1
Agreement attraction in Czech and what it tells
us about typology and processing
Jakub Dotlačil
(joint work with Jan Chromý and Radim Lacina)
FDSL, 2024
1
2
Take-home message
▶ Some syntactic phenomena are not simply present or absent in
a language.
▶ Rather, there is a continuum of how “strongly” a phenomenon
is present in language.
▶ The continuum arises not because of language
contact/bilingualism and is not caused by lexical variation.
2
2
Take-home message
▶ Some syntactic phenomena are not simply present or absent in
a language.
▶ Rather, there is a continuum of how “strongly” a phenomenon
is present in language.
▶ The continuum arises not because of language
contact/bilingualism and is not caused by lexical variation.
▶ Reading experiments on sentence processing can tap into the
continuum.
2
2
Take-home message
▶ Some syntactic phenomena are not simply present or absent in
a language.
▶ Rather, there is a continuum of how “strongly” a phenomenon
is present in language.
▶ The continuum arises not because of language
contact/bilingualism and is not caused by lexical variation.
▶ Reading experiments on sentence processing can tap into the
continuum.
▶ Reading experiments can inform us about the nature of an
effect beyond just the strength of the effect.
2
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
At times, the sheer weight of all these facts make them hard […] to understand.
from Reagan, Address to the nation, 1982, simplified
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
At times, the sheer weight of all these facts make them hard […] to understand.
from Reagan, Address to the nation, 1982, simplified
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
At times, the sheer weight of all these facts makes them hard […] to understand.
from Reagan, Address to the nation, 1982, simplified
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
At times, the sheer weight of all these facts make them hard […] to understand.
from Reagan, Address to the nation, 1982, simplified
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
At times, the sheer weight of all these facts make them hard […] to understand.
from Reagan, Address to the nation, 1982, simplified
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
3
The phenomenon
Agreement attraction:1
A verb agrees with a different noun than the canonical argument.
The noun causing the faulty agreement is called attractor.
I focus on number agreement attraction (for gender agreement
attraction, see Radim Lacina’s talk).
At times, the sheer weight of all these facts make them hard […] to understand.
from Reagan, Address to the nation, 1982, simplified
(1) The key to the cabinets is/are…
Around 10% of wrong agremeent forms in production tasks.2
1
(Jespersen, 1924; Zandvoort, 1961; Bock and Miller, 1991; Bock and Cutting, 1992)
2
(Bock and Miller, 1991)
3
4
The findings
Agreement attraction effects can be tested experimentally…
▶ …in production studies;
▶ …in reading studies on sentence processing.
3
(Wagers et al., 2009)
4
4
The findings
Agreement attraction effects can be tested experimentally…
▶ …in production studies;
▶ …in reading studies on sentence processing.
An example of a reading study:3
(2) a. The key to the cell unsurprisingly was/were rusty from
many years of disuse.
b. The key to the cells unsurprisingly was/were rusty from
many years of disuse.
The verb form either matches or mismatches the subject.
The attractor either matches or mismatches the verb form.
Does the attractor match affects reading profiles?
3
(Wagers et al., 2009)
4
5
Wagers et al, 2009: results
(3) a. The key to the cell unsurprisingly was rusty…
b. The key to the cell unsurprisingly were rusty…
c. The key to the cells unsurprisingly was rusty…
d. The key to the cells unsurprisingly were rusty…
5
5
Wagers et al, 2009: results
▶ Slowdown due to ungrammatical agreement on the regions
following the verb.
⇒ The effect of ungrammaticality.
▶ Relative speedup when the attractor agrees with the verb.
⇒ The effect of agreement attraction.
5
6
Agreement attraction: theory
Agreement attraction is explained theoretically as a consequence of
a (faulty) encoding4 and/or an incorrect recall.5
4
(Bock and Eberhard, 1993; Eberhard et al., 2005; Smith et al., 2021)
5
(Vasishth and Lewis, 2006; Vasishth et al., 2019)
6
6
Agreement attraction: theory
Agreement attraction is explained theoretically as a consequence of
a (faulty) encoding4 and/or an incorrect recall.5
(3) The key to the cells unsurprisingly were rusty…
The faulty-encoding approaches: the subject is wrongly marked as
plural due to the plural number of the attractor.
The wrong-recall approaches: when reading the verb, people recall
the attractor, instead of the actual subject.
4
(Bock and Eberhard, 1993; Eberhard et al., 2005; Smith et al., 2021)
5
(Vasishth and Lewis, 2006; Vasishth et al., 2019)
6
6
Agreement attraction: theory
Agreement attraction is explained theoretically as a consequence of
a (faulty) encoding4 and/or an incorrect recall.5
(3) The key to the cells unsurprisingly were rusty…
The faulty-encoding approaches: the subject is wrongly marked as
plural due to the plural number of the attractor.
The wrong-recall approaches: when reading the verb, people recall
the attractor, instead of the actual subject.
Both approaches: agreement attraction is an interaction of a
grammatical phenomenon with general cognitive capacities.
4
(Bock and Eberhard, 1993; Eberhard et al., 2005; Smith et al., 2021)
5
(Vasishth and Lewis, 2006; Vasishth et al., 2019)
6
7
Agreement attraction: theory
Is agreement attraction a linear order effect?
6
(Franck et al., 2002)
7
(Haskell and MacDonald, 2005)
8
(Tabor et al., 2004)
7
7
Agreement attraction: theory
Is agreement attraction a linear order effect?
No.6
(4) Is/are the helicopter for the flights safe?
(5) The threat to the president(s) of the company/ies was/were…
6
(Franck et al., 2002)
7
(Haskell and MacDonald, 2005)
8
(Tabor et al., 2004)
7
7
Agreement attraction: theory
Is agreement attraction a linear order effect?
No.6
(4) Is/are the helicopter for the flights safe?
(5) The threat to the president(s) of the company/ies was/were…
Yes.7
(6) The shirt or the socks was/were…
6
(Franck et al., 2002)
7
(Haskell and MacDonald, 2005)
8
(Tabor et al., 2004)
7
7
Agreement attraction: theory
Is agreement attraction a linear order effect?
No.6
(4) Is/are the helicopter for the flights safe?
(5) The threat to the president(s) of the company/ies was/were…
Yes.7
(6) The shirt or the socks was/were…
If yes: agreement attraction like local coherence effects:8
(7) The coach smiled at the player tossed a frisbee by the
opposing team.
6
(Franck et al., 2002)
7
(Haskell and MacDonald, 2005)
8
(Tabor et al., 2004)
7
8
Language variation and agreement attraction
Is agreement attraction universal?
1. The theory posits general cognitive mechanisms that are at fault.
2. Empirical studies show that it is widespread:
Agreement attraction was found in English,9 French10, Spanish11,
German,12 Arabic,13 Russian,14 Armenian,15 Turkish,16 Greek.17 The
presence of agreement attraction was confirmed in a meta-analysis
of agreement-attraction studies.18
9
(Wagers et al., 2009; Parker and An, 2018)
10
(Franck et al., 2015; Franck and Wagers, 2020)
11
(Acuña-Fariña et al., 2014; Lago et al., 2015)
12
(Lago and Felser, 2018)
13
(Tucker et al., 2015, 2021)
14
(Slioussar, 2018)
15
(Avetisyan et al., 2020)
16
(Lago et al., 2019)
17
(Paspali and Marinis, 2020)
18
(Jäger et al., 2017)
8
9
Czech and number agreement attraction
I will present four experiments (three published19) in Czech showing
null to negligible effects of agreement attraction. We will look at:
19
(Chromý et al., 2023)
9
9
Czech and number agreement attraction
I will present four experiments (three published19) in Czech showing
null to negligible effects of agreement attraction. We will look at:
▶ prepositional phrases;
▶ relative clauses;
▶ prepositional phrases with syncretism (nominative-genitive
form syncretic);
▶ (somewhat) complex structures;
19
(Chromý et al., 2023)
9
9
Czech and number agreement attraction
I will present four experiments (three published19) in Czech showing
null to negligible effects of agreement attraction. We will look at:
▶ prepositional phrases;
▶ relative clauses;
▶ prepositional phrases with syncretism (nominative-genitive
form syncretic);
▶ (somewhat) complex structures;
We will see:
▶ There is hardly any effect of number agreement attraction in
Czech.
▶ The effect can be detected under specific circumstances, but it
is very small.
▶ What’s going on? We will look into a possible explanation that
captures not just the absence of the effect, but why it is so
small.
19
(Chromý et al., 2023)
9
10
Czech and number agreement attraction, experiment I
We start with a simple design similar to the classic studies on
attractions:
(8) the postcard from the friend/friends was/were…
10
11
Czech and number agreement attraction, experiment I
1 2 3 4 5 6–8
Pohled od kamarád-a určitě bud-e probouzet krásné vzpomínky
postcard[nom.sg] from friend-gen.sg surely will-sg arouse nice memories
Pohled od kamarád-ů určitě bud-e probouzet krásné vzpomínky
postcard[nom.sg] from friend-gen.pl surely will-sg arouse nice memories
*Pohled od kamarád-a určitě bud-ou probouzet krásné vzpomínky
postcard[nom.sg] from friend-gen.sg surely will-pl arouse nice memories
*Pohled od kamarád-ů určitě bud-ou probouzet krásné vzpomínky
postcard[nom.sg] from friend-gen.pl surely will-pl arouse nice memories
‘The postcard from the friend(s) surely will-sg/pl form nice memories.’
Tested on 24 items, 120 fillers, 134 participants, an online self-paced reading study.
Predictions:
Ungrammaticality ⇒ slower RTs for the plural verb case.
Agreement attraction ⇒ faster RTs for the plural verb when the attractor is plural.
11
12
Reading study I: results
12
13
Reading study I: Bayesian analysis on the postcritical
region
Agreement attraction would make us expect a negative interaction
between VerbNumber and AttractorNumber.
Mixed-effects models with lognormal
link with the full subject,
item random effect structure.
Fixed effects:
▶ Verb number
▶ Attractor number
▶ Verb number:Attractor
number
13
14
Reading study I: Bayesian analysis on the postcritical
region
Bayes factor analysis comparing two models on the ungrammatical
subset:
▶ H1: a model with the attractor number as the fixed effect.
▶ H0: a model without the attractor number as the fixed effect.
Ratio of the likelihood of H0 and H1 ⇒ Values higher than 1 =
support (evidence) for the null model.
14
14
Reading study I: Bayesian analysis on the postcritical
region
Bayes factor analysis comparing two models on the ungrammatical
subset:
▶ H1: a model with the attractor number as the fixed effect.
▶ H0: a model without the attractor number as the fixed effect.
Ratio of the likelihood of H0 and H1 ⇒ Values higher than 1 =
support (evidence) for the null model.
Testing different a priori asusmptions about the size of the attractor
number effect in H1, all give evidence for H0.
µ σ BF01
0 0.01 1.53
0 0.03 3.81
-0.03 0.009 32.5
-0.015 0.0045 4.54
14
15
Czech and number agreement attraction, experiment II
It has been noted that number agreement attraction is common in
relative clauses.20
(9) The driver(s) who the runner wave(s) to each morning …
20
(Wagers et al., 2009)
15
16
Czech and number agreement attraction, experiment II
1 2 3 4 5 6 7–11
Závodník, kter-ého fanoušek nezřídka natáče-l kamerou, …
racer[nom.sg] who-acc.sg fan[nom.sg] often record-pst[sg] with camera …
Závodníci, kter-é fanoušek nezřídka natáče-l kamerou, …
racer[nom.pl] who-acc.pl fan[nom.sg] often record-pst[sg] with camera …
*Závodník, kter-ého fanoušek nezřídka natáče-l-i kamerou, …
racer[nom.sg] who-acc.sg fan[nom.sg] often record-pst-pl with camera …
*Závodníci, kter-é fanoušek nezřídka natáče-l-i kamerou, …
racer[nom.pl] who-acc.pl fan[nom.sg] often record-pst-pl with camera …
‘The racer(s) who the fan often recorded-sg/pl with a camera, was/were very popular in the country.’
28 items, 98 fillers, 212 participants, an online self-paced reading study.
Predictions:
Ungrammaticality ⇒ slower RTs for the plural verb case.
Agreement attraction ⇒ faster RTs for the plural verb when the attractor is plural.
16
17
Reading study II: results
17
18
Reading study II: Bayesian analysis on the postcritical
region
Agreement attraction would make us expect a negative interaction
between VerbNumber and AttractorNumber.
BF analysis:
µ σ BF01
0 0.01 1.47
0 0.03 3.44
-0.03 0.009 25.1
-0.015 0.0045 3.68
18
19
Czech and number agreement attraction, experiment III
It has been noted for some, but not all, languages (Russian,21 contra
Armenian22) that the morphological form matters (for agreement
attraction, it should be syncretic with nominative).
Nom.Pl kamarád-i
Acc.Pl kamarád-y
friends
▶ Exp 1: attractors as in the first column;
21
(Slioussar, 2018)
22
(Avetisyan et al., 2020)
19
19
Czech and number agreement attraction, experiment III
It has been noted for some, but not all, languages (Russian,21 contra
Armenian22) that the morphological form matters (for agreement
attraction, it should be syncretic with nominative).
Nom.Pl kamarád-i archivářk-y
Acc.Pl kamarád-y archivářk-y
friends archivers
▶ Exp 1: attractors as in the first column;
▶ Exp 3: attractors as in the second column (syncretic)
21
(Slioussar, 2018)
22
(Avetisyan et al., 2020)
19
20
Czech and number agreement attraction, experiment III
1 2 3 4 5 6–8
Složk-a pro archivářk-u nejspíš bud-e zahrnovat veškeré nálezy
file-nom.sg for archiver-acc.sg probably will-sg include all findings
Složk-a pro archivářk-y nejspíš bud-e zahrnovat veškeré nálezy
file-nom.sg orm archiver-acc.pl=nom.pl probably will-sg include all findings
*Složk-a pro archivářk-u nejspíš bud-ou zahrnovat veškeré nálezy
file-nom.sg orm archiver-acc.sg probably will-pl include all findings
*Složk-a pro archivářk-y nejspíš bud-ou zahrnovat veškeré nálezy
file-nom.sg orm archiver-acc.pl=nom.pl probably will-pl include all findings
‘A file for an archiver(s) will(-sg/pl) probably include all findings.’
32 items, 96 fillers, 263 participants, an online self-paced reading study.
20
21
Reading study III: results
21
22
Reading study III: Bayesian analysis on the postcritical
region
Agreement attraction would make us expect a negative interaction
between VerbNumber and AttractorNumber.
BF analysis:
µ σ BF01
0 0.01 0.58
0 0.03 1.25
-0.03 0.009 3.19
-0.015 0.0045 0.43
22
23
Czech agreement attraction in comparison to other
languages
Does Czech lack agreement attraction? Or does it have it?
In my view, that’s the wrong question to ask.
23
23
Czech agreement attraction in comparison to other
languages
Meta-analysis: comparison of reading profiles of agreement
attraction across 5 languages: reading times on the word after the
verb (just the attraction effect)
23
23
Czech agreement attraction in comparison to other
languages
Meta-analysis: comparison of reading profiles of agreement
attraction across 5 languages: reading times on the word after the
verb (full picture)
23
24
Czech, experiment IV
1 2 3 4 5 6–8
Zpráv-a pro archivářk-u, která pilně pracuje nejspíš bud-e …
report-nom.sg from archiver-acc.sg who works diligently probably will-sg …
Zpráv-a pro archivářk-y, které pilně pracuje nejspíš bud-e …
report-nom.sg from archiver-acc.pl=nom.pl who work diligently probably will-sg …
*Zpráv-a pro archivářk-u, která pilně pracuje nejspíš bud-ou …
report-nom.sg from archiver-acc.sg who works diligently probably will-pl …
*Zpráv-a pro archivářk-y, které pilně pracuje nejspíš bud-ou …
report-nom.sg from archiver-acc.pl=nom.pl who work diligently probably will-pl …
‘The report for the archiver/s, who work diligently, will (SG/PL) most likely contain all findings.’
24 experimental items, 322 participants, an online self-paced reading study.
24
25
Reading study IV: results
25
26
Interim summary
The phenomenon that seems universal reveals variation not wrt
presence or absence, but wrt how robust it is.
26
26
Interim summary
The phenomenon that seems universal reveals variation not wrt
presence or absence, but wrt how robust it is.
To uncover the variation, the methods like reading time studies,
which give us a continuous measure of the effect size, are useful.
26
26
Interim summary
The phenomenon that seems universal reveals variation not wrt
presence or absence, but wrt how robust it is.
To uncover the variation, the methods like reading time studies,
which give us a continuous measure of the effect size, are useful.
Such reading times studies need to be combined with
considerations of the size of the experiment.
26
27
Theoretical explanation
▶ Why is the effect in Czech so small?
▶ Why is syncretism relevant for number agreement attraction in
Czech (but not, say, Armenian)?
27
28
Syncretism and languages: Russian vs. Armenian
Recall: Syncretism is claimed to be relevant in Russian, but not in
Armenian.
Armenian: tested on attraction using relative clause structures
(proactive interference):
(10) The driver(s) who the runner wave(s) to each morning …
Russian: tested on attraction using PPs (retroactive interference):
(11) The key to the cabinets usually are rusty.
Thus, there might be differences in constructions, not languages.
28
29
Syncretism and languages
In fact, our Czech experiment on proactive interference included
nominative attractors:
Repetition item in one condition from Exp 2
1 2 3 4 5 6 7–
*Závodníci, kter-é fanoušek nezřídka natáče-l-i kamerou, …
racer[nom.pl] who-acc.pl fan[nom.sg] often record-pst-pl with camera …
‘The racer(s) who the fan often recorded-pl with a camera, were very popular in the country.’
29
29
Syncretism and languages
In fact, our Czech experiment on proactive interference included
nominative attractors:
Repetition item in one condition from Exp 2
1 2 3 4 5 6 7–
*Závodníci, kter-é fanoušek nezřídka natáče-l-i kamerou, …
racer[nom.pl] who-acc.pl fan[nom.sg] often record-pst-pl with camera …
‘The racer(s) who the fan often recorded-pl with a camera, were very popular in the country.’
▶ Exp 2: relative clause attraction ⇒ agreement attraction not found, although we had
nominative attractors
▶ Exp 3/4: attraction in PP + syncretism ⇒ agreement attraction found, albeit small
29
30
Czech explained
There are two sources of agreement attraction:
1. retrieval/encoding interference.
2. local coherence (linear order effect).
Retroactive interference:
(12) The key to the cabinets usually are rusty.
Both options derive agreement attraction, but the second only
under syncretism.
Proactive interference:
(13) The driver(s) who the runner wave(s) to each morning …
Only option 1 derives agreement attraction.
30
30
Czech explained
There are two sources of agreement attraction:
1. retrieval/encoding interference.
2. local coherence (linear order effect).
Retroactive interference:
(12) The key to the cabinets usually are rusty.
Both options derive agreement attraction, but the second only
under syncretism.
Proactive interference:
(13) The driver(s) who the runner wave(s) to each morning …
Only option 1 derives agreement attraction.
Detected Czech agreement attraction is due to Option 2.
30
31
Why is the effect so small in Czech?
Why do we observe speed-up in reading?
31
31
Why is the effect so small in Czech?
Why do we observe speed-up in reading?
31
31
Why is the effect so small in Czech?
Why do we observe speed-up in reading?
The distribution is shifted.
31
31
Why is the effect so small in Czech?
Why do we observe speed-up in reading?
The distribution is shifted.
The distribution has a shorter tail.
31
32
Shifted log normal distribution and RTs
We can study both changes using shifted log-normal distribution.
▶ Parameter shift: shift the whole distribution
▶ Parameter location: move the tail
A plea for shifted log-normal distribution
32
32
Shifted log normal distribution and RTs
We can study both changes using shifted log-normal distribution.
▶ Parameter shift: shift the whole distribution
▶ Parameter location: move the tail
A plea for shifted log-normal distribution
Log-normal distribution
32
32
Shifted log normal distribution and RTs
We can study both changes using shifted log-normal distribution.
▶ Parameter shift: shift the whole distribution
▶ Parameter location: move the tail
A plea for shifted log-normal distribution
Log-normal distribution Shifted log-normal distribution
32
33
Czech vs. English agreement attraction
We use shifted log-normal and inspect the parameters.
▶ Parameter shift
▶ Parameter location
Englisha
a
(Wagers et al., 2009)
33
33
Czech vs. English agreement attraction
We use shifted log-normal and inspect the parameters.
▶ Parameter shift
▶ Parameter location
Englisha
a
(Wagers et al., 2009)
Czech
33
34
Agreement attraction in Czech and other languages
(At least) two sources of agreement attraction.
Only one present/dominant in Czech (local coherence), which
explains:
▶ why syncretism relevant;
▶ why only retroactive interference;
▶ why agreement attraction only changes location, not shift.
The last point is the reason for the small size of the effect.
34
35
Summary
Agreement attraction,
35
35
Summary
Agreement attraction,
its almost null, but not quite so, effect in
Czech,
35
35
Summary
Agreement attraction,
its almost null, but not quite so, effect in
Czech,
its small effect in comparison to other
languages,
35
35
Summary
Agreement attraction,
its almost null, but not quite so, effect in
Czech,
its small effect in comparison to other
languages,
my explanation in terms of two sources
of agreement attraction, one of which absent/minimized
in Czech.
35
Thank you!
▶ Co-authors:
▶ Jan Chromý
▶ Radim Lacina
37
References I
Jespersen, Otto (1924). The philosophy of grammar. London: Allen and Unwin.
Zandvoort, R. M. (1961). “Varia syntactica”. In: Language and society: Essays
presented to Arthur M. Jensen on his seventieth birthday. Copenhagen: Liber
Amicorum.
Bock, J Kathryn and Carol A Miller (1991). “Broken agreement”. In: Cognitive
psychology 23.1, pp. 45–93.
Bock, Kathryn and J Cooper Cutting (1992). “Regulating mental energy:
Performance units in language production”. In: Journal of memory and
language 31.1, pp. 99–127.
Bock, Kathryn and Kathleen M Eberhard (1993). “Meaning, sound and syntax
in English number agreement”. In: Language and Cognitive Processes 8.1,
pp. 57–99.
Franck, Julie et al. (2002). “Subject-verb agreement errors in French and
English: The role of syntactic hierarchy”. In: Language and cognitive processes
17.4, pp. 371–404.
Tabor, Whitney et al. (2004). “Effects of merely local syntactic coherence on
sentence processing”. In: Journal of Memory and Language 50.4, pp. 355–370.
37
38
References II
Eberhard, Kathleen M et al. (2005). “Making syntax of sense: number
agreement in sentence production”. In: Psychological review 112.3, pp. 531–559.
doi: 10.1037/0033-295X.112.3.531.
Haskell, Todd R and Maryellen C MacDonald (2005). “Constituent structure
and linear order in language production: evidence from subject-verb
agreement.”. In: Journal of experimental psychology: Learning, memory, and
cognition 31.5, p. 891.
Vasishth, Shravan and Richard L. Lewis (2006). “Argument-head distance and
processing complexity: Explaining both locality and antilocality effects”. In:
Language 82.4, pp. 767–794.
Wagers, Matthew W et al. (2009). “Agreement attraction in comprehension:
Representations and processes”. In: Journal of Memory and Language 61.2,
pp. 206–237.
Acuña-Fariña, J Carlos et al. (2014). “Gender and number agreement in
comprehension in Spanish”. In: Lingua. International review of general
linguistics 143, pp. 108–128. doi: 10.1016/j.lingua.2014.01.013.
Franck, Julie et al. (2015). “Task-dependency and structure-dependency in
number interference effects in sentence comprehension”. In: Frontiers in
psychology 6, p. 349.
38
39
References III
Lago, Sol et al. (2015). “Agreement attraction in Spanish comprehension”. In:
Journal of Memory and Language 82, pp. 133–149. issn: 0749-596X. doi:
https://doi.org/10.1016/j.jml.2015.02.002. uRl: https:
//www.sciencedirect.com/science/article/pii/S0749596X15000212.
Tucker, Matthew A et al. (2015). “Representing number in the real-time
processing of agreement: Self-paced reading evidence from Arabic”. In:
Frontiers in psychology 6, p. 347.
Jäger, Lena A et al. (2017). “Similarity-based interference in sentence
comprehension: Literature review and Bayesian meta-analysis”. In: Journal of
Memory and Language 94, pp. 316–339.
Lago, Sol and Claudia Felser (2018). “Agreement attraction in native and
nonnative speakers of German”. In: Applied Psycholinguistics 39.3, pp. 619–647.
doi: 10.1017/S0142716417000601.
Parker, Dan and Adam An (2018). “Not all phrases are equally attractive:
Experimental evidence for selective agreement attraction effects”. In: Frontiers
in psychology 9, p. 1566. doi: 10.3389/fpsyg.2018.01566.
39
40
References IV
Slioussar, Natalia (2018). “Forms and features: The role of syncretism in
number agreement attraction”. In: Journal of Memory and Language 101, pp. 51
–63. doi: https://doi.org/10.1016/j.jml.2018.03.006. uRl: http:
//www.sciencedirect.com/science/article/pii/S0749596X18300305.
Lago, Sol et al. (2019). “Straight from the horse’s mouth: agreement attraction
effects with Turkish possessors”. In: Linguistic Approaches to Bilingualism 9.3,
pp. 398–426.
Vasishth, Shravan et al. (2019). “Computational models of retrieval processes
in sentence processing”. In: Trends in Cognitive Sciences 23.11, pp. 968–982.
Avetisyan, Serine et al. (2020). “Does case marking affect agreement attraction
in comprehension?” In: Journal of Memory and Language 112, p. 104087. doi:
10.1016/j.jml.2020.104087.
Franck, Julie and Matthew Wagers (2020). “Hierarchical structure and memory
mechanisms in agreement attraction”. In: Plos one 15.5, e0232163. doi:
10.1371/journal.pone.0232163.
Paspali, Anastasia and Theodoros Marinis (2020). “Gender agreement
attraction in Greek comprehension”. In: Frontiers in psychology 11, p. 717. doi:
10.3389/fpsyg.2020.00717.
40
41
References V
Smith, Garrett et al. (2021). “Encoding interference effects support
self-organized sentence processing”. In: Cognitive Psychology 124, p. 101356.
doi: 10.1016/j.cogpsych.2020.101356.
Tucker, Matthew A et al. (2021). “Attraction effects for verbal gender and
number are similar but not identical: Self-paced reading evidence from
Modern Standard Arabic”. In: Frontiers in psychology, p. 3774. doi:
10.3389/fpsyg.2020.586464.
Chromý, Jan et al. (Oct. 2023). “Number Agreement Attraction in Czech
Comprehension: Negligible Facilitation Effects”. In: Open Mind 7, pp. 802–836.
issn: 2470-2986. doi: 10.1162/opmi_a_00107. eprint:
https://direct.mit.edu/opmi/articlepdf/doi/10.1162/opmi\_a\_00107/2164880/opmi\_a\_00107.pdf.
uRl:
https://doi.org/10.1162/opmi\_a\_00107.
41