in the German of the late consecutive bilinguals than in the German of the
control group (see Figure 2.1). However, in the group analyses, the end of the
prenuclear rise did not occur earlier in the German of the bilinguals than in
that of the German control group (de Leeuw et al., 2012).
When the late consecutive bilinguals were analysed individually, it was
found that the alignment of both the start and the end of the prenuclear rise
were similar in the L1 and the L2 of most bilingual migrants. In particular,
two participants displayed a merged prenuclear rising accent in their native
German and English which was within the English monolingual norm: these
participants most clearly evidenced L1 attrition. In contrast, only one partici-
pant displayed a prenuclear rising accent which was conducive to the mono-
lingual norms in both her German and English. Moreover, in the analysis of
predictor variables, an earlier age of arrival significantly correlated with an
earlier alignment of the prenuclear rise: late consecutive bilinguals who
moved to Canada at a younger age were more likely to perform within the
English monolingual norm in their German than those who arrived to
Canada at a later age, although all participants arrived between 16 and 32
years of age. In sum, these findings revealed L1 attrition in the native German
Maturational Constraints in Bilingual Speech 31Figure 2.1 Bar chart of alignment of start of prenuclear rise (as measured from the
onset of the vowel) in milliseconds (ms) against group (GM = German monolingual con-
trols; GB = German of bilinguals; EB = English of bilinguals; EM = English of monolingual
controls). Positive values indicate a prenuclear rise which started after the onset of the
vowel, whereas negative values indicate a prenuclear rise which started before the
onset of the vowel