form tag

experimenting form tags:

<br />protected double CalcStepSize( double range, double targetSteps ) <br />{ <br /> // Calculate an initial guess at step size <br /> double tempStep = range / targetSteps; <br /> <br /> // Get the magnitude of the step size <br /> double mag = Math.Floor( Math.Log10( tempStep ) ); <br /> double magPow = Math.Pow( (double) 10.0, mag ); <br /> <br /> // Calculate most significant digit of the new step size <br /> double magMsd = ( (int) (tempStep / magPow + .5) ); <br /> <br /> // promote the MSD to either 1, 2, or 5 <br /> if ( magMsd > 5.0 ) <br /> magMsd = 10.0; <br /> else if ( magMsd > 2.0 ) <br /> magMsd = 5.0; <br /> else if ( magMsd > 1.0 ) <br /> magMsd = 2.0; <br /> <br /> return magMsd * magPow;s k y n a r e <br />main <br />linkdump <br />>junkdump <br /> +--config files <br />contact <br />Under constr.... <br /> <br /> <br /> <br />Can we say structure dependant grammar is not built in but derived from general intelligence? Can we exclude “built in” stuff when we discuss preference of structure dependency? Grammar is a generative system which produces “grammatical” outputs and none of ungrammatical ones. Binary distinction between grammatical and ungrammatical sentences leads to a confusion: grammatical and deviant sentences categorize children’s utterances better. Grammar is property of a language not of our mind. Some grammatical sentence in propositional calculus is banned for semantic reason: ex, excluded middle. There are semantically deviant sentences like p^~p though grammatically positive. Primitive truth: truth value of a proposition. <br /> <br />Interpretational schema: truth value production from compound propositions: and semantic are similar. Grammar of prop-calc is parallel to interpretational schema: though many diff ways to define grammatical: prop-calc has intrinsic grammar. Grammar is inductive definition of set of strings of prop-calc for which truth is defined. A^B is true if A and B are both true. This is semantic of A^B. Its grammaticality is defined in induction (II). The semantic fixates grammaticality if A^B, no brain involved. 1. Grammar is inductive definition of set of strings on which semantics are defined. Washoe has classes of words, a frame, new word projection into the frame, usage of conjunction: though she never had innate template of (human) language. Meaning of “A gives X to B” and knowledge that it’s a grammatical sentence are parallel. The intelligence, which makes Washoe know the above are parallel, makes Washoe learn not that she knows the meaning and knows the sentence is grammatical, but she knows that the meaning includes grammaticality. Children try to learn meaning of English and to do that, they must construct grammar..Grammar=semantic-semantic predicate. “Is she t coming?” H1 is possible but to know the meaning, a child should inverse and inversion is so complicated for H1 since H1 doesn’t consider structure. So, H1 drops. LT(O,D): Domain can be small and large. Exceptional musician’s ability is continuous to lay people’s. Piaget says extraordinary GI developed a language and taught others then evolution made our Language area bigger—this part is certainly innate but why evolution did such a thing is the question. Constructivism (Piaget): to have concept of conservation is to know expressions like “same amount” and use symbolism (in full-sense of mastery of conservation). If maturation schedule is innate, constructivism is now nativism. But it is not innate but reflective abstraction is: this can be thought of as usage of language: then Piaget essentially is saying language is innate. We can’t learn a thing unless we innately have something—is not tautology. To learn something, we have to have some thing makes us learn—is tautology but this doesn’t logically necessitates innatness of “something”. Perception is not necessarily has a predicate hard coded. Fodor says brain is like a computer and there’s an innate program which makes the brain learn. And, every candidate of learning should be converted into brain languageànothing new can be learned: for every x, p is true of x if(biconditional) q(x). Associationist, hypothetico-inductive. Cross-inductions: many inductions are weighed and scored credibility. <br /> <br />Putnam believes general intelligence/or multi purpose learning strategy attributes initial state of language faculty. But, he doesn’t explain what general intelligence is. H1, though simpler, needs to have vast amount of quantifiers, categories…but H2 doesn’t need. Chomsky: Grammar(universal grammar) is property of homo sapiens. Putnam: Grammar(of English) is property of English language. Grammar < Language < Brain. Even Putnam says language <br /> <br />Note: Apparently, we learn something we didn’t know. Like our grammar, each of them wasn’t there when we came to the world, but in a few years, we got it working. General notion of ‘learning’ is that there is some kind of learning device/or –s which takes data (learning material) and makes those data into a resultant knowledge. In case of language, interestingly, the data (learning material) isn’t like “No, you place a verb after the subject…” but the data is purely positive, meaning we get to hear many sentences, not instructions of how grammar works. So the question rises, “how do we get grammar out of those sentences?” Sentences are something already compiled by grammar. Since we don’t get the source code (the grammar itself, or description/or instruction of the grammar) but compiled real life results of the grammar, most people can’t help but say that we are pre-disposed to learn grammar (or language). One of those who say pre-disposition (or innateness, species specific) of our language is Chomsky. He gives examples of a Martian scientist. If it (the Martian) is to analyze our English samples, it would make ‘logical’ analysis and say something like H1 (non-structure dependent movement rules). But, Chomsky notes, our human children just don’t have fancy of playing with H1 but they strictly follows H2. Chomsky assumes that’s because there’s a language faculty innate to our brain, specializing in perception and analysis of human language, just like there’s a face recognition device in our brain. Some people, like Putnam, agree that our children should be pre-disposed to perceive the linguistic data in a certain way, but disagree that the pre-disposition isn’t specific to human language but it is more related to our species’ ability to learn, generally. One thing to note is Putnam disagrees with traditional behaviorists in a sense that he doesn’t say tabula rosa is the initial state of our children. Also, Putnam’s learning theory isn’t like that of behaviorists because while behaviorists’ LT is based on conditioning, Putnam’s is more related to our intelligence (Putnam calls his general learning theory). Many already argued that behaviorists’ tabula rosa doesn’t fully describe our learning system. Since a pigeon in the Skinner’s box already has some kind of color perception, memory, spectrum relatedness, induction (ability to add up ‘similar’ cases to yield a generalization)…etc, it’s hard to say blank slate is true. At least, we have to have an ability to be conditioned before we get conditioned. Putnam’s general intelligence isn’t defined well. He says he can explain how our children gets H2 instead of H1 without mentioning built-in-ness of preference to H2, but using general intelligence only. Putnam says grammar is property of our language, not our brain. Putnam: children must use abstract structure, so H2 is natural. Grammar is dependent on learning semantics. (Chomsky: grammar is presupposed before learning semantics). H2 has simpler inverse (Chomsky: inverse algorithm is similar to finding gap game). How can we explain general intelligence? LT(human, maze) < LT(rat, maze). If LT is the LT, it’s hard to define why rats are better at maze but not in language learning domain. In prepositional calculus, AB& is bad and doesn’t have semantics because it has bad syntax. A&B is good syntactically and semantic takes that and gives it a meaning. Since our children are trying to learn meaning of sentences (semantics), not syntax, (because getting what we’re talking is more urgent than to learn what NP and VPs are) and since semantic rules are necessarily structure dependent (false structures don’t get the meaning), it’s possible to say that children selects H2 naturally, without help of hard-coded selection in our brain. Grammar (set of rules and vocabulary) provides definition the notions ‘sentence in a language’. Having similar grammar (speaking in same language) doesn’t mean that we have same brain structure..i.e. lateralization. Innateness gives same genes, same structure to every human species. So, the result should be UG. But we can’t say UG implies we have same innate structures. Even innateness theory doesn’t say we have same structures; it only says that we have similar genes that will unfold in the course of development. And it’s possible to have different unfolded genes (structures) but have same UG. Evolutionary assumption is that in a particular time, in an environment, some members will reproduce more that others. That’ll lead to survival of those who had more massive reproduction. Some say those who are stronger have more reproduction; some say it’s not strong-ness but fit-ness to an environment that leads to survival. Language doesn’t make us strong but more fit to an environment. We can converse more and cooperate more with languages. That’ll make us survive. Not that language makes us super species, but the ability to cooperate with one another (and language certainly do facilitate that) give us more chance of survival. So, our some of ancestors had some genes related to language (in innateness view, for g-people, it’ll be genes related to general intelligence, not a language). Those ancestors cooperated better, and survived. So, we now have those inherited genes that make us cooperative via language. And the question re-circles: is it language or general intelligence that those genes are related to? Chomsky say LT(human, language) != LT(rat, maze): human’s ability to learn a language comes from a different source that gives a rat the ability to learn a maze. Also, Chomsky says LT(human, language) != LT(human, face recognition). This part is hard to digest because it says my ability to learn a language comes from different source that make me learn face recognition skills. I have only one brain. But, when I say that I have one general intelligence, it’s hard to explain if I’m applying H2 stuff when I learn face recognition skill. At the same time, I can say I have one multipurpose device that accommodates well enough in different situations so that when it faces linguistic data, it turns on H2 preference stuff and when it is inputted with facial data via occipital channel, it turns off H2 and turns on some other modules. Some might say mentioning different modules directly supports modular learning theory, but what if I’m saying different modules of a learning device? Not that we have different modules (each designated as a learning device; each can fully perform ‘learning’ in isolation without help of others), but we have different modules in a one giant intelligence device. In that case, LT(human, language) will need some part of LT(human, face recognition). Certainly, we’re not ready to talk whether we have one device with different modules in it or many independent devices housed in a brain. Math skills may not help us to survive. But, it is maybe a mandatory library to install a ‘language’ in our brain. Logic is ‘language’ seems to have selective advantage, but not ‘math’. How can we explain ‘math’ in evolutionary terms? How do we quantify intelligence? We need independent definition of ‘intelligence’; we have to know what we are measuring before we start to measure. May be intelligence can be quantified as complexity of neural network in brain, and ‘complexity’ itself can be quantified as # of nodes each neuron makes…etc. Given we quantified ‘intelligence’ and we measured that human has higher of it than other animals. How can we relate those higher score to H2 preference? Different sentences have two implications: different type of sentences and different sentences. ‘I <3 her’ and ‘I <3 her’ are two different sentences of same type. Can a thought be expressed in every language? Can every thought be expressed in every language? If a though can be expressed in every language, then the other thought from the set should be able to be expressed in every language. So, every thought can be expressed in every language. Problem is in the indefinite article ‘a’. ‘the thought can be expressed in every language’ is different from ‘a thought..’. Culture specific words can’t have simple counter part. Idioms are hard to translate. Art can’t be translated. Effability assumes there’s finite set of thoughts humans can have and every thought can be expressed in every human languages. This means that ‘every thought’ is universal. Given a extraordinary person who has the finite set of thoughts, and effability says if he jots down all of his thoughts, ordinary people can have that thought via their language. That makes semantics universal. But, semantics is not the thought itself. It’s meaning of sentences. Since we are mortal, how can we have infinite thoughts? We can generate infinite # of sentences because our finite set of grammar is recursive. Semantics gives meaning to those infinitely generated sentences. There should be finite semantic rules, just like grammar, to give the infinite material some meaning. If effability holds true, are the rules for the semantic level innate? Heliostrope <br /> <br />} <br /> <br /><img src="http://www.toothpastefordinner.com/010503/any-sort-of-device.gif"></img> <br />

Is that cool?