Hello everybody!

This research has got some limits;
- The ones who responded are not chosen at random, so you can't say that the results are representative of the group.
- There is an element of bias. Some of the questions are impossible to answer in an objective way. You have to score yourself and your children, and what may be "in between" for one could be classified as "very poor" by another.

This is true, but this survey is not meant to be particularly scientific, I was rather trying to get an overview of the group. Even if there are reasons to doubt the reliability of the research, I hope it will give us things to think about and to discuss within the families and on the list. By the way, I am not a scientist, I am just another parent who is curious about bilinguality. During night -time I'm working as a musician, and during day-time I take care of our 23 months old Swedish-Dutch monkey Max.

If you have got any questions you can mail me: mailto:henrik@knoware.nl  I can't answer questions as "Can you tell me what my answer of question 5 was?", because of privacy reasons I've separated the answers from names and emailadresses. But I will keep the answers for a while in the case that anything has to be rechecked.

You can read the questions of the survey at this page. 

I want to thank all the people who responded to my request, and especially I want to thank Susan Naves in Australia for helping me with the English language.

Henrik Holm, Amsterdam
Father of Max since febr. '97
m-Swedish M-Dutch

Index

1. Frequently Used Terms

2. Countries and languages

3. The Children

4. Scores

5.Comments of the participants

6. Summary

 1. Frequently Used  Terms

ml@h, ML@H, also mL@H, MLaH, etc.:
Minority language at home. Using the minority language, which is to say the language not spoken in the community, at home to create a bilingual environment for the children.

OPOL:
One Parent, One Language. Each parent uses his or her native language when speaking to the children, to create a bilingual environment for the children.

m=X, M=Y:
The lowercase m means minority language, the uppercase means the majority (community) language. So "m=English, M=Norwegian" means the person in question is part of a family that uses English, living in a Norwegian language environment. It doesn't say whether that person is himself/herself a native English speaker
or not, nor does it tell what sort of bilingual pattern the family uses. Many list members use this code in their signatures, and everyone is encouraged to do so.
 

 2. Countries and languages

2.1. Countries where the families are living:

 Twenty-nine families (47,5 %) are living in an English-speaking country.

Back to index
 

2.2. Language of the mother:

note: most of the persons with two languages mentioned that they are bilingual since childhood
 
 Back to index
 

2.3. Language of the father:

Further findings distilled from the first three questions are as follows:
 

• Thirty-one couples (50,1 %) are living in the country where the language of the man is spoken; 20 (32,8 %) are living in the country where the language of the woman is spoken.

• Six couples (9,8 %) are living in a country where none of the languages of the couple is spoken, so their children are growing up surrounded by three languages.

• Two couples (3,3 %) are living in a community which is bilingual.

• Two couples (3,3 %) are living in their native language area but choose to raise their children bilingual, so they teach the children a language which is not native to the parents. One of these couples does this with the ml@h-model, they communicate with the children in a "target language", and when the parents talk to each other they do it in their native language. The other couple uses the opol-model.  In this case, the mother speaks in the second, non-native, language to the child.

• One bilingual couple (father M, mother m), choose to give their child a third language. In this case the father speaks in the third, not native, language to the child.

2.4. Language of communication between mother and father (not between parents and children).

Of the couples who are communicating with each other in ONE language, 26 (56,5 %) are using the M-language (language of the community) which is also the native language of one of them. Thirteen (28,3 %) are using the m-language. Four couples (8,7 %) are communicating in a language that is not native to either one of the partners. Two of the couples do this to be able to understand each other.  The other two couples do it because they choose to communicate in the language of the country where they are living even if it's not one of their native languages. Three other couples (6,5 %) communicate in the language which is native to both of them. Two of those couples choose to raise their children bilingual, so they teach the children a language which is not native to the parents (see above). The third couple is a couple living in a bilingual community.

For the ones who are interested in the differences between languages spoken at home and the sexes: of the 46 couples who are communicating in ONE language, seventeen (37,0 %) are speaking the M-language native to the man, 9 (19,6 %) are speaking the M-language native to the woman, 10 (21,7 %) are speaking the m-language native to the woman and 3 (6,5 %) are speaking the m-language native to the man.
 
Back to index
 

 3. The Children

3.1. Number and age of the children

Thirty-four couples (55,7 %) has got one child, twenty-two couples (36,1 %) have two children, three couples (4,9 %) have three children, and two couples  (3,3 %) have four. The average age of the children at the single child families with one child is 30 months, or 2,5 years. The youngest is 8 weeks, and the oldest 11 years. Half are younger than 2 years and 88 % are younger than three years. The average age of the first child at the families with two children is 4,1 years old and the second is 21 months. The average age of children in families with three children is 8, 6, and 3 years old. The average age of children in families with four children is 8, 6, 5 and 2 years old.

In total there are 95 children in this survey with the average age of 3,2 years and 50 % of all the children are younger than 2,7 years old. We can conclude that we are talking about a young group of children.

Back to index
 

3.2. Model of communicating with the children

3.3. Who takes care of the children during daytime?

• In 37 of the families (60,7 %), the children go to a kind of daycarecenter or school or both.  In 17 of these families (27,9 % of the total) the child's attendance is fulltime.
• In 21 families (34,4 %) it is mainly the mother who takes care of the children. In 12 of these families (19,7 % of the total) the mother does this fulltime.  For the remaining nine families (14,8 %), there is also daycare or school involved.
• In 13 families (21,3 %) both parents take care of the children part-time, but mostly combined with school or daycare (9 = 14,8 %). Four of these families (6,6 %) do not send their children to any form of daycare or school.
• In six families (9,8 %) the grandparents are helping with taking care of the children.  In three families they do this fulltime, whilst another three families combine this with daycare, school and one or two of the parents.
• Three families have got the help or an au pair in the house.
• In another three families (4,9 %) it is mainly the father who takes care of the children. Two familes combine this with daycare or a babysitter, whilst one father is doing it fulltime.

 Back to index

 4. Scores

4.1. Ability of the children to understand the languages of the parents

For the following statistics I changed values as "good", "poor", "in between" to an equivalent numbering system;

5 = "very good", or "very willing"
4 = "good" or "rather willing"
3 = "in between"
2 = "poor" or "rather not"
1 = "very poor" or "not willing at all".

(I use the terms "very good" and "good", while correct English schould be "very well" and "well". Since the terms already were used in the questionnaire I also use them in this report).

If we look at the answers to questions 8a and 9a regarding understanding the language of the mother and father we get the following results:

Of the 61 replies, eight did not complete this section (this also applies to questions 8b,c, 9b,c, 10, a,b,c) because
the children where too young give a score. This leaves 53 families who completed this section. Thirty-two (60,4 %) answered "very good" to the understanding of both parent's language.  This I scored as "(5 - 5)". One family (1,9 %) answered "good" to both questions (4 - 4). The rest of  the families (20 = 37,7 %) reported a difference in the ability to understand both parents.

The following table reflects these conclusions:
 

Numbers of   families	%	Ability to  understand parents  a and b, "(a - b)" *
32	60,4 %	(5 - 5)
13	24,5 %	(5 - 4)
1	1,9 %	(4 - 4)
5	9,3 %	(5 - 3)
1	1,9 %	(5 - 2)
1	1,9 %	(4 - 3)

Table 5.

* In the parenthesis the score of the best understood language is first followed by the the score of the least understood language whether this is the language of the mother or the father.

We can conclude that most of the children, 86,8 %, understand both parents good to very good. In the replies
where there was a difference in understanding, the children understood the M-language better than the m-language in 12 families. In six of these families the M-language is the language of the mother, whilst in the other six families the M-language is the language of the father. In six families the children understood the m-language better. In all cases this was the language of the mother. In two families both parents speak a minority language, and the children of one family understood the father's language better, whilst the other children understood the mother's language better.
 
Back to index
 
 

4.2. Ability and willingness of the children to speak the languages of the parents

 

Numbers of   families	%	Ability to speak the   language of the parents   a and b, "(a - b)"
14	26,4 %	(5 - 5)
14	26,4 %	(5 - 4)
5	9,3 %	(5 - 3)
5	9,3 %	(5 - 2)
2	3,8 %	(5 - 1)
4	7,5 %	(4 - 4)
6	11,3 %	(4 - 3)
1	1,9 %	(4 - 2)
1	1,9 %	(3 - 2)
1	1,9 %	(2 - 2)

 Table 6.

I checked to see if the different models "OPOL", "not consistent OPOL", "ml@h", "OPOL/ml@h", and "no model", had any influence on the scores of understanding, ability to speak the languages, and willingness to speak the languages.  I came to the following conclusions: there is no significant difference between the models "OPOL", "ml@h" and "ml@h/OPOL", but if you use the "not consistent OPOL" or "no model" type of interaction, the scores of the less dominant language were generally lower than average.
You can see this in the following table:
 

MODEL	Ability to understand the parents	Ability to speak the language of the parents	Willingness to speak the language of the parents
OPOL	18 x (5 - 5)  5 x (5 - 4)  1 x (5 - 3)  1 x (4 - 4)  average: (5,0 - 4,7)          A	8 x (5 - 5)  5 x (5 - 4)  2 x (5 - 3)  3 x (5 - 2)  3 x (4 - 4)  3 x (4 - 3)  1 x (4 - 2)  average: (4,3 - 3,8)    B	12 x (5 - 5)  1 x (5 - 4)  4 x (5 - 3)  3 x (5 - 2)  1 x (5 - 1)  1 x (4 - 4)  2 x (4 - 3)  1 x (4 - 2)  average: (4,8 - 3,8)  C
not consistent OPOL	3 x (5 - 5)  2 x (5 - 4)  1 x (5 - 2)  average: (5,0 - 4,2)      D	1 x (5 - 4)  3 x (5 - 3)  1 x (5 - 2)  1 x (5 - 1)  average: (5 - 2,7)    E	2 x (5 - 5)  1 x (5 - 4)  1 x (5 - 3)  1 x (5 - 2)  1 x (5 - 1)  average: (5,0 - 3,2)  F
OPOL/ml@h	5 x (5 - 5)  2 x (5 - 4)  1 x (4 - 3)  average: (4,9 - 4,5)      G	2 x (5 - 5)  3 x (5 - 4)  1 x (4 - 4)  1 x (4 - 3)  1 x (2 - 2)  average: (4,4 - 3,9)  H	5 x (5 - 5)  1 x (5 - 4)  1 x (4 - 4)  1 x (4 - 3)  average: (4,8 - 4,5)    I
ml@h	6 x (5 - 5)  3 x (5 - 4)  2 x (5 - 3)  average: (5,0 - 4,4)      J	4 x (5 - 5)  4 x (5 - 4)  1 x (5 - 2)  1 x (5 - 1)  1 x (4 - 3)  average: (4,9 - 3,8)  K	7 x (5 - 5)  3 x (5 - 3)  1 x (4 - 3)  average: (4,9 - 4,3)      L
no model	1 x (5 - 4)  1 x (5 - 3)  1 x (4 - 3)  average: (4,7 - 4,3)  M	1 x (5 - 4)  1 x (4 - 3)  1 x (3 - 2)  average: (4,0 - 3,0)  N	1 x (5 - 2)  2 x (4 - 4)  average: (4,3 - 3,3)    O

Table 7.

This might look a bit complicated. Don't worry - it is! You should read this table as follows: If we look at the numbers in cell "A", we see for example "5 x (5 - 4). This means that there are 5 families using OPOL where the children understand one of the parents "very good" (5) the other parent "good" (4). The average score for the dominant language in this cell is 5,0 and for the less dominant language 4,7.

Looking at the average scores above, you can see reflected in all models that the understanding, ability and willingness to speak is consistently high for the dominant language (the first number).  However, looking at the scores for the less dominant language (the second number) spoken in a home where inconsistent OPOL or 'no model' is used, they are lower in the areas of "ability to speak" (E,N), and "willingness" (F,O). These scores vary from 2,7 to 3,3.
 
 Back to index
 

4.3. The language of the surroundings

4.4. Ability of the parents to understand and speak each other's languages

 Back to index
 

You should read this table as follows: If you look at the first numbers in cell "A", you can see "2 x (5 - 4)". This means that 2 couples are living in a country where the language spoken is that of the man, and that the women in these couples speak the language of her partner "very good" (5), whereas their partners speak the language of the women "good" (4). In all the cells te scores of the women is placed first in the parenthesis.

I find it interesting to compare cells A and E, where the most scores can be found. You can see that for 17 couples living in the language area of the man, the woman speaks the language of her partner better than the reverse. This equals 54,8 % of the replies. If we compare this with cell E we see that the percentage does not differe very much, (54,8 % in cell A, 50 % in cell E), but what differs is the scores "poor" to "very poor".  For the men in cell A this percentage is  38,7 %, (12 of 31). For the women in cell E this percentage is 25 %  (5 of 20).

I was curious to see if there was any relationship between the ability of the parents to speak each others languages and the ability of the children to understand and speak both languages. I divided the parents into two groups. Group "A" where both parents scored a "4" or "5" in "understanding and speaking each others languages".  This amounted to 25 couples. And group "B" where at least one parent scored a "3" or below.  This amounted to 28 couples. As before, 8 questionnaires couldn't be used because they were incomplete. Next I examined the children's scores and found the following average values:

 

	Understanding	Ability to speak	Willingness
Group A  (25 couples)	(5,0 - 4,7)	(4,7 - 3,8)	(4,9 - 4,4)
Group B  (28 couples)	(5,0 - 4,2)	(4,9 - 3,2)	(4,8 - 3,9)

Table 9.

In the parenthesis, the first number reflects the average score of the dominant language, and the second number reflects the less dominant language. At group B the average scores of the less dominant language is lower, but I don't know if the differences are big enough to say if there is a significant relationship between the ability of the parents to understand each other and that of the child's ability to speak and understand a language.  Maybe there are some scientists out there who could comment?

If we look at the children of group B, the group where at least one of the parents scores lower in the areas of understanding and speaking the other's language, we can see the same pattern in the chidren. The language that one of the parents scores lower is also the language that their children score lower, (in 91,7 % of the cases), and this is almost always the m-language (also in 91,7 % of the cases).

Back to index

4.5. Ability of the parents to understand and speak the language of the surroundings

 5.Comments of the participants

>You didn't allow for the
>highest level of bilingualism, bi-literacy.  That's the real breakpoint
>for older children's bilingual language development.  Also for the
>parents' relationship I would think.  The levels in that are: reads -
>reads at age level - writes - writes at age level - completely literate
>in second language.
____

>"You should have asked the willingness of the spouse to support the
>bilingual efforts in the family.  My husband is very supportive and that's why
>I'be been able to achieve as much as I have even though he does not know my language.
>Also, you need to allow for different levels of language ability in the spouse
>too. If the spouse understands the other language, it is possible to follow OPOL,
>but not ml@h. If both are fluent in both languages, then ml@h is the best
>method. "
____

Yes, you're right, I didn't ask that and it would definitely be interesting topics for another survey.  Anyone keen?  As far as the comment on 'best' method goes, I do not know which is better.  Does anyone know of any reports which discuss the theory postulated above?

>Comments:  answering with the categories you give (very good, good, etc.)
>is obviously subjective and also assumes a good knowledge of the
>age-group's abilities.  Also, especially in the early years kids have such
>varying abilities that it is hard for a layperson to know what the
>"standard" or "average" would be for any given age group.

Definitely true! That's one reason why we have to be critical when we look at the results of the research. Another reason to be critical is that the people who answered are not chosen at random. See this "research" more as a beginning of a discussion where we may learn things from.
 
Back to index
 

 6. Summary

• Sixty-one of 340 list-members (17,9 %) responded to this survey. Most of the respondents were living in the USA (21, or 34,4 %). Second came Germany (8 families, or 13,1 %), The Netherlands (7 families, or 11,5 %) and Australia (3 families, or 4,9 %). There were also answers from 16 other countries (2 or 1 family per country). Twenty-nine respondents (47,5 %) are living in an English-speaking country.

• 31 respondents (50,1 %) are living in the country where the language of the man is the language of the community; 20 (32,8 %) are living in the country where the language of the woman is the language of the community. Six couples (9,8 %) are living in a country where the langauge of the community is not the main language of either of the parents. As a result, their children are growing up surrounded by three languages. Two couples are living in a community which is bilingual and two couples are living in their native language area but choose to raise their children bilingual with a second non-native language.

• In most of the families the parents communicate with each other in ONE language  (46 couples or 75,4 %). Fifteen couples (24,6 %) use two languages in one way or another. Of the couples who are communicating with each other in ONE language, 26 (56,5 %) are using the M-language (language of the community) which is also the native language of one of them. Thirteen (28,3 %) are using the m-language.

• Thirty-four couples (55,7 %) have one child, twenty-two couples (36,1 %) have two children, three couples (4,9 %) have three children, and two couples  (3,3 %) have four. The average age of the children at the single child families with one child is 30 months, or 2,5 years. In total there are 95 children in this survey with the average age of 3,2 years and 50 % of all the children are younger than 2,7 years old. We can conclude that we are talking about a young group of children.

• Twenty-seven of the families (44,2 %) use OPOL. Nine families (14,8 %) use OPOL, but not consistently. Twelve families (19,7 %) use ml@h. Eleven families (18,0 %) use a mix of ml@h and OPOL. Three families (4,9 %) said that there was no model at all. An interesting discovery is that ml@h seems to be more popular in English speaking countries, while OPOL is more popular in European countries.

• We can conclude that most of the children, 86,8 %, understand both parents "good" to "very good", no matter which model was used. At the questions "ability to speak the language of the parents" and "willingness to ...."  there is no significant difference between the models "OPOL", "ml@h" and "ml@h/OPOL", but if you use the "not consistent OPOL" or "no model" type of interaction, the scores of the less dominant language were generally lower than average.

• For the questions regarding parents ability to speak each others languages and to speak the language of the surroundings the women scored better than the men.

• In the families where at least one of the partners speak the language of the other "in between", "poor" or "very poor" the children score a lower average  of the less dominant language for the questions "ability to speak...." and "willingness to speak...".