4c) The number of speakers of each language, within a household, can also tip the linguistic balance. "Elena" speaks Castilian in her father’s presence; she uses Catalan with her mother and sisters when he is not present. Thus, the "majority language" of that household is Catalan: it is in that language that most "linguistic transactions" take place. A larger family, or a more insistent father, can shift the linguistic mass from one side to the other. In general, however, the larger the family, the less the "paternal dominance" factor seemed to affect "mixed" households.

5. The guarderia as a linguistic influence

5a) This particular institution, known as the llar d’infants, remains the great unknown of the present study. If the senyora de fer feines were identified as not being Catalan-speaking, and the babysitter could be either Catalan- or Castilian-speaking (depending on the aforementioned factors), the llengua de la guarderia remains shrouded in mystery. Some parents chose a particular pre-school or daycare center because of the "vehicular" language used there; however, others selected the school based on the "methodology" it purportedly used. Others could only take their three-to-five year olds to the guarderia that fit their working schedules. Some schools had serious educational missions and excellent activities. Others merely acted as "parking lots" for children. Some guarderies that I visited were large, bright, modern and well-maintained. Others did not meet the basic government guidelines for size, play space, or sanitary installations. "We wanted to shut down our guarderia, one pre-school director told me, " because we didn’t comply with the law. But the Generalitat wouldn’t let us." The need for pre-schools is so great, (and with the current wave of migration even greater), that even "illegal" guarderies stay open (9).

5b) In one coastal town of almost 30,000 inhabitants, two public and five private guarderies operated in 1998. In some cities, the Catalan government operates pre-schools, but this city had only city-operated and private institutions. Private schools usually stayed open more hours, had a better "atmosphere," and were considerably more expensive than public guarderies. Some provided lunch for the children; in others, all pupils left during the lunch break. While all the pre-schools in this seaside city claimed to use Catalan as the "vehicular" language of communication, 60% of the city’s residents were born outside Catalonia, mostly in Extremadura, Andalucía, and different African countries. In one school, almost 14% (11 out of 80) children came from families in which neither Catalan nor Castilian was the native language. In the face of such discrepancies in quality, variety in physical installations, and linguistic compositions, what (if any) conclusions can we draw about the linguistic influence of this institution? While the majority of conversations I heard between children and employees were in Catalan, most workers and volunteers were keenly aware of the presence of a foreign visitor. Only hours of unobtrusive observation in a random sampling of schools could clarify the true linguistic influence of the guarderia.

6. Conclusions

6a) The vast majority of my informants agreed that the typical babysitter or domestic worker was female. Furthermore, senyores de fer feines, nannies, and other domestic employees were almost never catalanoparlants.

6b) My informants gave less emphatic answers about the linguistic composition of the cangur (‘babysitter’) group. They were often chosen from among one’s own ethnolinguistic band. Parents might choose their cangurs from a network of friends and family. The babysitter reflects this choice of the family’s personal network.

6c) Within the family, the father might remain the weak link in linguistic transmission, but he maintained an importance in linguistic bargaining that was not commensurate with his physical presence in the household (10). Perhaps the strongest position, within a linguistically "mixed" home, is that of the Castilian-speaking mother who wishes to have a Castilian-dominant household. If she combines her (probable) greater contact with the children with the presence of a senyora de fer feines who takes over (linguistically) when she is not present, Catalan is not likely to prevail. We must examine this situation with the knowledge that Catalan women have fewer and fewer children. The Catalan birth rate is 1.14 children per woman, well below the "replacement threshold" of approximately 2.1 children per woman (Idescat 1998: 128). Furthermore, the median age of the Catalan population is increasing: 39.3 years in 1995, compared to 33.7 in 1975 (Idescat 1998: 133). The average age of maternity attained 30.3 years by 1995 in Catalonia: Women postpone childbearing in the Principat, as they do elsewhere, to begin careers. If there were no immigration from outside Spain to Catalonia (principally from Africa, the Americas, and the European Community), Catalonia’s population would actually decrease.