Talk:Conlang/Advanced/Grammar/Government

What to do with Intermediate syntax

 * I've moved this thread here from Talk:Conlang. &mdash; Pi zero (talk) 13:30, 15 December 2008 (UTC)

Back in February, the fairly short Intermediate/Syntax section was mostly replaced with extensive materials including multiple subfiles (by a user Psygnisfive (Talk | contribs)). I have just altered the table of contents to bring this material out where we can see it. There seem to me to be several difficulties that need to be addressed. I'm not yet ready to suggest a strategy for addressing these problems &mdash; does anyone else have some thoughts on how to proceed? Putting the new files into the table of contents, so the new stuff can be seen in context, seemed like a reasonable first step. Myself, I'm going to meditate on it for a while, and perhaps I'll come up with a proposal. Pi zero (talk) 21:21, 25 August 2008 (UTC)
 * The material has a different style and tone than the other existing content of the wikibook.
 * The material goes into a level of technical detail that seems out of place where it is. Part of this is probably that some of it is Advanced rather than Intermediate, part of it goes back to the style/tone, and... I'm not entirely sure if part of it is something else.
 * There's a section on how to apply the material to conlanging, rather than each section being continuously grounded in the conlanging theme.


 * I agree. On the one hand, it's very useful information and something that any conlanger should be aware of. But on the other hand, it's not really appropriate for the book; it's more linguistic than conlinguistic, if you know what I mean. It's definitely more Expert material than Intermediate. I suggest we move it there and try to rewrite in a more accessible tone; in particular I think that the Applying Knowledge section should be integrated into each of the other chapters rather than being the after-thought that it is. Ingolemo (talk) 15:48, 16 December 2008 (UTC)


 * I'm not opposed to this suggestion, and certainly it would be good for the Intermediate section to move this out to make room for more level-appropriate content. I don't have a concept of how to square it with my proposed delineation of a mission for the Advanced section (on the main page or the Advanced introduction); and it seems likely to swamp the existing Advanced content, if not the anticipated Advanced content as well.  Admittedly, I have had in mind that the Advanced section ought to cover universals.  Pi zero (talk) 19:53, 19 December 2008 (UTC)


 * Yeah, it obviously needs to be trimmed. It could probably go into the Advanced section under the guise of a chapter (or possibly subsection) on formal systems of analysis ("Phrases and Trees"); from there we could extract and/or prune all the content that isn't particularly relevant to that topic ("Linguistic Universals" would become a separate chapter independent from this, and "Parts of Speech" could largely be retained in the intermediate syntax section, though probably without the emphasis on phrases at the bottom)
 * Or, if that's too much work, we could simply try to pawn it off to the Linguistics Wikibook or something :) Ingolemo (talk) 00:36, 21 December 2008 (UTC)


 * I'd like (if you've no objection to the delay) to try my hand at drafting a new overall outline for the Advanced section, into which this material could then be integrated with adjustments as you describe. This could take a few days, though honestly a week or three seems more likely; I'm not really set up for fast action on wikibooks, which is one of the reasons I've adopted a strategy, over the past six months or so, of plodding slowly along, shoring up and reinvigorating the overall structure of the book by inches.  It somehow hadn't registered on me that the outline of the Advanced section would (of course) need upgrading in parallel with upgrades to the Intermediate section.  Pi zero (talk) 17:46, 30 December 2008 (UTC)


 * Okay, I'm ready to admit that it's unlikely I would get to this for a long while yet. Pi zero (talk) 03:51, 7 January 2009 (UTC)

As an intermediate conlanger who stumbled across this book, I was completely lost in the technical explanation. The verbiage is not clear. It's talking about theory of English. It needs to tell me about how to recreate a different theory in MY language. That's my two cents. The section on applying the knowledge is more like what I was expecting. The former material should be incorporated more simply into this, I think, as it retains the tone better and doesn't become a linguistics textbook. --162.95.216.224 (talk) 00:03, 24 December 2008 (UTC)

Just so we're on the same level; PiZero, you're going to draft a plan and present it here within the next two or three months? If so, then that's fine by me &mdash; there's plenty of other tasks for me to complete in the meantime. And hey, changes will happen at whatever rate they happen. :) Ingolemo (talk) 05:01, 8 January 2009 (UTC)


 * I'll try, though my draft "plan" probably won't amount to much. A potentially long time to wait for a pretty small result.  I don't mean to stand in your way, though (but you say you've got plenty of other things to occupy you...). Pi zero (talk) 18:35, 9 January 2009 (UTC)


 * Here's roughly how I envision rearranging the Advanced phonology and grammar sections. For all its faults, I still think it's considerably better than what we've got; but I don't imagine being able to write the content of most of the modules in it until we know much more clearly just what will be in the Intermediate section.  At any rate, it clarifies how the government theory material would fit in.
 * Introduction to the Advanced level
 * Phonology
 * Common sounds, in depth &mdash; including subtleties of articulation, diphthongs, suprasegmentals
 * Unusual sounds &mdash; including pharyngeals, clicks, ingressives
 * Non-human phonologies
 * Grammar
 * Aligning arguments &mdash; morphosyntactic alignment &mdash; including theta roles, non-acc, abs-erg, split-erg, Austronesian, active-stative, trigger, dechticaetiative
 * Forming words &mdash; including polysynthesis, incorporation, oligosynthesis, etc.
 * Constraints on natural languages &mdash; intro (relation to conlanging, anadewism, controversy); universals; government theory
 * Destroying the noun/verb distinction
 * The link between a language and its culture
 * Metaphor and its applications for conlanging
 * The Grand Tome of Conlanging Vocabulary
 * Left to my own devices, all I'd do now is shift the modules about to fit this outline, including moving the government material into it to make room for Intermediate syntax. --Pi zero (talk) 16:45, 1 March 2009 (UTC)


 * That all sounds good to me.
 * On the subject of what to do in intermediate syntax, I'm begining to think it's not a good idea to enforce the division between syntax and morphology. It's a useful distinction linguistically, but I'm not so sure it helps when trying to actually explain grammatical concepts. The reason I say this is because I'm having quite a bit of trouble writing intermediate morphology without covering major topics in syntax at the same time.
 * I don't know if you've got any ideas, but with all that in mind, I'm seeing something like this;
 * Intermediate Grammar
 * Nouns (nominal morphology and noun phrases, maybe prepositions as well)
 * Adjectives & Adverbs (adjectives, adverbs, determiners)
 * Verbs (verbal morphology and basic sentence structure)
 * Clauses (subordinate clauses)
 * Forms (affixes, clitics, word order; how to actually show the distinctions expressed by the grammar, this may be better as the first article in the section)
 * It still needs a bit of work, but I think that covers most of what we need it to. We'll probably need/want to split the more complicated pages up a little (separate pronouns and prepositions from nouns, clauses -> nominal clauses and relative clauses). What do you think? Ingolemo (talk) 02:50, 3 March 2009 (UTC)


 * Very plausible. The division between morphology and syntax always has been rather messy.  Your "Forms" section probably will naturally want to go first; "Clauses" will naturally want to go last; and it may turn out that "Adjectives & Adverbs" will want to go after "Verbs", as well as after "Nouns".
 * This means, of course, that in shifting this material to the Advanced part, the Intermediate/Syntax taxon will simply vanish from the outline. --Pi zero (talk) 14:19, 4 March 2009 (UTC)


 * Okay. Unless there's any objections, I'll rename Intermediate Morphology to Intermediate Grammar and then start adapting from there. Ingolemo (talk) 17:13, 5 March 2009 (UTC)

NPOV (If That Exists Here)
Much of this is entirely theory-dependent, and a conlanger who follows these "instructions" will end up producing conlangs that can be analyzed by Chomskyan theories, and no others. This is not a good idea. The very point of natural language syntactic frameworks is to describe the languages that exist in the world, and to exclude all those that don't. The point of conlanging is different. At the very least, this should not be presented as *the* syntax section. Rather, it should be labeled "An Introduction to Chomskyan Syntax for Conlangers". Dedalvs (talk) 06:34, 26 January 2009 (UTC)


 * Oh, good, a reason to throw it out after all. Though I have doubts that Chomsky's purpose has ever been "to describe the languages that exist in the world, and to exclude all those that don't."


 * And yes, NPOV exists here. --Pi zero (talk) 00:13, 27 January 2009 (UTC)


 * The language "Spanish with /e/ swapped with /o/ and /i/ swapped with /u/" is not spoken in the world, but which Chomskyan theory would reject it? --Damian Yerrick (talk) 20:30, 30 June 2009 (UTC)


 * I take Dedalvs's point to be that such theories attempt to exclude features that the theorist believes cannot occur naturally, whereas the conlanger might or might not care whether a feature can occur naturally &mdash; and if they do care, that doesn't mean they're trying to obey such constraints. That's why we've moved this material to a place in the outline reserved for theories that try to exclude unnatural features.  --Pi zero (talk) 22:20, 30 June 2009 (UTC)

Parts of Speech
Has a conlang rejected all parts of speech? —Preceding unsigned comment added by 97.85.177.144 (discuss • contribs) 20:06, 13 April 2013‎


 * Parts of speech are the different roles that words can play in larger grammatical structures. It seems like you'd have to have different words playing different grammatical roles, otherwise you couldn't build grammatical structures and you wouldn't really have a language.  But you might be able to imagine a fundamentally different way for a language to structure its grammar, so that the parts of speech would be fundamentally different from the ones we're familiar with in human natlangs.  Some conlangs, as I recall, have some parts of speech in them that aren't quite like anything in any natlang.  To build a language whose parts of speech are all drastically different requires imagining some profoundly different sort of grammatical structure that can nevertheless express arbitrarily complex thoughts and, presumably, come across as plausible.  I've tried a few things in that direction, so I can say with confidence that it's possible.  --Pi zero (discuss • contribs) 15:36, 14 April 2013 (UTC)

Verb Subcategories
In the table, it says that "give" is a ditransitive 3 and takes a noun phrase plus a noun or complement phrase. What about in "The man gave the book to the dog."? Wouldn't the second part by a prepositional phrase? —Preceding unsigned comment added by 98.179.157.122 (discuss • contribs) 13:39, 11 January 2016


 * The English language is a relatively exotic language in how it treats ditransitive verbs; in the technical jargon, English is "secundative" rather than "indirective". Or at least, some linguistics would say it is.  I've been studying up on this kind of thing for our advanced page on alignment (/Advanced/Grammar/Alignment), and I'm coming increasingly to two conclusions:  one, it's not at all clear that anyone really understands this stuff; and two, a lot of the disagreements about how to describe grammatical structure are like arguing over how many angels can dance on the head of a pin.  That said, there may be some enlightenment to be found in the technical details of this, so I'll riff on it for a bit here.


 * There are supposed to be three roles in a ditransitive verb (or at least, in ones like "give"): agent, theme, and recipient. Here, the man is the agent, the book is the theme, and the dog is the recipient.  Since English is a nominative-accusative language (rather than an ergative-absolutive one), the three relevant cases are nominative, accusative, and dative.  The accusative case is for the direct object of the verb, the dative case is for the indirect object of the verb.  However, there is some question here as to which is actually the direct object and which is indirect.  We can say the sentence two different ways:
 * (1) The man gave the book to the dog.
 * (2) The man gave the dog the book.
 * In an indirective language, the theme is the direct object (accusative), the recipient the indirect object (dative). In a secundative language, the recipient is the direct object (accusative), the theme the indirect object (dative).  Some people will tell you that English doesn't mark the dative, others will tell you that English uses prepositions to mark all cases other than nominative/accusative.  So what is happening here?  In (1), it looks like English is indirective, with accusative book, and dative dog marked by a preposition.  In (2), it looks like English is secundative, with accusative dog and dative book.  And then both of these sentences valency-reduce to a transative (that is, monotransitive) verb as
 * (3) The man gave the book.
 * Where clearly the book is accusative (the direct object). But if English is indirective, as it seems in (1), then all you've done here is drop the dative, whereas if English is secundative as it seems in (2) then you've actually dropped the accusative (book) and promoted the dog from dative to accusative, which is kind of like the valency-reduction of passive voice, where you drop the nominative and promote the accuative to nominative (from "The man gives the book" to "The book is given").


 * I think it's more useful to describe English as secundative, because then you can easily explain how (1) is formed from (2): first, you valency-reduce by dropping the accusative and promoting the dative ("the man gave the book"), then you modify the sentence with a prepositional phrase ("to the dog"). You can't directly form a passive voice from (2) because passivization only works on a transitive sentence, not on a ditransitive one; so [see below] you have to valency-reduce to a transitive sentence first, which gives you "the man gave the book", with optional modifying "to the dog", and then passivize it to "the book was given", which with modifying phase would be "the book was given to the dog".  Another modifying prepositional phrase would give you "the book was given by the man" or "the book was given to the dog by the man" or "the book was given by the man to the dog".


 * If you try to claim English is indirective then you have to invent additional machinery to explain why it's possible to say "the man gave the dog the book". There is such terminology; the conversion from (1) to (2) is called "dative shift".  Form (1) is called the "oblique dative" form, (2) is the "dative complement construction".  But then you have to explain why it's okay to say
 * (4) The man gave a million dollars to a charity.
 * (5) The man donated a millon dollars to a charity.
 * but you can only do a dative shift on the first of these, "the man gave a charity a million dollars" but not "the man donated a charity a million dollars". Someone who favors this way of looking at things would explain it by saying verb "donate" canot ungo dative shift.  It seems to me much simpler to say that "donate" is a transitive verb, "give" is a ditransitive verb.  But it's, at least superficially, a matter of taste, just like the choice between claiming (I'm tempted to say pretending) that "I drive" and "I drive a car" are two different verbs, versus saying "drive" in both sentences is a single verb that takes an optional direct object.


 * The thing to keep in mind is, the language works the same way regardless of which way you describe it. --Pi zero (discuss • contribs) 16:09, 11 January 2016 (UTC)


 * Heh. It's been pointed out to me that the ditransitive sentence can be passivized, producing
 * The dog was given the book.
 * Seems to me that supports treating the dog as accusative in the ditransitive form, since it's what is promoted to nominative in the passive. But here's another variant to boggle at:
 * The dog was given to.
 * --Pi zero (discuss • contribs) 18:18, 11 January 2016 (UTC)


 * To me, it seems like give should be changed to a simpler verb in the example, unless this behaviour is the norm for ditransitive 3 verbs.
 * Also, would it be possible that the two gives are somewhat like overloading(?) functions in programming? Like, the word behaves differently depending on the number and types of parameters? 166.167.81.114 (discuss) 19:28, 11 January 2016 (UTC)


 * Actually, I believe give is an altogether typical ditransitive verb in English. It's not that give is unusually complicated compared to other English ditransitive verbs, but that ditransitivity is English is complicated... at least, when you describe it using conventional grammatical terms like cases and such.  If there's another way of describing how it all works that is lots simpler, I don't think anyone has figured that out yet. There are two ways of describing how the different valencies of give relate to each other, and these two are like overloading and parametric polymorphism.  Overloading means there are several functions whose behaviors aren't really related to each other, they just happen to all have the same name; parametric polymorphism is where there's one function that uses a uniform rule for handling different kinds of stuff that can be passed to it.  If you say there are several verbs give with different valencies, that's like overloading; if you say there's one verb give whose form varies in thus-and-such regular way (regular for English ditransitive verbs), that's like parametric polymorphism.  --Pi zero (discuss • contribs) 19:46, 11 January 2016 (UTC)

Universals Page
There should be a page full of linguistic universals for all to see. —Preceding unsigned comment added by 97.85.166.212 (discuss • contribs) 10:52, 8 April 2013‎


 * The whole idea that there are even human lingusitic universals is quite controversial. I highly recommend the book w:Don't Sleep, There Are Snakes: Life and Language in the Amazonian Jungle.  --Pi zero (discuss • contribs) 11:15, 8 April 2013 (UTC)

Finding problems
http://en.wikipedia.org/wiki/X-bar_theory should solve the big brown dog problem.

--88.70.111.91 (talk) 21:33, 20 May 2010 (UTC)

Overhaul
I've merged all the files into a single page. Somehow or other, imho, this needs to be thoroughly overhauled, to not tell conlangers this theory is the only way there is, and to greatly shorten it. The current outline for the parent section calls for a section on Constituency grammar, which would likely be the overhauled form of this. --Pi zero (discuss • contribs) 23:09, 7 February 2020 (UTC)