"Blue potatoes are ungainly things
As are red and purple lamb chops
Yet when we eat and creep and fall
We never ask a silent question."
Those are just a few words that Racter, an artifical intelligence computer program has composed. As a matter of fact, Racter (derived from raconteur) has actually "written" an entire book, called The Policeman's Beard Is Half-Constructed (read it here: http://ubu.artmob.ca/text/racter/racter_policemansbeard.pdf).
But is the program actually "writing"? Is it actually using language to communicate to us, or are Racter's writings just a random string of words and phrases that are grammatically correct but carry no real meaning? Some of Racter's written observations seem quite astute, as it concludes at one point that "more than iron, more than lead, I (Racter) need electricity...I need it for my dreams."
As Hayakawa taught us, language is used for cooperation, to add to the great pool of knowledge. Are Racter's ramblings truly "language" then? Racter may not be self-aware, but that doesn't necessarily make impossible its ability to communicate coherently to human beings (Google the "chinese room parable" for a good explanation of this.)
Is true language and literature solely the domain of the human mind, or will machines one day be able to match us?
(Eric Wei, who is suffering from a fever and is dreading midterms. PS: Did Mr. Lazarow say how many questions would be on the midterm or the content? I'd assume semantics, the American Dream, and the Puritan material would be on there...)
11 comments:
I don't think we can really classify this as language. My understanding that language was created for the use of communication. The computer is programmed to print out such words and phrases. It isn't really trying to communicate with us (unless it's trying to show that its acting according to its programming).
Something doesn't have to be writing or language just because there are words. The thing symbolized doesn't necessarily have to exist just because we see a symbol.
(And I understood that we were having two essays on the midterm and that there will be a final vocab test the Monday after midterms that will also count as part of our midterm grade. I assume all topics discussed so far are up for game.)
I agree with Mary about the words not necessarily representing language. The computer can say things but are they ever anything of value? Does he think things through and come to logical conclusions? This would be similar to a human brain. If he can just string words together, I don't believe that counts as language using the definition that we understand it.
(Kelley Volosin)
By the definition of language, I can see how this computer is using language. I think Racter is trying to send a message out although it is true that this computer is just programmed to jumble up a few words to make them into a sentence.
In this case, I think the language that this computer is utilizing is used for one-way communication, if such thing exists. The computer is giving us a message, and we understand what it is talking about. In a way, the computer and humans are cooperating, but in a strange form.
Jennifer Park
I'd say that Racter is capable of writing, but not of "communicating", and we all know that language = communication. Young children make noise, they (try to) verbalize their thoughts. They may not be coherent, but they are recognizable. (That makes sense, right?) A baby cries, signaling for the parent to provide it with milk. No "words" are exchanged, yet a message is passed along. Thus, we consider a child's cry to be a form of communication. Racter spits out jumbled words that we can read, but we are unable to exactly decode their meaning. Their production was therefore futile. This is where I meet a contradiction, however, because according to Hayakawa, a lot of our language is empty and pointless.
Just like everything else we have discussed, there are two sides to the argument. The ironic part, however, is that both sides can be defended by Hayakawa's reasoning.
1.) Racter does not use language because he is not "communicating".
2.) Racter is using language because a vast amount of speech/writing has no essential meaning (loaded words, ritualistic utterances, etc..).
I hope you understand what I was trying to say. I understand that it is a little bit vague.
(Sam Maliha)
I'm in agreement with mostly everyone here that Racter is not using language because it is in no way communicating. The way it’s programmed is the way it operates from my understanding, and therefore the computer is not capable of rational thought or language for that matter. Even if it is capable of creating sentences, I believe the key to language is meaning. Because a computer can never give meaning to the words it formulates, I believe the claim that the computer is “using language” to be incorrect.
Now I have a question of my own. Do animals use language? Such as when they communicate to one another whether or not there is danger? Although monkeys will never write works of literature, would you classify the screams and shouts they make as language? I mean after all, they do have meaning, do they not?
(Kevin Trainer)
I agree with a number of you. Because Racter's words do not have a meaning that we are able to process and respond to, they cannot really be considered "language." Language = cooperation. However, I also understand Sam's suggestion that perhaps Racter's use of language CAN be justified by the principle of the existence of both directive AND affective language, not simply directive language. In other words, not all language has to have "meaning" in the sense that a response must follow. This of course implies that Racter may actually be using "language." Now that I've utterly contradicted myself, I ask: is there possibly more than one way of looking at this? It seems so, if we remind ourselves of Hayakawa's principles.
I have one quick question which results from the fact that I lack much knowledge of AI computer programming: if Racter "wrote" a book, was it programmed specifically to write a book, or was this action unintended? I'm a little confused...
In response to Kevin's question, animals DO use language because they communicate through their own mechanisms, despite the fact that they do not use words. (If I remember correctly, I believe that Hayakawa implied this in the first chapter of LITAA.) When monkeys screech at each other, they are communicating in some way, whether this may be signaling family members of danger or telling other monkeys to beware of their territory. Thus their screeches are a form of language. Birds and a myriad of other animals have certain mating calls; these calls are forms of language as well.
(Janet Lee)
I think that the question is "Is Racter trying to communicate with us?" as much as it is "Can we (or perhaps, DO we) want to understand Racter?" Your example reminded me of an exercise we did in a writing class at nerd camp (yes, i know). You take a well-known poem or speech or whatever and you replace all the key words by looking the word up in the dictionary and picking the fifth word listed on the page. Afterwards, you try and analyze it.
I believe we ended up with a poem along the lines of "Cling fast to driftwood, There is a lightbulb hidden behind the fig leaf". Which we then analyzed for its religious overtones (fig leaf-->Garden of Eden, lightbulb-->God, you get the point). So I guess my response is that we can find meanings and intent behind almost anything.
If we wanted to find meaning in Racter's seemingly meaningless phrases, sentences, etc., as Grace implied, I'm sure we could. We are humans- we love to analyze things. It's hard to tell whether or not Racter actually is trying to communicate, per se. But, then again, how do we know we're communicating with each other...?
Emily T.
Emily brings up an interesting point. How do we know we are communicating? How do we know that we even exist? Are we dreaming, or is this reality? What IS reality? The list goes on and on. I think we are communicating, or under such an illusion, because we do manage to get work down. Language = cooperation, and we are utilizing it to its full effect. We're doing something right, I suppose.
(Sam Maliha)
Putting aside all of the more abstract concepts such as logic and reason, we can simply refer to Hayakawa's postulations on human language. Yes, animals have their own form of language which stimulates other animals into action - in effect, they are capable of transferring signals (i.e. There is a predator nearby). The distinguishing characteristic that sets our language apart from the animals' primitive cries is the idea that "Language... can be about language." (Hayakawa 6) We can try as hard as we want to find reason in Racter's "book", but one cannot deny that he would never be able to respond to any human language (i.e. written texts).
Post a Comment