ChapterPDF Available

04DeHouwerTrilingualInput

Authors:
  • Harmonious Bilingualism Network

Abstract and Figures

About the influence of parental input patterns on child language use (in children who hear at least three languages).
Content may be subject to copyright.
1
Ms. version of De Houwer, A., 2004, Trilingual input and children's language use in
trilingual families in Flanders’. In C. Hoffmann & J. Ytsma, eds., Trilingualism in the
Individual, Family and Society, Clevedon, UK: Multilingual Matters, 118-138.
sorry about the figures being in the wrong order!
Introduction
There are many children all over the world who grow up in a bilingual environment and
who themselves actively speak two languages. At the same time, there are also many
children who grow up hearing two languages in their environment but who in fact do not
end up speaking those two languages themselves (although they will generally be able to
comprehend two languages). While in itself this lack of active language use in bilingual
settings goes undisputed, it is not always clear what might be the explanations for it (see,
e.g., De Houwer, 1999).
We know even less about children growing up in a trilingual environment. Whether
young children who hear three languages in their environment normally end up speaking
three languages is an open question, and we know little about the reasons for children’s
active trilingual usage or their lack of it.
Bilingual homes or a combination of home and school environments where different
languages are used are typical settings in which children become bilingual. Given this
pivotal role of the home setting, I have suggested elsewhere (De Houwer, 1999) that one
major cause of children's active vs. passive bilingualism may lie in the specific patterning
of language use by parents in the home. The same reasoning may be applied to trilingual
situations. Specific parental input patterns may, to a large extent, determine whether
children who hear three languages on a regular basis actually end up speaking those three
languages or not.
The study presented in this chapter aims to investigate the relationship between parental
input patterns and child language use in trilingual settings. The data for this study come
from a large-scale survey on the home language use of 18,046 families with young
children in officially monolingual Dutch-speaking Flanders. The survey was not
specifically designed for the study presented in this chapter. Rather, in the absence of
language census data in Flanders, the overall aim was to explore the presence of languages
2
other than Dutch in the lives of school-age children growing up in Flanders. This was
accomplished by simply asking respondents to state which language(s) they and their
family members used at home.
As it turned out, in about 2% of the families surveyed children's global language input
(combining school and home) consists of three languages. Since the survey gives
information on parental input patterns as well as children's language use in the home, it
becomes possible to analyse what the relationship is between these two for those families
in which the children hear three languages on a regular basis.
After presenting some general information on the survey, I will give a characterization
of parental and child language use in the trilingual families in the sample. In the ensuing
analyses of possible relationships between parental input patterns and child language use I
will, on the whole, pay more attention to the possible relationships between parental input
patterns and the lack of active trilingual usage in the children rather than to those between
input patterns and children's active trilingualism.
Method
Setting
The data for this study were collected throughout Flanders, the officially monolingual
Dutch-speaking region of Belgium. Although Brussels is the official capital of Flanders, it
is not part of the Flemish region and as such was not covered in the data collection.
According to the Belgian Institute of Statistics, NIS, there were 5,898,824 individuals
living in Flanders on January 1, 1997. 288,307 of these or 4.9% have non-Belgian
citizenship. Like many other Western European countries, Belgium, including Flanders,
has a recent history of attracting 'guest-workers' from other countries. Many of these
workers from countries such as Italy, Greece, Morocco, and Turkey decided to stay in
Belgium and brought their families to join them in Belgium. Larger Flemish cities such as
Ghent and Antwerp now have many neighborhoods where people of very diverse cultural
backgrounds live in close proximity. They also have much higher proportions of non-
Belgian citizens than more rural areas do. Today, many young people are 'second
generation' immigrants.
3
The fact that Flanders is officially Dutch-speaking means, amongst other things, that all
non-private schools, colleges and universities, other educational services, health services
(including speech therapy) and public administration have to use Dutch and only Dutch.
Television and radio programs broadcast out of Flanders are officially Dutch-speaking
although about half of all television programs are imports from English-speaking countries
and transmitted in English (with Dutch subtitles). Service encounters tend to take place in
Dutch. On the whole, then, public life in Flanders is very much oriented towards Dutch.
Data collection: the survey
Basic aim
In Belgium (and thus also in Flanders) it has been illegal since 1947 to ask about home
language use in official census questionnaires. The prohibition was a result of the Belgian
'language wars' between the Dutch- and French-speaking communities in Belgium and
continues to be in place because knowing the linguistic reality might upset the carefully
constructed political balance between the separate language-defined regions (cf. De
Houwer, to appear; see also Verdoodt, 1996). The lack of empirically founded information
or even informed estimates on home language use in Flanders makes it impossible to
verify to what extent educational or social services should take into account the possibly
bilingual or even trilingual backgrounds of their clients, whether they be children or
adults.
The survey undertaken in this study was designed as a very first step towards gaining
some information on home language use in families with young children in Flanders so
that there would at least be some idea of the incidence of multilingualism and of the
languages involved. The focus of the survey was on children in mainstream schools where
a Dutch-speaking home background is typically taken for granted and where
mulitilingualism is not expected by teachers or the local community. This means that
children attending international or European schools or schools with a large proportion of
children from non-Flemish backgrounds were not part of the targeted population. The
reason for excluding these schools was that there is a public acknowledgement in Flanders
that children in these schools have a home language that is different from the school
language (any precise statistical information is lacking, however).
4
Recruitment
In a series of three data collection phases spanning twelve months the principals of Dutch-
medium primary schools throughout Flanders were contacted by telephone and asked
whether they were prepared to cooperate with the survey. If they were, a set of
questionnaire forms were sent to the principals, who then organized the distribution of the
forms to each student in the school in first, second and third grade (children aged 6 to 9).
The children were asked to take the questionnaire home, have their parents fill it out, and
return the form to the school. Schools then sent the completed questionnaires back to the
investigator.
All returned and completed questionnaires thus concern families in Flanders with at
least one child aged 6 to 9, and all school-age children in the sample hear Dutch at school.
Most children in Flanders start going to (pre-)school at age two and a half.
The questionnaire and yield
The questionnaires consisted of a single page in Dutch. They were designed to require as
little knowledge of written and spoken Dutch as possible (see the Appendix for a copy of
the original instrument and a translation into English).
The information requested on the questionnaire concerned (1) the family's place of
residence, and (2) for mother, father and each child living in the same household (space
for up to 5 children): the language(s) spoken at home, family members' ages and
citizenship.
The questions about age, place of residence and citizenship were asked to allow later
analyses of possible relationships between home language use and any of these variables.
None of these factors will be discussed in the present chapter, which will focus exclusively
on language use within the family.
It should be noted that the questionnaire gives only a picture of one particular moment
in time and that the information requested does not furnish any information about previous
and possibly different language use patterns or family composition.
The total number of returned questionnaires suitable for data analysis amounted to
18,046. These 18,046 questionnaires give information on the home language use,
5
citizenship, age, family position and place of residence of 74,690 individuals residing in
Flanders (children and adults combined).
Analyses and results
Trilingual families: a characterization
For the purposes of the present analysis, only the language use in the trilingual families in
the sample is taken into account. An overview of language use in the entire sample can be
found in De Houwer, to appear.
Families are here considered to be trilingual if within the family unit at least two
languages X and Y are spoken either in addition to or instead of Dutch. So even if within
the family unit only two languages X and Y are present, the family in question is still
considered trilingual. The reason for this is that at least one child in such a family will
have trilingual input, viz. Dutch at school and X and Y at home. All school age children in
the families counted as trilingual thus hear at least three languages in their day-to-day lives
(of course, children may hear many other languages at school or from other sources but
this possibility is not considered here any further).
Under the above definition there are 308 trilingual families in the sample (1.7% of all
families surveyed). These families represent a total of 1333 individuals (not counting
preverbal infants under age 1). Not all of these individuals speak a language other than
Dutch at home, though: 15% of them speak only Dutch. However, only 7% of all the
parents in the trilingual families are monolingual Dutch-speaking, whereas three times as
many, or 22%, of all the children in trilingual families are monolingual Dutch-speaking.
Thus it is mostly the children rather than the parents in the trilingual families that do not
speak a language other than Dutch at home, in spite of being exposed to trilingual input.
The actual languages other than Dutch that are spoken in the trilingual families cover a
very large range. There are about 25 different languages represented in the trilingual
sample, ranging from Western European languages, such as French and English, to non-
Indo-European languages, such as Kiswahili, Berber and Urdu.
In the following analyses the data from 64 of the 308 families first counted as trilingual
will be excluded. There are two main reasons for this. First, in 24 families counted as
trilingual the parental input consisted of only one language X other than Dutch. Yet at
6
least one child in each of these families was reported to speak two languages other than
Dutch at home. If the information was filled out correctly, it is clear that the source of the
third language spoken by any of the children cannot lie in parental usage at the time of
completion of the questionnaire. The source of the input in a third language, if any, cannot
be traced. Hence these 24 families will be discarded from further analysis.
Then there are 40 families who reportedly offer quadrilingual input. For 12 of these
there is a strong suspicion that the parents indicated which languages they are skilled in,
rather than which languages they actually speak at home. All families concerned have
Belgian citizenship, and the languages listed besides Dutch are those that university-
educated Flemings are typically expected to know, viz., French, English, and German or
Spanish. In none of these 12 families do the children speak anything but Dutch. In the
other 28 multilingual families there are nearly as many different parental input patterns as
there are families. In addition, the patterns of language use for the children vary
considerably as well: the children in these 28 multilingual families either speak just Dutch,
two languages, three languages, or four languages at home. Because of the low numbers
and the huge variability here any further analyses of relationships between parental input
patterns and children's language use seem rather pointless.
For the remaining 244 families, the global language input to the children is trilingual.
Either the children receive trilingual input at home from their parents, including Dutch, or
they get bilingual parental input at home but the school provides Dutch, the third
language. Families where there is trilingual input at home account for 69% of the 244
trilingual families. Those with bilingual input at home in languages X and Y that is
supplemented by Dutch at school account for the remaining 31%. There are thus more
than twice as many families offering trilingual input in comparison to bilingual input.
All in all there are 14 individual input patterns present in the data. A parental input
pattern is a configuration of reported spoken home language use by mother and father
combined (the parent pair), or by either mother or father in single parent families. The 14
patterns are all the patterns that are theoretically possible. As is to be expected on purely
mathematical grounds, there are twice as many patterns for the families with trilingual
input than there are for the families with bilingual input. The most frequent individual
pattern is the one where both parents speak the same three languages at home, viz. Dutch
7
and two languages other than Dutch. A close second is the pattern where both parents
speak two languages other than Dutch at home but no Dutch. Table 1 lists all the patterns
that are present in the data in their order of frequency of occurrence.
INSERT TABLE 1 ABOUT HERE
The children in the 244 families exhibit five different home language use patterns: (1)
Dutch and two other languages X and Y, (2) two languages X and Y but no Dutch, (3)
Dutch and one other language X, (4) one language X only, and (5) only Dutch (note,
however, that we only know that children are reported to speak a particular language; we
have no information on children's levels of proficiency). Pattern (1) is the only trilingual
pattern, but children exhibiting home pattern (2) are most likely also actively trilingual,
since they must use their third language, Dutch, at school. Patterns (3) and (4) both mean
that children use two languages X and Dutch, whether the latter is used at home or not.
Pattern (5) means that children are at best passively bilingual, but that as far as active
language use goes they are monolingual.
If we look at the children's total language use across home and school, we find that
even though the children all receive trilingual input they are certainly not all actively
trilingual: not even half of the 608 children in the 244 trilingual families are active
trilinguals (see Figure 1). More than a third are actively bilingual rather than trilingual,
and more than a fifth speak only one language, viz. Dutch.
INSERT FIGURE 1 ABOUT HERE
Only about two fifths of the children in the survey who could have been speaking three
languages actually do so. Thus, trilingual input clearly is no guarantee for actually
speaking three languages. However, it may matter quite a bit in what particular fashion the
trilingual input is offered to children. There might well be particular parental input
patterns that foster active trilingual usage, and others that do not. Given the large variation
in the individual parental input patterns (cf. above) this is an issue that can certainly be
explored on the basis of the available data.
8
Parental input patterns and children's language use
In order to investigate the possible relationship between parental input patterns and
children's language use, the analyses in the following will be focused on the family unit
rather than on individual children. The main reason for this is that parental input patterns
are considered to operate on the family level. In order not to over-complicate the analyses,
children's language use will be categorized as falling into either of two categories: (a)
actively trilingual and (b) not actively trilingual. Actively trilingual children speak Dutch,
and two languages X and Y (patterns (1) and (2) above). They represent 42% of the
children studied. Not actively trilingual children speak Dutch, and may also speak one
other language X. They account for 58% of the children (patterns (3), (4), and (5) in the
previous subsection).
Failure or success at being actively trilingual will be decided on the basis of the 'best'
behavior of any one child in the family. Thus, for instance, a family will be considered as
having actively trilingual children if at least one child in the family speaks three
languages, and regardless of the language use of the other children in the family. This
analytical decision does not unnecessarily boost or lower the rates of successful language
transmission for a particular parental input pattern, since earlier analyses have shown that
on the whole, all children in the same family exhibit similar language use patterns (cf. De
Houwer, 2001). In addition, the approach here gives all parental input patterns a similar
"chance" regardless of the number of children that parents happen to have.
Within this 'family-oriented' approach, 45% of the 244 families have actively trilingual
children, and 55% of them do not. The small differences between these figures and the
ones for individual children above (45% instead of 42%, and 55% instead of 58%) are
attributable to the fact that families have different numbers of children.
As noted before, there is a multitude of individual parental input patterns (cf. Table 1
above). Not one of them is uniquely associated with either of the two child language use
patterns (i.e., active trilingual usage or not). However, the 14 individual patterns can be
put into broader categories which may help to address the issue of whether and/or to what
extent children's trilingual language use is correlated with parental input patterns.
9
INSERT FIGURE 2 ABOUT HERE
A first obvious division in the input patterns is the one between families where the
parental input includes Dutch (in addition to two languages X and Y) and those where it
does not (cf. the distinction between 'trilingual' and 'bilingual' home input above). As
Figure 2 shows, the families with actively trilingual children are about equally divided
between those that have Dutch in the parental input and those that do not. However,
families in which the children do not actively speak three languages are almost exclusively
families in which the trilingual parental input includes Dutch: more than four fifths of the
families where the children do not actively speak three languages have Dutch in the
parental input.
We can also look at the data starting from the difference in parental input. Of the
families in which the parental input includes Dutch 69% have children who are not
actively trilingual. Only 31% of the trilingual families where at least one of the parents
speaks Dutch at home have children who speak three languages. On the other hand, three
quarters of the families in which the parental input does not include Dutch have children
who are actively trilingual. The difference between the two types of input in relation to
child active trilingual usage is statistically highly significant (ChiSquare = 39.416;
p<.001).
There is a strong correlation, then, between the presence of Dutch in the parental input
('the Dutch-factor') and children's active trilingualism or, rather, the lack of it. The
presence of Dutch in the parental input is strongly associated with a lack of active
trilingualism in the children, whereas the absence of Dutch in the parental input is strongly
associated with child active trilingualism.
We find a small confirmation of the role of the "Dutch"-factor even in the limited
subsample of the 15 single parent families that are part of the 244 trilingual families. In 12
of these 15 single parent families the single parent speaks Dutch and two languages X and
Y at home. In 7 (that's just more than half) of these 12 families no child is actively
trilingual. Two of the three families in which the single parent just speaks two languages
X and Y at home have actively trilingual children. These data, then, follow the general
trend for the Dutch factor as outlined above.
10
The "Dutch"-factor can account for the child language use patterns in 71% of the 244
trilingual families. This percentage combines on the one hand families with Dutch parental
input but no children who speak three languages, and on the other hand families without
Dutch in the parental input who do have actively trilingual children. In 29% of the
families, however, parents do not have the type of children that might be expected on the
basis of the "Dutch"-factor: here either the parents who speak Dutch at home do have
actively trilingual children or the parents who do not use Dutch at home do not. It is clear,
then, that the "Dutch"-factor is not the only one that can help explain the absence or
presence of active trilingualism in the children.
Another parental input factor that might help explain the absence or presence of active
trilingualism in the children is whether both parents in the parent pair use two languages X
and Y or not. In 39% of the families with two parents the parents both speak languages X
and Y, regardless of whether they use Dutch or not. In the other 61% of dual parent
families the parents do not both speak X and Y at home. There are three subpatterns here:
the main subpattern (108 families) is the one where one parent speaks languages X and Y
and the other parent just language X. The second most frequent subpattern (31 families) is
represented by those parent pairs in which only one parent speaks languages X and Y (the
other parent speaks Dutch). In the third group of families (22 in total) one parent speaks
language X and the other parent language Y. For the 15 single parent families in the
trilingual sample we can of course make no comparisons between the language use of the
parents.
The number of dual parent families amounts to 229. When we look at the factor 'two
languages X and Y used by both parents or not', we find a picture that is rather similar to
the one for the factor 'Dutch in the parental input or not'.
INSERT FIGURE 3 ABOUT HERE
As Figure 3 shows, the families with actively trilingual children are equally divided
between those where both parents speak both languages X and Y and those where they do
not. However, families in which the children do not actively speak three languages are
mostly families in which not both parents speak both languages X and Y: in 70% of the
11
families where the children do not actively speak three languages the parents differ from
each other in their use of languages X and Y.
When we look at the same data in another way and consider the families in which the
parents differ in their use of languages X and Y, 64% have children who are not actively
trilingual. Only 36% of the trilingual families where parents differ from each other in their
language use have children who speak three languages. Families in which both parents
speak both languages X and Y have actively trilingual children in 57% of the cases. The
difference between the two types of input in relation to child active trilingual usage is
statistically significant (ChiSquare = 8.7718; p<.005).
Again, then, we find a correlation between a characteristic of the parental input patterns
and children's active trilingualism or the lack of it. In this case, the fact that parents differ
from each other in the X and Y language choices they make at home is associated with a
lack of active trilingualism in the children. The converse is not quite true: the fact that both
parents both speak two languages X and Y at home shows no strong association with
active trilingualism in the children. As indicated above, in 57% of the families in which
both parents each speak languages X and Y the children are actively bilingual. However,
in the other 43% of these families the children are not. This rather small (14 points)
difference does not warrant any statements regarding any strong beneficial effects of the
use of the same two languages X and Y by both parents. Rather, this factor appears to be
fairly neutral.
So far we have identified two separate aspects of parental input patterns that are
associated with a lack of active trilingualism in the children, namely, the use of Dutch by
at least one of the parents, and the use of different language choices for X and Y in the
parent pair. If we combine both these inhibiting factors, 93% of the cases where there is no
active trilingualism in the children are accounted for (cf. total of boxed percentages in
Table 2). That is, in 93% of the 135 families in which in spite of trilingual input the
children do not speak three languages, at least one of the parents either speaks Dutch
and/or the parents differ in their use of the two languages X and Y. In nine of the ten
remaining families (the other 7%) without actively trilingual children the parents both
speak exclusively X and Y at home. One might expect the children in these nine families
to be actively trilingual, and one can only guess at the possible reasons for why they are
12
not. In the 27 other families where the parents both speak exclusively X and Y at home the
children do speak three languages (cf. Table 2).
INSERT TABLE 2 ABOUT HERE
The validity of the two factors associated with the lack of active trilingualism in
children, viz. the presence of Dutch in the parental input and the use of different languages
X and Y in the parent pair, is further supported by the association of their counterparts
with active trilingualism. Those families where Dutch is not present in the parental input
and/or in which both parents use both languages X and Y account for 74% of all families
with actively trilingual children (total of shaded numbers in Table 2).
Active child trilingualism is thus not expected when there is Dutch in the parental input
and when parents do not both speak languages X and Y. Indeed, when this pattern obtains
(viz. D + NOT 2x (X+Y) - see top row in Table 2), 77% of the 104 families do not have
actively trilingual children. For the 24 families with this input pattern that do have actively
trilingual children obviously other features must play a role.
For the entire group op 244 trilingual families, the parental input patterns identified
above that are associated with children's active use of three languages, or the lack of it, can
account for 84% of the variation in whether children speak three languages or not. For the
other 16% (39 families) the type of data available through the survey cannot be used to
explain either the absence or the presence of active trilingualism in the children. A first
group of families consists of the 24 families with actively trilingual children in spite of the
fact that one of the parents speaks Dutch and both parents do not both speak two
languages X and Y (input pattern 'D + NOT 2x (X+Y)'; cf. previous paragraph). A second
group consists of the 9 families with no actively trilingual children in spite of doubly
favourable input conditions, namely, no parent speaks Dutch at home, and both parents
each speak X and Y (input pattern 'no D + 2x (X+Y)' - cf. the fifth pattern in Table 2 and
discussion previously). Finally, there is the third group consisting of 6 single parent
families (see discussion previously).
13
Discussion and conclusion
Trilingual usage by children growing up in a trilingual environment in Flanders is not the
'default' option: in even less than half of the 244 families surveyed in which children could
have grown up to speak three languages they actually do. Since children need language
models to learn from, and since their primary socialization context typically is that of the
family, it makes sense to investigate to what extent language input patterns as used by the
children's parents might be correlated with children's language use (cf. also De Houwer,
1999). In this case, even though of course we can always only measure correlations and
never observe causation, it can also be assumed that we are dealing with a potential effect
situation in which parental input patterns actually have an influence on children's language
use.
As shown in the analyses, there are two main conditions that do not favor the active use
of three languages in children growing up in a trilingual environment: the fact that either
one of their parents speaks Dutch at home, and the fact that not both their parents speak
the same two languages X and Y. On the other hand, when parents do not speak Dutch at
home and/or both speak two languages X and Y there is a three in four chance that they
have at least one child who speaks three languages. Parental input patterns, then, certainly
seem to affect child language use: the factors 'presence of Dutch' and 'usage of two
languages X and Y by each parent' can explain a large proportion of the data.
Why the parental input patterns identified here should play a role in children's use of
three languages or the lack of it can only be guessed at. However, parental input that
includes Dutch might be an inhibiting factor because of the fact that if parents use Dutch
at home it becomes a strong competitor for the other home languages: after all, Dutch in
Flanders has the advantage of being used in the environment at large and at school. When
home language choices have to be made, children have three possibilities rather than just
two, and there may be no real communicative need for the use of languages X and Y since
the parents speak Dutch themselves, and so allow it for home language use. Another
argument could be that given the same total amount of talk (not a realistic proposition
though!), there is less home input in each of three languages (Dutch, X and Y) than there
is in two languages X and Y. Thus there might be less opportunity for children to learn
languages X and Y if in addition Dutch is spoken by the parents at home. Conversely,
14
children who hear both their parents speak languages X and Y might have more balanced
and varied opportunities to learn and use these languages than children whose parents each
speak different languages. Also, as I pointed out in De Houwer (1999: 79), it might matter
to children whether their parents project a similar or a dissimilar linguistic identity.
Parents who both speak two languages X and Y at home (regardless of whether they use
Dutch as well) project a more similar linguistic identity than parents who do not. This
might be another underlying reason for the supportive nature of the 2x (X+Y) factor.
Other potentially important factors that might play a role in the establishment of active
trilingual usage in children may be the relative frequencies with which the home languages
are spoken. Differences here might account for the differences in failure and success rates
for the very same individual parental input patterns. Another very important factor could
be the interactional strategies (cf. Lanza, 1997) that parents use in communicating with
their children.
The exclusive focus here on parental language use does not imply that children's home
language input is fully dependent on parental language: children's language input at home
may also come from siblings, visitors, relatives and many other people besides. The
totality of this input will play a role as well, of course.
Doubtlessly there is a multitude of factors that can help to explain children's active
trilingual usage under trilingual input conditions. It is clear that the factors identified so far
do not tell the whole story. However, more than four fifths of the variation in the survey
presented in this chapter can be attributed to just two main factors, viz. the use of Dutch
(the environmental language) in the parental input, and the use of the non-environmental
languages X and Y by both parents. More large-scale surveys and analyses of the type
presented here are needed to further substantiate the results of this study. If indeed more
evidence from populations in different regions confirms the results obtained in Flanders,
this will be good news for parents wishing to promote active trilingualism for their
children raised in a trilingual environment: while a 100% success rate cannot be
guaranteed, their children will stand a very good chance of speaking three languages if
parents do not use the environmental language at home but instead each use their two
languages X and Y.
15
Acknowledgements
This research is based on data collected under direct supervision of the author by L. Cox,
L. Debouck, S. De Preter, S. De Witte, M. Marti, S. Stockmans, P. van Aarle, I. Van
Luchene & A. Wagemans. Financial help for the main portion of data collection was made
available by the Department of Political and Social Sciences, UIA/University of Antwerp.
Dr. J. Van Borsel, University of Ghent, was instrumental in helping this project 'take off'.
Many thanks to all, and to the many schools and individuals who invested their time and
effort.
I would also like to thank Wolfgang Wölck and the editors for helpful comments on a
draft version of this chapter.
16
APPENDIX: RELEVANT PART OF ONE PAGE QUESTIONNAIRE USED FOR
SURVEY
ORIGINAL VERSION (in Dutch)
Geachte ouders,
Wij doen een onderzoek naar de taalachtergronden van jonge kinderen in Vlaanderen. Zouden
wij u daarom mogen verzoeken onderstaand formulier in te vullen?
U kan dit formulier terug meegeven aan uw kind. De school zorgt er dan voor dat het terug bij
ons komt.
Heel hartelijk dank!
(name of student helping in data collection)
Uw woonplaats: ........................................................
Gezinsleden die in huis wonen:
thuis gesproken taal/talen
leeftijd
nationaliteit
moeder
vader
kind 1
kind 2
kind 3
kind 4
kind 5
17
ENGLISH TRANSLATION
Dear parents,
We are doing a study of the language backgrounds of young children in Flanders. We would
greatly appreciate it if you could help us and fill out this form.
The form can go back to school with your child. The school will then see to it that the form gets
back to us.
Very many thanks!
(name of student helping in data collection)
Place of residence: ........................................................
Family members living at home:
age
citizenship
mother
father
child 1
child 2
child 3
child 4
child 5
18
References
De Houwer, A. (1999) Environmental factors in early bilingual development: the role of
parental beliefs and attitudes. In G. Extra & L. Verhoeven (eds.) Bilingualism and
Migration (pp. 75-95). Berlin: Mouton de Gruyter.
De Houwer, A. (2001) When do children speak their parents' language? A macro-level
investigation of language use in bilingual families in Flanders. Paper presented at the
Third International Symposium on Bilingualism, Bristol, United Kingdom, April 18-20.
De Houwer, A. (to appear) Home languages spoken in officially monolingual Flanders: a
survey. In K. Bochmann, P. Nelde & W. Wölck (eds.) .) Conflict linguistics.
StAugustin: Asgard Verlag.
Lanza, E. (1997) Language mixing in infant bilingualism. A sociolinguistic perspective.
Oxford: Clarendon Press.
Nationaal Instituut voor de Statistiek (1997) Bevolkingsstatistieken. Ministerie van
Economische Zaken, Koninkrijk België.
Verdoodt, A. (1996) Belgique. In H. Goebl, P. Nelde, Z. Stary & W. Wölck (eds.) Contact
linguistics. An international handbook of contemporary research (pp. 1107-1123).
Berlin: Walter de Gruyter.
19
Table 1. Parental input patterns in 244 trilingual families
__________________________________________________________________
parent1 parent2 no. of families home input B or T? *
D+X+Y** D+X+Y 38 T
X+Y X+Y 36 B
D+X+Y D+X 34 T
D+X X+Y 27 T
D+X+Y D 23 T
X+Y X 23 B
D+X+Y X+Y 15 T
X Y 13 B
D+X+Y - *** 12 T
D X+Y 8 T
D+X D+Y 6 T
D+X+Y X 3 T
D+X Y 3 T
X+Y - *** 3 B
__________________________________________________________________
* B= Bilingual input at home (no Dutch); T=Trilingual input at home (includes
Dutch)
** X and Y refer to any two languages spoken at home excluding Dutch; D stands for
Dutch
*** single parent family
20
Table 2. Parental input patterns and child active trilingualism in 244 trilingual families
______________________________________________________________________
input pattern no actively trilingual actively trilingual
children children
D + NOT 2x (X+Y) 59% 22%
D + 2x (X+Y) 22% 22%
D + X + Y (single parent) 5% 4%
no D + NOT 2x (X+Y) 7% 25%
no D + 2x (X+Y) 6% 25%
no D, only X + Y (single parent) 1% 2%
total 135 109
__________________________________________________________________
D = Dutch used by at least one parent
2x (X+Y) = both parents speak languages X and Y
21
Figure 2. Relationship between use of Dutch in parental
input and incidence of active trilingualism in children
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
active trilingualism no active trilingualism
no Dutch input
at home
Dutch input at
home
22
Figure 3. Relationship between use of X and Y in parental
input and incidence of active trilingualism in children
0%
10%
20%
30%
40%
50%
60%
70%
80%
active trilingualism no active trilingualism
both parents X
and Y
not both parents
X and Y
23
Figure 1. Children's global active
language use under trilingual input
conditions
42%
36%
22%
Dutch + X + Y Dutch + X Only Dutch
ResearchGate has not been able to resolve any citations for this publication.
Chapter
Full-text available
This chapter aims to explore some possible environmental factors that help determine whether very young children will grow up to speak two languages from a very tender age rather than just one. In this exploration, I will be focussing particularly on the possible role of parental beliefs and attitudes. As I will try to explain, such beliefs and attitudes can be seen to lie at the basis of parents' language behavior towards their children, which in turn is a powerful contributive factor in children's patterns of language use.
When do children speak their parents' language? A macro-level investigation of language use in bilingual families in Flanders
  • A De Houwer
De Houwer, A. (2001) When do children speak their parents' language? A macro-level investigation of language use in bilingual families in Flanders. Paper presented at the Third International Symposium on Bilingualism, Bristol, United Kingdom, April 18-20.
Language mixing in infant bilingualism. A sociolinguistic perspective
  • E Lanza
Lanza, E. (1997) Language mixing in infant bilingualism. A sociolinguistic perspective. Oxford: Clarendon Press.