categories, which includes national identities such as Venezuelan, Dominican,
and Cuban-American. These differences between the Spanish-English and
Welsh-English datasets were statistically significant (χ^2 = 126.47, df = 3,
p < 0.0001). Overall it seems that the uniform Welsh identity is related to the
choice of Welsh ML in Wales whereas in Miami a more diverse range of iden-
tities is related to the more diverse choice of Spanish (66%) and English (34%)
ML shown in Figure 6.1.
Languages of social network
This result is based on participants’ reports of the languages used with
their five closest contacts. Figure 6.8 shows a clear difference between the
two communities. In the figure a score of three means the participants have
an English-speaking social network, two means that both languages are
used, and one means that primarily Spanish or Welsh is used. The overall
mean scores (Spanish-English = 2, Welsh-English = 1.5) from the analysis
134 Part 3: Bilingual Language Use
Cuban
22%
American
Cuban- 28%
American
Venezuelan 12%
4%
Dominican
4%
Nicaraguan
6%
Puerto
Rican
4%
Columbian
8%
Ecuatorian
4% Other
8%
Welsh
90%
English
3%
Brish
3%
Other
4%
Figure 6.7 Self-reported national identity in the Miami and Wales corpora