5 ms·

A common paradigm in the 1990s version of universal grammar (I have not kept up) is the principles and parameters framework. In this framework, the universal g

A common paradigm in the 1990s version of universal grammar (I have not kept up) is the principles and parameters framework. In this framework, the universal grammar consists of a collection of principles (such as "syntactic movement is always sensitive to hierarchical structure") that can be configured via a collection of parameters (traditionally, these were thought of as binary or ternary switches). A summary of this framework is available at http://web.uconn.edu/snyder/papers/CELS.pdf http://web.uconn.edu/snyder/papers/CELS.pdf . The theory is that the principles of universal grammar arise directly from the structure of the brain, while the parameters are set during language acquisition. From the perspective of an adherent of the principles and parameters variety of universal grammar, the examples of cross-linguistic variation you mention are not all that radical. A much stronger challenge to universal grammar is the existence of the Pirahã language, which (it has been argued) cannot express recursion, which would otherwise be a linguistic universal.

jgg 13y ago
I'm a guy who likes abstract stuff, and that definition seems extraordinarily abstract to me, to the point of obfuscating the actual meaning of what we're talking about. Am I right that seems that the definition of grammar there is an almost-meta abstraction of structure? I still can't help but believe that, in practice the 'less radical' points I mentioned, like tense, mood and case (which I think would be switches, right?) arise differently in different contexts, if at all, owing to differences in cognition in various groups, and thus reinforces that mode of cognition in someone who learns the language. The difference in morphology between, say, Ancient Greek and Modern English is pretty drastic, so what gives? In fact, I don't think English has ever been as morphologically-complex as Ancient Greek, and yet some hypothesize that they even descended from the same Indo-European language...if both languages start out from the same point, but evolve with a completely-different grammatical structure that makes different concepts explicit or notable, why would that happen in light of a true universal grammar that is inherent to everyone? It's one thing to say that everyone at a certain intellectual level have a capacity to understand and use certain grammatical constructs, but to claim them as inherent to all people seems sketchy... >The theory is that the principles of universal grammar arise directly from the structure of the brain, while the parameters are set during language acquisition. Wouldn't that prove that language inherently affects cognition, and thus proves linguistic relativism? I don't think you actually made a point in favor of linguistic relativism, but that was the theme of this whole discussion thread.
- stan_rogers 13y ago
  English and the protolanguages preceding it certainly were morphologically complex. English, though, has gone through a couple-three near-pidginizing events along the way that have made it a much simpler, more telegraphic language than it, in most ways, should be. Some aspects of the Celtic languages that were on the ground before English hit Great Britain have worked their way into the grammar (present progressive and do-support). The evidence seems to show that much of the inflective complexity was on its way out after the Vikings made the scene (the languages, Old Norse and Old English, were close enough to allow a degree of mutual comprehension, but differences in case markings, verb endings and the always-arbitrary grammatical gender made a hard-enough task more difficult than it needed to be). And one can hardly say that the influence of Norman French on the language was slight. Remember that the purpose of language is to communicate, so the long-term linguistic homogeneity of the group largely controls how complex the language can become. "City" languages tend to be much simpler (on average) than "forest" languages because outsiders rarely have to learn that 700-speaker language. The surface grammar — what actually comes out of the speaker's mouth — doesn't directly reflect the underlying grammar/syntax, though. The 25-syllable Cree word will tree in pretty much the same way as the 18-word English sentence it replaces. The tonal variations a native English speaker uses convey the same information as the pragmatic particles a Mandarin speaker has to include at the end of every utterance. (We literates are often prone to forgetting that these squiggly marks are to language what choreographic notation is to dance or staff notation is to music.) We might not always be aware of the mechanics of including things like evidential marking, ergativity or pragmatics, but we do it nonetheless. And we've replaced (for the moment, at least) most of our tense and case markings with entire words. The syntax is structurally the same, we just add the word "from" instead of adding the ablative case ending. And really, does anybody actually need grammatical gender? (See Twain's The Awful German language for a reductio ad absurdum.)
  - jgg 13y ago
    Well, I was making the point that, at some point, Ancient Greek was more morphologically complex than any variation of English (which is true to my knowledge). I agree that we can accomplish specific meanings without explicit grammatical structure, but the rabbit hole we were tumbling down was the simple idea that the prominence of a grammatical construct might facilitate a different kind of thinking than in a language that doesn't contain said construct. Grammatical gender is a rarely-useful concept, I agree.

jgg 13y ago

I'm a guy who likes abstract stuff, and that definition seems extraordinarily abstract to me, to the point of obfuscating the actual meaning of what we're talking about. Am I right that seems that the definition of grammar there is an almost-meta abstraction of structure? I still can't help but believe that, in practice the 'less radical' points I mentioned, like tense, mood and case (which I think would be switches, right?) arise differently in different contexts, if at all, owing to differences in cognition in various groups, and thus reinforces that mode of cognition in someone who learns the language. The difference in morphology between, say, Ancient Greek and Modern English is pretty drastic, so what gives? In fact, I don't think English has ever been as morphologically-complex as Ancient Greek, and yet some hypothesize that they even descended from the same Indo-European language...if both languages start out from the same point, but evolve with a completely-different grammatical structure that makes different concepts explicit or notable, why would that happen in light of a true universal grammar that is inherent to everyone? It's one thing to say that everyone at a certain intellectual level have a capacity to understand and use certain grammatical constructs, but to claim them as inherent to all people seems sketchy... >The theory is that the principles of universal grammar arise directly from the structure of the brain, while the parameters are set during language acquisition. Wouldn't that prove that language inherently affects cognition, and thus proves linguistic relativism? I don't think you actually made a point in favor of linguistic relativism, but that was the theme of this whole discussion thread.

stan_rogers 13y ago
English and the protolanguages preceding it certainly were morphologically complex. English, though, has gone through a couple-three near-pidginizing events along the way that have made it a much simpler, more telegraphic language than it, in most ways, should be. Some aspects of the Celtic languages that were on the ground before English hit Great Britain have worked their way into the grammar (present progressive and do-support). The evidence seems to show that much of the inflective complexity was on its way out after the Vikings made the scene (the languages, Old Norse and Old English, were close enough to allow a degree of mutual comprehension, but differences in case markings, verb endings and the always-arbitrary grammatical gender made a hard-enough task more difficult than it needed to be). And one can hardly say that the influence of Norman French on the language was slight. Remember that the purpose of language is to communicate, so the long-term linguistic homogeneity of the group largely controls how complex the language can become. "City" languages tend to be much simpler (on average) than "forest" languages because outsiders rarely have to learn that 700-speaker language. The surface grammar — what actually comes out of the speaker's mouth — doesn't directly reflect the underlying grammar/syntax, though. The 25-syllable Cree word will tree in pretty much the same way as the 18-word English sentence it replaces. The tonal variations a native English speaker uses convey the same information as the pragmatic particles a Mandarin speaker has to include at the end of every utterance. (We literates are often prone to forgetting that these squiggly marks are to language what choreographic notation is to dance or staff notation is to music.) We might not always be aware of the mechanics of including things like evidential marking, ergativity or pragmatics, but we do it nonetheless. And we've replaced (for the moment, at least) most of our tense and case markings with entire words. The syntax is structurally the same, we just add the word "from" instead of adding the ablative case ending. And really, does anybody actually need grammatical gender? (See Twain's The Awful German language for a reductio ad absurdum.)
- jgg 13y ago
  Well, I was making the point that, at some point, Ancient Greek was more morphologically complex than any variation of English (which is true to my knowledge). I agree that we can accomplish specific meanings without explicit grammatical structure, but the rabbit hole we were tumbling down was the simple idea that the prominence of a grammatical construct might facilitate a different kind of thinking than in a language that doesn't contain said construct. Grammatical gender is a rarely-useful concept, I agree.