00:01:42 <ehird> i wanna write some haskell
00:01:49 <psygnisfive> i hope they put some sort of easter egg in this program
00:02:26 <psygnisfive> the markov chains for "ring ring ring ring ring ring ring banana phone" are disproportionately strong
00:02:46 <ehird> also I don't have anything to write
00:03:26 <psygnisfive> write a subsumption architecture framework.
00:05:50 <ehird> psygnisfive: THINK OF SOMETHING
00:06:32 <kerlo> ehird: why don't you want to write an AI?
00:06:46 <ehird> kerlo: too hard. too boring. too many Friendly concerns.
00:07:07 <ehird> psygnisfive: I'm not particularly inclined to die.
00:07:20 <kerlo> Yeah, I agree with ehird on that point.
00:07:57 <kerlo> Yeah, well, keeping it from doing... anything = friendly AI.
00:08:28 <psygnisfive> friendliness comes from life experience, as does evilness
00:08:38 <ehird> extending Dasher to a whole UI would be fun
00:08:50 <ehird> more common tasks are closer
00:09:04 <ehird> you could move in any direction (for more freedom)
00:09:18 <kerlo> In humans, sure. In AI, no; an AI's main goals do not change except by mistake.
00:09:29 <ehird> kerlo: that's a narrow definition of AI
00:09:36 <ehird> simulating a human = AI = goal changing.
00:10:07 <psygnisfive> write the ZUI combined with a kit for everything-is-a-service application-less type stuff!
00:10:16 <kerlo> ehird: generally, "I want to accomplish X, so I should forget about X and instead accomplish Y instead" is not good reasoning.
00:10:28 <ehird> kerlo: I don't want to write an AI.
00:10:44 <ehird> Oh, you mean in the AI.
00:10:46 <ehird> That doesn't matter.
00:10:48 <ehird> simulating a human
00:10:52 <ehird> and that human would change goals
00:10:54 <ehird> psygnisfive: too hard.
00:11:08 <kerlo> "I want to accomplish X, so I should accomplish Y in a way that assists with X", on the other hand, is plenty good, and it doesn't change the main goals.
00:11:30 <ehird> kerlo: whether it's logical or not:
00:11:37 <ehird> and you could simulate a human
00:11:39 <ehird> and that would be AI
00:11:41 <kerlo> Anyway, if AI means simulating a human, what do you call artificial things that act intelligently?
00:11:46 <ehird> so saying that AIs never change goals is wrong.
00:12:07 <ehird> a simulated human is artificial, and intelligent.
00:12:30 <ehird> psygnisfive: yes, yes, har, har.
00:12:34 <kerlo> That doesn't mean that a simulated human is the easiest, simplest or most typical example of an artificial and intelligent thing.
00:12:52 <ehird> kerlo: that's irrelevant
00:12:58 <ehird> you said that [all] ai does not change goals
00:13:48 <kerlo> Okay. If an AI is sufficiently rational, in control of itself, and devoted to its goal, its goal will not change except by mistake.
00:14:08 <ehird> congrats, what you just said is totally unrelated to what we were saying!
00:14:59 <psygnisfive> kerlo's goal of expressing this point is not changing
00:15:03 <kerlo> My point in saying that was to note that what I said earlier is... not very true.
00:15:15 <ehird> psygnisfive: that explains it!
00:16:01 <psygnisfive> if Kerlo wrote chat bots, i bet he'd code them in organic systems. because for him, since hes an AI, the physical world is the program. :o
00:16:30 <ehird> kerlo: so if you're an AI, why don't you singularitize?
00:16:45 <psygnisfive> ehird: you could try coding some sort of cell-like agent bot system thing
00:17:02 <psygnisfive> and experiment with self organizing complex systems
00:17:06 <kerlo> I'm only partially self-modifying.
00:17:09 <psygnisfive> the code itself should be relatively simple actually
00:17:33 <psygnisfive> he can only modify things that arent part of his goal
00:17:36 <kerlo> Essentially, my self-modification ability is limited to the ability to think about things because I want to.
00:18:12 <ehird> kerlo: who made you
00:20:00 <Gracenotes> fly requiremendous sounds were is a large as the past our past light continued in the could on the other than the royousuck you suck you suck you suck you suck you suck you suck you suck
00:23:49 -!- M0ny has quit ("PEW PEW").
00:24:15 <ehird> i love how there's a guy called Havoc Pennington
00:24:16 <ehird> who works on linux
00:24:19 <ehird> (he didn't change his name)
00:26:00 <kerlo> ehird: the future offspring of Eliezer Yudkowsky, Marcus Hutter, and Douglas Adams.
00:26:28 <kerlo> Which will be possible in the future due to more lenient laws regarding marriage. And time travel, of course.
00:26:40 <ehird> kerlo: More lenient biology, I'd assume, also...
00:27:02 <kerlo> The Constitution was amended to give it jurisdiction over biology, so yes.
00:27:15 <kerlo> The Constitution of the United States of the Entire World, that is.
00:27:25 <ehird> kerlo: that the singularity's doing?
00:28:21 <kerlo> Yeah, the Singularity hasn't happened yet. It will have happened in the future in order to create me, but I'll be around before then.
00:28:45 <kerlo> All this is possible because of quantum mechanics.
00:29:01 <ehird> kerlo: I either hate you or my brain; I'm too confused to work out which
00:29:27 <ehird> as I said, I don't know.
00:30:10 <oklopol> why would ehird understand nonsense
00:30:31 <kerlo> The future doesn't make any sense!
00:31:17 <oklopol> time travel is kinda like differentials, the first thing you automatically do when starting to think about them is prove they make no sense, end of story
00:31:23 <psygnisfive> the singularity is the point at which our models break down
00:31:51 <kerlo> Anyway, I will now answer interview questions from all of you.
00:32:09 * kerlo watches people form a remarkably dispersed line.
00:32:15 <oklopol> The Instruction Set – a Critical Interface
00:32:24 <psygnisfive> HOW DO YOU GO TO THE BATHROOM IN 6 DIMENSIONS AND SPACE AND TWO DIMENSIONS OF TIME?!
00:32:41 <kerlo> oklopol: stuff is going pretty well.
00:32:44 <oklopol> yeah and how do you aim with a hundred penises?
00:32:57 <ehird> i've always wanted to walk in time.
00:33:07 <ehird> it'd be confusing as FUCK
00:33:20 <kerlo> psygnisfive: just concatenate your source code with someone else's, wait a while, square your "happiness" variable, and split up again.
00:33:38 <ehird> actually if you say that time is like the other 3 dimensions and everything's always going forwards in it, then going back in time will put you and your time machine in a void deplete of matter
00:33:43 <kerlo> psygnisfive: we don't, silly; we use simulators for that.
00:34:09 <ehird> psygnisfive: what's not confusing
00:34:12 <kerlo> oklopol: well, our brains are easily a hundred times as powerful as pre-Singularity humans'.
00:34:28 <ehird> if everything's going ... vinn in Time, then if you go vout to go to the past, there's no other matter there with you
00:34:33 <ehird> same with vinn to go forwards in time
00:34:52 <psygnisfive> ehird: this feels like the Langoliers theory of time
00:35:00 <ehird> 00:33 kerlo: psygnisfive: just concatenate your source code with someone else's, wait a while, square your "happiness" variable, and split up again.
00:35:35 <kerlo> ehird: how happy are you right now, as a number?
00:36:28 <kerlo> Tell me being--wait.
00:36:33 <oklopol> i can choose any happiness value.
00:36:51 <kerlo> psygnisfive: love is an intense feeling of affection and care towards another person.
00:37:32 <ehird> Had to fall over there.
00:38:21 <psygnisfive> lets build a subsumption architecture library.
00:39:59 -!- nooga has quit (Read error: 104 (Connection reset by peer)).
00:40:56 <oklopol> you do realize if you keep asking me to program something with you, i will do it at some point.
00:41:17 <oklopol> ...that was a threat, mister.
00:41:38 <oklopol> i'm kinda busy with university atm.
00:41:44 <ehird> oklopol: you cannot escape his logical snare!
00:42:45 <oklopol> can't say i'd mind, working over irc.
00:43:17 <oklopol> for all i know you could be fellating right now
00:44:43 <oklopol> i was thinking you know fellating nothing in particular.
00:45:07 <oklopol> you know just fellating away.
00:45:49 <psygnisfive> i could sit in the corner of your room, just fellating the air, and you could use me as a fellatio machine whenever you were in the mood!
00:45:55 <ehird> 00:43 oklopol: for all i know you could be fellating right now
00:45:55 <ehird> 00:43 psygnisfive: :o
00:46:19 <oklopol> psygnisfive: how would you get in my room?
00:46:29 <oklopol> i don't even let my closest friends in here really.
00:47:21 <psygnisfive> oklopol: presumably i'd walk through your door.
00:48:20 * Sgeo judges em instead of me
00:49:31 <oklopol> finnish uses one of its rarer cases, the instructive, for "on foot"
00:50:04 <psygnisfive> english does weird things with number in certain places
00:52:42 <ehird> http://canonical.org/~kragen/sw/js-calc.html
00:52:50 <ehird> this is a really nice programming language/calculator UI
00:55:04 <ehird> psygnisfive: you should try i
00:55:10 <ehird> For a good time, try "40", Enter, Enter, "i", Tab, "/1a4*4**s2^".
00:55:16 <ehird> shows the ui is really nice
00:56:17 <psygnisfive> now you said i should do this in firefox yes?
00:56:34 <ehird> load it and you have a text field
00:56:37 <ehird> look at the right hand for instructions
00:56:42 <ehird> but I recommend trying that key combo first
00:56:49 <ehird> it's a stack-based graphing calculator thingy with a nice UI
00:57:30 <ehird> look at the graph next to it
00:57:32 <ehird> also, do it step by step
00:57:37 <ehird> you see each step of the calculation being done
00:57:38 <oklopol> yes look beneath the surface
00:57:50 <oklopol> you have to *feel* the computation
00:58:03 <ehird> what do you mean it's nonsense
00:58:15 <psygnisfive> 40 enter enter prouces two lines saying 40 = 40
00:58:29 <ehird> then you type /1a4*4**s2^
00:58:54 <ehird> / divides each element of the seq by 40
00:59:00 <ehird> a calculates the arctangent
00:59:04 <ehird> 4* times it by four
00:59:19 <ehird> tht's your problem
00:59:24 <ehird> psygnisfive: do you have the input field focused?
00:59:28 <ehird> and this is in FF?
00:59:42 <ehird> then * timses the sequence by it
00:59:45 <ehird> then s calculates the sine
00:59:51 <ehird> then 2^ squares it
00:59:58 <ehird> and you get a sequence with a pretty graph next to it
01:00:20 <kerlo> I think the mathematical "times" is a preposition that was originally a noun.
01:01:13 <kerlo> Well, the "times" in "5 times" is a noun, isn't it?
01:01:45 <psygnisfive> yes. i suppose it actually wasnt a postposition at all
01:01:53 <kerlo> But then we have usages like "multiply it times four" that demonstrate its prepositional usage.
01:01:54 <psygnisfive> since im thinking weird math linguistics now
01:02:31 <kerlo> "Times" is a verb there?
01:03:10 <kerlo> I'm not seeing it.
01:03:27 <kerlo> In "times it by four", it's certainly a verb.
01:03:58 <ehird> i love how that calculator keeps track of the formula
01:03:58 <ehird> (((((-1), 1), (-1.5)), 3) * (5 ^ (iota 4))) + reduce
01:04:01 <psygnisfive> it could be prepositional tho. we'd need to do more research and find more data, to be honest.
01:04:17 <kerlo> But hey. "Multiply it how?" "Times four." Put it before a noun, and you get an adverb. That sounds like a preposition to me.
01:04:51 <kerlo> Maybe our Englishes are different.
01:04:51 <psygnisfive> listen kerlo, im a linguist. dont argue with me.
01:04:56 <ehird> mine isn't equivalent
01:05:15 <kerlo> Are you a linguist?
01:06:03 <psygnisfive> no i only speak english, thats a common misconception
01:06:14 <kerlo> I'm aware that "Englishes" is not the most proper thing to say.
01:06:32 <psygnisfive> <psygnisfive> no i wont correct your grammar dont worry
01:08:07 <kerlo> I'll ask *my* linguistics student whether "times" is a preposition. :-P
01:08:47 <psygnisfive> also, "your" student will have no better answer that i do. like i said, we really need more data
01:09:22 <kerlo> I'm using the possessive rather loosely.
01:09:41 <kerlo> Alternatively, a helpful response:
01:10:04 <psygnisfive> kerlo, let me show you something interesting about english
01:10:13 <psygnisfive> "a friend of mine" contains a double possessive :D
01:10:44 <psygnisfive> well, a possessive pronoun and then a possessive preposition
01:11:09 <psygnisfive> infact, its all over the place, and actually has interesting semantic differences
01:11:15 <kerlo> I kind of expected you to sarcastically say "sentences have to have verbs in them in order to make any sense".
01:11:17 <psygnisfive> a student of chomsky (someone who studies chomsky's work)
01:11:33 <psygnisfive> a student of chomsky's (someone who chomsky teaches)
01:12:04 <kerlo> I would have contradicted you in an equally sarcastic manner.
01:13:13 <kerlo> Something like "Really? New to me; never really a big fan of complete sentences."
01:13:44 <psygnisfive> that last sentence is a slight bit baffling, tho i know what you intend to convey. :P
01:14:31 -!- FireFly has quit ("Later").
01:14:32 <psygnisfive> english "be" in that sentence is semantically vacuous anyway.
01:14:38 <kerlo> Anyway, I suppose things like "a friend of mine" exist to fill the possessive-as-an-adjective hole in English where "a friend of me" is inadequate for some reason.
01:14:55 <psygnisfive> actually i dont think thats what the difference is at all
01:14:55 <kerlo> Yeah, that sentence doesn't tell you whether it's present or past.
01:15:38 <psygnisfive> "student of chomsky" sentence behaves very much like a binary predicate like this: studies(x,chomsky)
01:16:02 <psygnisfive> i think "student of chomsky's" uses the same predicate, but uses it differently:
01:16:46 <kerlo> I think that "student of X", where X is a noun, generally means one that studies X, which blocks the "Chomsky's student" meaning of "student of Chomsky".
01:16:47 <psygnisfive> studies(x,y) & x in { s | s is associated with chomsky }
01:17:00 -!- BeholdMyGlory has quit (Remote closed the connection).
01:17:18 <psygnisfive> kerlo, i dont think "X's student" can have the "student of X" reading at all.
01:17:36 <kerlo> The latter half of that conjunction is just a long way of saying (x is associated with chomsky), isn't it?
01:17:48 <psygnisfive> I think that "X's student" and "student of X's" are merely alternative forms of the same thing
01:18:11 <kerlo> psygnisfive: I never said it could have that meaning; I said that that meaning was blocked.
01:18:12 <psygnisfive> its a long winded way of saying that x is a member of those things that are associated with chomsky.
01:18:28 <psygnisfive> kerlo: blocking implies that it COULD have that meaning tho
01:18:55 <psygnisfive> plus, blocking means something specific in linguistics.
01:19:09 <psygnisfive> especially in that sense, where one reading blocks another.
01:19:14 -!- GregorR has quit ("Leaving").
01:20:14 <ehird> "our consciousness is like a Lisp dialect, interpreted by Prolog on a Java VM running in a VMware instance simulated by Erlang on an array of FPGA’s emulating Apple ][’s within the Quantum Computer we call the universe"
01:20:15 <oklopol> <psygnisfive> english "be" in that sentence is semantically vacuous anyway. <<< how?
01:20:24 <kerlo> "I am a member of those things whose name is Bob. You are a member of those things who are doing how?" "I'm a member of those things who are doing fine. Hey, be a member of those things who look! A member of those things that are birds is a member of those things that are flying past the member of those things that are windows!"
01:20:47 <ehird> kerlo: KEEP TALKING LIKE THAT SO I CAN HATE YOU
01:20:50 <kerlo> oklopol: "be" just lets you use an adjective or a noun as a verb, really.
01:21:34 <psygnisfive> "red hat" conveys something like red(x) & hat(x)
01:22:38 <psygnisfive> but the primary sentential structures in english convey different relations
01:23:03 <psygnisfive> "john runs" conveys something more like "running(e) & agent(e,John)"
01:23:06 <oklopol> "I was never really a big fan of complete sentences." has a vacuous be?
01:23:14 <psygnisfive> the verb describes the type of event, and so forth.
01:23:49 <psygnisfive> the vacuous "be" is a dummy verb that essentially acts to connect its complement, the adjective or noun, to the thing that the adjective or noun predicates on
01:24:19 <kerlo> oklopol: "I never really liked complete sentences", if "to like" were the same as "to be a fan".
01:24:45 <psygnisfive> ignoring tense, "never", and "really", it means something like
01:24:56 <oklopol> kerlo: thank you i understand english; now translate the linguisticsary psygnisfive said there.
01:26:03 <psygnisfive> past(rare(actual(a_big_fan_of_complete_sentences(Me))))
01:26:25 <psygnisfive> the a_big_fan_of_complete_sentences is a complex predicate, ofcourse. perhaps some lambda like
01:26:44 <kerlo> Look up the word "copula" on Wikipedia, maybe. :-P
01:27:25 <kerlo> I kind of like the idea of copulas carrying semantic meaning. Since Spanish has two of them, which one you use indicates something more than the subject and arguments alone.
01:27:29 <oklopol> psygnisfive: so vacuous as in it isn't really a proper verb because it isn't an event but just structure?
01:27:56 <psygnisfive> but its not the kind of semantics thats relevant to what we were talking to.
01:28:10 <kerlo> I'd like a language with, say, five different copulas.
01:28:28 <psygnisfive> oklopol: vacuous in that aside from the generic properties that verbs have, like conveying tense and person agreement, the empty copula contributes nothing else in english.
01:28:49 <psygnisfive> in spanish it conveys the same as english, plus semantics about the essentialness of the predicate
01:29:15 <kerlo> Lojban has about ten different types of infinitives, which is fun. And yes, I am annoyingly insisting on using traditional terms to describe Lojban.
01:29:23 <psygnisfive> im designing a language that has a number of copulas. or copula-like items.
01:29:28 <oklopol> psygnisfive: so just structure? and by that i meant it serves only a syntactic purpose
01:29:46 <psygnisfive> however, you could reconstrue it differently
01:29:52 <oklopol> yay i got a partial success!
01:30:11 <psygnisfive> if you dont mind me lapsing into formalism here
01:30:53 <psygnisfive> actually no, real formalism, using CCG. :D
01:31:46 <psygnisfive> thats the core of the content of vacuous be.
01:32:54 <kerlo> "lo fagri" means "a fire", "lo nu fagri" means "for there to be a fire", and there are things like "what it's like to be a fire", "there indeed being a fire", and a generic one that's simply "to be a fire" in a generic way.
01:32:56 <oklopol> so basically it's an identity combinator
01:33:49 <oklopol> psygnisfive: so i-combinator.
01:34:01 <kerlo> And as any Haskell programmer knows, an application combinator and an identity combinator are the same thing. :-P
01:34:10 <psygnisfive> i mean, the issue is really that some predicative items arent of the right syntactic category for use in a sentence
01:34:24 <kerlo> Slereah_: figure out how to make Reader an arrow so that definitions of the class Arrow can be used as sets of combinators.
01:34:39 <psygnisfive> so BEx takes x of some incorrect type to the correct type
01:35:08 -!- GregorR has joined.
01:35:39 <oklopol> psygnisfive: yep, alternatively me.
01:36:01 <psygnisfive> "be" could be (S\NP)/NP (S\NP)/AdjP with semantics \p.\x.state(s) & attributed_property(s,p) & possessor_of_property(s,x)
01:36:15 <psygnisfive> so that it really does convey some amount of semantics
01:36:24 <kerlo> So, irony, anyone? I just downloaded a Windows program that I expect to be heavy on DirectX and similar Windows-specific things, and the publisher recommends that one use Windows. So far, the installation has gone better under Linux.
01:36:30 <psygnisfive> namely, that the thing the sentence talks about is a state of affairs, not an event
01:36:47 <kerlo> However, the hard drive is clicking away, which suggests that something will happen any minute now...
01:36:57 -!- Slereah_ has set topic: /prog/ except COOL FREE RINGTONES | http://tunes.org/~nef/logs/esoteric/?C=N;O=D.
01:37:14 <psygnisfive> but yes, oklopol, one way of viewing vacous BE is as an identity combinator
01:37:25 <psygnisfive> only it does type alternation in the process.
01:37:49 <psygnisfive> i prefer the non-vacuous interpretation, to be honest.
01:41:08 <psygnisfive> the issue is partly that english (and perhaps most or all languages) has a relatively large number of types that are arbitrarily assigned to words.
01:41:19 <oklopol> okay i really need to sleep now, cuz reason
01:41:26 <psygnisfive> it might not be entirely arbitrary, but we dont have a full understanding of that, so.
01:42:19 <oklopol> just checking, since open issues are quite intriguing
01:42:49 <kerlo> I'm waiting for nouns and adjectives to become the same thing.
01:42:52 <psygnisfive> there actually does seem to be some genuine difference between the conceptual content of nouns versus verbs
01:42:57 <kerlo> I'll happen if we stop using the plural form of our noun.
01:43:23 <psygnisfive> but loosing plural wont be enough to cause it, kerlo.
01:43:36 <kerlo> It seems like the only difference between noun and adjective is that noun can become plural while adjective can't.
01:43:40 <psygnisfive> some languages dont distinguish nouns and adjectives
01:43:46 <psygnisfive> others dont distinguish verbs and adjectives
01:43:58 <kerlo> I mean, the plural form of adjective are the same as the singular form.
01:44:46 <oklopol> <kerlo> I'll happen if we stop using the plural form of our noun. <<< do you happen often?
01:44:47 <kerlo> But noun can be used as adjective, more or less. Put a noun before a noun, and it modifies it as an adjective would.
01:44:49 <psygnisfive> but the corest of the differences is is simply that adjectives are adjectives and nouns are nouns
01:45:07 <psygnisfive> kerlo: nominal modification is not the same thing tho.
01:45:31 <psygnisfive> adjectives are actually a lot more complex than you'd think too
01:45:33 <kerlo> But it looks the same.
01:45:50 <psygnisfive> it only "looks the same" in that both come before a noun.
01:45:59 <kerlo> In "a brick house", I can't even tell whether "brick" is being a noun or an adjective.
01:46:10 <oklopol> stop bikeshedding, this is the real issue.
01:46:36 <kerlo> It means that someone might say "a brick house", intending for "bring" to be a noun, and someone else might look at that and say, "oh, 'brick' is an adjective".
01:46:41 <psygnisfive> i mean, you could interpret it as a complex adjective, this is true
01:46:59 <psygnisfive> you might say that brick is a noun with a 0 derivational morpheme that turns it into an adjective
01:47:34 <psygnisfive> but this isnt going to make adjectives and nouns merge in english
01:47:44 <psygnisfive> atleast not without a LOT of further erosion of the differences
01:47:49 <psygnisfive> because there are still a LOT of differences.
02:01:01 <psygnisfive> http://www.youtube.com/watch?v=p-hG6YJQ0YI
02:03:36 <Slereah_> psygnisfive, you are terrible people
02:06:57 <psygnisfive> the name shakespeare has historically been written no fewer than fifty five different ways
02:07:07 <psygnisfive> before the normalization of english spelling
02:10:17 <kerlo> My favorite spelling:
02:11:44 <kerlo> S-with-circumflex e-with-acute k s p e-with-macron r.
02:12:22 <psygnisfive> but that looks like a phonetic alphabet being used
02:12:23 <kerlo> Pronounce the s-with-circumflex as in Esperanto, the e-with-acute as in French, and the e-with-macron as in Wiktionary's enPR.
02:12:39 <kerlo> Or as in any American dictionary's pronunciation key.
02:12:56 <kerlo> IPA is superior if there's only one way to pronounce each word.
02:13:22 <kerlo> Unfortunately, words like "superior", "there's", and "word" have multiple pronunciations.
02:13:39 <kerlo> And you're American, I believe, so the British are wrong.
02:13:51 <kerlo> You'll die for this!
02:13:58 * kerlo stabs psygnisfive with a Speare
02:15:32 <psygnisfive> someone should make a tee-shirt with a picture of the panchen lama sort of like this: http://en.wikipedia.org/wiki/File:Khedrup_Je.jpg
02:15:51 <psygnisfive> only where he's sitting in the position of a comma in some phrase
02:16:20 <psygnisfive> or Dalai Comma, if noone knows who the Panchen Lama is
02:17:07 <kerlo> Anyway, esoteric syntax idea: the program is composed of sentences, which are composed of words, and each word has an affix indicating its case and an affix indicating whether it's indicative or infinitive.
02:17:29 <psygnisfive> kerlo: you're describing like half the words languages there.
02:18:02 <kerlo> psygnisfive: can you say that again using either more punctuation or fewer typos?
02:18:45 <kerlo> Yeah, human languages tend to act like that.
02:18:54 <kerlo> Which, clearly, means we should make a computer language that acts like that.
02:19:27 <kerlo> like-head-indicative i-subject-indicative eat-object-infinitive taco-object-indicative
02:20:09 <psygnisfive> also, whats this head thing? and object on eat? what?
02:20:33 <kerlo> I like to eat tacos.
02:20:59 <kerlo> "Like" is the head of the sentence, "I" is the subject of "like", "to eat" is the object of "like", "taco" is the object of "to eat".
02:21:33 <kerlo> Then I reject your linguistics and substitute my own.
02:21:39 <psygnisfive> also, head? head-ness isnt something that is marked.
02:21:54 <psygnisfive> if you want to turn it into some sort of predicate system
02:22:29 <psygnisfive> but the thing that you like is not _just_ eating.
02:23:22 <kerlo> That's why there's taco-object-indivative there. It makes it "eating tacos" instead of just "eating".
02:23:40 <psygnisfive> but i suppose in some sense you might MARK "eat"
02:24:04 <psygnisfive> assuming "eat" is the head of subclause (its not), you could mark the head of the clause for this sort of thing.
02:24:43 <psygnisfive> eat should get indicative, and taco should get infinitive
02:24:47 <kerlo> Is a clause the same thing as a phrase?
02:25:05 <psygnisfive> on the grounds that the marking indicates the "mood" of the clause its in.
02:25:18 <kerlo> In English, in "I like to eat tacos", "to eat" is in the infinitive, is it not?
02:25:56 <psygnisfive> the clause is infinitive, yes. or you might say the verb is. but were not talking about the form of the verb, we're talking about the form of the marker that gets affixed to the words.
02:26:27 <psygnisfive> if X is the head of an argument of an indicative clause, X concordializes by showing "IND"
02:26:47 <kerlo> Whether this system is like human languages or not, and whether I'm using terminology correctly or not, it's a system.
02:27:06 <psygnisfive> your system is poorly defined, and potentially inconsistent. :P
02:28:13 <kerlo> It's poorly defined in that I haven't told you much.
02:28:31 -!- MouD has quit (Read error: 110 (Connection timed out)).
02:29:29 <kerlo> Okay, let's say that every predicate receives an M tag, which is either D or F, and a C tag, which can be any of many things.
02:30:28 <kerlo> The main predicate of the sentence is the one whose C tag is H. Every other predicate is an argument; the predicate that it's an argument to is the nearest predicate that can accept an argument of its C tag.
02:31:51 <kerlo> Finally, D and F change the meaning of the predicate: a D predicate acts normally, while an F predicate essentially states that its subject is the predicate.
02:34:57 <kerlo> So, for example, in the sentence "like-H-D i-S-D eat-O-F taco-O-D", the main predicate is "like", "i" is an argument to the nearest predicate that can accept an S (which is "like" for some unspecified reason), "eat" is an argument to the nearest predicate that can accept an O (which is definitely "like"), and "taco" is an argument to the nearest predicate that can accept an O (which is now "eat").
02:36:26 <psygnisfive> Subject and Object are not properties of words, you know. also, EAT can accept subjects. as in, "i eat taco"
02:37:02 <kerlo> Indeed; that's why I said "for some unspecified reason".
02:37:26 <psygnisfive> it could be order specific. e.g., "nearest preceeding"
02:37:51 <psygnisfive> honestly tho, i dont see the point of making the subject/object distinction
02:38:23 <psygnisfive> theres no such distinction in natural languages, in terms of argument structure, only in terms of some vague, poorly specified way that noone really has defined, let alone provide clear evidence for.
02:38:51 <kerlo> What do you mean by "in terms of argument structure"?
02:39:03 <psygnisfive> structure of the arguments of the verbal predicate
02:39:22 <psygnisfive> in simple terms, something like: like(liker, liked), eat(eater, eaten), etc.
02:39:48 <kerlo> I still don't see what you mean.
02:40:29 <kerlo> In "I love her", "I" is the subject and "her" is the object, and you can tell both by the positioning of the words and the words themselves.
02:40:34 <psygnisfive> we actually think that arguments to predicates can be one of any number of kinds of things. doers, doees, goals, sources, etc.
02:40:56 <psygnisfive> yes, kerlo, but theres a serious flaw in that argument
02:41:14 <psygnisfive> "subject" and "object" convey, by your definition, nothing but positional notions.
02:41:45 <psygnisfive> secondly, this isnt always true. sometimes the subject follows the verb in english, and there are a number of cases in italian where it follows.
02:42:14 <kerlo> When you come back, tell me what I said that is incorrect, if anything.
02:52:39 -!- bsmntbombdood_ has joined.
02:53:17 -!- bsmntbombdood_ has quit (Client Quit).
03:09:30 <psygnisfive> its that the notions of subject and object are essentially meaningless
03:09:46 <psygnisfive> subject might be construable as the thing being talked about, but even then its shakey
03:16:12 <psygnisfive> whats more, the position of the "subject" isnt always the same in english
03:17:04 -!- Ilari_ has joined.
03:17:16 <psygnisfive> in that the rule for that language isnt universal.
03:17:29 <psygnisfive> in italian, for instance, some "subjects" look like objects, or vice versa
03:17:30 -!- mental has joined.
03:19:04 -!- lament has quit (Nick collision from services.).
03:19:08 -!- mental has changed nick to lament.