Friday, May 7, 2010

Build-Yer-Own Language (Part 6)

Part 1 (phonetics and phonology) Part 2 (more phonology) Part 3 Part 4 (both morphology) Part 5 (syntax) Recap

At the end of the morphology "lesson", I gave two sentences*:
Your boys sit - klen-bal-xmal tuk-laubu-ba
They become childish - klen-bal-laubu-mifs
These are going to serve as my launching point on Pamak syntax. More specifically, we're going to use first one, because we can't get much word order out of a single word.

If you compare the sentence "klenvalxmal tuklaububa" with the lists of morphemes I gave, you'll notice that the word for "sit", xmal, is the root of the first word, and the word for "boy", laubu, is the root of the second. This means that Pamak is either a Verb Subject Object (VSO) language or a Verb Object Subject (VOS) language**. To make it simple on everyone, I'm going to say it's VOS.

Now that we know that, we know more about what the x-bar tree for Pamak will look like.

Remember the tree?
Remember how I said any of those nodes could be flipped the other way? And how I said we could move bits of the sentence up and down in the tree? Good. Keep that up.

Let's assume that top XP is a Tense Phrase. This means that the Spec position is the subject of the sentence. Since we know that in Pamak the subject follows the object, and the object follows the verb, we can reason that Spec has to go to the right of X', instead of the left like in the diagram. That way, we don't have to assume any internal rearranging to get Verb + Object-as-Complement-to-Verb + Subject, in that order.

Languages are generally internally parallel. If you've got a voiceless consonant, you're likely to have a voiced one. If you've got a word for "he", you're probably going to have a word for "she". In terms of syntax, if you've got one X-bar node on the right, you're likely to have all X-bar nodes on the right.

This means that Pamak will have the following rules:

  1. adjectives, prepositional phrases, and relative clauses follow nouns (car blue, cars with windows, cars that rock)
  2. adverbs and prepositional phrases follow verbs (said happily, said with vigor)
  3. prepositions come before what they refer to (before noon, over mountains)
  4. possessives follow nouns (car Kevin's, mountains Canada's)
  5. tense markings precede verbs, at least before movement (is seeing, rather than seeing is)
If memory serves, languages with right-sided nodes are more likely to have suffixes than prefixes. There'll still be both, of course, but you'll see more suffixes than anything else.

Note that no human language is perfectly symmetrical on the surface, either because there's an exception to the "deep structure" (X-bar) rules, or because the language moves pieces around ("transformation"). To this end, I'm adding three more rules for Pamak:

  • prepositions are suffixes, not individual words, so transform downwards in the tree
  • Pamak has no word for "the" and no word for "a"
  • possessive prefixes move from after the noun to before the noun

And now I think we're ready for some sentences.

klen-bal-xmal tuk-laubu-ba 
he/she-plural-sit you-boy-plural.noun
Your boys sit. 

They become childish.

stax-mifs bim pa-vlask 
red-verb roof I-house
The roof of my house is red.

no vlask-tal bim stax
is/are house-on roof red
The red roof is on the house.

"tuk-laubu-mis-mifs?" klen-lavan-gar bimf-mis-san
you-boy-adj.-verb he/she-ask-past anger-adj.-adverb
"Are you childish?", she asked angrily

(This format's used by linguists to make it easier to match meaning to morpheme. Hope you don't mind.)

I could go on, but I think you probably get the basic idea. Pamak, like most agglutinative languages, has a lot of words with complicated structure. You probably also noticed that it can get away with "verbing" nouns and adjectives more frequently than English does ("is red" and "becoming childish"), which makes it both easier and harder from a learner's perspective. There are real languages that do that as well, just like there are languages without gender and languages without "the".

Next up is either semantics or tense and aspect. Haven't decided yet. Does anyone care either way?

* I know, I know, I gave them in a recap too, but they were originally in morphology.
** or that it has some other basic word order, but shuffles things around, and that's unlikely with such a simple sentence
*** don't worry, that's still coming up

No comments: