Txeumé

[ná'oolkiłí]

Yes, this would only work if you'd be able to slot everything else before the verb into one giant NP (or whatever it's called in your framework). So one solution would be to have Txeumé be only partially head-final.

Well the standard analysis for German is that in unembedded clauses the verb undergoes head movement to C, and C has an EPP feature (meaning it needs some XP to move into its specifier, maybe the subject, maybe an adverb, etc). In embedded clauses with overt complementizers, though, for whatever reason there's no verb movement and no EPP feature. These trees illustrate that (let DP1 = Obj, DP2 = Subj, angle brackets = movement trace / silent copy).

Code: Select all

(1)  CP         (2)  CP
     /\              /\
  DP2  C'           C  TP
       /\              /\
    C+V  TP         DP2  T'
         /\              /\
     <DP2> T'           T  VP
           /\              /\
          T  VP         DP1  V
             /\
          DP1 <V>     'C S O V'

      'S V O'      

If this were what's going on in Txeumé, you'd have to stipulate that CP is head-initial when there's no overt complementizer (for V2) but head-final when there is one (for clause-final complementizers). But like you said, variable headedness isn't off the table. Or you could come up with some fun VP-remnant movement story...

Cool examples, Astraios. Head-final languages are neato.

So I wonder whether the fact that Lakota and Basque postpone adjectives, numerals, and demonstratives may have something to do with the fact that they allow grammatical prefixes?

Yeah, maybe. That's a really interesting topic of research in syntax/morphology these days — are words put together in the same way that sentences are? If so, it's hard to explain why a head-final language can have prefixes at all (or why in a lot of cases (eg verbal) suffixes don't appear in the order that the syntax would predict). If not, though, it remains a challenge to explain the fairly robust correlation that exists between direction of affixation and headedness.

Do you think you could do a "historical gloss" of the "The North Wind and the Sun" text?

Ja, sure. The syntax changes so much that just seeing things side by side really doesn't help, so I'll give an explanation too. I've changed some things since my first post, so it's a little different.

El viento del norte y el sol disputaban sobre sus poderes, y decidieron conceder la palma al que despojara a un viajero de sus vestidos.
Og spóigé sotz degpudand stav, notxéd éunn sór ma sintzog, é raiucheiga seutiurag depochaigócq, og se resa paumau conseigd.
[ʔɔˈspuːdzɨsədz dɨpʊˈdɐ̃ʊ̃stə | ˈnɐʊtʃɨd ˈʔøn ˈsormə ˈsɐ̃ɪ̃dʒə | ʔɨrœˈʔɐɪdzə sœˈtør dɨpəʔɐˈdzok | əsɪˈrɐɪsə ˈpɔɒmə kɔ̃ˈsɛst]

First the correspondences of the lexical elements:

• viento → éun /ɵ̃/
  norte → notx /notʃ/
  sol → só /sʊ/
  disputar → degpudaig /dɛpudadz/
  sobre → sotz /sodʒ/
  poder → póig /pʊdz/
  decidir → desig /desidz/
  conceder → conseig /kɔ̃sedz/
  palma → paum /pɒm/
  despojar → depochaig /depoʔadz/
  viajero → aiucheig /œʔedz/
  vestido → eutiv /øty/

Next let's look at one clause at a time. (SC = subject clitic, OC = object clitic)

• Og spóigé sotz degpudand stav
  [ʔɔˈspuːdzɨsədz dɨpʊˈdɐ̃ʊ̃stə]
  /ɔ=s=ˈpʊdz=ɪ=sodʒ dɛpud-ˈɐ̃=stɒ/
  3PL.SC=3.POSS=power=DEF=about fight-GER=AUX:PART
  'Them having been fighting over their powers...'

Instead of coordinating the two major clauses like Spanish does, Txeumé uses a converb construction. Here we have a subordinated clause (so no V2) where whose final verb is a participle (well, in a morphological form that is etymologically derived from the Spanish past participle; whether this is synchronically a participle is another question). This is roughly equivalent to English adverbial clauses of the form Having Xed... I wanted to preserve the imperfective aspect of the original disputaban, I've used the progressive auxiliary + 'gerund' ( ← Sp. ando/iendo form).

• notxéd éunn sór ma sintzog ... conseigd.
  [ˈnɐʊtʃɨd ˈʔøn ˈsormə ˈsɐ̃ɪ̃dʒə ... kɔ̃ˈsɛst]
  /ˈnotʃ=ɨ=d ˈɵ̃=r ˈsɔ=r=ma ˈsɪ̃dʒ=ɔ ... kɔ̃ˈsedz=d/
  north=DEF=GEN wind=DEF sun=DEF=and decide:3PL.PST=3PL.SC ... award.INF=PFX
  'The north wind and the sun decided to award...'

Here's a nice, unembedded clause. Our verb desintz 'they decided' splits in half: the root moves into second position, after the subject DP and before the subject clitic, while the prefix d(e) is left behind and cliticizes to the infinitive. Notice that coordination looks like X Y and; the enclitic conjunction m(a) is derived form Sp. más.

• é raiucheiga seutiurag depochaigócq, og se resa paumau conseigd.
  [ʔɨrœˈʔɐɪdzə sœˈtør dɨpəʔɐˈdzok | əsɪˈrɐɪsə ˈpɔɒmə kɔ̃ˈsɛst]
  /ɪ=r=œˈʔedz=a s=øtʏ=r=a depoˈʔadz ˈʊ=k | ɔ=se=r=es=a pɒm=ɒ kɔ̃sedz=d/
  3SG.SC=3SG.OC=traveler=DAT 3.POSS=clothing.PL=DEF=ACC strip:INF COND=COMP | 3PL.SC=3.IO=3SG.OC=3SG=DAT prize=DEF.ACC award.INF=PFX
  'Who would strip a traveler of his clothing, they would award the honor to him'

The biggest complication here is that I haven't thought of a good way to do headless relative clauses in Txeumé (does anyone know if/how head final langs do headless relatives?), so I had to use a different construction: a correlative. Correlatives differ from regular relatives in that they (usually) contain the relativized head noun, they are at the left edge of the matrix clause, and a pronoun doubles the head in the matrix clause (I can give a better explanation if anyone wants; I just don't want this post to be too long). They don't really exist in English, but I've kind of approximated what they look like in my translation.