Turkic copula strategies:
Grammaticalisation → Typology → Universal Dependencies

Jonathan Washington
Swarthmore College
UDTW2, 2025-10-26
Ljubljana

Overview

My talk in a nutshell

Overview copula stratgies found throughout Turkic,
show grammaticalisation trajectories, and overview approaches that are sensible given the limitations of UD

The next ~14 minutes:

  • Overview of Turkic copula strategies
  • Grammaticalisation trajectories of each
  • Examples from along this spectrum
  • Proposals for the annotation of each
  • A question I haven't figured out yet
  • A note on *bār/*yōq
  • Concluding thoughts

Turkic copula strategies

Strategy free examples bound examples
Subj Pred Эпӗ пӳртре.chv
Men yüti.ybe
Сен бичикчи кижи.alt
Arkadaşım evde.tur
Subj Pred Agr Bän äbdä män.otk
Мен бажыңда мен.tyv 		Мен үйдөмүн.kir
Ben evdeyim.tur
Subj Pred ER Bän äbdä ärür män.otk
Ben evde idim.tur 		Seŋ yü qayta par-i?ybe
Ben evdeydim.tur
Subj Pred BOL Мен үйдө болчумун.kir
Ben evde olacağım.tur
Subj Pred TUR Pādšāhïnïŋ qïzï tururmän.chg
Po ǰerlerde irey göʰhey turə.dkh 		Sen juɫaguçitır.slr
Arkadaşım evdedir.tur
Subj Pred [other verbs] Мен қалада жүрмін.kaz
Sag-gurgun otyrmysyňyz?tuk

Subj Pred

Root


Case=Nom
Person=3
Number=Sing
Person[psor]=1
Number[psor]=Sing Case=Loc
Person=3
Number=Sing
    
‘My friend is in the house.’

Root


Case=Nom
Person=1
Number=Sing Case=Loc
Person=3
Number=Sing
‘I am in the house.’

Subj Pred ER

Full ER verb: AUX/copotk


Case=Nom
Person=1
Number=Sing Person=1
Number=Sing Lemma=är
Aspect=Imp
‘I am in the house.’
 loses forms

Defective ER verb: AUX/copkaa


Case=Nom
Person=1
Number=Sing Lemma=e
Tense=Past
Person=1
Number=Sing
‘I was in the house.’
 cliticises

Cliticised defective ER: AUX/cop subtokentur


Case=Nom
Person=1
Number=Sing Lemma=i
Tense=Past
Person=1
Number=Sing
‘I was in the house.’
  • always AUX/cop, even when cliticised
  • sometimes lemma not visible in forms (e.g., çocuktumtur)
  • leverages UD subtokens
  • unifies languages regardless of level of grammaticalisation

Subj Pred BOL

  • VERB when lexical meaning, AUX/cop when copula
  • ambiguity: not always clear to annotator which meaning was intended
  • consistent with UD guidelines
  • consistent across languages: decision depends entirely on meaning

Full BOL verb: VERB, xcomptur


Case=Nom
Person=1
Number=Sing Lemma=ol
Tense=Past
Person=1
Number=Sing
‘I became a doctor.’
 semantically bleached

BOL verb as copula: AUX/copkir


Case=Nom
Person=1
Number=Sing Lemma=бол
Tense=Past
Aspect=Imp
Person=1
Number=Sing
‘I was at the house.’
other
meanings

BOL verb, other meanings: VERBtur


Case=Nom Lemma=ol
Tense=Past
Person=3
Number=Sing
‘What happened?’

Subj Pred Agr

Echoed pronoun: AUX/coptyv


Case=Nom
Person=1
Number=Sing Lemma=??
Tense=Pres
Person=1
Number=Sing
‘I am in the house’
 cliticisation

Cliticised PRON-like Agr: AUX/cop subtokenuzb


Case=Nom
Person=1
Number=Sing Lemma=??
Tense=Pres
Person=1
Number=Sing
‘I am in the house’
 phonol. changes

Cliticised Agr: AUX/cop subtokentur


Case=Nom
Person=1
Number=Sing Lemma=??
Tense=Pres
Person=1
Number=Sing
‘I am in the house’
  • always AUX/cop, even when cliticised and not similar to pronoun
  • leverages UD subtokens
  • unifies languages regardless of level of grammaticalisation

Lemma=ER (ärotk/euzb/itur)

  • ER lemma despite distinct etymology
  • lemma consistent with other copula strategies
  • Already in use in some Turkic treebanks

Lemma=PRON (menotk,uzb/bentur)

  • forms don't resemble PRON in some languages (e.g., -(y)Im ≉ bentur)
  • forms bear clear resemblance to PRON, even in some more highly grammaticalised languages (e.g., -MIn ≈ menkaz)
  • consistent with & highlights etymology
  • origins ≠ contemporary structure

Subj Pred TUR, etc.

Full TUR verb: VERBtur


Case=Nom
Person=1
Number=Sing Lemma=dur
Aspect=Prog
Person=1
Number=Sing
‘I am standing in the house.’
 semantically bleached

TUR verb as copula: AUX/copchg


Case=Nom
Person=1
Number=Sing Lemma=tur
Aspect=Imp
Person=1
Number=Sing
‘I am the Padshah's daughter.’
 cliticises

Cliticised defective ER: AUX/cop subtokentur


Case=Nom
Person[Psor]=1
Number=Sing Lemma=dur
Person=3
‘My friend is in the house.’
  • VERB when lexical meaning, AUX/cop when copula
  • ambiguity: not always clear to annotator which meaning was intended (e.g., үйде тұрмынkaz)
  • leverages UD subtokens
  • consistent with UD guidelines
  • consistent across languages: depends entirely on meaning
  • unifies languages regardless of level of grammaticalisation
  • note: cliticised TUR in some languages not inherently 3rd person (e.g., sen juɫaguçitırslr)

Note on *bār/*yōq

Copula followstur


‘You were there.’ (lit. ‘You were present.’)

Can be nominal dependentskir


‘tourists without euros’ (lit. ‘their euros being-absent tourists’)

Functions as copula?sah


‘I am at the house.’ (lit. ‘I am present to the house.’)
  • Generally ADJ predicates ‘present’, ‘absent’:
    • Copulas follow (e.g., sen var idintur)
    • May be nominal dependents (in some languages)
      (e.g., евросу жок туристтерkir)
  • In Sakha acts like copula as well:
    (e.g., Мин дьиэҕэ баарбын.sah)

Hvalaslv,
Təşəkkür edirəmazj,
Тавтапуҫchv, Raxmetkaa,
Рақметkaz, Ракматkir,
Баһыыбаsah, Sagoluňtuk
Teşekkür ederimtur,
Четтирдимtyv, Rahmatuzb,
Thank youeng!


slides available at

https://jonorthwash.github.io/2025-TurkicCopulas-UDTW2/presentation.html