All this leads us to think that it is the non-assimilated form that is becoming the prestige form once again for school-age speakers, since it is the form reflected by the spelling, the spoken norm used by all the media in Catalan, the form used by those teachers who are not from the area, and of course, the form that coincides with Spanish.

2.3 Maintenance or otherwise of open vowels

The traditional language model of the county of Baix Vinalopó uses open /o/ and /e/ in the normal way. It is only the most recent generations that are beginning to show a change in this respect. Figure 8 shows this in the case of the open /o/ phoneme: 

Some 10% of the young adults between 18 and 36 throughout the county show the Spanish inspired/ tendency > [o] that we refer to here. Distribution is not uniform across the county: preservation of the distinction, generally speaking, is strongest where Catalan has most vitality.

The factor Eval seems not to influence this trend (see figure 9) and this is a first instance in which such a contradiction is evident in our data. Our expectation was that Eval (education in Catalan) would not be associated with elimination of normative or standard forms, or that this factor could cause them to be reintroduced if they had a tendency to be lost. This contradiction has to do with the fact, above all, of the particular linguistic behaviour of the school age speakers of Elx. In Crevillent, on the other hand, where Catalan has great vitality, the Eval factor is weakly associated with maintenance of the open v. closed distinction. Lastly, at Santa Pola the education programme there is not proving effective in recovering the distinction between open and closed, since the pronunciation has gone to ; here the completion of the sound change is impeding the return to the distinction, for the moment.

Figure 8. Percentage by age of open o: dorm (sleeps)		Figure 9. Percentage by school syllabus of open o: dorm (sleeps)

2.4 Yeism

Yeism is not a feature of the pronunciation in this county, yet it has made rapid progress. This is one of the most striking developments. Among the elders, only 16% of the age group across the county have Yeism, while among school students the sound [] scarcely exists. What is more, the gap between the elders and the adults is very marked —58 %— the transition between the two younger generations is a mere 8 %. (See figure 10.)

Figure 10. Percentage by age of realisations of ll: cavall (horse)

There is no notable relation between education syllabus and the preservation of the phonetic distinction under consideration here, given that Yeism is absolutely categorical among school students. What we have here, then, is clearly a generational change that cannot be stopped by Eval.

2.5 Palatalisation of the implosive before [k]

In the Baix Vinalopó, the traditional pronunciation for palatal sibilants is found only among the elders and to some extent the adults (see figure 11). Selection of the form falls sharply among the young adults and is reduced to a mere 14% among the school-aged, replaced basically by the general Catalan form and coinciding with the Spanish pronunciation for mosca ("fly") which supplied more than 60% of the replies.

Figure 11. Percentage by age of implosive s followed by [k]: mosca (fly)

In this instance too, the traditional form is in sharp decline. The two factors most likely to be invoked here, standard Catalan and Spanish, provide support for this change. Additionally, there is a quite vigorous innovation, mo[h]ca, evidently having its source in Spanish, and clearly linked to the fronterizo Spanish of the Baix Segura and Murcia. This incipient innovation is heard occasionally among the adults but more habitual among the school-aged of Guardamar and Elx. The Spanish media are similarly a factor reinforcing this realisation.

As can be seen in figure 12, Eval subjects quite frequently give the Spanish-influenced reply [h]ca —around 20% of the total number of replies —a frequency which compares with the level evinced by the Lval group, though the latter produce slightly more. The reason for this is that the phenomenon is more usual in areas of heavy Spanish influence, and for that reason frequently elicited for the Eval respondents in Elx. The school-age speakers in Crevillent and Santa Pola have not advanced far with this change.

Figure 12. Percentage by syllabus of implosive s followed by [k]: mosca (fly)

On the other hand, the traditional form mo[]ka is still present in the speech of school children in the Lval category, occurring 47% of the time), while it has all but disappeared among Eval pupils in favour of standard mosca.

2.6 The s in syntactic phonetics

The only traditional and current model for the county (with the single exception of Guardamar) is that which you find in practically all Catalan speech: the voicing of –s before a vowel (see figure 13). The advance of a voiceless variant and the retreating of the traditional voiced sibilant is generational: while there is no sign of the change from [z] > [s] in the two oldest age groups, half the young adults produce it and two thirds of the school-agers. As figure 14 shows there are no relevant differences according to the medium of education.

Figure 13. Percentage by age voiced s in syntactic phonetics: dos amics (two friends)		Figure 14. Percentage by syllabus of voiced s in syntactic phonetics: dos amics (two friends)

2.7 The -ix group

In intervocalic position the variant without the semi-vowel is overwhelmingly the preferred form throughout the county and for all the adult generations. This can be observed in figure 15. It is only among the school-age speakers that there any appreciable occurrences —at 22%— of the semi-vowel form, occurring mostly with Eval subjects (see figure 16).

Figure 15. Percentage by ages of realisation of semi-vowel in group ix: caixa (box; savings bank)		Figure 16. Percentage by school syllabus of realisation of semi-vowel in the group ix: caixa (box; savings bank)

2.8 Betacism

Betacism was something totally alien to the everyday speech of the county, until comparatively recently. The influx of Spanish, historically associated with betacism, has started to have real impact in recent years.

Our study detected more than 30% with betacism, distributed with gradual increase over the generations (see figure 17): with schoolchildren it reaches 62%. It affects all areas in a similar way, although again it is less prevalent in the conservative rural areas and more usual in the coastal areas and in Elx.

The Eval group (see figure 18) once again show a slightly more Spanish-affected pattern, as we have seen several times before. The Lval group distinguish the two sounds rather more frequently. We note that in Santa Pola, betacism has become categorical and the Eval factor shows no corrective influence. At Crevillent, Eval subjects are considerably less betacist than the rest, and this, as we have seen before, is the expected pattern. In Elx, however, betacism is more prevalent among the Eval group and adds to the overall effect that the Eval results are not the expected ones. We return to this question in section 2.14 where we attempt to explain this development.