psyg-12-742127 No embe 3, 2021 Time: 18:15 # 1
ORIGINAL RESEARCH
published: 10 No embe 2021
doi: 10.3389/ psyg.2021.742127
Edi ed by:
Ped o Guija o-Fuen es,
Uni e si y o he Balea ic Islands,
Spain
Re iewed by:
Nu ia Saga a,
Ru ge s, The S a e Uni e si y
o New Je sey, Uni ed S a es
Ma ia Ga a a,
Uni e si y o Eas Anglia,
Uni ed Kingdom
*Co espondence:
Adam Zawiszewski
[email p o ec ed]
Special y sec ion:
This a icle was submi ed o
Language Sciences,
a sec ion o he jou nal
F on ie s in Psychology
Recei ed: 15 July 2021
Accep ed: 07 Oc obe 2021
Published: 10 No embe 2021
Ci a ion:
Ma ínez de la Hidalga G,
Zawiszewski A and Laka I (2021)
Going Na i e? Yes, I Allowed by
C oss-Linguis ic Simila i y.
F on . Psychol. 12:742127.
doi: 10.3389/ psyg.2021.742127
Going Na i e? Yes, I Allowed by
C oss-Linguis ic Simila i y
Gillen Ma ínez de la Hidalga, Adam Zawiszewski*and I zia Laka
Depa men o Linguis ics and Basque S udies, Facul y o A s, Uni e si y o he Basque Coun y, Vi o ia-Gas eiz, Spain
Can na i e compe ence be achie ed in a second language? He e, we ocus on he
Language Dis ance Hypo hesis ha claims ha ea ly and p o icien bilinguals can
achie e na i e compe ence o g amma ical p ope ies sha ed by hei wo languages,
whe eas unsha ed g amma ical p ope ies pose a challenge o na i e-like syn ac ic
p ocessing. We p esen a no el beha io al and E en -Rela ed Po en ial (ERP) s udy
whe e ea ly and p o icien bilinguals beha e na i e-like in hei second language when
p ocessing (a) a gumen s uc u e al e na ions in in ansi i e sen ences in ol ing agen
s. pa ien subjec s and (b) subjec e b ag eemen , bo h o which a e g amma ical
p ope ies sha ed by hei wo languages o hese bilinguals. Compa ed o na i e
Basque bilinguals (L2Spanish) on he same asks, non-na i es elici ed simila sen ence
p ocessing measu es: (a) in he accep abili y ask hey eac ed as e and mo e
accu a ely o unaccusa i e sen ences han o une ga i es and o pe son han numbe
iola ions: (b) hey gene a ed a la ge P600 o ag eemen iola ions in unaccusa i e
sen ences han une ga i es; (c) hey gene a ed la ge nega i i y and posi i i y e ec s
o pe son han o numbe iola ions. P e ious s udies on Basque-Spanish bilinguals
ind ha ea ly and p o icien non-na i es display e ec s dis inc om na i es in
bo h languages when p ocessing g amma ical p ope ies whe e Basque and Spanish
di e ge, such as a gumen alignmen (e ga i e/nomina i e) o wo d o de ype (OV/VO),
bu hey pe o m na i e-like o sha ed p ope ies such as subjec ag eemen and
wo d meaning. We con end ha language dis ance, ha is, he deg ee o simila i y
o he languages o he bilingual is a c ucial ac o ha dese es u he and de ailed
a en ion o ad ance ou unde s anding o when and how bilinguals can go na i e in a
second language.
Keywo ds: non-na i e language p ocessing, e en - ela e po en ials, une ga i e s. unaccusa i e p edica es,
subjec - e b ag eemen , phi- ea u es, bilingualism
INTRODUCTION
Can non-na i e speake s a ain na i e-like compe ence in g amma ical p ocessing? Resea ch
ca ied ou h oughou he las decades has iden i ied key ac o s o ake in o accoun when
s udying non-na i e syn ac ic p ocessing, namely age o acquisi ion (AoA), p o iciency, simila i y
be ween L1 and L2 and ac i e use o he language (Ca a a e al., 2015;Ha sho ne e al., 2018;
B ice e al., 2019).
In second language acquisi ion, syn ax is epo ed o be ha de o acqui e han o he aspec s
o language (Ojima e al., 2005;Ko z, 2009;Vandenbe ghe e al., 2019). I has also been shown
ha AoA and he le el o p o iciency play a big ole in a aining na i e-like pe o mance
F on ie s in Psychology | www. on ie sin.o g 1No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 2
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
(i.e., Webe -Fox and Ne ille, 1996;Wa enbu ge e al., 2003).
Webe -Fox and Ne ille (1996), o ins ance, used ERPs o es
Chinese-English adul bilinguals exposed o English a di e en
age du ing he li e span (1–3, 4–6, 7–10, 11–13, and a e
16 yea s o age) and asked he pa icipan s o ead sen ences
con aining syn ac ic and seman ic anomalies. Resul s e ealed
signi ican AoA e ec s o syn ac ic p ocessing (ph ase s uc u e,
speci ici y and subjacency cons ain s), ha is, in compa ison
o English monolinguals, beha io al and elec ophysiological
measu es o Chinese-English bilinguals we e a ec ed by a delay in
L2 exposu e as sho as 1–3 yea s. By con as , ega ding seman ic
anomalies, only subjec s exposed o English a e 11–13 yea s
showed di e ences as compa ed o na i es. Wa enbu ge e al.
(2003), in u n, used he MRI me hod o es he e ec s
o AoA and p o iciency in h ee g oups o I alian-Ge man
bilinguals who lea ned hei L2 a di e en ages and had di e en
p o iciency le els (ea ly AoA (= a bi h), high p o iciency g oup;
la e AoA (>6 yea s), high p o iciency g oup and la e AoA
(>6 yea s), low p o iciency g oup). Pa icipan s ead sen ences
in hei L1 and L2 con aining syn ac ic (gende , numbe o
case disag eemen ) and seman ic anomalies (i.e., “The dee
shoo s he hun e ”). Resul s e ealed ha di e ences in he
MRI pa e n epo ed o he syn ac ic ask we e due o he
delay in AoA, while he pa e n o b ain ac i i y o seman ic
judgmen depended on he le el o p o iciency, suppo ing he
indings o Webe -Fox and Ne ille (1996).
On he o he hand, se e al s udies a i m ha high p o iciency
L2 speake s can a ain na i e-like pe o mance (F iede ici e al.,
2002;Rossi e al., 2006;He nandez e al., 2007;Ko z e al., 2008)
ega dless o hei la e AoA, hus challenging he C i ical Pe iod
Hypo hesis (Lennebe g, 1967). F iede ici e al. (2002) used ERP
measu es o es adul Ge man lea ne s (AoA = 24.1 yea s) o
an a i icial language and showed a simila ERP pa e n o ha
epo ed o na i e speake s o Ge man on a simila ask in
na u al language. Acco ding o he au ho s, hese esul s indica e
ha a language lea ned la e can be p ocessed in a na i e-like way.
Simila ly, Rossi e al. (2006) showed ha high-p o iciency la e L2
I alian-Ge man and Ge man-I alian lea ne s (AoA >10 yea s)
display he same ERP componen s as na i e speake s when
p ocessing wo d ca ego y and subjec - e b ag eemen syn ac ic
iola ions, sugges ing ha wi h a high p o iciency L2 lea ne s can
show na i e-like esponses ega dless o he la e AoA.
Finally, some s udies epo di e ences be ween na i e and
non-na i es ega ding ce ain syn ac ic phenomena bu no
o he s (Zawiszewski e al., 2011;Fouca and F enck-Mes e,
2012;E docia e al., 2014;Díaz e al., 2016;Zawiszewski and
Laka, 2020). Zawiszewski e al. (2011),Díaz e al. (2016), and
Zawiszewski and Laka (2020) examined he p ocessing o case
mo phology and E docia e al. (2014) he p ocessing o wo d
o de compa ing na i e and non-na i e speake s o Basque
(L1Spanish) and ound ha non-na i es, despi e an ea ly AoA
and high compe ence in hei L2 did no p ocess hese wo
aspec s o Basque g amma like na i es. Since case alignmen
and basic wo d o de a e wo main g amma ical ea u es ha
Basque and Spanish do no sha e (Basque is e ga i e and OV,
Spanish is nomina i e and VO), hey concluded ha linguis ic
dis ance, ha is, he deg ee o simila i y o he bilingual’s
g amma s was a ele an ac o in inal a ainmen in second
language p ocessing.
These di e en se s o indings epo ed in he L2 p ocessing
li e a u e ha e been accoun ed by many heo e ical p oposals.
Some posi ha L2 acquisi ion s ongly depends on he
L1 and hus he esul s can be in e p e ed in e ms o
a posi i e o nega i e ans e (i.e., The Uni ied Model o
Language Acquisi ion, He nandez e al., 2005;MacWhinney,
2005). Schwa z and Sp ouse (1996) also sugges ha L2
acquisi ion hinges on he ea u es a ailable in he L1 ( he Full
T ans e /Full Access model) and his iew is also compa ible wi h
he In e p e abili y Hypo hesis (Tsimpli and Dimi akopoulou,
2007), which posi s ha only in e p e able ea u es a e accessible
o he L2 lea ne s while he unin e p e able ones a e subjec
o c i ical pe iod cons ain s and, consequen ly, inaccessible
o L2 lea ne s. The possibili y o syn ac ic in o ma ion being
sha ed be ween bo h languages has been also sugges ed by
Ha suike e al. (2004) (Sha ed Syn ax Accoun ): i sugges s ha
g amma ical ules ha a e he same in L1 and L2 a e ep esen ed
once. In o he wo ds, L2 lea ne s would ely on hei L1 whene e
using a g amma ical s uc u e p esen in he wo languages
(see also Zawiszewski and Laka, 2020, o simila assump ions),
Con e sely, Clahsen and Felse (2006) pu o wa d he Shallow
S uc u e Hypo hesis and sugges ha la e L2 lea ne s a e no able
o p ocess syn ax in a na i e-like way and ha e o ely o a la ge
ex en on seman ic/p agma ic in o ma ion (see also S einhaue
e al., 2009 o a discussion).
THE PRESENT STUDY
The p esen s udy sough o examine he p ocessing o
in ansi i e p edica es in Basque by ea ly and p o icien
L1Spanish—L2Basque bilinguals. To his pu pose, we used
g amma ical and ung amma ical pe son and numbe ag eemen
manipula ions and compa ed he esul s o hose p e iously
epo ed by Ma inez de la Hidalga e al. (2019) o na i es.
The dis inc ion be ween in ansi i es whose sole a gumen
is an agen (une ga i es) and in ansi i es whose sole a gumen
is a heme (unaccusa i es) is a gene al p ope y o g amma s
[Unaccusa i e Hypo hesis (UH), Pe lmu e , 1978], and bo h
Basque and Spanish di e en ia e hese wo ypes o p edica es.
Mo e p ecisely, he UH claims ha unaccusa i e in ol e
mo e complex de i a ions han une ga i es because hemes a e
p omo ed o subjec s o unde go mo emen and lea e a ace
(Bu zio, 1986), whe eas agen s a e bo n as subjec s. Impo an ly,
some au ho s p opose o he unaccusa i e e bs in Basque
he same de i a ion as s a ed by he UH (O iz de U bina,
1989), while o he s claim no need o he ex a de i a ional
s ep (Laka, 2006a,b;Le in, 1983). Mo e complex de i a ions
a e usually ela ed o a g ea e p ocessing cos (longe eading
o eac ion imes, la ge ERP signa u es) as compa ed o less
complex s uc u es (i.e., Ma zke e al., 2002). Consequen ly, la ge
p ocessing cos is expec ed o unaccusa i es in compa ison o
une ga i es. Subjec ag eemen is also a sha ed p ope y be ween
Basque and Spanish, and bo h g amma s ep esen i by means
o pe son and numbe ea u es. Unaccusa i e e bs ha e been
F on ie s in Psychology | www. on ie sin.o g 2No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 3
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
ound o be ha de o lea n han une ga i es o second language
lea ne s a ini ial s ages (Yuan, 1999;Oshi a, 2001;Mon ul,
2005;in e alia). Oshi a (2001) o ins ance, pu o h he
Unaccusa i e T ap Hypo hesis (UTH), a guing ha L2 lea ne s
assume a i s all in ansi i e p edica es o be une ga i es. As
p o iciency inc eases, howe e , lea ne s no ice ha unaccusa i es
unc ion di e en ly and s a making di e ences be ween he wo
p edica es, and a highe le els o p o iciency hey a e ound o
pe o m na i e-like ega ding his linguis ic dimension.
In a p e ious s udy ca ied ou in Basque, Ma inez de la
Hidalga e al. (2019) in es iga ed Basque-Spanish bilinguals in
o de o es he UH hypo hesis and phi- ea u e p ocessing.
Resul s e ealed ha in he accep abili y ask he pa icipan
eac ed as e and mo e accu a ely o unaccusa i e sen ences
han o une ga i es and o pe son han numbe iola ions
and hey gene a ed a la ge P600 o ag eemen iola ions in
unaccusa i e sen ences han in une ga i es. Fu he mo e, hey
gene a ed la ge nega i i y and posi i i y e ec s o pe son
han o numbe iola ions. O e all, he esul s e ealed g ea e
p ocessing cos s o une ga i es han o unaccusa i es and
he au ho s in e p e ed hese indings as e idence p o iding
suppo o di e en s uc u al ep esen a ions o bo h ypes
o p edica es. Howe e , he p edic ion o highe p ocessing
cos o unaccusa i es han o une ga i es was no con i med,
suppo ing he idea o an inhe en a he han s uc u al na u e
o case in Basque (Le in, 1983;Laka, 2006a,b). Rega ding
ag eemen ea u es, na i e speake s p ocessed pe son and
numbe ea u es sepa a ely, he pe son being a mo e salien
han he numbe (see Ca mina i, 2005;Zawiszewski e al., 2016;
Mancini, 2018, o mo e in o ma ion on he p ocessing o pe son
and numbe ea u es).
Hypo heses and P edic ions
Ou wo king hypo hesis is he Language Dis ance Hypo hesis
(LDH, a e Zawiszewski and Laka, 2020): no di e ences a e
expec ed o p ocessing ai s o L2 ha a e p esen in L1, whe eas
e en a an ea ly AoA and high p o iciency in L2, na i e s.
non-na i e di e ences will a ise in he p ocessing g amma ical
p ope ies o L2 no p esen in L1. P e ious s udies in Basque
in es iga ing e ga i e case mo phology (Díaz e al., 2011;
Zawiszewski e al., 2011;Zawiszewski and Laka, 2020) and wo d
o de p ocessing (E docia e al., 2014) in na i e and ea ly and
highly p o icien Spanish-Basque bilinguals ound di e ences
be ween bo h popula ions, a ibu ed by he au ho s o he
di e ging g amma ical cha ac e is ics o Basque and Spanish.
In he p esen s udy, he expe imen al manipula ions in ol e
g amma ical ai s sha ed by bo h Basque and Spanish, namely
he dis inc ion be ween unaccusa i e s. une ga i e p edica es,
and pe son s. numbe ea u es in subjec - e b ag eemen .
Howe e , despi e he ac ha bo h Basque and Spanish
dis inguish be ween unaccusa i e and une ga i e p edica es, in
Basque agen s bea an e ga i e case ma king and hemes a e
mo phologically unma ked. In con as , in Spanish all subjec s
a e mo phologically indis inguishable.
We en a i ely hypo hesize ha , gi en he ea ly AoA and
high p o iciency o he non-na i e speake s unde s udy, a
simila pa e n o esul s o ha epo ed in Ma inez de la
Hidalga e al. (2019) will eme ge: (a) as e and mo e accu a e
esponses o unaccusa i e sen ences han o une ga i e ones and
o pe son iola ions han numbe iola ions in he accep abili y
ask; (b) a gene al N400–P600 pa e n as an ERP esponse
o e b ag eemen iola ions, unaccusa i e iola ions gene a ed
a la ge posi i i y as compa ed o une ga i es and pe son
ea u e iola ions gene a ed a la ge nega i i y as compa ed o
numbe ea u e iola ions in he ea ly ime window; and (c)
pe son iola ions in he une ga i e condi ion gene a ed a la ge
posi i i y as compa ed o numbe iola ions, and la ge posi i i y
ob ained o numbe iola ions in he unaccusa i e condi ion
as compa ed o numbe iola ions in he une ga i e condi ion
in he la e ime window (P600 e ec ). The p edic ions made by
he LDH a e also compa ible wi h he Sha ed Syn ax accoun
(Ha suike e al., 2004).
Pa icipan s
In his expe imen 26 ea ly and highly p o icien non-na i e
speake s o Basque, whose L1 was Spanish1 ook pa in
he expe imen ( i e males; mean age 20.5 yea s, SD = 2.67;
AoA = 3.31 yea s, SD = 1.3). Da a om wo pa icipan s we e
excluded as a esul o excessi e eye mo emen s and o he
a i ac s. All pa icipan s we e schooled in Basque om ea ly
childhood and we e he e o e highly p o icien in Basque (see
Table 1 o de ails) as e ealed by he ac ha 21 pa icipan s
had a ce i ied C1 le el in Basque and he emaining 3 we e
comple ing hei unde g adua e deg ee in Basque.
MATERIALS AND METHODS
416 sen ences dis ibu ed in ou lis s (256 expe imen al and 160
ille s) we e c ea ed. The ma e ials we e o ganized acco ding o
he manipula ions used in he expe imen (2 ×2×2 design):
p edica e ype (unaccusa i e s. une ga i e), ea u e (pe son and
numbe ), and g amma icali y (g amma ical and ung amma ical)
(see Table 2). Fo pe son condi ions 2nd pe son was used in
he g amma ical condi ion and o 1s pe son was used in he
ung amma ical manipula ion. Fo numbe condi ions, he design
used in Mancini e al. (2011) was ollowed: 3 d singula s. plu al
manipula ions. The c i ical wo ds we e he auxilia y e bs, always
p eceded by he main e bs and ollowed by h ee wo ds all e bs
we e con olled o leng h and equency.
P ocedu e
Pe sonal compu e s (Windows 7 ope a ing sys em) and
P esen a ion so wa e ( e sion 16.3) we e used o p esen he
s imuli on sc een. Be o e he expe imen s a ed, pa icipan s
we e old abou he EEG p ocedu e and sea ed com o ably in
a quie oom in on o a 24 inch moni o . The expe imen
was conduc ed in he Expe imen al Linguis ics Labo a o y a
he Uni e si y o he Basque Coun y (UPV/EHU) in Vi o ia-
Gas eiz. Pa icipan s conduc ed an accep abili y judgmen
ask, we e bo h accu acy and eac ion imes we e eco ded.
Sen ences we e displayed in he middle o he sc een wo d by
wo d o 350 ms (ISI = 250). A ixa ion c oss (+) indica ed he
1One pa icipan had Ca alan as mo he ongue; gi en he ypological simila i y
be ween Ca alan and Spanish, hose da a we e no disca ded om he inal
analysis.
F on ie s in Psychology | www. on ie sin.o g 3No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 4
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
TABLE 1 | The ollowing se en-poin scale was applied o measu ing he ela i e
use o language: 1 = I speak only Basque, 2 = I speak mos ly Basque, 3 = I speak
Basque 75% o he ime, 4 = I speak Basque and Spanish wi h simila equency,
5 = I speak Spanish 75% o he ime, 6 = I speak mos ly Spanish, 7 = only
Spanish.
Rela i e use o language
Be o e p ima y school (0–3 yea s) 6.54 (0.72)
P ima y school (4–12 yea s)
School 2.88 (1.42)
Home 6.54 (0.72)
O he s 5.86 (1.08)
Seconda y school (12–18 yea s)
School 3.58 (1.07)
Home 6.67 (0.56)
O he s 5.92 (0.93)
A ime o es ing
Uni e si y/wo k 4.5 (1.69)
Home 6.63 (0.58)
O he s 5.46 (1.02)
Sel - a ed p o iciency Basque Spanish
Speaking 5.92 (0.58) 6.88 (0.34)
Comp ehension 6.42 (0.5) 6.92 (0.28)
Reading 6.46 (0.51) 6.83 (0.38)
W i ing 6 (0.59) 6.67 (0.48)
P o iciency le el was de e mined by using he ollowing ou -poin scale: 7 = na i e-
like p o iciency,6 = ull p o iciency, 5 = wo king p o iciency, 4 = limi ed p o iciency.
SDs alues a e in pa en heses.
beginning o each sen ence ial. A e each ial he wo ds zuzen?
“co ec ?” o o e ? “inco ec ?” we e displayed in he sc een,
and pa icipan s had o judge he accep abili y o he p e iously
shown sen ence as ei he co ec o inco ec . Hal o pa icipan s
used he le hand o co ec esponses (le C l) and he o he
hal he igh hand ( igh In o).
All sen ences we e andomly dis ibu ed in ou blocks. Each
block las ed app oxima ely 10 min each and pa icipan s had a
sho b eak be ween each block, o as long as hey needed. Be o e
he expe imen began, pa icipan s an a sho aining session
consis ing o h ee ials. They we e ins uc ed o a oid blinking
o mo ing while he sen ences we e being displayed and o make
he accep abili y judgmen as as and accu a ely as possible.
The whole expe imen , including elec ode-cap applica ion and
emo al, las ed abou 1 h 15 m.
EEG Reco ding
The EEG was eco ded om 32 ac i e elec odes secu ed in an
elas ic cap (Ac icap Sys em, B ain P oduc s). Elec odes we e se
on s anda d posi ions acco ding o he ex ended In e na ionals
10–20 sys em acco dingly: Fp1/Fp2, Fz, F3/F4, F7/F8, FC5/FC6,
FC1/FC2, T7/T8, C3/C4, Cz, CP5/CP6, CP1/CP2, P7/P8, P3/P4,
Pz, O1/02, Oz, LM, VEOG and HEOG. All eco dings we e
e e enced o igh mas oid posi ion and e- e e enced o -line
o he linked mas oids. Ve ical and ho izon al eye mo emen s
and blinks we e moni o ed by means o wo elec odes posi ioned
benea h and o he igh o he igh eye. Elec ode impedance
was kep below 5 kOhm a all scalp and below 10 kOhm o he
eye elec odes. The elec ical signals we e digi ized online a a
a e o 500 Hz by a B ain Vision ampli ie sys em and il e ed
o line wi hin a band pass o 0.1–35 Hz. A e he EEG da a
we e eco ded, he ocula co ec ion p ocedu e (G a on e al.,
1983) as well as he a i ac ejec ion p ocedu e we e applied
(o line). T ials wi h o he a i ac s wi h any ol age exceeding
150 µV and ol age s eps be ween wo sampling poin s exceeding
35 µV we e emo ed.
TABLE 2 | Expe imen al condi ions wi h examples o expe imen al ma e ials.
Condi ions Sen ence examples
P edica e ype Fea u e G amma icali y
Unaccusa i e Pe son G amma ical 1. Zu gau goizean buel a u za a Bilbo ik.
you-ABS oday mo ning.in e u ned 2SG.ABS-be Bilbao- om
“You ha e come back om Bilbao his mo ning.”
Ung amma ical 2. *Zu gau goizean buel a u naiz Bilbo ik.
you-ABS oday mo ning.in e u ned 1SG.ABS-be Bilbao- om
Numbe G amma ical 3. Hu a gau goizean buel a u da Bilbo ik.
3.SG-ABS oday mo ning.in e u ned 3SG.ABS-be Bilbao- om
Ung amma ical 4. *Hu a gau goizean buel a u di a Bilbo ik.
3.SG-ABS oday mo ning.in e u ned 3PL.ABS-be Bilbao- om
Une ga i e Pe son G amma ical 5. Zuk goizean biziki su i u duzu au kezpenean.
you-ERG mo ning a.lo su e ed ha e-2SG.ERG p esen a ion- he-a
“You ha e su e ed a lo his mo ning a he p esen a ion.”
Ung amma ical 6. *Zuk goizean biziki su i u du au kezpenean.
you-ERG mo ning a.lo su e ed ha e-1SG.ERG p esen a ion- he-a
Numbe G amma ical 7. Ha k goizean biziki su i u du au kezpenean.
3.SG-ERG mo ning a.lo su e ed ha e-3SG.ERG p esen a ion- he-a
Ung amma ical 8. *Ha k goizean biziki su i u du e au kezpenean.
3.SG-ERG mo ning a.lo su e ed ha e-3PL.ERG p esen a ion- he-a
*s ands o ung amma ical sen ences.
F on ie s in Psychology | www. on ie sin.o g 4No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 5
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
Da a Analysis
Fo he da a analysis ou ypes o subjec ag eemen iola ions
we e compa ed: unaccusa i e pe son iola ions (za a “be.2SG”
s. ∗naiz “be.1SG”; condi ions 1 s. 2 in Table 1, espec i ely);
unaccusa i e numbe iola ions (da “be.3SG” s. ∗di a “be.3PL”;
condi ions 3 s. 4 in Table 1, espec i ely); une ga i e pe son
iola ions (duzu “ha e.2SG” s. ∗du “ha e.1SG”; condi ions 5
s. 6 in Table 1, espec i ely); une ga i e numbe iola ions (du
“ha e.3SG” s. ∗du e “ha e.3PL”; condi ions 7 s. 8 in Table 1,
espec i ely).
Fo he ERP measu es, segmen s we e c ea ed om 200 ms
be o e and 1,000 ms a e he onse o he c i ical wo ds ( he
auxilia y) in he sen ences. The ials associa ed wi h each
sen ence ype we e a e aged o each pa icipan . The EEG
200 ms p io o he onse was also used as a baseline o all
sen ence ype compa isons.
Th ee hund ed o ou hund ed milliseconds and ou
hund ed o se en hund ed milliseconds empo al windows we e
selec ed o s a is ical analysis in all condi ions based on he
li e a u e and isual inspec ion o he da a. A e he s imuli
we e eco ded and a e aged, analyses o a iance (ANOVA) we e
ca ied ou in nine egions o in e es ha we e compu ed ou
o 27 elec odes: la e al elec odes: le on al (F7, F3, FC5), le
cen al (T7, FP5, C3), le pa ie al (P7, P3, O1), igh on al
(F4, F8, FC6), igh cen al (C4, FP6, T8), and igh pa ie al
(P8, P4, O2); midline elec odes: on al (Fp1, Fz, Fp2), cen al
(FC1, Cz, FC2), and pa ie al (CP1, Pz, CP2). Repea ed-measu es
ANOVAs we e conduc ed in all expe imen al manipula ions and
ials (co ec ly and inco ec ly judged ials) o each window
o ime using i e wi hin-subjec s ac o s: g amma icali y (2
le els: g amma ical, ung amma ical), ype (2 le els: unaccusa i e,
une ga i e), ea u e (2 le els: pe son, numbe ), hemisphe e (2
le els: le , igh ), and egion (3 le els: on al, cen al and
pa ie al). Midline ( on al, cen al, and pa ie al) elec odes we e
analyzed independen ly. Whene e he sphe ici y o a iance was
iola ed (G eenhouse and Geisse , 1959) co ec ion was applied
o all he da a wi h g ea e han one deg ee o eedom in he
nume a o . Finally, u he s a is ical compa isons we e ca ied
ou (spli by he g amma icali y condi ion) whene e we ound
a s a is ically signi ican in e ac ion. We only conside e ec s o
he ype, ea u e, hemisphe e o egion ac o s when he e is an
in e ac ion wi h g amma icali y.
Fo he beha io al esul s, e o a es and esponse la encies o
all he ials epea ed measu es ANOVAs we e pe o med wi h
g amma icali y ( wo le els: g amma ical, ung amma ical), ype
( wo le els: unaccusa i e, une ga i e) and ea u e ( wo le els:
pe son, numbe ) condi ions as wi hin-subjec ac o s. Subsequen
compa isons (by subjec and by i em) we e ca ied ou whene e
a g amma ical in e ac ion was signi ican .
RESULTS
Beha io al Resul s
He e, esul s conce ning he accep abili y ask and eac ion imes
a e p esen ed. Pa icipan s we e e y accu a e in he accep abili y
ask (mean accu acy o 91.84%, SDE = 1.3), as was o be expec ed
gi en hei high p o iciency in Basque (see Figu e 1).
Rega ding accep abili y judgmen e o s, he analysis showed
a ma ginally signi ican GRAMMATICALITY e ec in he
analysis by i em [F1(1, 23) = 1.8, p= 0.193; F2(1, 253) = 3.03,
p= 0.083] e ealing highe accu acy o he ung amma ical
sen ences as compa ed o he g amma ical ones (92.74% s.
90.95%). The analysis o accu acy also e ealed a main FEATURE
e ec [F1(1, 23) = 23.4, p<0.001; F2(1, 253) = 24.62, p<0.001]
indica ing ha pa icipan s we e mo e accu a e wi h condi ions
con aining pe son ea u e (94.04%) compa ed o condi ions
con aining numbe ea u e (89.65%).
The GRAMMATICALITY∗FEATURE in e ac ion u ned ou
o be s a is ically signi ican as well [F1(1, 23) = 5.34, p= 0.03;
F2(1, 253) = 3.22, p= 0.074]. The analyses by g amma icali y
ac o showed ha pa icipan s we e signi ican ly less accu a e
wi h g amma ical pe son (92.45%) han wi h ung amma ical
pe son (95.63%) [F1(1, 23) = 5.97, p= 0.023; F2(1, 253) = 7.63,
p= 0.006], whe eas he e we e no di e ences be ween
g amma ical numbe (89.46%) and ung amma ical numbe
(89.85%) [F1(1, 23) = 0.06, p= 0.81; F2(1, 253) = 0.01, p= 913].
The analyses by ea u e ac o showed ha pa icipan s we e
mo e accu a e wi h g amma ical pe son (92.45%) han wi h
g amma ical numbe (89.46%) [F1(1, 23) = 6.31, p= 0.02; F2(1,
253) = 5.74, p= 0.017], and hey we e signi ican ly mo e accu a e
wi h ung amma ical pe son (95.63%) han wi h ung amma ical
numbe (89.85%) [F1(1, 23) = 35.2, p<0.001; F2(1, 253) = 23.56,
p<0.001].
Finally, a iple TYPE∗GRAMMATICALITY∗FEATURE was
signi ican in he analysis by subjec [F1(1, 23) = 8.09,
p= 0.009; F2(1, 253) = 2.77, p= 0.097]. The analyses by
g amma icali y ac o showed ha in unaccusa i es g amma ical
pe son condi ion (94.54%) did no di e om ung amma ical
pe son condi ion (95.16%) [F(1, 23) = 0.19, p= 0.667], and
nei he did g amma ical and ung amma ical numbe (89.98%
s. 90.37%) [F(1, 23) = 0.07, p= 0.788]. In he une ga i e
condi ions pa icipan s we e signi ican ly mo e accu a e wi h
sen ences con aining ung amma ical pe son (96.1%) han wi h
g amma ical pe son (90.37%) [F(1, 23) = 13.32, p= 0.001], bu
no di e ences we e ound be ween g amma ical (88.93%) and
ung amma ical numbe (89.32%) [F(1, 23) = 0.03, p= 0.861].
The analyses by ype ac o e ealed pa icipan s we e mo e
accu a e wi h sen ences con aining g amma ical pe son ea u e
in unaccusa i es (94.54%) han in une ga i es (90.37%) [F(1,
23) = 12.38, p= 0.002], whe eas no di e ences we e ound
be ween sen ences con aining ung amma ical pe son ea u e
in unaccusa i es (95.16%) and in une ga i es (96.1%) [F(1,
23) = 0.66, p= 0.423]. Wi h ega d o numbe ea u e,
no di e ences we e ound be ween g amma ical unaccusa i e
(89.98%) and (88.93%) une ga i e p edica es, and nei he
be ween ung amma ical unaccusa i e (90.37%) and une ga i e
(89.33%) p edica es. Finally, he analyses by ea u e ac o
showed ha pa icipan s we e signi ican ly mo e accu a e wi h
g amma ical unaccusa i e sen ences con aining pe son ea u e
(94.54%) han wi h numbe ea u e (89.98%) [F(1, 23) = 15.89,
p= 0.001], and simila ly ung amma ical unaccusa i e sen ences
con aining pe son ea u e (95.16%) we e judged mo e accu a ely
F on ie s in Psychology | www. on ie sin.o g 5No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 6
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
FIGURE 1 | Pe cen age o co ec esponses (%) and s anda d de ia ion e o (SDE) o non-na i e speake s o Basque.
han numbe ea u e (90.37%) [F(1, 23) = 13.65, p= 0.001].
Rega ding une ga i e p edica es, no di e ences we e ound
be ween g amma ical sen ences con aining pe son and numbe
ea u e [F(1, 23) = 0.74, p= 0.397], bu ung amma ical sen ences
con aining pe son ea u e (96.1%) we e judged signi ican ly mo e
accu a ely han ung amma ical sen ences con aining numbe
ea u e (89.33%) [F(1, 23) = 29.67, p<0.001].
Rega ding esponse imes (see Figu e 2), he analyses e ealed
a main TYPE e ec [F1(1, 23) = 16.41, p= 0.001; F2(1,
253) = 4.21, p= 0.041] indica ing pa icipan s eac ed as e
o unaccusa i e p edica es (668.37 ms) han o une ga i e
p edica es (707.25 ms). A main GRAMMATICALITY e ec
[F1(1, 23) = 71.51, p<0.001; F2(1, 253) = 207.2, p<0.001]
e ealed ha pa icipan s we e signi ican ly as e eac ing
o ung amma ical sen ences (597.93 ms) compa ed o hei
g amma ical coun e pa s (777.68 ms). A FEATURE e ec
[F1(1, 23) = 11.16, p= 0.003; F2(1, 253) = 9.61, p= 0.002]
e ealed ha pa icipan s we e signi ican ly as e esponding o
sen ences con aining pe son ea u e (665.47 ms) han numbe
ea u e (710.15 ms).
ERP Resul s
A e he baseline co ec ion, epochs wi h a i ac s we e
ejec ed, which esul ed in he exclusion o app oxima ely 6.91%
(SD = 2.43) o he ials. Simila ly o he p ocedu e epo ed in
Ma inez de la Hidalga e al. (2019), 300–400 ms. ime window
was selec ed o an ea ly ime window and a 400–700 ms. ime
window was chosen as a la e ime window.
Rega ding he ea ly ime window (300–400 ms), he analysis o
he la e al elec odes e ealed a main GRAMMATICALITY e ec
[F(1, 23) = 18.92, p<0.001] indica ing a la ge nega i i y o he
ung amma ical condi ions as compa ed o he g amma ical ones
(1.08 µV s. 2 µV).
Rega ding he midline elec odes, a main e ec o
GRAMMATICALITY showed ha o e all ung amma ical
condi ions (2.04 µV) displayed a la ge nega i i y han
g amma ical condi ions (2.93 µV) [F(1, 23) = 11.13, p= 0.003].
A signi ican TYPE∗GRAMMATICALITY in e ac ion was
ound [F(1, 23) = 4.9, p= 0.037]. Fu he analysis (by
g amma icali y) showed no signi ican di e ences be ween
ung amma ical (2.36 µV) and g amma ical unaccusa i es (2.79
µV) [F(1, 23) = 2.25, p= 0.147] bu a la ge nega i i y o he
ung amma ical une ga i e condi ion (1.73 µV) in compa ison
o he g amma ical une ga i e condi ion (3.07 µV) [F(1,
23) = 12.67, p= 0.002] was ound. The compa ison by ype
e ealed no di e ences be ween he g amma ical unaccusa i e
(2.79 µV) and une ga i e (3.07 µV) condi ions [F(1, 23) = 0.76,
p= 0.394], and nei he be ween ung amma ical unaccusa i e
(2.36 µV) and une ga i e (1.73 µV) condi ions [F(1, 23) = 1.66,
p= 0.211].
The analysis o he la e al elec odes in he la e ime window
(400–700 ms) e ealed a main GRAMMATICALITY e ec [F(1,
23) = 60.25, p<0.001] indica ing a la ge posi i i y o he
ung amma ical condi ions as compa ed o he g amma ical ones
(2.08 µV s. −0.03 µV). In addi ion, a signi ican main e ec o
FEATURE eme ged [F(1, 23) = 13.47, p= 0.001], indica ing ha
o e all pe son ea u e gene a ed a la ge posi i i y as compa ed
o numbe ea u e (1.44 µV s. 0.61 µV).
A signi ican TYPE∗GRAMMATICALITY in e ac ion was
ound [F(1, 23) = 9.34, p= 0.006]. Fu he analysis (by
g amma icali y) showed a signi ican ly la ge posi i i y o he
ung amma ical unaccusa i e condi ion (2.32 µV) in compa ison
o he g amma ical one (−0.18 µV) [F(1, 23) = 64.23, p<0.001]
and a la ge posi i i y o he ung amma ical une ga i e
condi ion (1.83 µV) in compa ison o he g amma ical
une ga i e condi ion (0.12 µV) [F(1, 23) = 35.3, p<0.001].
The compa ison by ype e ealed no di e ences be ween he
g amma ical unaccusa i e (−0.18 µV) and une ga i e (0.12 µV)
condi ions [F(1, 23) = 1.21, p= 0.282] and no di e ences eme ged
o ung amma ical unaccusa i e manipula ions (2.32 µV) in
F on ie s in Psychology | www. on ie sin.o g 6No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 7
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
FIGURE 2 | Mean eac ion imes (ms) and s anda d de ia ion e o (SDE) o non-na i e speake s o Basque.
compa ison o he une ga i e manipula ions (1.83 µV) [F(1,
23) = 2.28, p= 0.145].
Rega ding he midline elec odes, a main e ec o
GRAMMATICALITY showed ha o e all ung amma ical
condi ions (3.59 µV) displayed a la ge posi i i y han
g amma ical condi ions (0.63 µV) [F(1, 23) = 59.63, p<0.001].
In addi ion, a signi ican FEATURE e ec eme ged [F(1,
23) = 20.94, p<0.001], indica ing ha o e all pe son ea u e
gene a ed a la ge posi i i y as compa ed o numbe ea u e (2.78
µV s. 1.44 µV).
A signi ican TYPE∗GRAMMATICALITY in e ac ion was
ound [F(1, 23) = 13.45, p= 0.001]. Fu he analysis (by
g amma icali y) showed a signi ican ly la ge posi i i y o he
ung amma ical unaccusa i e condi ion (4.03 µV) in compa ison
o he g amma ical one (0.36 µV) (F(1, 23) = 56.96, p<0.001)
and a la ge posi i i y o he ung amma ical une ga i e
condi ion (3.14 µV) in compa ison o he g amma ical numbe
condi ion (0.9 µV) [F(1, 23) = 37.99, p<0.001]. The compa ison
by ype e ealed no di e ences be ween g amma ical une ga i es
(0.9 µV) and unaccusa i es (0.36 µV) [F(1, 23) = 3.02, p= 0.096],
bu a sligh ly la ge posi i i y eme ged o ung amma ical
unaccusa i e manipula ions (4.03 µV) in compa ison o he
une ga i e manipula ions (3.14 µV) [F(1, 23) = 3.91, p= 0.06].
See Figu e 3 o he g and a e age pa e ns, Figu e 4 o
he mean ol age di e ence maps and Table 3 o he
summa y o he esul s.
Na i e and Non-na i e Compa ison
In o de o be e unde s and he simila i ies and di e ences
be ween he non-na i es and he na i e speake s es ed in
Ma inez de la Hidalga e al. (2019), we pe o med an addi ional
analysis compa ing bo h g oups di ec ly.
Beha io al Resul s
Rega ding accu acy, no di e ences be ween bo h g oups we e
ound. A ma ginal main e ec o TYPE eme ged [F1(1, 46) = 2.85,
p= 0.098; F2(1, 252) = 4.18, p= 0.041] indica ing ha o e all,
bo h na i e and non-na i e pa icipan s we e mo e accu a e
wi h condi ions con aining unaccusa i e p edica es (92.76%)
compa ed o une ga i e p edica es (91.82%). The analysis o
accu acy e ealed a signi ican main GRAMMATICALITY
e ec [F1(1, 46) = 4.89, p= 0.032; F2(1, 252) = 13.49,
p<0.001] e ealing ha o e all bo h na i e and non-na i e
pa icipan s we e mo e accu a e wi h condi ions con aining
ung amma ical sen ences (93.41%) compa ed o g amma ical
sen ences (91.16%). The analysis o accu acy also e ealed a
main FEATURE e ec [F(1, 46) = 41.51, p<0.001; F2(1,
252) = 41.5, p<0.001] sugges ing ha bo h na i es and non-
na i es we e mo e accu a e wi h condi ions con aining pe son
ea u e (94.17%) compa ed o condi ions con aining numbe
ea u e (90.4%).
A GRAMMATICALITY∗FEATURE in e ac ion u ned ou
o be ma ginally signi ican in he by subjec analysis [F1(1,
23) = 3.82, p= 0.057; F2(1, 252) = 2.4, p= 0.118]. The analyses by
g amma icali y ac o showed ha pa icipan s we e signi ican ly
less accu a e wi h g amma ical pe son (92.61%) han wi h
ung amma ical pe son (95.73%) [F(1, 46) = 11.57, p= 0.001],
whe eas he e we e no di e ences be ween g amma ical numbe
(89.72%) and ung amma ical numbe (91.08%) [F(1, 46) = 1.18,
p= 0.283]. The analyses by ea u e ac o showed ha pa icipan s
we e mo e accu a e wi h g amma ical pe son (92.61%) han wi h
g amma ical numbe (89.72%) [F(1, 46) = 16.23, p<0.001],
and hey we e signi ican ly mo e accu a e wi h ung amma ical
pe son (95.73%) han wi h ung amma ical numbe (91.08%)
[F(1, 46) = 37.67, p<0.001].
F on ie s in Psychology | www. on ie sin.o g 7No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 8
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
FIGURE 3 | (A) Pe son ea u e unaccusa i e p edica e condi ion; (B) numbe ea u e unaccusa i e p edica e condi ion; (C) pe son ea u e une ga i e p edica e
condi ion; (D) numbe ea u e une ga i e p edica e condi ion.
Finally, a iple TYPE∗GRAMMATICALITY∗FEATURE
in e ac ion u ned ou o be signi ican [F1(1, 46) = 9.28,
p= 0.004; F2(1, 252) = 5.72, p= 0.017]. The analyses by
g amma icali y ac o showed ha pa icipan s we e accu a e
when pe o ming he ask wi h g amma ical and ung amma ical
unaccusa i es con aining pe son ea u e (94.01% s. 95.36%)
[F(1, 46) = 2.32, p= 0.134], and simila ly wi h unaccusa i es
con aining numbe ea u e (89.72% s. 91.93%) [F(1, 46) = 2.14,
p= 0.15]. In he une ga i e condi ions pa icipan s we e
signi ican ly mo e accu a e wi h sen ences con aining
ung amma ical pe son (96.1%) han wi h g amma ical pe son
(91.21%) [F(1, 46) = 15.93, p<0.001], bu no di e ences
we e ound be ween g amma ical (89.72%) and ung amma ical
numbe (90.24%) [F(1, 46) = 0.13, p= 0.716]. The analyses
by ype ac o e ealed ha pa icipan s we e mo e accu a e
wi h sen ences con aining g amma ical pe son ea u e in
unaccusa i es (94.01%) han in une ga i es (91.21%) [F(1,
46) = 9.41, p= 0.004], whe eas no di e ences we e ound
be ween sen ences con aining ung amma ical pe son ea u e
in unaccusa i es (95.36%) and in une ga i es (96.1%) [F(1,
46) = 1.11, p= 0.296]. Wi h ega d o numbe ea u e, no
di e ences we e ound be ween g amma ical unaccusa i e
(89.72%) and (89.72%) une ga i e p edica es [F(1, 46) <0.01,
p= 1], and nei he be ween ung amma ical unaccusa i es
(91.93%) in con as o ung amma ical une ga i e (90.24%)
p edica es [F(1, 46) = 2.16, p= 0.148]. Finally, he analyses by
F on ie s in Psychology | www. on ie sin.o g 8No embe 2021 | Volume 12 | A icle 742127
psyg-12-742127 No embe 3, 2021 Time: 18:15 # 9
Ma ínez de la Hidalga e al. Going Na i e I Allowed by Language Dis ance
FIGURE 4 | Mean ol age di e ence maps (g amma ical minus ung amma ical).
ea u e ac o showed ha pa icipan s we e signi ican ly mo e
accu a e wi h g amma ical unaccusa i e sen ences con aining
pe son ea u e (94.01%) han wi h numbe ea u e (89.72%)
[F(1, 46) = 22.79, p<0.001], and simila ly ung amma ical
unaccusa i e sen ences con aining pe son ea u e (95.36%)
we e judged mo e accu a ely han numbe ea u e (91.93%)
[F(1, 46) = 12.01, p= 0.001]. Rega ding une ga i e p edica es,
no di e ences we e ound be ween g amma ical sen ences
con aining pe son (91.21%) and numbe ea u e (89.72%) [F(1,
46) = 2.18, p= 0.147], bu ung amma ical sen ences con aining
pe son ea u e (96.1%) we e judged signi ican ly mo e accu a ely
han sen ences con aining numbe ea u e (90.24%) [F(1,
46) = 37.89, p<0.001].
The analysis o esponse imes e ealed a main TYPE
e ec [F1(1, 46) = 18.21, p<0.001; F2(1, 254) = 7.68,
p<0.006] indica ing ha pa icipan s eac ed as e o
unaccusa i e p edica es (637.29 ms) han o une ga i e
p edica es (673.08 ms). A main GRAMMATICALITY e ec
[F1(1, 46) = 122.46, p<0.001; F2(1, 252) = 453.86, p<0.001]
e ealed ha pa icipan s we e signi ican ly as e eading
ung amma ical sen ences (565.16 ms) compa ed o hei
g amma ical coun e pa s (745.22 ms). A FEATURE e ec
[F1(1, 46) = 15.48, p<0.001; F2(1, 254) = 11.22, p= 0.001]
e ealed ha pa icipan s we e signi ican ly as e eading
sen ences con aining pe son ea u e (636.77 ms) han numbe
ea u e (673.61 ms). A signi ican TYPE∗GRAMMATICALITY
in e ac ion eme ged [F1(1, 46) = 5.04, p= 0.03; F2(1, 252) = 4.3,
p= 0.039]. The analyses by g amma icali y ac o showed ha
pa icipan s eac ed as e o ung amma ical unaccusa i es
(557.11 ms) han o g amma ical unaccusa i e (717.47 ms)
p edica es [F1(1, 46) = 88.49, p<0.001; F2(1, 252) = 190.11,
p<0.001], and simila ly pa icipan s esponded as e o
ung amma ical une ga i es (573.21 ms) compa ed o hei
g amma ical coun e pa s (772.96 ms) [F1(1, 46) = 105.37,
p<0.001; F2(1, 252) = 274.53, p<0.001]. The analyses by
ype ac o e ealed signi ican di e ences be ween g amma ical
unaccusa i e and une ga i e p edica es [F1(1, 46) = 16.78,
p<0.001; F2(1, 252) = 0.02, p= 0.004], indica ing ha
pa icipan s eac ed as e o g amma ical unaccusa i es
(717.47 ms) han o g amma ical une ga i es (772.96 ms),
bu no di e ences we e ound be ween ung amma ical
unaccusa i es (557.11 ms) and ung amma ical une ga i e
p edica es (573.21 ms) [F1(1, 46) = 2.45, p= 0.124; F2(1,
252) = 0.77, p= 0.3.25].
Finally, a iple TYPE∗GRAMMATICALITY∗FEATURE
in e ac ion u ned ou o be ma ginally signi ican in he by
subjec analysis [F(1, 46) = 3.25, p= 0.078; F2(1, 252) = 3.62,
p= 0.058]. The analyses by g amma icali y ac o showed ha he
unaccusa i e ung amma ical pe son condi ion (550.82 ms) was
ead as e han he g amma ical pe son condi ion (705.5 ms)
[F(1, 46) = 60.54, p<0.001], and simila ly o numbe
(563.4 ms s. 729.44 ms) [F(1, 46) = 54.99, p<0.001]. In
he une ga i e condi ions pa icipan s we e signi ican ly as e
wi h sen ences con aining ung amma ical pe son (532.75)
han wi h g amma ical pe son (757.99 ms) [F(1, 46) = 109.43,
p<0.001], and simila ly hey we e as e wi h ung amma ical
numbe (613.67 ms) han wi h g amma ical numbe (787.93 ms)
[F(1, 46) = 59.2, p<0.001]. The analyses by ype ac o
e ealed pa icipan s we e as e wi h sen ences con aining
g amma ical pe son ea u e in unaccusa i es (705.5 ms) han
in une ga i es (757.99 ms) [F(1, 46) = 8.4, p= 0.006], whe eas
no di e ences eme ged be ween ung amma ical pe son ea u e
in unaccusa i es (550.82 ms) and une ga i es (532.75 ms)
[F(1, 46) = 1.23, p= 0.273]. Wi h ega d o numbe ea u e,
pa icipan s esponded as e o g amma ical sen ences in
unaccusa i es (729.44 ms) han in une ga i es (787.93 ms) [F(1,
F on ie s in Psychology | www. on ie sin.o g 9No embe 2021 | Volume 12 | A icle 742127
