1. Regional languages and dialects in decline 

In the Netherlands, the standard language is Standard Dutch, but there are also three regional languages which are recognized by the European Charter for Regional or Minority Languages (Frisian, Limburgish, and Low Saxon, which are spoken in parts of the provinces of Friesland; Limburg; and Groningen, Drenthe, Overijssel and Gelderland, respectively), and in addition to this there are many local dialects. Since a number of decades, the usage of regional languages and dialects is in severe decline. In a large-scale study (Driessen & Withagen, 1999), which analyzed data collected in 1994 from 7730 primary school pupils and their parents it was found that: 13% of the children spoke a regional language or dialect, while 26% of their parents did so; large differences in language choice exist between provinces; higher educated parents have a better command of Dutch than lower educated parents; the correlation of the children’s language and mathematics achievement and their language choice is weak at best.

2. Research questions and design 

The present study (Driessen, 2016) is a replication of the 1994 study, i.e. comparable data were analyzed with the same variables and computations. There were two main questions: (1) How often do young children and their parents speak a regional language or dialect in 2014 and does this differ from the situation in 1994? (2) Is there any correlation between the children’s language choice (i.e. speaking a regional language or dialect versus Standard Dutch) and their academic language and math achievement?

The data are from the 2014 wave of the national large-scale COOL5-18 cohort study. A total of 437 primary schools participated with 3639 native-born pupils in second grade (6-year-olds) and their parents. (Immigrant pupils and their parents were excluded.)

The pupils’ parents completed a questionnaire with questions related to, among other things, social background and language choice patterns and language proficiency.

From the child’s perspective, four questions were related to the language they generally spoke in conversations in four domains, with: (a) mother; (b) father; (c) siblings; (d) friends. The parents were also asked which language they generally spoke with their partner. In both cases the relevant answer categories were: (1) Dutch; (2) a regional language or dialect. For the children a sum score was computed relative to the number of domains they spoke Dutch, ranging from 0% (in none of the four domains) to 100% (in all of the four domains).

Each of the parents were asked about their command of the Dutch language, distinguishing four language modalities: (a) understand; (b) speak; (c) read; (d) write. The ultimate answer categories were: (1) poor; (2) fairly good; (3) very good.

The parents also provided information regarding their highest level of education.

Two multiple choice achievement tests were administered to the children, language and mathematics. The results were expressed as so-called skill scores.

3.  Results

The first two columns of Table 1 show the relative number of times the children speak Dutch in the four domains discerned for 1994, and twenty years later in 2014. For the Netherlands as a whole, in 1994, the children generally spoke Dutch in 87% of the domains, while in 2014 this had increased to 93%. This means a 6% point loss of regional language or dialect speaking. Columns three and four show that for the parents this loss is even more dramatic, viz. 17% points. There are significant differences according to province, or, perhaps better, regional language. Especially in Limburg and Friesland there still are relative many children and – even more – parents who speak their regional language.

Table 1 - Language mostly spoken by child is Dutch and language mostly spoken by parents is Dutch, by year and province (mean percentages)

(1) Frisian; (2) Low Saxon; (3) Limburgish

Table 2 shows the correlation between the children’s language and mathematics test scores and a number of background characteristics. For this analysis of variance the background characteristics have been collapsed into a number of (ordinal) classes. What is important here is the eta statistic, which can range from 0.00 to 1.00 (and which is comparable to the Pearson r correlation statistic).

Table 2 – Language and mathematics proficiency and background characteristics (mean scores)

* p<0.01

The data in Table 2 make it clear that for the children’s language and mathematics proficiency it does not matter whether they themselves or their parents speak Dutch or a regional language or dialect. Of the background characteristics analyzed here the most important is the level of education of the parents, although it should be mentioned that a correlation of eta = 0.18 still is not very strong.

To answer the two research questions, (1) the data show that the usage of regional languages and dialects indeed is in severe decline, but that there are differences between provinces, and (2) that children who generally speak a regional language or dialect or who have parents who generally speak these language varieties do not perform worse on a Dutch language and mathematics test than children who generally speak Standard Dutch or who have parents who speak Standard Dutch. 

Note

This item is based on: Driessen, G. (2016) Taalvariatie en onderwijsprestaties van autochtone kleuters. Een replicatiestudie na twee decennia. Taal en Tongval, 68(1), 1-12.