Internally headed relative clauses in Trurian

I have recently relaunched Trurian, my first conlang (the first sketch dates to around 10 years ago). It’s a really pleasant feeling to be working on something fresh, and it’s also somewhat easier than my High Eolic work – both due to its already sizable vocabulary (2000+ entries) and since much of the grammatical legwork (especially basic morphology) is already present, which gives me more time to concentrate on the interesting stuff, such as syntax and pragmatics.

One feature which I’ve decided to revamp considerably from previous incarnations of the language is the way relative clauses are handled. In a burst of inspiration, I’ve decided to make them internally headed: this means that the head of the relative clause only occurs within the clause itself. The clause is marked by a relativizer. If the clause acts in the role of a subject of a sentence, the relativizer is ner:

ner Aenith woni ilnerono larneí
REL Aenith man.ACC kill.NONPAUS dangerous
“The man that Aenith killed [was] dangerous.”

Notice how the relativizer marks the beginning of the relative clause (underlined), whereas the matrix clause (including the verb – Trurian is a verb-final language) follows. The relativized noun, won ‘man’, is only present within the relative clause itselfunlike in (for example) English, where it exists as a head outside of the relative clause (the man that), and again as a gap within the relative clause (that Aennith killed [0]…).

What happens when the role of the relativized noun in the matrix clause is something other than subject? Well, the relativizer ner also inflects for case, and its different case forms vary precisely for the role of the entity referred to in the matrix clause. So, for example, if you want to say I had met the man that Aenith killed – which would put ‘the man’ into the accusative case of the matrix clause in the case system of Trurian – the appropriate relativizer is no longer ner, but rather ni (which I gloss as REL.ACC):

elrer ni Aenith woni ilnerono in emeyon
earlier REL.ACC Aenith man.ACC kill.NONPAUS I meet.PAUS
“I had met the man that Aenith killed.”

The form of the relativizer is independent of the role of the relativized noun within the relative clause. If you want to say I met the man that killed my father, the relativizer is still ni:

ni won in-idhoni ilnerono in emeyon
REL.ACC man my-father.ACC kill.NONPAUS I meet.PAUS
“I met the man that killed my father.”

Along with the accusative, Trurian has three other noun cases: the genitive, the dative, and the ablative. The relevant relativizer forms for the role of the relativized noun in the matrix clause are nel (for the genitive), neth (for the dative), and ana (for the ablative), as you can see in the examples below:

nel Aenith woni ilnerono fes lärtheís
REL.GEN Aenith man.ACC kill.NONPAUS house big.FEM
“The house of the man that Aenith killed is big.”

neth Aenith woni ilnerono wechebi entayon
REL.DAT Aenith man.ACC kill.NONPAUS money.ACC take.PAUS
“Aenith took money from the man that he killed.”

feser ana Aenith woni ilnerono yonéth
house.ADV REL.ABL Aenith man.ACC kill.NONPAUS live.3.MASC.SING.PAUS
“Aenith [now] lives in the house of the man that he killed.”

Finally, there is an issue that I still haven’t managed to resolve fully satisfactorily, namely: how do speakers “know” which constituent within the relative clause is the relativized noun? With the examples above, it’s pretty simple to figure this out, since the non-relativized constituents are very well-known/specific (Aenith, my father) and thus unlikely to be modified by a relative clause in any case. I’m quite prepared to put it all down to pragmatics and possible ambiguity, but maybe there’s some way around the problem.

Share and Enjoy:
  • Print
  • Digg
  • StumbleUpon
  • del.icio.us
  • Facebook
  • Yahoo! Buzz
  • Twitter
  • Google Bookmarks
This entry was posted in Project updates, Trurian and tagged , , , . Bookmark the permalink.

4 Responses to Internally headed relative clauses in Trurian

  1. One idea for marking which constituent is the head of the relative clause would be to front the constituent to directly after the relativizer, if that would work for your syntax would allow that.

    Also, a question- can only core arguments be relativized? Or can obliques be relativized as well? Forgive me if I missed it in your description.

    • jonafras says:

      Anything down to an indirect object in the dative can definitely be relativized – not sure about arguments within adverbial phrases though, or genitives. Probably yes, but we’ll see.

      Constituents could definitely be moved – that’s a very good idea! But I’m thinking of having it more as a disambiguation mechanism.

  2. Amanda says:

    Internally headed clauses in my Toma Heylm have the relativization marker directly before the relativized NP:

    I see the woman who gives a book to the man
    eff tisanot [min-u syin anya datil mind-u syin ay-alam]
    I(NOM) see(f) [REL_ACC the woman(NOM) give(f) book-ACC the DAT-man]

    I see the book which the woman gives to the man
    eff tisanot [syin anya datil min-u mind-u syin ay-alam]
    I(NOM) see(f) [the woman(NOM) give(f) REL_ACC book-ACC the DAT-man]

    I stole this usage wholesale from the example in Describing Morphosyntax – even including the actual relativizer min!

Leave a Reply

Your email address will not be published. Required fields are marked *