1
Exploring the Syntax-Phonology Interface:
An Experimental Study on Czech Pronominal Clitics
Anna Poĺomská (Masaryk University in Brno, Czech Republic)
Budapest • May 8
2
Outline
̶ How to define clitics?
̶ Syntactic properties of clitics
̶ Phonological properties of clitics
̶ Experiments
̶ Conclusion
How to define clitics syntactically?3
Wackernagel´s clitics SYNTAX
̶ after the 1st syntactic constituent
‘A dog bit him.’
Kous ho dnes pes.
bit.PRET him.ACC today dog.NOM
Pes ho dnes kous.
dog.NOM him.ACC today bit.PRET
Dnes ho pes kous.
today him.ACC dog.NOM bit.PRET
1st constituent CLITIC the rest of the sentence
How to define clitics phonologically?4
Enclitics PHONOLOGY
̶ integrated into the preceding prosodic word (= host)
[kous]ω ho
bit.PRET him.ACC
[pes]ω ho
dog.NOM him.ACC
[dnes]ω ho
today him.ACC
[kous ho]ω
[pes ho]ω
[dnes ho]ω
host CLITIC clitic integration
Clitics on the interface between syntax and phonology.5
Syntax + phonology
Question to be answered experimentally:
Is the degree of phonological integration influenced
by syntactic properties of the prosodic host?
Clitics on the interface between syntax and phonology.6
Syntax + phonology
Question to be answered experimentally:
Is the degree of phonological integration influenced
by syntactic properties of the host?
Clitics on the interface between syntax and phonology.7
Syntax + phonology
Question to be answered experimentally:
Is the degree of phonological integration influenced
by syntactic properties of the prosodic host?
Syntactic properties of clitics8
Syntax
̶ ordering is a result of MOVEMENT
Matushansky (2006)
movement-type 1 movement-type 2
verbal constituent non-verbal constituent
[Kous] 𝑉 ho dnes pes. [Pes] 𝑁 ho dnes kous. [Dnes] 𝐴𝑑𝑣 ho kous pes.
Syntactic properties of clitics9
Syntax
movement-type 1 movement-type 2
verbal constituent non-verbal constituent
[Kous] 𝑉 ho dnes pes. [Pes] 𝑁 ho dnes kous. [Dnes] 𝐴𝑑𝑣 ho kous pes.
Syntactic properties of clitics10
Syntax
movement-type 1 movement-type 2
verbal constituent non-verbal constituent
[Kous] 𝑉 ho dnes pes. [Pes] 𝑁 ho dnes kous. [Dnes] 𝐴𝑑𝑣 ho kous pes.
verb and clitic form one constituent
Phonological properties of clitics11
Phonology
̶ how to evaluate the degree of phonological integration?
= how to measure the degree of enclisis?
→ let´s consider phonological processes
and their domain of application in phonology
Phonological properties of clitics12
Phonology
̶ some processes apply only within the prosodic word
̶ in Czech: regressive voicing assimilation in obstruent clusters
pro[s]it ~ pro[zb]a ‘to ask’ ‘a request’
suffix is fully integrated
Phonological properties of clitics13
Phonology
̶ some processes apply only within the prosodic word
̶ in Czech: regressive voicing assimilation in obstruent clusters
pro[s]it ~ pro[zb]a ‘to ask’ ‘a request’
pe[s] ~ pe[zɦ]o kous ‘a dog’ ‘a dog bit him’
the clitic is fully integrated
Experimental part 114
Experiment 1
̶ we measured the degree of voicing of the obstruent before the clitic
pe[s] ~ pe[?ɦ]o dnes kous ‘a dog’
dne[s] ~ dne[?ɦ]o kous pes ‘today’ ‘a dog bit him today’
kou[s] ~ kou[?ɦ]o dnes pes ‘bit.PRET’
c
Experimental part 115
Experiment 1
̶ degree of voicing corresponds to degree of integration
pe[s] ~ pe[zɦ]o dnes kous ‘a dog’
dne[s] ~ dne[zɦ]o kous pes ‘today’ ‘a dog bit him today’
kou[s] ~ kou[zɦ]o dnes pes ‘bit.PRET’
voiceless [s] changes fully into voiced [z]: full integration
Experimental part 116
Experiment 1
̶ degree of voicing corresponds to degree of integration
pe[s] ~ pe[s] [ɦ]o dnes kous ‘a dog’
dne[s] ~ dne[s] [ɦ]o kous pes ‘today’ ‘a dog bit him today’
kou[s] ~ kou[s] [ɦ]o dnes pes ‘bit.PRET’
voiceless [s] stays voiceless [s]: no integration
Experimental part 117
Experiment 1
̶ degree of voicing corresponds to degree of integration
pe[s] ~ pe[s⇒z ɦ]o dnes kous ‘a dog’
dne[s] ~ dne[s⇒z ɦ]o kous pes ‘today’ ‘a dog bit him today’
kou[s] ~ kou[s⇒z ɦ]o dnes pes ‘bit.PRET’
voiceless [s] changes into partially voiced [s⇒z]: partial integration
Define footer – presentation title / department18
Design of Exp1
̶ 20 Czech native speakers read 14 sentences → 280 utterances
̶ pronominal clitic ho „him/it.ACC“ control items: [word]ω + [word]ω
[stem + suffix]ω
type
V
(verb)
N
(obj)
N
(subj)
Adv
(adv)
control
items
number of
sentences
2 2 3 3 2
Define footer – presentation title / department19
Design of Exp1
̶ what is measured
→ degree of voicing (word_ending_with_C + ho)
̶ how it is measured
→ program Praat: fraction of voiced frames
(based on Pitch values)
Experimental part 120
Results of Exp1
-C ho degree of voicing in front of the clitic ho
[V-head]ω + clitic 91%
[non-V-phrase]ω + clitic 83%
21
Results of Exp1:
Clitics between words and suffixes – VOICING ASSIMILATION
-C ho degree of voicing
[V-head]ω + clitic 91%
[non-V-phrase]ω + clitic 83%
-C-ba degree of voicing
[stem + suffix]ω 95%
-C h- degree of voicing
[word]ω + [word]ω 76%
Phonological properties of clitics22
Phonology
̶ some processes indicate the boundary of the prosodic word
̶ in Czech: obstruent devoicing
mra[z]it ~ mrá[s] ‘to freeze’ ‘frost’
Phonological properties of clitics23
Phonology
̶ some processes indicate the boundary of the prosodic word
̶ in Czech: obstruent devoicing
mra[z]it ~ mrá[s] ‘to freeze’ ‘frost’
mra[z]it ~ mrá[s] [ɦ]o spálil ‘to freeze’ ‘frost burnt him’
the clitic is NOT integrated
Experimental part 224
Experiment 2
̶ we measured the degree of devoicing of the obstruent before the clitic
mra[z]it ~ mrá[?ɦ]o spálil ‘to freeze’ ‘frost burnt him’
Experimental part 225
Experiment 2
̶ degree of devoicing corresponds to degree of integration
mra[z]it ~ mrá[s] [ɦ]o spálil no integration
mra[z]it ~ mrá[zɦ]o spálil full integration
mra[z]it ~ mrá[z⇒s ɦ]o spálil partial integration
‘to freeze’ ‘frost burnt him’
Define footer – presentation title / department26
Design of Exp2
̶ 20 Czech native speakers read 11 sentences → 220 utterances
̶ pronominal clitic ho „him/it.ACC“ control items: [word]ω + [word]ω
[stem + suffix]ω
type
V
(verb)
N
(obj)
N
(subj)
Adv
(adv)
control
items
number of
sentences
2 2 2 3 2
Define footer – presentation title / department27
Design of Exp2
̶ what is measured
→ degree of devoicing (word_ending_with_C + ho)
̶ how it is measured
→ program Praat: fraction of unvoiced frames
(based on Pitch values)
Experimental part 228
Results of Exp2
-C ho degree of devoicing in front of the clitic ho
[V-head]ω + clitic 8%
[non-V-phrase]ω + clitic 20%
29
Results of Exp2:
Clitics between words and suffixes – OBSTRUENT DEVOICING
-C ho degree of devoicing
[V-head]ω + clitic 8%
[non-V-phrase]ω + clitic 20%
-C-ba degree of devoicing
[stem + suffix]ω 5%
-C h- degree of devoicing
[word]ω + [word]ω 24%
Experimental part 2
Data correlation30
Correlation of the two processes
Exp2 degree of devoicing
[V-head]ω + clitic 8%
[non-V-phrase]ω + clitic 20%
Exp1 degree of voicing
[V-head]ω + clitic 91%
[non-V-phrase]ω + clitic 83%
31
Conclusion
̶ phonology mirrors syntax
̶ prosodic integration hierarchy:
stem+suffix ˃ verbal-head+clitic ˃ non-verbal phrase+clitic
32
Future research
main goal: MORE DATA
̶ other pronominal clitics
̶ verbal clitics (auxiliaries)
̶ other phonological processes (e. g. degemination)
33
References.
[1] Bjorndahl, C. 2022. Voicing and frication at the phonetics-phonology interface: An acoustic study of
Greek, Serbian, Russian, and English. Journal of Phonetics 92.
[2] Booij, G. 1996. Cliticization as prosodic integration: The case of Dutch. The Linguistic Review 13.
[3] Franks, S. et al. 2004. Pronominal Clitics in Slavic. Journal of Slavic Linguistics 12.
[4] Lenertová, D. 2004. Czech Pronominal Clitics. Journal of Slavic Linguistics 12.
[5] Matushansky, O. 2006. Head Movement in Linguistic Theory. Linguistic Inquiry 37(1).
[6] Palková, Z. 1997. Fonetika a fonologie češtiny.