This paper is an updated and expanded version of the final chapter, “The Dynamics of Language maintenance,” in German language maintenance: A handbook (Sherwood Park, 2004).

As we have seen, a number of personal variables have been shown to influence language retention or loss. [1] However, some of them are interrelated and it is therefore not always clear from the literature whether or not a particular effect can be attributed to a specific variable.

The personal variables examined here are

• gender
• age
• period of immigration
• birthplace in Canada or abroad
• religious affiliation
• rural vs. urban residence
• level of education
• occupational choice
• endogamy.

Figure 5 below shows a visualization of the relationships between these variables and language maintenance. Gender, age, period of immigration, birthplace, and endogamy may be considered individually, but there are strong interrelationships among religious affiliation, rural/urban residence, level of education, and occupational choice (affiliation with a certain religious grouping determines place of residence in urban or rural areas; residence, in turn, has an influence on the level of education obtained, which largely determines occupational choice).

Figure 5 Interrelationships between personal variables and language maintenance

Of course, there may also be interaction effects between gender and residence in an urban versus a rural locality, between period of immigration and occupational choice, and so forth.

In other words, the prediction of the dynamics of language maintenance is complex and subject to the interaction among many variables, e.g.,

• The “new ethnics” arrived in Canada after World War II (“period of immigration”); many of them were young (“age”) and got married in Canada to a non-German spouse (“exogamy”). They settled in an urban area (“rural/urban residence”), some went on to secondary and higher education (“level of education”) and chose white-collar work commensurate with their qualifications (“occupational choice”).

• Mennonite returnees from Latin American countries (“religious affiliation”) came to Canada in the last decade or so (time of immigration) and settled in rural areas (“rural/urban residence”) among other members of their religious group. Their families have many young children (“age”) who more often than not finish high school (level of education) and perform work in the agricultural field (“occupational choice”).

In the discussion of the personal variables and their relationship to language maintenance below, these interaction effects should be kept in mind.

6.1 Gender
There is some evidence in the literature that females tend to retain their heritage language longer than do males. As early as 1921, the Canadian census came to the conclusion that

females of nearly every racial origin appear to have acquired the lan­guage of the country of their adoption more slowly than males. This is probably accounted for by the fact that men more largely than women are employed in pursuits which bring them more intimately in touch with persons speaking English or French. [2]

This observation was reiterated in 1931, when the Census report stated that “in almost all classifications more males are bilingual than fe­males.” [3] Ten years later, it was reported that a larger percentage of fe­males than males was unable to speak either English or French. [4]

Prokop found that gender did not have a large effect on language maintenance among Albertans with German mother tongue and home language. Slightly more females spoke German in the home than males – 30% vs. 27% – after having acquired German as their mother tongue. But this difference was accentuated by the rural/urban distinction: 46% of females in rural areas who spoke German as their mother tongue (vs. 41% of the males) retained German as their home language. [5] It is likely that affiliation with a religious group lay at the basis of this urban/rural residence distinction.

6.2 Age
The category “age” is of special interest because it reveals to what degree language maintenance can be expected to extend into the future. If lan­guage maintenance is low among the young, compared with language maintenance among the older speakers of German, certain predictions can be made as to the future of German as a mother tongue and as a home language. The 1931 census, for example, reported that

the number of school age children was found to be the largest single factor in promoting the learning of English, which implies that the school and the associations that go with it are the most potent social agencies in this phase of assimilation. [6]

Furthermore, census researchers established that the proportion of the race which was of school age appeared to be “the most important single fac­tor in explaining the differences in the extent to which the several origins acquired English outside the home.” [7] The following four factors were singled out as significant predictors of the rate of acquisition of English:

1. percentage of children between 10 and 20 years of age,
2. segregation (almost as strong as the first predictor),
3. percentage of people in urban residence (two-thirds as strong), and
4. percentage of North-American born members of the ethnic group (about 15% as strong).

Prokop’s examination of the dynamics of German language maintenance in Alberta and the present investigation demonstrated a strong relationship between age and language maintenance: increasing age and decreasing language maintenance, but accentuated by the rural urban distinction: In the rural areas, many more young people percentage spoke German at home, and so did older people. [8] Of course, it was also determined that rural residence and German language maintenance in the second half of the 20th century was conditioned by religious affiliation.

6.3 Period of immigration
Presumably, persons having immigrated more recently will have retained their mother tongue more fully than those who immigrated many years ago.

A 1971 Census Report distinguished between the so-called “old eth­nic” and the “new ethnic,” the criterion being the time of their (or their ancestors’) immigration, namely either before or after World War II. This distinction proved to be effective in determining the extent to which per­sons of German ethnic origin had retained German as their home lan­guage. For Canada as a whole, only a quarter of those ethnic Germans who claimed German mother tongue and had immigrated before 1945 reported using German as the predominant language of the home in 1971; conversely, while three-quarters of the same group had adopted English as their home lan­guage. On the other hand, about 47% of those ethnic Germans with German mother tongue who immigrated after the end of World War II still spoke German at home. [9]

Clearly, those who had been in Canada longer had switched more extensively to English as their home language than had the more recent arrivals. However, the fact that the “new” ethnics tended to be occupationally different (fewer farm-related and unskilled workers) and preferred to settle in urban rather than rural areas may have confounded the effect of “period of immigration” alone.

In a study of ethnicity, immigration and language transfer, Kralt and Pendakur (1991) found a close link between heritage language use and recency of immigration. In other words, heritage languages are used primarily by immigrants and their children. The continued use of a heritage language by third and subsequent generations is probably relatively rare. [10] They also surmised that the relatively high transmission rate of German to children may be a function of the relatively large Mennonite component of this population who have retained the use of German to a large extent. [11] Bringing the issue to a point, Pendakur (1990) concluded that language maintenance is related to the relative age of the linguistic group and the proportion of immigrants within that group, as well as the age structure of this group – a conclusion which has certainly been confirmed by the findings of this study.

The 2006 Census found that length of time in Canada affects the language that allophones [i.e., persons having a mother tongue other than English or French] speak most often at home. [12] The longer allophone immigrants have been in the country, the more they are exposed to the predominant language of the host society. This tends to impact the language spoken most often at home. In 2006, 19% of immigrants to Canada since 2001 speak English or French most often at home. This proportion increases to one-third for those arriving in the 1980s. Among allophone immigrants who arrived between 1961 and 1970, half reported that they speak one or the other of the official languages most often at home. Most of the children and grandchildren of these immigrants have English or French as their mother tongue. As such, they no longer contribute to the growth of the language group of their parents or grandparents.

Stadler found that immigrants who had spent more than ten years in Canada spoke German significantly less fre­quently at home than did those with Canadian residence of less than ten years. [13]

Prokop established a strong relationship between recency of immigration and German language maintenance in Alberta: The language maintenance ratio rose from .21 for persons who had arrived in Canada before 1945, to about .30 for those who immigrated between 1946 and the late sixties, and to .66 for the immigrants in the late seventies. [14]

6.4 Birthplace in Canada or abroad
The Canadian censuses have frequently reported on differences between Canadian-born and foreign-born ethnics, such as use of mother tongue, home language, age or citizenship. Frequently, the objective was to determine whether persons born outside of Canada were in the process of becoming “like” Canadians.

In his 1990 study of German language retention in Alberta, Prokop reported that twice as many persons with German mother tongue spoke German at home in the rural areas than in the urban areas of Alberta (44% vs. 21%). But when the additional criterion “born in Canada or abroad” was introduced there was no difference in language retention between urban and rural residents who were born abroad (30%). [15]

6.5 Religious affiliation
Membership in certain religious groups has been shown to be a strong determinant of the extent to which a language was preserved over years and generations. This was certainly the case among the settlers in the German colonies in Central and Eastern Europe, in particular the Mennonites, Hutterites, Moravians, and German Baptists, but also among Lutherans and Catholics. The desire to retain German continued when the immigrants came to North America. After all, they had left, among other reasons, because they were no longer guaranteed then right to free practice of their religion, and for most this meant the continued use of German in church, in the family and the community. Not surprisingly, the cultivation of the German language in many of its variants represented a high priority in the conservative Mennonite settlements from Ontario to Alberta, and then to British Columbia.

German also played an important role among those immigrants who did not live in block settlements, especially in the Prairies. When they arrived in the Canadian West, they soon discovered that, because of the pattern of settlement required by homesteading, living in Canada meant isolation from other farmers, and that life in the early years meant unexpected hardships. Understandably, the immigrants yearned for a place where they felt that they belonged, and the churches provided just such a place. Religious beliefs offered the community an emotionally satisfying, common unifying bond; church services were held in the mother tongue, and church activities offered an escape from the toil of the workday. The cohesive effect of religion was a facet of pioneer life which profoundly affected the structure and social content of the community. It was the opinion among the scholars of the 1920s and 1930s – and in fact many contemporary sociologists [16] – that “… the strongest ties which bind German-speaking people in Alberta together and the strongest active forces encouraging the up-keep of German in the homes lie in their religious adherence.” [17]

The ethnic churches were usually the strongest and the most active of all the institutions supporting the survival of distinctive ethnic cultures. [18] Moreover, it has been claimed that the churches, and particularly their publications, frequently were advocates of national ideology and tended to interpret events occurring in Canada in terms of survival of the interests of the ethnic community. [19] The settlers’ reminiscences abound with references to their great desire to have a church of their own and to be able to “hear the word of God in their mother tongue.”

In spite of the initial desire to have church services offered in German, the churches soon felt the need to offer additional services in English. This was not necessarily due to an increase in the overall anglophone church membership, but mainly because the old people had died, and the young people had learned English and preferred their services in the English language. At first, the churches were reluctant to make the change, [20] and in some churches, such as the Mennonite, the change to English, on a regional as well as a local basis, was very controversial. This process of Anglicization began as early as the 1930s as the first generation of German-speaking immigrants began to die out, and has continued up to the present. In the non-conservative Mennonite and Hutterite areas, German is still used for the occasional Sunday service in formerly German-language churches, and social and other events are once in a while being organized for speakers of German, especially the old and the very young. With few exceptions, the churches have abandoned efforts to retain the children in church by encouraging parents to send them to their language school because numbers no longer warrant it. Virtually all Protestant, Catholic and German Church of God language schools across the country have either closed or merged with other language schools.

The situation is entirely different, as this investigation has shown, in the regions of Canada where Hutterites and conservative groups of Mennonites live. The Hutterites traditionally have large families, and in the case of the Mennonites, their natural growth rate has been enhanced by a steady stream of Low German-speaking returnees from Mexico, Belize, Bolivia, Uruguay, and other countries. It has been shown in the provincial chapters that German language maintenance and growth is very pronounced in southwestern Ontario, a number of isolated localities in Manitoba and Saskatchewan as well as in northern Alberta and British Columbia. It was concluded that in the not-too-distant future a dialectal variant of German will be acquired mainly as a mother tongue and used as a home language by the Hutterites, the Old Colony and Old Order Mennonites and the Amish.

6.6 Residence in urban or rural areas
It may be assumed that people in rural areas, where the official language is likely to be less overwhelming at work or at leisure than it is in urban areas, retained their ancestral language longer. The 1931 Census did, in fact, make a very clear statement about the effect of residence in a rural area:

Segregation is a powerful impediment to linguistic assimilation. The more cosmopolitan commercial life of urban centers, on the other hand, favors it. [21]

In 1941, it was reported that “twice as many persons in rural than in urban localities were unable to speak either of Canada’s of­ficial languages.” [22] Conversely, “urban residence favours English more than does rural, urban females favour it more than do urban males while among rural persons there appears to be no sex differentiation.” [23]

In 1961, the next time that the Census addressed the question of age as a predictor for the acquisition of English, the Report concluded:

The extent to which children … learn in early childhood the corre¬sponding mother tongue depends on many factors. Among those that appear to be most important is the relative proportion between foreign-born and native-born segments within an ethnic group. An¬other is the ratio of rural to urban residents. ... Since a substantial number of post-war immigrant families are resident in urban centers it would appear that Canadian-born children of these families will tend to learn English or French as the language first taught in the home to a greater degree than in the past when a larger number of immigrant families settled in small farm communities, often in blocks of the same linguistic group. [24]

More recently, Prokop’s (1990) study of German language maintenance in Alberta and the data reported in this investigation make it clear that residence in urban or rural areas – nowadays reflecting a religious affiliation – only indirectly impacts the linguistic vitality of a language. [25] Schrauf (1999) also concluded that in geographically bounded ethnic communities the practice of native religious forms is significantly associated with the retention of the mother tongue into the third generation.

6.7 Level of education
Intuitively, two different statements of a relationship between level of ed­ucation and language maintenance appear to be equally meaningful: Mainte­nance of the ancestral language might be higher among those who have a higher level of education because they have a better appreciation of the need for retaining their mother tongue and its culture for themselves and their children; after all, it appears to be an accepted fact in Canadian so­ciety, especially among the “professionals,” that knowledge of more than one language is a “good thing” for children from a developmental as well as a practical point of view. Alternatively, it could be that the more highly educated will move in circles where the appearance of being an immi­grant and having a foreign accent might be viewed as being detrimental to their social position. They, and their children, may also make a greater personal effort to become as fluent and correct in their use of English as possible.

It appears from the literature that there is an inverse relationship between cultural and linguistic maintenance and the level of education reached.

Anderson found Mennonites with up to high school education were generally in favour of preserving their identity but were resigned to its loss, while those with university education were largely indifferent. [26] O’Bryan et al. also reported a negative relationship between level of education and the desire to retain German: 29% of those who had eight or fewer years of education considered it very desirable, but only 18% of those with 13 or more years of education did so. [27] Similarly, Borhek concluded that formal education was the most powerful predictor of assimilation and in-group choice among his sample of Ukrainians in Alberta. He came to the conclusion that there was a marked relationship between high occupational status and high assimilation among those having high school education or more. Among the less highly educated, high or low occupational status was unrelated to preparedness to assimilate. It seemed that only in the presence of higher formal education did the ef­fects of workplace, occupational status, and place of residence lead to decreasing ethnic loyalties and involvement. [28]

O’Bryan et al. [29] inferred from a review of the available research that education was negatively related to current level of knowledge of the language; in­deed, fluency was found to be lower among the better educated respon­dents (this is in addition to the observation that the better educated tend to be second- and third-generation Canadians who, in turn, tend to have a lesser knowledge of their ethnic language). They concluded cautiously that “there is some tendency for respondents having more formal educa­tion not to know their ancestral language.” They gave the following as possible reasons for this state of affairs: those who know the ethnic lan­guage may not want to obtain a high degree of formal education because they hold a job which is not tied to their level of education, or because the time and energy required for higher formal education might take time away from the family and the ethnic community and undermine efforts at retaining the ancestral language; schools may downgrade and suppress an interest in ethnic ties; and there might be a lack of opportunity to study and use the language at school and university.

In a multivariate analysis of the 2002 Ethnic Diversity Survey, Turcotte found that the parents’ education made a difference in passing on an ancestral language. [30] “For example, individuals whose mother had a postsecondary education were less likely than those whose mother’s highest level of education was elementary school to have learned their parents’ mother tongue as their first language: 61% vs. 70%.” He concluded, following other studies, that individuals with a higher level of education have a greater tendency to shift to the host country’s official language, even for home use. In a related findings, individuals with household incomes of $20,000 or less used the ancestral language at home, compared with 27% of those with household incomes above $100,000. [31] Of course, household income is not only related to educational background, but also to occupational choice.

Prokop established that Albertans with German mother tongue who had less than a Grade 9 education – irrespective of whether they lived in a rural or urban area – were substantially more likely to have retained the mother tongue as their home language than were those who finished high school or had some post-secondary education. [32] Most of these persons are Hutterites and Mennonites; so religious affiliation and its implications may interact with the “level of education” criterion

6.8 Occupational choice
This criterion is related to “level of education,” of course, and the im­plied hypothesis may again be stated both ways. Some tentative evidence regarding this hypothesis is provided by Stadler, who determined that the use of English in her sample increased with socio-economic status among both parents and children. [33] O’Bryan et al. concluded that, according to the available research, income differences should be strongly and nega­tively related to the current level of language knowledge, although their own research did not reveal such relationships. It could be that in their study positive and negative relationships tended to cancel each other out in the overall statistics. They speculated that

it may be that those achieving economic success are less inclined to have retained the language, but among those who are so inclined, their higher income may be a positive factor in reacquisition or primary acquisition, since currently [1976] almost all costs of such language activities are borne by the individual.

In a later analysis, Kalbach and Richard suggested that “ethnic con­nectedness” was related negatively to measures of socio-economic status achievement, but only for the first generation among whom ethnic visibility is highest. They speculated that “the persistence of ethnic be­haviour appears to prejudice [the immigrants’] chances for higher status achievement.” [35]

Using 1971 Census data, Wolowyna determined in a study of speakers of Ukrainian across Canada that families in which both parents spoke Ukrainian as their mother tongue were more rural and had lower levels of education and income; the husbands usually worked in primary and blue-collar occupations. [36]

In Prokop’s investigation of German language maintenance in Alberta, the highest language retention rates were obtained in the primary occupations (farming, horticulture, and animal husbandry) where 40% of those who had learned German as their mother tongue still used it in the home. Persons in machining and processing occupations and the service industry had substantially lower language retention rates (.23, .21 and .19, respectively). In the other occupational categories, the extent to which a German mother tongue was still used as the home language varied between 10% and 16%. However, an inspection of the distribution of the language maintenance ratio by occupational category and level of education revealed that these differences were due to educational level rather than occupational choice: those with less than Grade 9 education, with only minor exceptions, had the highest language retention rates, with a drop in the rate for virtually all higher educational levels, irrespective of occupation. [37]

6.9 Endogamy and mother tongue retention
An important factor in the retention of the ancestral language and traditions is the extent to which immigrants intermarry within their own ethnic group. It may be surmised that the ancestral language will less likely be spoken at home and old traditions cultivated if the spouses belong to different ethnic groups, and even less likely if one of them is the speaker of a minority ethnic language.

In the early part of the 20th century, Canadian censuses did indeed consider intermarriage “at once an index and a method of assimilation” [38] and charted the “progress made in intermarriage with those of British and French origin.” [39] Census researchers in 1931 concluded that

in practically every instance, a high percentage speaking one of the official languages of Canada in the home is associated with a large amount of intermarriage with the British and French and vice versa. The two phenomena are closely connected, statistically as well as logically. [40]

As late as 1961, “the degree to which the various ethnic groups intermarry is of considerable interest and importance as a measure of cultural assimilation.” [41]

In the first half of the 20th century, endogamy was highest among members of the Slavic ethnic group (see Table 13) and lowest among the Scandinavians. After World War II, the percentage of husbands married to wives of the same origin was highest among the Italians.

The percentage of German men married to women of German origin declined gradually from 75% in 1921 to about 60% in 1941 and to less than 50% after the War.

Table 13 Percentage of husbands married to wives of the same origin [42]

The percentage of women of German origin married to men of the same origin was comparable. In 1921, the average of female endogamy was about four percent higher than male endogamy; [43] in 1951, 52.0% of German men were married to wives of German origin, and conversely, 52.3% of German ethnic females were married to German males. [44]

The Royal Commission on Bilingualism and Biculturalism examined the relationship between endogamy and mother tongue retention rate in the 1941 Census data. Both endogamy and mother tongue retention rate were highest among Japanese and French origins (see Table 14), but in most instances the retention rate was higher than endogamy (e.g., Ukrainian, Polish, Italian, the Scandinavian languages). In German, the endogamy rate was 58%, very similar to the mother tongue retention rate of 54%.

Table 14 Endogamy and mother tongue retention rate for selected ethnic origins, 1941 [45]

Kalbach [46] concluded that the rate of intermarriage depended on nativity and the period of immigration of the foreign-born. Most ethnic groups showed evidence of increasing ethnic mixing as the length of residence increased for the foreign-born (see Table LM.13). But based on 1961 Census data, Kalbach demonstrated clearly that aggregate measures of endogamy are misleading as indicators of assimilation. He argued that since between one half and three quarters of all immigrants entering Canada since 1933 were married, the degree of ethnic mixing was already limited. If there are strong pressures for endogamous marriage in a given ethnic group, the evidence should be found among the native-born as they could still choose whether to marry within or without their ethnic group.

Table 15 shows that among native-born Canadians of German origin the rate of endogamy was consistently lower than for either pre- or post-war immigrants. Among the native-born living in metropolitan areas, about 31% had wives of German origin, compared to pre-war immigrants in metropolitan areas (endogamy rate=ca. 54%) and particularly post-war immigrants (endogamy rate=ca. 82%). The pattern is similar for those living in non-metropolitan urban areas and in rural areas. In other words, for the German ethnic group endogamy among the native-born is relatively low – especially when compared to corresponding rates for Ukrainians where ca. 44% of the native-born men living in metropolitan areas had wives of Ukrainian origin, Jews (89%), and Asiatic groups (72%). Endogamy among native-born Canadians of Scandinavian origin living in metropolitan areas origin was lowest at 13%. [47]

Table 15 Percentage of family heads of normal families with wives of the same origin, by area of residence, for German origin of native-born, pre-war and post-war immigrants, for Canada, 1961 [48]

It has been demonstrated amply that endogamy has a strong effect on mother tongue retention. Stadler, in her 1983 examination of language use and attitudes among German immigrants in Vancouver, showed convincingly that endogamy had a significant positive effect on language maintenance. None of the parents in the ethnically mixed marriages spoke German in the home, and children in these families never spoke German at home, although they were learning German in a German-language school. [49]

In a custom analysis of 1981 Census data for Alberta, Prokop showed that in families where both parents were of German origin, the likelihood of children acquiring German as their mother tongue was greatest (17%); in mixed marriages, the chance of children learning German as their mother tongue was practically nil: certainly when the husband was of non-German origin, and very low (2%) when the wife was non-German origin. Furthermore, in only 9% of the families where both parents were of German origin did the children use German as the home language, and in all other family types, German was virtually not used as the home language at all. [50]

When the parents’ mother tongue – rather than their ethnic origin – is used as the criterion, practically none of the children learned German as their mother tongue in those families where one of the parents did not have German as his or her mother tongue (particularly not where the husband did not have German as his mother tongue): in none of these families did the children use German as their home language. [51]

An analysis of 1996 Census data provides insights into recent trends in endogamy in Canadian society.[52] It comes to the conclusion that

[f]ar fewer children have the heritage language as their home language than as their mother tongue; in other words, although the heritage language may be the first language they learned, they do not use it as their main language in the home. Even in endogamous marriages, fewer than half of he children use the heritage language as their home language, except in Polish, Chinese, Spanish or Vietnamese heritage language families. When only one of the parents has the heritage language as mother tongue, its use as the home language is very rare – less than one in ten children. The only exceptions are children of exogamous marriages where one parent’s mother tongue is Chinese, Punjabi or Vietnamese… [Yet the children are often able to speak the heritage language.] In seven of the 13 largest language groups, at least 90% of children of endogamous marriages knew the heritage language well enough to conduct a conversation. Similarly, the children of exogamous marriages had a far greater tendency to know the heritage language. It is apparent that many children learn their parents’ mother tongue as a second language.

Table 16 shows that 61% of children between 5 and 14 years of age learned German as their mother tongue if both parents had German as their mother tongue, but only two percent did so if only one parent spoke German as his/her mother tongue. When both parents had acquired German as their mother tongue, 42% also used it as their home language, but only one percent did so when only one parent had a German mother tongue.

Table 16 Percentage of children learning a heritage language and using it as the home language in endogamous and exogamous marriages, for Canada, 1996 [53]

Time also appears not be on the side of language continuity. A study of the determinants of heritage language continuity using standard regression analysis and family data from the 1981 and 1991 Censuses [54] not only confirmed many of the earlier findings, but also found that over the ten-year period the probability that children in families in which both parents have non-official mother tongues would inherit the parental mother tongue was 50.9 percent; by 1991 this probability had fallen to 44.9 percent. Moreover, the proportion of families where such transmission is potentially possible also decreased since 1981. Marriage in which at least one partner had an official mother tongue, whether exogamous or not, accentuated sharply the process of language shift, and such marriages are increasingly common.

Regrettably, more recent data on German are not available. But in 2006, the Canadian census reported that when an allophone lives in a couple with an Anglophone or a Francophone, the language other than English or French is seldom the primary language used at home. For 97% of the cases where the allophone's spouse or partner has English as mother tongue, English is the predominant language at home outside Quebec; In Quebec, the proportion reaches 92%. [55]

Does it make a difference whether the mother or the father is the person providing language input? While it is argued by others that women speak to their children differently than men the results of Boyd’s investigation show that there is no difference in the level of proficiency attained if the amount of time for which mothers and fathers were at home is held constant. [56]

Clearly endogamy plays an enormously significant role in language maintenance. However, with little immigration from German-speaking countries and forty years of residence in Canada the children of the German immigrants of the 1950s and 60s will likely continue to intermarry with members of other ethnic groups, which will lead to a further decline in the number of children who are using German in the home.

6.10 Discussion
In addition to the sociocultural reasons for language retention, language maintenance in the German ethnic group is influ­enced by a number of factors – recency of immigration, rural or urban re­sidence, level of education and type of occupation, sex, age, the rate of endogamy, and the extent to which German is spoken in the home envi­ronment.

The use of the German language in Canada is strongly related to the recency of time of immigration of its speakers. In an all-Canadian study one third of the immigrants from the Federal Repub­lic of Germany reported that they spoke English as their home language after as little as six months’ residence in Canada. By extension, the likeli­hood of language maintenance among those German-Canadians who were already born in this country is even lower.

Whether a person resides in an urban or a rural area of Canada has had a significant impact on language maintenance from the time when such issues were first investigated. Fifty years ago, the census found per­sons residing in the rural areas to be twice as likely not to speak either of Canada’s official languages when compared to urban dwellers. The early Canadian censuses explained this dif­ference by the rural dwellers’ lack of contact with the English-speaking world and the vastly greater opportunities for intermingling of the ethnic groups with English speakers in the urban areas. This argument still ap­plies, especially so in the communities which encourage a strict separa­tion of their members from the outside world.

A similar observation was made for sex differences in language maintenance. In the first half of the century, there were a number of re­ports that foreign-born females were slower to learn English than males were; this was attributed to the fact that most speakers of German in those days lived in rural areas and had little contact with the outside world, and among them women remained in the home even more of the time than men did. Recent census data show the effect of sex on language maintenance to be small, but in the same direction as previously ob­served.

Another major factor in the retention of German was found to be the level of education. Persons with less than a Grade 9 education showed a rate of language preservation two to three times higher than the rate of preservation of German by persons with any higher level of schooling. The relationship was not linear in nature, however. It is not possible to maintain that the higher the level of formal schooling, the lower the rate of language retention; instead, a plateau appears to be reached with entry into high school, beyond which there are only small differences in lan­guage loss and language retention. Once more, the rural/urban distinc­tion made a significant difference: more than twice as many rural than urban residents with less than a Grade 9 education spoke German in the home in 1981.

Occupational criteria are clearly a function of educational back­ground: not surprisingly, persons in the primary occupations, such as farming, and in horticultural and animal husbandry occupations, dis­played a higher level of language retention than did persons in any other occupational field. Traditionally, rural residents have tended to value formal education less highly than urban residents. To the present day, certain religious groups, especially the Hutterites, do not generally en­courage their children to obtain formal schooling beyond the level of compulsory school attendance.

Age is another extremely important predictor for language loss or lan­guage retention. While some 85% of small children of German origin are reported to speak German in the home, this rate declines drastically after the age of ten and reaches a plateau of about 25% after the age of 20. Again, the residence criterion plays a major role in determining the rate of language maintenance: across all age groups, the rural residents use German as their home language considerably more frequently than do urban residents: by age 15, some 15% of the urban residents with Ger­man background still speak German as their dominant home language; in the country, as many as two thirds claim German to be the first lan­guage of the home.

An association between the desire to marry within one’s own ethnic group and the rate of language preservation was noted frequently as early as the first part of the century. It was determined that “the Germans” tended to select a partner outside their own ethnic group more often than most other ethnic groups. This observation holds still today as about three quarters of men and women of German origin marry partners with a different ethnic background; the rate of endogamy is only slightly higher in rural than in urban areas.

The rate of endogamy proves to be a significant factor in determining whether German will be used as the dominant language of the home and, consequently, whether the children will learn German as their mother tongue and the language which they use at home. It was found that the chances for children to acquire German as their mother tongue and as their home language was greatest if both parents spoke German in the home. In one-parent families where the father or the mother was either of German origin, had a German mother tongue, or used German as his or her home language, only a small number of children picked up German as their first language, and virtually none learned it long and intensively enough to be able to use it as the home language. In ethnically mixed families where one partner was unable to use German as the home lan­guage, virtually no children appeared to acquire German as the mother tongue or to learn and use it as the dominant home language.

This finding is of great relevance because another part of the present study determined that the skill level to which a child is brought at an early age predicts very strongly the level at which the youngster is able to function later in life. In particular, those who as small children reached a level of proficiency in German at which they understood and spoke Ger­man freely and spontaneously, and also read and wrote German at their level as if they lived in a totally German-speaking environment, retained their original high level of proficiency best.

Small wonder that the fluency with which German is spoken by the immigrant generation vs. the second and third generations differs consid­erably. About two-thirds of the first-generation immigrants perceived themselves to be fluent in the language, but only five percent% of the second generation lid so and virtually none in the third generation. In Edmonton, about half of the immigrant population claimed to use German every day; only quarter of the second and a miniscule one percent of the third generation said that they used German every day.

The implications of this tremendous rate of language loss are obvious: if fewer children are learning German as their home language, fewer people in the next generation will be able to use German as the home lan­guage in endogamous families. Moreover, there will be more families with only one German-speaking partner; at the same time, we know that practically none of the children in ethnically and linguistically mixed Families acquire German either as their mother tongue or as the home language.

[1] For a review of recent studies of ethnic identity maintenance and heritage language retention see Henry P. H. Chow (2001).

[2] Canada. Dominion Bureau of Statistics, CC 1921, Vol. II, p. xvii.

[3] Canada. Dominion Bureau of Statistics, CC 1931, Vol. I, p. 246.

[4] Canada. Dominion Bureau of Statistics, CC 1941, Vol. I, p. 258.

[5] Prokop (1990), p. 86.

[6] Canada. Dominion Bureau of Statistics, CC 1931, Vol. XIII, p. 552.

[7] Canada. Dominion Bureau of Statistics, CC 1931, Col. XIII, p. 681.

[8] Prokop (1990), pp. 86-87.

[9] Canada. Statistics Canada, CC 1971, Vol. V, Part 2 (Bull. 5.1-9).

[10] Kralt and Pendakur, p. 13.

[11] Kralt and Pendakur, p. 12.

[12] Corbeil and Blaser, op. cit.

[13] Stadler, p. 39.

[14] Prokop (1990), p. 90.

[15] Prokop (1990), p. 85.

[16] As recently as 1999, Kalbach and Kalbach (1999) observed in their analysis of 1991 Census data that there was a strong positive relationship between ethnic identity and “ethnic connectedness”: Almost 19% of German Lutherans, Mennonites and Hutterites reported to speak German at home, compared to .7 and 1.4% of ethnic Germans belonging to the two major Canadian churches, the Anglican and the United Church (p. 89). It could be argued, however, that pooling urban and rural Lutherans together with the Mennonites and even the Hutterites is an obvious weakness in the design of the study. Interestingly, Kalbach and Kalbach found that greater ethnic connectedness among the “Germans” was associated with a low level of educational achievement, low occupational status, and low family income – all three could be said to be characteristics of Mennonites and certainly of Hutterites.

[17] Gerwin, p. 146. See also R. Breton, “Institutional completeness of ethnic communities and the personal relations of immigrants,” in B. R. Blishen (ed.), Canadian society: Sociological perspectives (Toronto: Macmillan, 1968), pp. 77-94.

[18] David Millet, “Religion as a source of perpetuation of ethnic identity,” in Migus (1975), p. 105.

[19] O’Bryan et al., p. 18.

[20] Millett (1975) suggested that the introduction of linguistic diversity in the churches may well have been intentional, for it justified their further existence (pp. 106-107).

[21] Canada. Dominion Bureau of Statistics, CC 1931, Vol. XIII, p. 552.

[22] Canada. Dominion Bureau of Statistics, CC 1941, Vol. I. 258.

[23] Canada. Dominion Bureau of Statistics, CC 1931 Census, Vol. I, p. 252.

[24] Canada. Dominion Bureau of Statistics, CC 1961, Bull. 7.1-9, pp. 9-29.

[25] Prokop (1990), 1990.

[26] Anderson, 1972.

[27] O’Bryan et al., pp. 106-107.

[28] J.T. Borhek, “Ethnic group cohesion,” American Journal of Sociology, 76 (1), 1970, 32-46.

[29] O’Bryan at al., p. 166.

[30] Martin Turcotte, “Passing on the ancestral language.” Canadian Social Trends (Spring 2006), p. 22. Statistics Canada – Catalogue No. 11-008.

[31] Turcotte, p. 25.

[32] Prokop (1990), pp. 90-92.

[33] Stadler, p. 48.

[34] O’Bryan et al., p. 166.

[35] Warren E. Kalbach and Madeline A. Richard, “Ethnic-connectedness: How binding is the tie?” in Tova Yedlin (ed.), Central and East European ethnicity in Canada: Adaptation and preservation (Edmonton: CEESSA, 1985), pp. 99-109.

[36] Oleh Wolowyna, “The effects of intermarriage on bilingual education among Ukrainian Canadians,” in Lupul (1985), p. 208.

[37] Prokop (1990), pp. 92-93.

[38] Canada. Dominion Bureau of Statistics, CC 1931, Vol. XIII, p. 546.

[39] Origin, birthplace, nationality and language of the Canadian people. A Census study based on the Census of 1921 and supplementary data (Ottawa, 1929), p. 121.

[40] Canada. Dominion Bureau of Statistics, CC 1931, Vol. XIII, p. 679.

[41] Canada. Dominion Bureau of Statistics, CC 1961, Vol. 7.1-6, p. 35.

[42] Origin, birthplace, nationality and language of the Canadian people. (A Census study based on the Census of 1921 and supplementary data (Ottawa, 1929), Table 60; CC 1931, Vol. 13, pp. 546; Enid Charles, The changing size of the family in Canada (Census Monograph 1) [based on 1941 Census data] (Ottawa, 1948), p. 55, Table XXIII; CC 1951, Vol. III, Table 144; CC 1961, Vol. 7.1-6, p. 36; CC 1971, Vol. 5.3-3, Table 8. Notes: 1. The definition of endogamous marriage changed over time. 1921: endogamous marriages among parents of children “born in the registration area;” 1931: percentage of males who married within their race; 1941: percentage of fathers of legitimate children married to mothers of the same racial origin; 1951-71: husbands married to wives of the same ethnic status. 2. Slavic. 3. Scandinavian.

[43] Origin, birthplace, nationality and language of the Canadian people, p. 118.

[44] Canada. Dominion Bureau of Statistics, CC 1951, Vol. III, Table 144.

[45] Origin, birthplace, nationality and language of the Canadian people. (A Census study based on the Census of 1921 and supplementary data (Ottawa, 1929), Table 60; CC 1931, Vol. 13, pp. 546; Enid Charles, The changing size of the family in Canada (Census Monograph 1) [based on 1941 Census data] (Ottawa, 1948), p. 55, Table XXIII; CC 1951, Vol. III, Table 144; CC 1961, Vol. 7.1-6, p. 36; CC 1971, Vol. 5.3-3, Table 8. Note: 1. The definition of endogamous marriage changed over time. Note 2: 1941: percentage of fathers of legitimate children married to mothers of the same racial origin.

[46] W.E. Kalbach, The impact of immigration on Canada’s population (Ottawa, 1970), pp. 332-335.

[47] Kalbach, p. 335.

[48] Based on Table 5.46 in W. E. Kalbach, The impact of immigration on Canada’s population (Ottawa, 1970), p. 335.

[49] Stadler, p. 31.

[50] Prokop (1990), p. 100.

[51] Prokop (1990), p. 101.

[52] Brian Harrison, “Passing on the language: Heritage language diversity in Canada,” Canadian Social Trends (Statistics Canada, Cat. 11-008, Autumn 2000), p. 17.

[53] Ibid. Note: Children aged 5 to 14, single responses only.

[54] Robert Swidinsky and Michael Swidinsky, “The determinants of heritage language continuity in Canada: Evidence from the 1981 and 1991 Census,” Canadian Ethnic Studies, 29(1), 1997.

[55] Corbeil and Blaser, op. cit.

[56] See Sally Boyd for a review of the more recent literature on the contribution of mothers and fathers to the children’s acquisition of a minority language (“A minority language as mother tongue or father tongue. Does it make a difference? In Tom Ammerlaan, Madeleine Hulsen, Heleen Strating, and Kutlay Yagmur (eds.), Sociolinguistic and psycholinguistic perspectives on maintenance and loss of minority languages (Münster: Waxmann, 2001), pp. 33-45.