h a l f b a k e r yThe mutter of invention.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
There's been a lot of debate in the past few years about restoring dinosaurs to life using samples of their DNA. But what if we wanted to restore people to life? Genetic factors play some role in personality, but environment is far more important. What then would allow us to reconstitute personalities?
Nothing could be better than their writings.
My idea, then, is to develop a computer system that will be able to take in an author's entire works, and construct a model of that person with which anyone will be able to converse. This would require a far better model of the human personality, and the relationship between personality and literary output, than already exists. But work on the relationship between literature and linguistics, and psychoanalytic literary criticism could potentially guide the way. I know this is far beyond existing artificial intelligence research, but it must be a possible goal for that research.
My first thought on who to do this for was Anne Frank, since she obviously deserves more life, and produced some highly personal writing, whilst being psychologically a fairly normal teenage girl, not a lunatic Byronic type or a nutty Emily Dickinson-a-like (who spent most of her life in an attic by choice). However, someone like Wordsworth, who did write about himself at astonishing lengths might be another candidate.
John Lennon Artificial Intelligence Project
http://triumphpc.co...hnlennon/chat.shtml
Database of quotes, indexed via natural languge ... something. I'm not having good conversations with it, but then, maybe I wouldn't have gotten very far with John Lennon, either. [jutta, Feb 12 2002]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Annotation:
|
| |
Hmm. I'd like to see the sort of person produced by this magical process if you fed in the works of, say Will Self, Iain Banks, or Irvine Welsh. |
|
| |
//What then would allow us to reconstitute personalities? Nothing could be better than their writings...work on the relationship between literature and linguistics, and psychoanalytic literary criticism could potentially guide the way. |
|
| |
So you don't hold with reader-response theory, whose central tenet is that no one meaning can be drawn from any text? So no one personality could be divined from any one set of writings. |
|
| |
please leave Mr Wordsworth where he is thank you very much. croissant. I'd like Sylvia Plath to have another shot at it though. |
|
| |
Why would you want to talk to someone who spent all their time writing about themselves? (not that I'm any expert on Wordsworth) I would have thought that they'd be a crushing bore. |
|
| |
DrBob: I really don't want Wordsworth, I just think an egotist would be easier to reassemble from their works, compared to someone like Wallace Stevens, who wrote very intellectual and impersonal poetry while working as an insurance clerk. Although I could ask Willy W if the rumours about him and his sister were true (if the answer really is hidden in his writing.) |
|
| |
The most interesting outcome of this would be how these resurrected writers would react to this period of time. I'm sure Jane Austen would be shocked, but Oscar Wilde would be loving it. Chaucer would have a right old chuckle, Pepys would need to lie down in a darkened room, and Dickens would join a trade union. But would it really be ethical to resurrect Anne Frank and have to break it to her exactly how bad was the overall scale of things that she saw one little part of? How to tell her that people still think that way? But if we're identifying favourites, I'd like to meet Iris Murdoch, of course. Wonder if she'd be a halfbaker? |
|
| |
I thought this was going to be about a group of pompous old farts sitting around discussing Plato's works.... |
|
| |
If created, could this new AI also write "new" works by the author? Could we then scan in all the works of Van Gogh and give the AI free rein to create new masterpieces? I'll take a new Dali for my living room.... |
|
| |
can I call you runf? You can have the Dali if I can have the Escher. |
|
| |
you can call me anything you want ;-) me and [po], we are like peas and carrots. |
|
| |
sorry, shall I assume you are not a fan of the movie "Forrest Gump"? "Peas and carrots" is one of his sayings. |
|
| |
not a huge fan, you old soft centre you. |
|
| |
Dot Parker so that I may take wit lessons from her--if she'd do that which, of course, she wouldn't. |
|
| |
A dream lies dead here. May you softly go
Before this place, and turn away your eyes,
Nor seek to know the look of that which dies
Importuning Life for life. Walk not in woe,
But, for a little, let your step be slow.
And, of your mercy, be not sweetly wise
With words of hope and Spring and tenderer skies.
A dream lies dead; and this all mourners know:
|
|
| |
Whenever one drifted petal leaves the tree-
Though white of bloom as it had been before
And proudly waitful of fecundity-
One little loveliness can be no more;
And so must Beauty bow her imperfect head
Because a dream has joined the wistful dead!
|
|
| |
(Dorothy is almost my favorite poet) |
|
| |
space is short for almost favourite poets here, sorry. |
|
| |
Thank you [rfr]. po, take solace in the fact that it wasn't Homer. |
|
| |
even more powerful than the Godfather |
|
| |
//deliberately a literary spin on their lives for consumption by others, I don't know how useful the books themselves are for reconstructing the author's character. |
|
| |
Exactly Mr Sealy. I'd go further (and I'm only doing this because pottedstu brought up literary criticism in his idea) - the author plays only a very limited part in the interpretation of the text. How can he have a greater impact? He has no idea who will read it and even if he did, he can never have a complete understanding of how someone else would interpret the text. Textual analysis with computers is a well understood discipline, but it is not possible to contrive an algorithm that could perfectly understand the text as intended by the author. Even if you could, you could still debate how much of the impact/meaning that the texct contains was actually intended by the author in the first place.
Its an interesting idea pottedstu (I'd love to have a blether with James Hogg or to go for a beer with R.L. Stevenson) but I can't see how this could be done without the dreaded "magic" word being used? |
|
| |
I'd like to hear Mark Twain, Samuel Langhorne Clemens, again. |
|
| |
Didn't they do that already? |
|
| |
The issue of a reader's reaction or interpretation is (at least to a degree) separate from the idea of restoring a writer's personality based on their work. It's disingenuous to say that because people can have widely differing interpretations of a work, the work can tell us nothing about the author. If you applied the same logic, then psychoanalysis could tell you nothing about a patient (which is another debate, I admit). |
|
| |
Both reader response theory and radical deconstructivist "all readings are equally valid" viewpoints are fringe areas in academic practice, largely because they conflict with common sense, and seem to deny the possibility of human communication. I do not believe that literature is fundementally different from other discourse, which is capable of conveying a clear and unambiguous meaning (although it need not always do so). |
|
| |
Insofar as literature (or texts in general) are produced by people in a common society with common experiences, they are capable of communicating shared meaning. And similarly, the difference between the way people express the same thing reveals differences in that person's environment and personality. |
|
| |
But you're maybe right to be cautious: the amount of information in even the most verbose author's corpus will not be enough to determine the full information contents of their brain. (But please don't mention the Intentional Fallacy, or I'll give you my 3000 word Why Monroe Beardsley was nuts essay.) |
|
| |
I think there is fault with the idea that an author's works, even the works of a prolific one, necessarily give any indication of the mind of the person therein. I don't think it would at all. Depending on the author, of course - undoubtedly some existed whose minds floated right below the surface of their work. |
|
| |
But overall, I'd say it's not so. Especially of authors of fiction. Processing all 64 Sherlock Holmes stories is not going to tell you much about Conan Doyle other than give you his take on the environment in which he existed. And certainly there would not be sufficient data to extrapolate anything valid. |
|
| |
And I'd add to that, the more exotic and obscure the author's works, the less one would learn of the author. I think you'd learn more about Conan Doyle than, say, Lovecraft. |
|
| |
I agree pottedstu - reader response theory is not a "mainstream" view of lierary interpretation, but its not as radical or as marginal as you are stressing (its a staple of most literature degrees for example). However, lets not argue about the popularity or lack of of one flavour of literary criticism, but I think you mention the failings of the type of investigation you propose in your last annotation. For example:
//I do not believe that literature is fundementally different from other discourse, which is capable of conveying a clear and unambiguous meaning - neither do I. If I were being devils advocate I'd say 100% clear and unabiguous meaning in any media is not possible, since the use of any media is an interpretation in itself. It doesn't deny the possibility of human communication, it just reduces the chances of conveying a meaning *perfectly*. The more complex the meaning, the harder the communication.
//texts in general) are produced by people in a common society with common experiences - so what is the common measure? Western society? Eastern society? Secular society? Religious society? How we define common? We have no measurable average - which undermines any attempt to use a perceived median.
That aside - I'm almost tempted to mention the Intentional Fallacy just to see a "Why Monroe Beardsley was nuts" essay. |
|
| |
waugs - believe it or not, an attempt has been made at this sort of anaylis, and one of the authors picked was Connan Doyle (partly because his work sits in a literary canon where works were generally more allegorical, partly because it is from a time where literature wasn't overly influenced by religion, nor was literatry theory itself an influence, and party because its all out of copywrite). The only conclusions (as far as I could see) were that every statistics the computer came up with needed interpretation by someone knowledgable about Connan Doyles's work - i.e. you have to interpret the interpretation. And how wothwhile is that? |
|
| |
The obvius way to test it would be to build the machine, use it to "revive" a living author and then see if we can tell the difference. We could use both fans of the author and people who know the person but have not read his/her books. Start with just a few books by the author and keep adding more until a close match is achieved. |
|
| |
ah, come on, this isn't a bard idea. |
|
| |