additional clusters, including /sl/,/sm/,/sn/,/skw/,/spl/,/spr/ and /str/,
had not yet reached the 75% acquisition threshold by 5;0. Comparatively
late acquisition is also apparent in other studies on cluster acquisition in
monolingual English-speaking children (see McLeod et al., 2001 for an over-
view). What these findings suggest then is that the bilingual environment
in Wales may have had a facilitative effect on phonological acquisition, per-
haps because children who routinely need to differentiate two distinct pho-
nological codes require enhanced phonological knowledge. The findings
obtained here tie in with Grech and Dodd’s (2008) study of Maltese-English
bilingual children which found that even Maltese children using only one
language at home had a significantly higher percent consonant correct score
than comparable children in monolingual settings from countries where
English is the dominant language.
Cluster acquisition and language dominance
The results revealed better performance and earlier acquisition of the
Welsh clusters by the Welsh-dominant bilinguals than the English-dominant
bilinguals, with significant differences on /sɡr/,/ɡwr/ and /ɡwn/. This sug-
gests that the comparatively greater amount of Welsh-language input has had
a facilitative effect on cluster acquisition in the Welsh-dominant children.
On the basis of these findings, one would expect the English-dominant
bilinguals to outperform their Welsh-dominant counterparts on the English
clusters. Indeed, they were significantly more accurate on /sp/,/sw/ and
/skw/. However, close inspection shows that the difference between the
groups was marginal, with close to 100% accuracy in both. Moreover, the
Welsh-dominant bilinguals were significantly more accurate than the
English-dominant ones on /fr/ and /skr/, and attained the acquisition
threshold earlier on many English clusters. Together, the findings suggest an
asymmetry in cluster acquisition with considerably better performance of
the Welsh-dominant bilinguals on the Welsh clusters, without, however, a
concomitant lag on the English ones. These results are comparable to Munro
et al.’s (2005) study of singleton consonant acquisition in Welsh-English
bilingual children which yielded similar findings. Asymmetries in phono-
logical acquisition were also found in Law and So’s (2006) study of
Cantonese-Putonghua bilinguals.
A possible explanation for these findings is that the Welsh-dominant
children shared certain characteristics, such as socioeconomic background,
that might have put them at an advantage over their English-dominant coun-
terparts. Previous studies have indeed found delayed acquisition patterns in
children from deprived social backgrounds (Templin, 1957). However,
according to the second author who is native to the rural community in
Ceredigion from which the participants were recruited, there are no system-
atic differences in socioeconomic background across the Welsh-dominant
and English-dominant children in the study.
Speech Learning in Bilinguals 21