Chap e 3
How di e en ypes o complexi y can
accoun o di icul s uc u es in
bilingual and monolingual language
acquisi ion
Es he Rinkea, C is ina Flo esb& Jacopo To eg ossaa
aGoe he Uni e si y F ank u bUni e sidade do Minho
Ce ain linguis ic s uc u es a e mo e challenging han o he s o bilingual speak-
e s. This is ue ac oss di e en languages and language combina ions. In his pa-
pe , we p opose an accoun in e ms o di e en ypes o linguis ic complexi y. Ou
a gumen a ion de i es om he esul s o a s udy based on a cloze es including
40 di e en linguis ic s uc u es o Eu opean Po uguese (EP). 180 child en pa ici-
pa ed, all o hem acqui ing EP as a he i age language in Swi ze land wi h di e en
en i onmen al languages (60 F ench-EP, 60 (Swiss) Ge man-EP, and 60 I alian-EP
bilinguals). The esul s show ha he s uc u es wi h he lowes accu acy a es a e
he same ac oss he h ee g oups. We single ou ou o hese s uc u es, namely,
(i) que as a subjec ela i e p onoun and as a consecu i e conjunc ion, (ii) hi d
pe son cli ic p onouns in di e en o ms and syn ac ic cons ella ions, (iii) simple
and con ac ed o ms o p eposi ions, and (i ) he in lec ed in ini i e in a conces-
si e cons uc ion. We show ha he di icul y o hese s uc u es e lec s di e en
o ms o linguis ic complexi y: de i a ional complexi y, memo y-based lea ning,
con ex dependency o ules and mul iple o m- unc ion mappings. These o ms
o complexi y cause di icul ies also in monolingual acquisi ion.
Es he Rinke, C is ina Flo es & Jacopo To eg ossa. 2024. How di e en
ypes o complexi y can accoun o di icul s uc u es in bilingual and mono-
lingual language acquisi ion. In Ma ia Polinsky & Michael T. Pu nam (eds.),
Fo mal app oaches o complexi y in he i age languages, 43–71. Be lin: Lan-
guage Science P ess. DOI: 10.5281/zenodo.12090434
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
1 In oduc ion
In his pape , we add ess he ques ion whe he and in which way he di icul-
ies ha he i age speake s (HSs) show wi h ce ain linguis ic s uc u es can be
ela ed o di e en ypes o linguis ic complexi y.
As a s a ing poin , we p o ide he esul s o a s udy based on a cloze es
ocussing on a numbe o di e en s uc u es o Eu opean Po uguese (EP) (To -
eg ossa e al. 2023). The es was comple ed by 180 bilingual child en in he age
span be ween 8 o 16 yea s wi h EP as hei he i age language (HL) and di e en
en i onmen al languages (F ench, Ge man, I alian).1The esul s show ha ce -
ain s uc u es a e pa icula ly di icul o he bilingual child en, whe eas o he s
a e unp oblema ic.
Because we ind a e y simila hie a chy o di icul y ac oss he di e en lan-
guage combina ion g oups, we assume ha he di icul ies encoun e ed by he
child HSs a e, in gene al, independen o he en i onmen al language. The e-
sul s o he abo emen ioned s udy challenge p e ious accoun s which assign
g ea impo ance o c oss-linguis ic in luence as a ac o de e mining de iances
in bilingual p oduc ion (see an Dijk e al. 2022 o a ecen me a-analysis on
c oss-linguis ic in luences in bilingual mo phosyn ac ic acquisi ion o di e se
language pai s).2
Al hough we know ha indi idual child en’s gene al p o iciency is dependen
on age and he amoun o inpu ha hey ecei e in hei HL (in e ms o “quan i y
o language exposu e”), i is s ill an open ques ion why ce ain s uc u es a e
mo e di icul o s abilize han o he s among bilingual as well as monolingual
child en.
1Th oughou he pape , we use he e ms (simul aneous o ea ly) bilingual child en and he i age
speake s o e e o he pa icipan s in ou s udy. By “simul aneous bilinguals”, we e e o he
acquisi ion ype, by “he i age speake s” we e e o he socio-poli ical con ex o acquisi ion.
HLs a e mino i y languages spoken wi hin amilies wi h a mig a ion backg ound. HLs a e, hus,
acqui ed in a bilingual con ex whe e ano he language is he o icial language o he socie y
(majo i y/socie al/en i onmen language). No mally, as consequence o he acquisi ion se ing,
he majo i y language becomes he HSs’ dominan language, bu his is no always he case,
i.e. language (im)balance is no aken as c i e ion o classi y HSs.
2No e ha his is no o say ha CLI does no play any ole in bilingual language acquisi ion.
The a gumen goes he o he way a ound: i we show ha he same s uc u es a e complex
and di icul o monolinguals and bilinguals wi h di e en language combina ions, i becomes
a he unlikely ha CLI is he (one and only) ele an ac o de e mining he di icul ies in he
acquisi ion o hese s uc u es by bilinguals. In any case, i CLI is a gued o be a de e mining
ac o in HL de elopmen , his has o be unequi ocally shown. I does no su ice o poin o
ypological di e ences be ween he wo languages o a bilingual speake .
44
3 How di e en ypes o complexi y in language acquisi ion
We a gue ha he complexi y o he a ge syn ac ic s uc u es is c ucially in-
ol ed in de ining he abo e-men ioned hie a chy o di icul y. Howe e , i is
e y di icul o de ine wha linguis ic complexi y ac ually means, because di -
e en no ions and unde s andings o complexi y exis in he li e a u e. In o de
o app oach ou hypo hesis, we will conside he ou s uc u es ha caused he
mos di icul ies o he child en es ed in To eg ossa e al.’s (2023) s udy when
comple ing he cloze es in hei HL. In pa icula , we will ocus on i) que as a el-
a i e p onoun and consecu i e conjunc ion, ii) cli ic p onouns in di e en o ms
and syn ac ic cons ella ions, iii) simple and con ac ed o ms o p eposi ions and
i ) he in lec ed in ini i e in concessi e sen ences.
In o de o show ha he di icul y o he bilingual child en indeed lies in he
complexi y o he s uc u es (and is no ela ed o bilingualism pe se o c oss-
linguis ic in luence), we will i s demons a e ha he espec i e s uc u es ha
a e di icul o bilingual child en a e also di icul o monolingual ones. In pa -
icula , we assume ha la eness o a linguis ic phenomenon in monolingual ac-
quisi ion indica es i s complexi y o he lea ning/acquisi ion p ocess. Based on
p e ious p oposals abou complexi y in monolingual language acquisi ion, we
a gue ha complexi y is a mul i ace ed no ion. Ou da a allow us o iden i y he
ollowing ypes o complexi y:
i. de i a ional complexi y (laye s o embedding, numbe o mo emen ope -
a ions, ins ances o me ge, e.g., in ela i e clauses)
ii. i egula and lexical o ms ha a e memo y-based (and no ule-based, e.g.,
lexically de e mined selec ion o “ e b+p eposi ion”)
iii. con ex dependen ules (in eg a ion o syn ac ic and discou se knowledge,
allomo phy dependen on phonological con ex , e.g., cli ic allomo phy de-
pending on he phonological con ex , con ac ed o ms o p eposi ions in
combina ion wi h de ini e a icles)
i . mul iple o m- unc ion mappings (e.g., di e en unc ions o que, po , he
use o he in lec ed in ini i e in ce ain concessi e clauses)
2 Empi ical da a: A hie a chy o di icul y in he i age EP
I is a well known ac ha ce ain linguis ic s uc u es cause mo e di icul ies
o bilingual speake s han o he s, pa icula ly in hei non-dominan language
(which is o en, hough no always, he HL). Fo example, bilinguals may show
45
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
mo e p oblems han monolingual speake s wi h phenomena like gende assign-
men and ag eemen (Mon ul e al. 2008), case ma king (Polinsky 2006, 2008),
p onoun ealiza ionand omission (To eg ossa e al. 2019, 2021), cli ic allomo phs
(Rinke & Flo es 2014), subjunc i e (Flo es e al. 2017), and a icle ealiza ion (Mon-
ul & Ionin 2010), jus o men ion a ew.
In o de o de elop a p o iciency assessmen ins umen o EP as HL, we con-
s uc ed a cloze es , p esen ed in de ail in To eg ossa e al. (2023). In gene al,
cloze- es s a e conside ed o be in eg a i e assessmen ools, because he pa ic-
ipan s ha e o access hei linguis ic knowledge o econs uc he missing gap
in he es (Chung & Ahn 2019).
The s udy was conduc ed in Swi ze land, wi h bilingual child en wi h di e -
en language combina ions (Po uguese-F ench, Po uguese-Ge man and Po u-
guese-I alian), as depending on he Swiss can on o esidence.
2.1 Pa icipan s
The s udy included 180 child HSs, 60 child en o each language combina ion.
Mos o he child en we e bo n in Swi ze land o emig a ed he e ea ly in li e. All
pa icipan s acqui ed Po uguese om bi h and he en i onmen al language as
a second i s o ea ly second language. Thei age anged om 8;6 o 16 yea s (M:
11;7; SD: 1;10). The s udy was conduc ed in coope a ion wi h he Camões Ins i u e,
whe e all pa icipan s a ended HL classes weekly. The cloze es was conduc ed
as an un imed w i en ask du ing a HL class.
Swi ze land is an ideal place o conduc his ype o s udy, because he e li es a
ai ly la ge communi y o Po uguese-speaking mig an amilies. Thei child en
acqui e he he i age language, EP, in he con ex o h ee di e en dominan en-
i onmen al languages: F ench/Ge man/I alian. In addi ion, Swi ze land has a
igh ne wo k o Po uguese HL classes, o e ed by he Po uguese Ins i u e o
he main enance and de elopmen o Po uguese ab oad (Ins i u o Camões, see
de Lou des Gonçal es & Vinzen in 2021).
In addi ion o he da a p esen ed in To eg ossa e al. (2023), we collec ed da a
om 23 monolingual Po uguese child en in he ages o 12-13 yea s (M: 12;3; SD:
0;5) o he sake o he p esen discussion. They comple ed an online e sion o
he same cloze es .
2.2 Tes and coding me hodology
The cloze- es is based on a sho na a i e modelled a e he B3 s o y o he
Edmon on Na a i e No ms Ins umen (ENNI; Bonga z & To eg ossa 2020,
46
3 How di e en ypes o complexi y in language acquisi ion
Schneide e al. 2005). The es includes 40 gaps wi h a a ie y o s uc u es ap-
ping in o di e en linguis ic domains: nominal mo phology, e bal mo phology,
(con ac ed and non-con ac ed) p eposi ions, di e en ypes o complemen iz-
e s, (cli ic) p onouns in di e en syn ac ic cons ella ions, de ini e and inde ini e
a icles, and lexical knowledge. Fo unc ional wo ds, we dele ed he whole wo d
o p o ided he ini ial le e in o de o acili a e comple ion and es ic he num-
be o possible answe s. Fo con en wo ds, we p o ided he i s hal o he wo d
(as is usually done in c- es s) o he same easons. The esul s we e coded acco d-
ing o he ollowing ou op ions: co ec , inco ec , missing, o no expec ed bu
co ec . Fo he analysis, we conside ed he co ec and unexpec ed (bu co ec )
answe s as “co ec ” (1) and he inco ec and missing answe s as “inco ec ” (0).
Needless o say, di e en s uc u es can be di icul o di e en easons. In o -
de o be able o di e en ia e be ween s uc u e- ela ed ac o s and o he causes,
we also collec ed in o ma ion on he language backg ound o he child en (age
o onse o he second language, quan i y o inpu , leng h o a endance o HL
classes, e c.). Conce ning he ele ance o hese ac o s we e e he eade o
To eg ossa e al. (2023).
2.3 Resul s
In Table 1 we epo he o e all esul s o monolingual and bilingual child en.
Since he da a collec ion me hod is di e en (online s. in pape o m) and he
monolingual child en’s age ange is mo e limi ed, he esul s ha e o be in e -
p e ed wi h cau ion. None heless, hey p o ide us wi h addi ional e idence e-
la ed o a hie a chy o di icul s uc u es, which we a gue o hold o all child en,
independen ly o hei being monolingual o bilingual.
Table 1: Accu acy a es o monolingual and bilingual child en
Bilinguals Monolinguals
O e all accu acy a e 4635/7200 (64.4%) 843/920 (91.6%)
Max 170/180 (94.4%) 23/23 (100.0%)
Min 51/180 (28.3%) 9/23 (39.1%)
Ac oss all 40 a ge s uc u es, he 180 bilingual child en show an accu acy
a e o 64.4% (4635/7200; max. 170/180 (94.4%)/min. 51/180 (28.3%)). The accu acy
a e o he monolinguals is 91.6% (843/920; max. 23/23 (100%)/min. 9/23 (39.1%)).
47

Es he Rinke, C is ina Flo es & Jacopo To eg ossa
A close look a he esul s e eals ha some o he s uc u es a e indeed pa ic-
ula ly challenging o he child en. The ollowing s uc u es ecei ed he lowes
accu acy a es:
2.3.1 que as a ela i e p onoun and consecu i e conjunc ion
The elemen que has a numbe o di e en unc ions in EP and occu s in di e en
ypes o subo dina ing cons uc ions. I may se e as a complemen ize in oduc-
ing a complemen clause (1a), a ela i e p onoun (1b) o a consecu i e ad e bial
conjunc ion (1c).
(1) a. [i em 18]
Ele
he
pensa
hinks
que
ha
pode
can
i
go
busca
b ing
um
a
balão
balloon
pa a
o
a
he
sua
his
amiga.
iend
‘He hinks ha he can b ing a balloon o his iend.’
b. [i em 12]
Mas
bu
sem
wi hou
que e ,
wan ing
o
he
coelhinho
abbi
la ga
eleases
o
he
balão,
balloon
que
ha
oa
lies
pa a
o
longe.
a away
‘Wi hou wan ing i , he abbi eleases he balloon ha lies away.’
c. [i em 14]
A
he
cadelinha
li le dog
es á
is
ão
so
zangada
ang y
que
ha
começa
s a s
a
o
g i a
shou
(...).
‘The li le dog is so ang y ha she s a s o shou (…).’
Table 2 shows ha he cons uc ions men ioned in (1b) and (1c) ecei ed low
accu acy a es in he bilinguals’ cloze es , which indica es ha hey a e di icul
o he child en.
2.3.2 Thi d pe son cli ic p onouns in di e en o ms and syn ac ic
cons ella ions
Cli ic p onouns in EP a e ma ked o a numbe o di e en mo phological ea-
u es (e.g. gende /numbe /case) and can occu as simple cli ics (2a) o con ac ed
o ms (cli ic allomo phs) (2b–c). Cli ic allomo phs a e allomo phic o ms o he
cli ic p onouns -o(s)/-a(s) ha change o phone ic easons due o he ending o
he e b o m which hey a ach o. Fo ins ance, in examples (2b) and (2c), he
cli ic -o (singula , -os plu al) changes i s o m o -lo(s) because he e b o m ends
wi h he consonan / /.
48
3 How di e en ypes o complexi y in language acquisi ion
Table 2: Accu acy a es o cons uc ions wi h di e en ypes o que-
subo dina o s
Example wi h gap expec ed
i em
g amma ical
ca ego y
accu acy
(bilingual’s)
Mas sem que e , o coelhinho
la ga o balão, oa pa a longe.
(see 1b)
[que] subjec ela i e
p onoun
28.3%
(51/180)
A cadelinha es á ão zangada
começa a g i a e a discu i em
oz al a com o seu amigo. (see
1c)
[que] ad e bial
consecu i e
conjunc ion
49.4%
(89/180)
(2) a. [i em 15]
Assus ado,
sca ed
es e
his-one
ou e-a
hea s-he
a
o
g i a .
c y
‘Sca ed, he hea s he c ying.’
b. [i em 7]
O
he
coelhinho
li le abbi
que
wan s
i á-lo.
ake-i
‘The li le abbi wan s o ake i .’
c. [i em 33]
e
and
pe gun a-lhe,
asks-him
se
i
pode ia
could
ajudá-los.
help- hem
‘And he asks him whe he he could help hem.’
Table 3 (p. 51) shows he accu acy a es associa ed wi h he men ioned s uc u es.
2.3.3 Simple and con ac ed o ms o p eposi ions
Many Po uguese p eposi ions can occu in a con ac ed o m wi h a de ini e
de e mine . Examples a e he p eposi ions em +a/o (in + he em./ he masc.) =
na/no and po +a/o ( o / h ough + he em./ he masc.) = pela/pelo (3a). Besides
p eposi ional ph ases wi h an ad e bial con ibu ion, he p eposi ion po ma ks
he agen o a passi e e b in EP, as shown in (3b). P eposi ions in combina ion
wi h e bs can also lead o a new e b meaning, which is seman ically opaque,
in he sense ha i does no de i e composi ionally om he meaning o he e b
49
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
and he one o he p eposi ion. In example (3c), he combina ion o he e bs i
and e (go + ha e) wi h he p eposi ion com (wi h) leads o he in e p e a ion
‘go o see’.
(3) a. [i em 2]
decidem
decide
i
go
passea
walking
pela
h ough- he
lo es a
o es
‘They decide o go o a walk h ough he o es .’
b. [i em 34]
A
he
mãe
mo he
ou e
lis ens
com
wi h
a enção
a en ion
o
he
ela o
epo
ei o
made
po
by
ele
him
‘The mo he lis ens wi h a en ion o his epo .’
c. [i em 35]
ai
goes
e
ha e
com
wi h
o
he
coelho
abbi
endedo ,
salesman
e
and
pe gun a-lhe
asks-him
pelo
o - he
p eço
p ice
do
o - he
balão.
balloon
‘He goes o see he salesman abbi and asks him o he p ice o he
balloon.’
Table 4 shows ha p eposi ions a e di icul o he bilingual child en, in pa ic-
ula in con ex s like (3a) and (3b).
2.3.4 In lec ed in ini i es in concessi e cons uc ions
EP possesses a special syn ac ic cons uc ion: he in lec ed in ini i e. The con-
s uc ion is ela i ely equen , especially in inal clauses in oduced by he
p eposi ion pa a as in (4a). The in lec ed in ini i e occu s also in concessi e
clauses in oduced by apesa de (‘al hough’, as in 4b).
(4) a. Os
he
pais
pa en s
o am
wen
à
o+ he
li a ia
book s o e
pa a
o
comp a em
buy+3PPl
os
he
li os
school
escola es
books
no os.
new
‘The pa en s wen o he book s o e o buy he new school books.’
b. [i em 28]
Apesa
despi e
de
o
eles
hey
pedi em
ask+3PPl
com
wi h
mui a
much
educação,
educa ion,
...
...
‘Al hough hey asked delica ely, ...’
50
3 How di e en ypes o complexi y in language acquisi ion
Table 3: Accu acy a es o (cli ic) p onouns in di e en o ms and syn-
ac ic cons ella ions
Example wi h gap expec ed
i em
g amma ical ca ego y accu acy
(bilingual’s)
Assus ado, es e
ou e-_ a g i a . (see 2a)
[a] cli ic p onoun
( eminine, singula ,
accusa i e)
40%
(72/180)
O coelhinho que
i á-__ (see 2b)
[lo] cli ic p onoun
(allomo ph, masculine,
singula , accusa i e)
38.8%
(70/180)
e pe gun a-lhe se
pode ia ajudá- (see
2c)
[los] cli ic p onoun
(allomo ph, masculine,
plu al, accusa i e)
44.4%
(80/180)
Table 4: Accu acy a es o (simple and con ac ed o ms o ) p eposi-
ions
Example wi h gap expec ed
i em
g amma ical ca ego y accu acy
(bilin-
gual’s)
decidem i passea p
lo es a (see 3a)
[pela] p eposi ion (con ac ion:
po + a)
40%
(72/180)
A mãe ou e com a enção
o ela o ei o ele, (see
3b)
[po ] p eposi ion (passi e
agen )
42.7%
(77/180)
ai e o coelho
endedo , e pe gun a-lhe
pelo p eço do balão. (see
3c)
[com] p eposi ion (in ixed
e bal exp ession)
53.3%
(96/180)
51
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
p eposi ion aand ha he con ac ion o he p eposi ion po + de ini e a icle
is he mos di icul o acqui e. Fu he mo e, he acquisi ion pa h seems o be
independen o he speake s’ L1.
As o he p eposi ion po , in addi ion o a spa ial meaning, i has also he
pu e g amma ical unc ion o in oducing he agen in passi e sen ences (as by-
ph ase), ei he in con ac ed o m o no , depending on he p esence o absence
o a de ini e a icle, espec i ely. I has been a gued ha by-ph ases o passi e
sen ences a e gene ally p oblema ic o child en (Fox & G odzinsky 1998). This
di icul y may be due o se e al ac o s, including he ype o passi e sen ence
(e.g., long o sho -ac ional passi es; see A mon-Lo em e al. 2016), he agen i i y
o he p edica e (Es ela 2015) and he abo e-men ioned di icul y o child en o
use seman ically acuous p eposi ions.
3.4 In lec ed in ini i es in concessi e cons uc ions
We know om s udies on spon aneous child speech ha in lec ed in ini i es
eme ge ea ly in EP (San os 2017), i.e., by he age o wo yea s. Howe e , a an
ini ial phase, hey only occu in inal clauses in oduced by pa a (San os e al.
2013). Only la e (i.e., by he age o h ee yea s), hey occu in complemen s o
pe cep ion e bs (San os e al. 2016). This means ha e en hough he in lec ed
in ini i e is a ailable o EP child en om ea ly on, he di e en con ex s ha al-
low i s use a e acqui ed g adually, which depends on bo h syn ac ic and lexical
cons ain s. In ac , some con ex s equi ing he use o an in lec ed in ini i e a e
acqui ed e y la e, i.e. in school age. This is he case o he concessi e s uc u e
apesa de (‘al hough’).
Acco ding o Cos a (2006), he concessi e connec o apesa de is s abilized
e y la e in EP (i.e. only by he age o en yea s, simila o he s abiliza ion o al-
hough o whe eas in English, see Diessel 2004). Cos a (2006) a gues ha his la e
acquisi ion is caused by h ee di e en , bu in e ac ing ac o s. The i s ac o
is equency: The connec o s apesa de and embo a a e p oduced signi ican ly
less by adul s han he ad e sa i e connec o mas. Howe e , equency pe se
does no explain he la e acquisi ion o his s uc u e. The la e s abiliza ion o
concessi e connec o s may be ela ed o he ac ha hey occu only in subo -
dina e clauses and mos o hem equi e he use o he subjunc i e, which is also
s abilized la e in EP.
58

3 How di e en ypes o complexi y in language acquisi ion
4 The ole o linguis ic complexi y
The discussion in §3 has shown ha he di e en s uc u es unde conside a ion
a e no only di icul o bilingual child en, bu a e also mas e ed ela i ely la e
by monolinguals. I hese s uc u es ake ime o be acqui ed in monolingual lan-
guage acquisi ion, we expec o ind an e ec o age in he bilingual g oup as
well. Thus, we an a s a is ical analysis o assess he e ec o he bilingual chil-
d en’s age on he acquisi ion o he mos di icul s uc u es. We conside ed he
nine s uc u es which a e ele an o he p esen pape (see i–i in §2). As we
men ioned in §1, he age ange o he pa icipan s is ela i ely la ge (i.e., om
8;6 o 16 yea s; M: 11;7; SD: 1;10). We an a binomial logis ic eg ession wi h accu-
acy as dependen a iable (0 = inaccu a e, 1 = accu a e) and age as ixed e ec .
The model showed a signi ican e ec o age (𝛽 = 0.64, SE = 0.06,𝑧 = 10.30,
𝑝<0.001). This shows ha bilingual child en’s knowledge o di icul s uc u es
imp o es wi h age. In his sense, bilinguals beha e jus like monolinguals, e en
i hey may need mo e ime o acqui e di icul s uc u es. In his sense, i is pos-
sible ha he s uc u es ha a e no mas e ed by younge bilingual child en a e
exac ly he s uc u es ha eme ge la e in monolingual language acquisi ion. In
o he wo ds, hese s uc u es a e ‘complex’ o bilinguals and monolinguals alike,
as shown by hei la e iming o acquisi ion ac oss he boa d. Since i is o en ob-
se ed ha bilinguals show a mo e p o ac ed de elopmen , i.e., hey acqui e
some s uc u es in la e age spans han monolinguals, we assume ha bilinguals
jus need some mo e ime o ca ch up wi h hei monolingual pee s (see Schulz
& G imm 2019, Tsimpli 2014 o simila conside a ions). In he emainde o his
pape , we in end o discuss why ce ain s uc u es a e associa ed wi h a mo e
complex lea ning ask han o he s.
4.1 No ions o linguis ic complexi y
In he li e a u e, complexi y in acquisi ion has been explici ly de ined and implic-
i ly assumed in many di e en ways. F om a syn ac ic pe spec i e, i has been
assumed ha child en ini ially p e e mo e syn ac ically economical s uc u es
o e less economical ones; i.e. s uc u es in ol ing less laye s o embedding o e
s uc u es in ol ing mo e laye s o embedding, o s uc u es in ol ing less mo e-
men ope a ions o e s uc u es in ol ing mo e mo emen ope a ions (Hamann
2006, Rizzi 1990, 2000). Jakubowicz (2003) p oposes ha compu a ional complex-
i y a ec s child language de elopmen , leading child en o p oduce less complex
s uc u es in a a ge -like way ea lie han mo e complex s uc u es (see also
Jakubowicz & Nash 2001). The au ho de elops he ollowing De i a ional Com-
plexi y Me ic.
59
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
(5) De i a ional Complexi y Me ic (DCM, Jakubowicz 2011)
• Me ging 𝛼𝑖𝑛 imes gi es ise o a less complex de i a ion han
me ging 𝛼𝑖(𝑛+1) imes.
• In e nal Me ge o α gi es ise o a less complex de i a ion han
In e nal Me ge o 𝛼+𝛽.
Fo example, wi h espec o wh-ques ions, he DCM p edic s, “ ha he child
is sensi i e o he numbe o imes ha a copy o he wh-elemen mus be me ged
o sa is y a compu a ional equi emen and o he numbe o cons i uen s ha
may (o mus ) unde go In e nal Me ge (he e unde : IM)” (Jakubowicz 2011: 340;
see also Soa es 2003 wi h espec o he acquisi ion o wh-ques ions in EP).
The no ion o complexi y p esen ed so a is mo i a ed syn ac ically. Ano he
way o de ining complexi y is mo e mo phologically o ien ed and based on he
obse a ion ha child en end o o e egula ize mo phological endings. Clahsen
e al. (2002) a gue o a dual-mechanism model be ween ule-based ( egula ) and
memo y-based (i egula ) ep esen a ions o mo phologically complex wo ds.
In hei s udy, child en acqui ing Spanish e b mo phology o e apply egula
pa adigms o e bs ha equi e i egula o ms bu no ice e sa. The au ho s
a gue ha “... he onse o o e egula iza ions is syn ac ically igge ed, by he
equi emen o gene a e a ully speci ied ini e e b o m in e e y sen ence, in
conjunc ion wi h lexical gaps o e ie al ailu es o i egula s. O e egula iza-
ions g adually dec ease o e ime when child en ge olde and memo y aces
o i egula s a e becoming s onge and he child en’s abili y o e ie e hem
is becoming mo e eliable” (Clahsen e al. 2002: 618). Coming back o he issue o
complexi y in acquisi ion, hese esul s sugges ha egula syn ac ic o mo pho-
logical ules a e less complex han i egula o ms, which ha e o be memo ized
and s o ed in he lexicon based on indi idual o ms (and hei equency) in he
inpu . Hence, he acquisi ion o ules ha a e applied egula ly seems o be less
cos ly han memo y-based lexical lea ning.
The mo phological ule men ioned in he p e ious example is based on a syn-
ac ic equi emen (namely o gene a e a ully speci ied ini e e b o m) ha
applies independen ly o he con ex (i.e., he si ua ion in which he sen ence
is u e ed) and, in p inciple, conce ns e e y sen ence. Howe e , his is no he
case o each mo phological o syn ac ic ule. We would like o add ano he ype
o complexi y which lies in-be ween ule-based egula and memo y-based i -
egula ep esen a ions, namely cases in which a ule is applied depending on a
speci ic (discou se o phonological) con ex . We sugges ha his also in ol es a
wo-s ep lea ning/acquisi ion p ocess: acqui ing he ule and unde s anding in
which con ex i applies and in which con ex i does no .
60
3 How di e en ypes o complexi y in language acquisi ion
One example ha has been men ioned in a numbe o s udies is con ex depen-
dency o a o m which is ela ed o p e ious discou se. In languages wi h null and
o e p onouns, his conce ns, o example, he decision whe he a p onoun has
o be o e ly ealized o can emain phonologically null. I has been sugges ed
ha in null subjec languages, bilingual speake s end o o e ealize p onouns
compa ed o monolingual speake s and may ail o accu a ely di e en ia e be-
ween he wo o ms in in e p e a ion asks. So ace e al. (2009: 464) a gue ha
his is a esul o he complexi y o he ask: bilinguals ha e mo e di icul ies in
in eg a ing di e en sou ces o in o ma ion. Acco ding o So ace (2011), he di -
e ences be ween monolingual and bilingual popula ions ela e o bilingualism
pe se and, in pa icula , o he alloca ion o gene al cogni i e esou ces o bilin-
gual p ocessing. Howe e , he complexi y o in eg a ing syn ac ic in o ma ion
and discou se in o ma ion ep esen s a complex lea ning ask also o young
monolinguals (as shown o Po uguese by Lobo & Sil a 2016, Rinke & Flo es
2018) and may esul in a p o ac ed de elopmen o such phenomena. Fo exam-
ple, Tulle e al. (2011) obse e ha in F ench, 3 d pe son accusa i e cli ics a e
di icul among young TD (= ypically de eloping) child en and AD (a ypically
de eloping) speake s a e childhood. The au ho s claim ha he
complexi y o objec cli ics is he esul o a combina ion o se e al p op-
e ies, he i s o which is hei non-canonical posi ion. […] Summa iz-
ing, he p oduc ion o accusa i e cli ics includes he ollowing p ope ies:
mo emen o a non-a gumen posi ion, clus e ing wi h nomina i e cli ics,
and e e ence o a non-local an eceden . P oduc ion o a hi d pe son ac-
cusa i e cli ic in ol es he ollowing addi ional p ope ies: es ablishing non-
discou se-dependen e e ence, ag eemen in bo h numbe and gende , bu
no animacy, and, po en ially, licensing o a null cli ic (condi ional on bo h
lexical and discou se es ic ions). They a e hus complex (mo pho)syn ac-
ically, in e ms o mo emen (whiche e analysis o cli ic cons uc ions is
adop ed) and ag eemen , and mas e ing hei usage (knowing whe he hey
can be null o no ) equi es adhe ing o lexical idiosync asies and discou se/
p agma ic condi ions. (Tulle e al. 2011: 427 .)
A simila obse a ion applies o 3 d pe son objec cli ics in EP, whose p oduc-
ion is associa ed wi h he same deg ee o complexi y as asc ibed by Tulle e al.
(2011) o F ench cli ics. In addi ion, EP allows o 3 d pe son null objec s in sim-
ila syn ac ic and discou se con ex s as cli ics. The e o e, he acquisi ion o he
a ge -like dis ibu ion o objec cli ics and null objec s in a null objec language
61
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
like EP and, hence, o he discou se-app op ia e p oduc ion o cli ics is mo e chal-
lenging han he acquisi ion o cli ics in non-null objec languages (Cos a & Lobo
2007, Flo es e al. 2020, Va lokos a e al. 2016).
In addi ion, EP cli ics show allomo phy in ce ain phonological con ex s, as de-
sc ibed abo e. Allomo phic a ia ion ep esen s ano he o m o linguis ic com-
plexi y. I has been shown, o example, ha allomo phic a ia ion o English
pas ense o ms (e.g. “- o e bs such as chase, -d o o ms such as c a e and
/əd/ o e bs such as eci e”) slows down mo phological de elopmen (O’G ady
e al. 2010: 369). O’G ady e al. (2010) also men ion homophony as a ac o de e -
mining mo phological de elopmen in i s language acquisi ion.
Whe eas he wo d he unc ions only as a de e mine in English, he su ix
-s can be used o ma k any one o h ee hings: plu al numbe in nouns,
hi d pe son singula in e bs, o possession. The esul ing complica ion in
he ela ionship be ween o m and meaning may impede acquisi ion.
(O’G ady e al. 2010: 369)
We assume ha in gene al, mul iple o m- unc ion mappings (e.g. allomo phy,
homophony) gi e ise o complexi y in acquisi ion and may cause di icul ies o a
slow down in de elopmen . To conclude, we iden i ied he ollowing ypes o lin-
guis ic complexi y in i s language acquisi ion: i) de i a ional complexi y (laye s
o embedding, numbe o mo emen ope a ions, ins ances o me ge); ii) i egula
and lexical o ms ha a e memo y-based (and no ule-based); iii) con ex depen-
den ules (in eg a ion o syn ac ic and discou se knowledge o allomo phy de-
pending on phonological con ex ) and i ) mul iple o m- unc ion mappings. In
he nex sec ion, we will discuss how hese no ions o complexi y apply o he
“hie a chy o di icul y” discussed in §2.
4.2 Towa ds an explana ion o he hie a chy o di icul y
In his sec ion, we would like o come back o he phenomena men ioned in §2
and §3 ha we e he mos challenging linguis ic s uc u es o he child en and
explo e o wha ex en hei di icul y can be ela ed o he abo e men ioned
no ions o linguis ic complexi y.
As al eady discussed in §2, he i em que as a ela i e p onoun and as a con-
secu i e complemen ize belonged o he cons uc ions wi h he lowes accu acy
a es ac oss he di e en language combina ion g oups. I is in e es ing o con-
as hese wo s uc u es wi h he decla a i e complemen ize que illus a ed
in (1a), which is associa ed wi h a high accu acy a e o 70.5% ( s. 28.3% o he
62
3 How di e en ypes o complexi y in language acquisi ion
ela i e p onoun and 49.4% o he consecu i e complemen ize ). As shown in
§3, he di e en accu acy a es o he di e en ypes o que co espond o he
o de o acquisi ion o he di e en ins an ia ions o que in monolingual EP: he
decla a i e complemen ize is acqui ed i s in child EP, ollowed by que in o-
ducing ad e bial clauses, ollowed in u n by ela i e clauses, some o which
may also eme ge a school age. E en he 12–13 yea s-old monolingual child en
showed low a es o accu acy in associa ion wi h he ela i e p onoun que. In
§3, we men ioned ha he di e ence be ween he a ious ypes o subo dina e
clauses (complemen clauses selec ed by he e b, adjoined ad e bial clauses and
ela i e clauses) can be accoun ed o in e ms o deg ees o de i a ional complex-
i y, in ol ing, o example, embedding (in concessi e clauses) o embedding and
mo emen (e.g., in ela i e clauses). An addi ional ac o con ibu ing o he com-
plexi y o he s uc u es a s ake is he mul iple o m- unc ion mapping o que
in EP (one o m wi h se e al unc ions), namely he homophony o que as a con-
junc ion o complemen and ad e bial clauses, as an in e oga i e o a ela i e
p onoun o an in e oga i e de e mine .
Thi d pe son cli ic p onouns ep esen ano he a ea o di icul y in he cloze
es s among he bilingual child en. As men ioned in §3.2, hese s uc u es a e also
e y challenging o EP monolingual child en and acqui ed success ully only a
school age. Cli ics a e di icul o a numbe o di e en easons. In addi ion o
po en ial (syn ac ic) de i a ional complexi y (i we assume a mo emen analysis
o cli ics), cli ics a e mo phologically complex because hey in ol e allomo phy
in EP. Depending on he phonological con ex , he o m o he cli ic may change.
Fo example, ollowing he – ending o in ini i es, he cli ic -o (acc. masc. sing.)
is ealized as –lo (see example 2b, c); a e he nasal –m (e.g., 3 d pe son plu-
al ini e e b o ms), osu aces as –no. As discussed in he p e ious sec ion,
such ules a e complex o di e en easons: hey a e con ex dependen ( he e-
o e in ol ing a wo s ep lea ning p ocess) and he e is no di ec o m- unc ion
mapping (because di e en o ms ha e he same unc ion and ealize he same
mo phological ea u es). A hi d ac o con ibu ing o he complexi y o cli ic
p onouns is hei discou se dependency, since he app op ia e use o cli ics (as
well as null objec s and ull noun ph ases) is dependen on discou se cons ain s
(Flo es e al. 2020).
The hi d phenomenon discussed in §2 and §3 a e p eposi ions in di e en
shapes and cons ella ions. We saw ha in he cloze es , bilingual child en, bu
also monolinguals, show low accu acy wi h con ac ed o ms o p eposi ions.
In addi ion, he bilingual child en ha e also p oblems wi h he p eposi ion po
in oducing passi e agen s and he lexically selec ed p eposi ion com. Fi s o
all, con ac ed o ms o p eposi ions a e de i ed based on a con ex -dependen
63

Es he Rinke, C is ina Flo es & Jacopo To eg ossa
ule (only in combina ion wi h de ini e a icles, no wi h inde ini e ones o ba e
nouns). Assuming a Dis ibu ed Mo phology app oach, Ximenes (2004) s a es
ha con ac ions o p eposi ions a e he esul s o a wo-s ep mo phological p o-
cess: “ wo ope a ions happening in he mo phological componen : me ge ol-
lowed by usion.” (Ximenes 2004: 182). As al eady discussed in §3, po as a p eposi-
ion ma king he agen o a passi e sen ence is gene ally p oblema ic o younge
child en and complex, because i is a unc ional and, hence, a seman ically ac-
uous p eposi ion. The homophony wi h he lexical p eposi ion po ma king a
spa ial meaning leads o mul iple o m- unc ion mapping and may con ibu e o
he complexi y o his p eposi ion as well. The complexi y o he p eposi ion com
in combina ion wi h he e b e (see example 3c) has a di e en sou ce. In his
con ex , he p eposi ion con ibu es o he o ma ion o a new e b – a p ocess
ha is e y p oduc i e in EP (e.g. acaba de ‘ inish (o )’, acaba com ‘des oy’,
acaba po ‘end up by’). C ucially, he combina ion o a e b and a p eposi ion is
seman ically opaque and can only be acqui ed h ough a memo y-based lexical
lea ning p ocess.
The ou h phenomenon ha was associa ed wi h some di icul y o he bilin-
gual child en in he cloze es was he in lec ed in ini i e in combina ion wi h he
concessi e connec o apesa de. As men ioned in §3, EP monolingual child en do
no exhibi any di icul y in he use o in lec ed in ini i es. Howe e , concessi e
connec o s a e acqui ed la e and exhibi a simila deg ee o complexi y as o he
conjunc ions in oducing ad e bial ph ases. When in oducing a clause, apesa
de occu s only in combina ion wi h in lec ed and unin lec ed in ini i es. In mo e
o mal egis e s, we ind he (mo e complex) cons uc ion apesa de que, which
in oduces ini e subo dina e clauses ha equi e he indica i e o he subjunc-
i e mood, which is ano he p ope y o EP which is acqui ed ela i ely la e in
L1 acquisi ion. In addi ion o hese o ms ha belong o he s anda d egis e , we
ind also he occu ence o apesa que in associa ion wi h he indica i e mood
in colloquial speech. Fu he mo e, apesa de may also in oduce a NP wi h con-
cessi e meaning, ins ead o a clause (e.g., Apesa da chu a, eles o am passea .
‘Despi e he ain hey wen o a walk.’). Hence, he di icul y ela ed o he use
o he in lec ed in ini i e in he cloze es does no depend on he s uc u e i sel ,
bu esul s om i s combina ion wi h he concessi e connec o apesa de, which
is acqui ed la e and can in oduce di e en s uc u es. In pa icula , i s al e na-
ion wi h a ini e e b in he same con ex , as in he use o he indica i e wi h he
non-s anda d apesa que, may inc ease he di icul y o he acquisi ion ask. Ac-
ually, he mos equen e o commi ed wi h his i em was he eplacemen o
he in lec ed in ini i e wi h he ini e 3 d pe son plu al indica i e o m pedi am.
64
3 How di e en ypes o complexi y in language acquisi ion
5 Summa y and conclusions
The p esen line o a gumen a ion de i es om he obse a ion ha some linguis-
ic s uc u es cause mo e di icul ies o bilingual speake s han o he s, especially
in hei (non-dominan ) HL. We aimed o show ha he di icul y o ce ain s uc-
u es is ela ed o di e en ypes o linguis ic complexi y. A cloze es conduc ed
wi h 180 EP he i age child en in Swi ze land – di ided in o h ee g oups wi h 3
di e en en i onmen al languages (F ench, Ge man and I alian) – e ealed ha
he child en exhibi ed pa icula di icul ies wi h some s uc u es, including el-
a i e p onouns and consecu i e conjunc ions, cli ic p onouns in di e en o ms
and syn ac ic cons ella ions, some simple and con ac ed o ms o p oposi ions
and in lec ed in ini i es in concessi e cons uc ions.
T iangula ing hese indings wi h he exis ing li e a u e on he L1 acquisi ion
o he s uc u es a issue, we we e able o conclude ha he s uc u es ha child
HSs ound he mos di icul we e exac ly he s uc u es ha usually eme ge la e
(o e y la e) in monolingual language acquisi ion. This was also con i med by
a small scale s udy conduc ed on Po uguese monolingual child en anging be-
ween 12 and 13 yea s, based on he same cloze- es as he one adminis e ed o he
bilinguals. Also o he monolinguals, ela i e p onouns, con ac ed p eposi ions
and cli ics we e associa ed wi h he lowes accu acy sco es.
O e all, hese esul s sugges ha he challenging s uc u es o bilingual chil-
d en ep esen a complex lea ning ask also in monolingual language acquisi ion.
In o he wo ds, child HSs acqui e mo phosyn ac ic s uc u es h ough he same
miles ones as hei monolingual pee s, al hough hey may lag behind in some
linguis ic domains ha equi e mo e inpu o be success ully acqui ed. No ably,
we also ound ha he accu acy in he use o hese s uc u es imp o ed wi h
age, highligh ing a de elopmen al end among he bilinguals. In addi ion, hese
esul s do no sus ain he assump ion ha CLI is he main ac o con ibu ing o
de elopmen al di e ences be ween he i age and monolingual child en.
The main con ibu ion o he p esen pape consis s in showing ha he di -
icul ies exhibi ed by he bilingual child en canno be accoun ed o in e ms o
a single no ion o complexi y. Ra he , di e en s uc u es may be complex in
acquisi ion o di e en easons. In pa icula , we iden i ied ou main no ions
o complexi y, as ela ed o he di e en s uc u es analysed in his con ibu-
ion, i.e., de i a ional complexi y, memo y-based lexical o ms, ules dependen
on phonological o discou se con ex s and mul iple o m- unc ion mappings. In
his sense, we mo ed away om he a emp o p o ide a single de ini ion o com-
plexi y, bu a he p oposed a mul i ace ed iew o his no ion, which ma ches
wi h ex ensi e esea ch on language acquisi ion.
65
Es he Rinke, C is ina Flo es & Jacopo To eg ossa
Re e ences
A mon-Lo em, Sha on. 2005. The acquisi ion o subo dina ion: F om p econjunc-
ionals o la e use. In Do i Ra id & Ha a Ba -Zee Shyldk o (eds.), Pe spec-
i es on language and language de elopmen , 192–204. Do d ech : Kluwe .
A mon-Lo em, Sha on, Ewa Haman, K is ine Jensen de López, Magdalena
Smoczynska, Kazuko Ya sushi o, Ma cin Szcze binski, Angeliek an Hou ,
Ine a Dabašinskienė, Anna Ga a ó, E in Hobbs, Lau a Kamanduly ė-
Me eldienė, Napoleon Ka sos, Sa i Kunna i, Ch isa Ni siou, Lone Sundahl
Olsen, Xa ie Pa amon, Uli Saue land, Reeli To n-Leesik & Hea he an de
Lely. 2016. A la ge-scale c oss-linguis ic in es iga ion o he acquisi ion o pas-
si e. Language Acquisi ion 23(1). 27–56. DOI: 10.1080/10489223.2015.1047095.
(1080).
Ba alha, Joana. 2018. Relações en e o conhecimen o explíci o da língua e a com-
pe ência de lei u a. No a Uni e si y o Lisbon. (Doc o al disse a ion).
Bloom, Lois, Ka in Li e & Je emie Ha i z. 1980. Seman ics o e bs and he de-
elopmen o e b in lec ion in child language. Language 56. 386–412.
Bonga z, Ch is iane & Jacopo To eg ossa. 2020. The e ec s o balanced bili -
e acy on G eek-Ge man bilingual child en’s seconda y discou se abili y. In-
e na ional Jou nal o Bilingual Educa ion and Bilingualism 23. 948–963. DOI:
10.1080/13670050.2017.1355888.
Bowe man, Melissa. 1979. The acquisi ion o complex sen ences. In Paul Fle che
& Michael Ga man (eds.), Language acquisi ion, 285–306. Camb idge: Cam-
b idge Uni e si y P ess.
B i o, Ed an P. 2018. The a ia ion o obliga o y p eposi ion-a icle con ac ions
in he in e language o adul lea ne s o Po uguese. Re is a do GEL 15. 241–
256.
Cha neca Ca alão, M. Flo en ina. 2011. Os pad ões de uso dos p onomes pessoais
à onos em po uguês eu opeu. Uni e si y o É o a. (MA hesis).
Chung, Eun Seong & Soojing Ahn. 2019. Examining cloze es s as a measu e o
linguis ic complexi y in L2 w i ing. Language Resea ch 55. 627–649.
Clahsen, Ha ald, F aibe A eledo & Iggy Roca. 2002. The de elopmen o eg-
ula and i egula e b in lec ion in Spanish child language. Jou nal o Child
Language 29(3). 591–622. DOI: 10.1017/s0305000902005172.
Cos a, João, Alexand a Fiéis & Ma ia Lobo. 2015. Inpu a iabili y and la e acqui-
si ion: Cli ic misplacemen in Eu opean Po uguese. Lingua 161. 10–26. DOI:
10.1016/j.lingua.2014.05.009.
66
3 How di e en ypes o complexi y in language acquisi ion
Cos a, Ana Luísa. 2006. Complexidade es u u al de conec o es concessi os. Tex-
os Selecionados do XXII Encon o da Associação Po uguesa de Linguís ica.
Lisboa: APL.
Cos a, João & Ma ia Lobo. 2007. Cli ic omission, null objec s o bo h in he ac-
quisi ion o Eu opean Po uguese. In Se gio Baauw, F ank D ijkoningen &
Manuela Pin o (eds.), Romance Languages and Linguis ic Theo y 2005, 59–72.
Ams e dam, The Ne he lands: John Benjamins. DOI: 10.1075/cil .291.06cos.
Cos a, João & Ma ia Lobo. 2009. Cli ic omission in he acquisi ion o Eu o-
pean Po uguese: Da a om comp ehension. In Ac isio Pi es & Jason Ro h-
man (eds.), Minimalis inqui ies in o child and adul language acquisi ion: Case
s udies ac oss Po uguese, 63–84. Be lin, Ge many: Mou on de G uy e . DOI:
10.1515/9783110215359.1.63.
Cos a, João, Ma ia Lobo & Ca olina Sil a. 2009. Null objec s and ea ly p agma ics
in he acquisi ion o Eu opean Po uguese. P obus 21(2). 143–162. DOI: 10.1515/
p bs.2009.005.
Cos a, João, Ma ia Lobo & Ca olina Sil a. 2012. Which ca ego y eplaces an omi -
ed cli ic? The case o Eu opean Po uguese. In P. La añaga & P. Guija o-
Fuen es (eds.), P onouns and cli ics in ea ly language, 105–129. Be lin, Ge many:
Mou on de G uy e .
Cos a, João, Ma ia Lobo & Ca olina Sil a. 2011. Subjec –objec asymme ies in he
acquisi ion o Po uguese ela i e clauses: Adul s s. child en. Lingua 121(6).
1083–1100. DOI: 10.1016/j.lingua.2011.02.001.
de Lou des Gonçal es, Ma ia & Mi iam Vinzen in. 2021. Po uguese in Swi ze -
land: In es iga ing he ep esen a ions o amilies and eache s. In A. Souza &
S. Melo-P ei e (eds.), Po uguese as a he i age language in Eu ope: A plu icen-
ic pe spec i e, 119–150. Campinas, B azil: Pon es Edi o es.
Diessel, Holge . 2004. The acquisi ion o complex sen ences. Camb idge, UK: Cam-
b idge Uni e si y P ess.
D omi, Es he & Ru h A. Be man. 1986. Language-speci ic and language-gene al
in de eloping syn ax. Jou nal o Child Language 13. 371–381.
Es ela, An ónia. 2015. The acquisi ion o he passi e s uc u e in Eu opean Po -
uguese. Jou nal o Child Language Acquisi ion and De elopmen 3(3). 153–168.
Flo es, C is ina, Es he Rinke & Aldona Sopa a. 2020. Acqui ing he dis ibu ion
o null and o e di ec objec s in Eu opean Po uguese. Jou nal o Po uguese
Linguis ics 19(5). 1–20. DOI: 10.5334/jpl.239.
Flo es, C is ina, Ana Lúcia San os, Alice Jesus & Rui Ma ques. 2017. Age and
inpu e ec s in he acquisi ion o mood in he i age Po uguese. Jou nal o Child
Language 44(4). 795–828. DOI: 10.1017/S030222.
67