes from an emotional relationship to the content that would be invisible to an entity that did not know what it was like to stand alone on an empty flat area as the sun has begun to set. This is probably accentuated by hardwired evolutionary associations of danger that come from being caught out in the open as the sun goes down, making one a target of predators or in danger from the elements.

So I was thinking you could scan a document for emotional content by creating a lookup table of the relationship of single words or word groups to a particular emotion, perhaps FEAR. Then use CSS to attach each word group to a background shade of perhaps red, from 0% to 100%, depending on how FEAR-related they were. Then by quickly scanning through the lookup table a browser could extract the content, apply new CSS rules, and the user could see the amount of FEAR in a document immediately, without having to read it, the way we feel an emotion by looking at an image. If the lookup tables were extracted from an individual’s writings algorithmically, then the extraction could be personalized. You could probably do that with Google glass, overlaying shades of color as a user reads, to personally accentuate the emotional content.

Comment » | consciousness, design, language

American != English

May 18th, 2010 — 2:29pm

Computer languages are specific ways of speaking that rely more heavily on process than content. And at this time, all major computer languages are dialects of English. But there are interesting cultural differences encoded into the structure of the different languages. I would say, to be even more specific, that most computer languages are dialects of American English and that some of our national personality comes through in the grammar and syntax.

The equals sign is a good example, as it has a unique meaning in languages like ObjectiveC. When you say ‘if (self = [super init])’ you aren’t really asking a question. This is exactly the same sort of question as “Could you pass the butter?” You know for a fact that the person you’re talking to can almost certainly pass the butter and you are actually requesting that he or she do it; right now. It’s a polite command. The equals sign is like that. In C, ‘=’ is used to make something become something else. If you are actually making an inquiry you have to repeat yourself like this “==”. All of which is second nature to the programmer, but programmers tend to be socially awkward, so the differences between commands, requests, and inquiries may not be particularly noticeable to them.

‘if (self = [super init])’ means, “You should initialize yourself right now. And if you can do that, I want you to continue with the rest of my request…”

Americans drape their orders in layers of false politeness, and their computer programs do the same.

Comment » | C, language

My New Painting(s): parthenogenesis

June 16th, 2008 — 11:53pm

Desperate to come up with new art, I pulled one of my unsuccessful paintings from the garage and decided to turn it into a triptych. I scratched out with a pencil the areas that will become the three new paintings I hope to make from this single crappy one. It’s a close up of a woman’s face, with agrammatical text (which seemed more interesting when I wrote it) “fruits of an eye where mouth can’t speak.”

I am fascinated with agrammatical text, but I seem to be the only person in LA who sees it as anything other than a gimmick. This non-sentence was supposed to define the concept for an audience, but it seems cloying in retrospect. The point of it all is that we are bound by the strictures of our thinking, which itself is bound by the structures of our language. So by breaking that structure and forming sentences that are incorrect in standard English, we can on occasion reach thoughts that would be unattainable in any other fashion. For this to make sense you have to believe that thought is composed of language or at least that language is a requirement for conscious thought to exist. Most people don’t believe this.

Most people assume that language is a way of describing an inner life. They think they have thoughts that are TRANSLATED into language. This, I believe, is bullshit. There is a reason why transcendental meditation requires you to clear your head of words. Without language, it would be impossible to understand the concept behind this very sentence. Your understanding of that sentence is different than the thought OF the sentence. In other words, you have a feeling that you sense as you understand a sentence like the one earlier, but that feeling is the result of conscious thought composed of words, it could not give rise to the words that created that feeling.

All of this doesn’t make my original painting any less crappy. I created it on a computer and then had it painted in oil in Vietnam (along with 12 more of varying degree of success). An interesting point, or criticism, that no one has bothered to make because no one really likes the idea, is that this concept—that words CREATE thought—is denied by the manufacture of the painting itself. That’s because with this series of paintings the ideas did come first; the idea (or Platonian original form) was created in the computer. This expression of the idea (created by Vietnamese laborers) is merely the playing out of that generative force.

None of this has any bearing on the fact that the original painting is lame. But now it will become three paintings. Hopefully one of them won’t suck.

Comment » | art, consciousness, language, Whining

Back to top