When discussing the results of our research of 1994 and of 1999–2000 with colleagues, we were often asked why we had not included a group of Western children as a comparison or “control.” The easy answer, of course, is that we all know that Western children do not use a geocentric FoR. On the other hand, quite a lot is already known about the cognitive and linguistic development of Western children, although not on the particular language elicitation and encoding tasks of our study. As to encoding, Levinson's team had only studied Dutch adults, and as far as children are concerned, Troadec (2007) is to our knowledge the only one to have carried out a study with French children, but he did not use the encoding tasks without some verbal manipulation in the instructions.

We therefore decided to include a sample of children in Geneva, although a somewhat smaller one than in the other locations (N = 75, age 4 to 12). The characteristics of this location are described in chapter 3. The sample characteristics are shown in Table 9.1.

The first language spoken in the family was recorded, and is presented in Table 9.2. It will be noted that most of the bilingual children know another European language, so bilingualism should make no difference to the topics we are studying. We will nevertheless examine this question when analyzing the data.