Attitudes towards the English influx in the Nordic countries : A quantitative investigation

This paper aims at investigating two questions: The first question is whether some Nordic populations really are more purist than others or if this commonly accepted truth is a mere myth. The second issue is whether a society's official purism or laissez faire is generally acknowledged ideology within the society or, alternatively, if they are parts of particular elite discourses. The paper comprises two independent analyses. The first attempts to empirically investigate the stereotypical image of the sociolinguistic environments of Nordic linguistic communities, here expressed as their attitudes towards English influx. The second attempts a more fine-grained analysis of the attitudes towards English as a product or emblem of belonging to certain social classes. "Social class" is here operationalised as "degree of formal education". In the first study we find the prevalent stereotypical image of the societies roughly reproduced. In the second we find, on the whole, that positive attitudes towards English (and hence anti-purist attitudes) correlate with high levels of formal education irrespective of the linguistic climate of the country. Iceland makes for a noteworthy exception to this trend.

The paper comprises two independent analyses.The first attempts to empirically investigate the stereotypical image of the sociolinguistic environments of Nordic linguistic communities, here expressed as their attitudes towards English influx.The second attempts a more fine-grained analysis of the attitudes towards English as a product or emblem of belonging to certain social classes."Social class" is here operationalised as "degree of formal education".
In the first study we find the prevalent stereotypical image of the societies roughly reproduced.In the second we find, on the whole, that positive attitudes towards English (and hence anti-purist attitudes) correlate with high levels of formal education irrespective of the linguistic climate of the country.Iceland makes for a noteworthy exception to this trend.

Background
This paper sets a rather ambitious goal.It wishes to confirm or reject the stereotypical images Nordic people have of each other's linguistic "climates".Among linguists and others with an interest in language these images are common knowledge: In Iceland strong purist sentiments are prevalent.In Denmark foreign words are accepted with little distress.
When the stereotypical images are presented and defended however, this is done by means of anecdotes, never by means of representative empirical data, since such data are not easy to come by.Through the inter-Nordic "Moderne Importord i Norden"-project however, attitude data on a representative scale has become accessible.The Moderne Importord i Norden-project is a large-scale, inter-Nordic project initiated by Nordisk Språkråd to investigate the English influence on the Nordic languages, the Nordic languages' adaptation of this influence, i.e. phonetic and orthographic adaptation of loan words, and the populations' attitudes towards this influence.The project is led by Helge Sandøy in Bergen, and is still underway.Below a small part of the empirical data of this project are compared with the tentative stereotypical images mentioned above.A comparison like this could add support to our common sense notions, or it could pose important questions to, and maybe in the long run even lead to a redefinition of, our notions.

Study 1
In the first study, the independent Nordic societies are viewed as homogeneous wholes.The aim is to see whether the stereotypical image the Nordic peoples have of each other's linguistic environments can also be found in a large-scale survey investigation of the populations.Or stated differendy, do the official attitudes, as it were, seep down through the general population, or do they remain ideological phrases shifted between language planners?

Linguistic consciousness
To my knowledge, no one has carried out a representative, comparative empirical study of attitudes in the Nordic countries towards the influx of English.A common knowledge state-of-affairs is often cited, in academic but especially in more popular discourse (e.g.Venås 1986, Lund 1990, Rask 1995, 1999, Davidsen-Nielsen & Herslund 1997, Phillipson 2000, Thorsen 1999).No well-defined theoretical framework has been proposed however.Hence, the best ad-hoc theoretical framework on which to base a comparison with empirical data, seems to be the somewhat abstract notion of "linguistic consciousness" posed by Lund (1986), and adopted by Vikør (1993).In effect "linguistic consciousness" can be utilised as an explication of the prevalent stereotypes of linguistic environments cited above.The term is not explicitly defined by Lund.Apparently it could cover a wide range of aspects of language policy, e.g.language purism, pro-neologism, pro-dialect sentiments, anti-English sentiments, etc.For practical purposes it can be thought of as the general linguistic climate of a society which underlies and supports official purism.Lund (1986: 35) lists the Nordic countries according to their "linguistic consciousness": "Least linguistically conscious is the Danish linguistic society.The Swedes are probably a bit more conscious; then follows, in order of ascending consciousness, the Finns, the Finland-Swedes, the Norwegians, and the Faroese" [my translation].A schematic presentation would thus look like this: Table 1.Linguistic consciousness of the Nordic societies.

"Linguistic consciousness" in order of descent:
The Faroes

Denmark
The Sami, Iceland, and Greenland are treated separately and not included in the list.Since the Sami and Greenland are not included in the "Moderne Importord i Norden" data either, their omission is not significant for this study.However, it is also part of the common-knowledge notion that Iceland has no (meaning few) un-adapted English loanwords, and that the Icelandic language policy strongly promotes the invention of words on the basis of native roots to substitute for English loanwords (Rask 1999;Davidsen-Nielsen & Herslund 1997).Thus it is suggested that Iceland should be placed high on the list of "linguistic consciousness".
The present study 1 is conducted to investigate whether a pattern similar to Lund's hypothesised (and the general common-knowledge one) can be found in an empirical material viewing the societies as wholes.The hypothesis is that the theoretical notion of "linguistic consciousness" is a good indicator of purism towards English.Hence we would expect to find a pattern similar to the one in table 1: the Faroes (or maybe Iceland) being the most purist, Denmark being the most laissez faire.

Survey questions in the studies
Two survey questions from the attitude investigation part of the "Moderne Importord i Norden" project are used to measure the populations' attitudes towards English influx in their languages.
The attitude investigation in "Moderne Importord i Norden" in itself consists of three parts.The first is a large-scale telephone poll conducted by Gallup or similar institutes.The second is comprised of qualitative interviews with fewer informants, representing different areas of society.Finally the third is a reaction test, a so called "matched guise test" as described in the now classic article by Lambert et al.( 1960:44-46) which investigates people's unconscious attitudes through their response to more or less English-influenced speech.For the present study, the interest is in the survey data.Respondents are here a representative sample of the populations, and the questions are of core interest for the notion of "linguistic consciousness".The other types of attitude data of course also shed light on the "linguistic consciousness" of the Nordic peoples, but they do it in ways more difficult to use for the comparison at hand.The researcher carrying out the survey asked respondents a total of 10 questions regarding their use of English, their attitude towards specific word pairs ("bodyguard" vs. "livvagt" to name a Danish example), and their attitude towards English influence on language domains such as "corporate languages".Finally respondents were asked about their attitudes towards English influx in more abstract and general terms.These, the abstract questions, are the ones used in the analyses at hand.The reason for their rather odd numbering (4a and 4b) is their insertion in the questionnaire, a numbering I have chosen to preserve for easier comparison with other analyses of the same survey data.The two questions are reproduced below in their original Norwegian wording and in my translation: 1 4a.Det brukes altfor mange engelske ord i språket i dag, Far too many English words are being used in the language these days.
New words should be created to substitute for the English words entering into the language.
A few things should be noted about the questions: 1) They concern the respective languages, not domains within the societies.That is, one could very well imagine that more internationally oriented countries (and more internationally oriented persons within the countries) use English in their everyday life, and that this influences their attitudes towards English.This however falls beyond the scope of the investigation which deals specifically with English influence on the Nordic languages, not on the Nordic societies in a broader sense.22) Of the two questions, one is posed as an agenda question; "new words should be created..." while the other concerns the current state of affairs.For someone embedded in the Danish language policy discourse, these two perspectives would be perceived to correlate highly.In other words, if one thinks that there are too many English words, one will promote neologisms.If one promotes neologisms, one perceives even a low degree of influx from English as being too much.However this need not be the case.In a highly purist society, the perception might be that "new words should be created", but that this is being done to such a high degree that "too many English words are not being used".Hence the questions would correlate negatively and express a positive evaluation of the countries' purist policies.

Method
As mentioned in Section 2.2, the questions were asked in a telephone poll conducted by professional survey institutes.The only exception is the Faroese data which were gathered by the university, using students as interviewers.Answers were given on a scale with the labels "agree fully", "agree somewhat", "neither agree nor disagree", "disagree somewhat", "disagree fully", and "don't know".For analysis the answers have been coded so that "fully agreeing" answers are scored as 1 and "fully disagree" are scored as 5, the intermediate answers being scored as 2, 3 and 4 respectively.In other words, because the questions are phrased in terms of hostility to English influx, the higher the score, the more positive the attitude towards English influx.
The sampling was done at random and was representative of the populations as wholes on background variables such as gender, age, income etc.The total number of respondents, excluding "don't know's", were 5,663, comprised of approx.1,000 respondents from Denmark, Sweden, Norway and Finland, 700 from Iceland, and 500 from the Faroes and the Swedish speaking part of Finland.

Comparison
In order to enhance the immediate intelligibility, the responses scored in the tests are here presented as arithmetical means in order of ascending positive attitude towards English influx.Arithmetical means are strictly speaking not the proper way to represent data of an ordinal scale type such as these, but they make for easier overview.The significance testing was done in SPSS ver.10.1 using the Kruskal-Wallis test which is based on ordinal scale data and operates with "mean rank". 3s one can readily see, the results are highly statistically significant (p< .001).A remarkable thing about the results is how closely answers to the second question (4b) resemble the hypothesised order of "linguistic consciousness".This is a strong empirical claim in support of the validity of the stereotypical images suggested above.The populations viewed as wholes in fact do reproduce the hypothesised order, and, as it were, reproduce the stereotypes of linguistic purism and laissez faire.
Even more interesting is the fact that the correlations in Table 2b are even stronger than for Table 2a.The question in Table 2b is the agenda question, the question about which policy the language community should adopt.The proposed policy need not correlate with the perception of the current state of affairs (as expressed in Table 2a) -although a correlation would be expected (cf the discussion in Section 2.2).The responses to the question in Table 2b count as added support for the "linguistic consciousness" notion since this question is most directly concerned with linguistic purism and a pro-neologism policy.
However, the "state-of-affairs" question (4a), regarding the influx from English, also reproduces the hypothesised order to a high degree.Yet there are a few significant exceptions.
Sweden is somewhat higher on the list than would be expected, which I find no immediate explanation for.One can only interpret the answers to the two questions to mean either that Swedes find the English influx too large but that they do not want to replace English words for Swedish -which leaves the Swedish language community with an unsolvable dilemma of how to fill the semantic gaps.Alternatively, the solution to the dilemma lies in an interpretation which could claim that both the labels "too English" and "purist" are laden with negative connotations in Sweden.This would lead the respondents to oppose themselves to both labels, and would trigger the illogical connection between the two.However, exposing a questionnaire-type question to this type of analysis is to open a door to relativism.No statement can then be taken at face value, and the study loses its meaning.Thus we will leave the interpretation as an enigma and conclude that something is going on which cannot be explained without further study.
The other important reordering shown when we compare Tables 2a and  2b is that the Faroes are in the opposite end of the list from what would be expected if one presumes that there is a positive correlation between the two questions.This however is to some extent anticipated in the discussion of the questions in Section 2.2.It is likely that the Faroese have indeed answered the question by stating that there are not too many English loan-words, thus reproducing a general belief that there are indeed none or very few.In other words, the Faroese wish for a purist policy, and are happy with what they have.
On the other hand, this type of reasoning could indicate that Norway and Iceland have unsuccessful purist policies.People do in fact state that there are "too many English words", that is a higher number of English words in Norwegian and Icelandic than people find desirable.The results could also be interpreted in accordance with the "positive correlation"interpretation proposed.The respondents might understand the two questions to be two sides of the same coin.The latter interpretation would claim that Norwegians and Icelanders find that there may not be many English words, but even a few are too many.The policy is not in itself unsuccessful, people merely share the belief that a purist policy is needed.
Solving the correlation between the two questions, as well as throwing some light on the Swedish dilemma, will demand further studies, preferably of a qualitative kind.Hopefully the "Moderne Importord i Norden" project's qualitative analyses will help to understand how the populations may interpret the questions differently and which aspects of the issues they base their answers on.

Study 2
The second study follows up on the first and aims at expanding on it for two reasons.The first reason is that it is very likely that the wording of the questions posed is perceived differendy in the different societies.Not only are translations never exact representations, but it is also very likely that an ideology of purism has an innately positive tone in the more purist societies and vice versa for the more laissez faire ones.The result of these tendencies would be that the answers to the questions do not in fact express peoples' "attitudes".Rather than comparing attitudes we may be comparing the informants' understanding of the questions and positive or negative connotations associated with the labels.Analysing the answers from an intra-societal perspective avoids both the problem of translation and the problem of labels having innately more positive or negative values in one society than another.
The other reason for attempting to expand on the first study with an intra-societal study, is that it will enable us to get a closer look at the nature of and the ownership of the "discourses" of linguistic purism or laissez faire.We wish to establish which social groups and classes are the purist ones and to suggest what part the discourse of linguistic purism plays in societal struggles for symbolic power.

Hypothesis
The second study is based on a hypothesis founded on a Bourdieu-inspired view of society as a field of constant power struggles between groups (Bourdieu 1991(Bourdieu , 1998)).In this view, groups are constructed and remain in constant conflict with each other.This has a number of theoretical implications.
To begin with, all conflict between groups is based on the fight over resources, capital in Bourdieu's terms.But the capital at stake in the social struggle is a multidimensional entity, not to be equated merely with money or material goods.Capital can also be "symbolic capital", e.g. in the form of knowledge, education, etc. Intellectuals possess a high degree of symbolic capital, they have the diplomas of long education, they are regarded as knowledgeable, their interpretation of the world is taken to be, in a sense, more objective and rational than everybody else's.However, their share of material capital, money, does not correlate with their share of symbolic capital.They are not as wealthy as e.g. a business executive, who on the other hand possesses a lower share of symbolic capital.Thus Bourdieu's "capital" is a complex entity.
An aspect of the conflict between groups, and maybe the more fundamental one, is that the groups only exist in that they identify themselves in opposition to other groups.This implies that group membership is not so much a matter of members sharing certain features, as it is a matter of symbolically marking "not-belonging" to some other group.The characteristics, such as style of clothes, brand of car or political affiliation, of any two groups (e.g.intellectuals and business executives) are therefore constructed as being in opposition, but also in opposition to some common third party (e.g.unskilled workers).This situation makes for the complex symbolic oppositions found in modern society.
The basic assumption for this second study is that "attitude towards English influx" can be viewed as one such symbolic emblem of group membership on a par with brand of car, political affiliation etc.Furthermore, it is assumed that it is a symbolic emblem of "high" vs. "low" status in the official hierarchy (here operationalised by length of education) rather than e.g. an emblem of "type of education".If the latter was the case, one might find large differences between e.g.people with a long education within the human sciences and others with a long education within business, differences which might be concealed when these two groups are treated together under "long education".It is likely that such differences do exist, but unfortunately the only information about education given in the survey is its length, so such an effect cannot be evaded.On the other hand, in grouping all types of long education together, one combines the groups which have the most capital in Bourdieu's terms.The business executives are regarded as having the most "material capital", but the intellectuals (i.e.people within the humanities) are regarded as having the most "symbolic capital".Thus in combining the two one can claim to get an understanding of the elite of society, though elite here counts as a somewhat larger percentage of the population than would usually be counted as elite.
The hypothesis tested in this study is that the official policies, as expressed by the notion of "linguistic consciousness", are in line with the sentiments of those having higher social strata, the elite, since they are formulated by language professionals in the countries and sanctioned by the political elite.Promotion of the official language policy would, in other words, act as a symbolic emblem of belonging to the elite.Such a finding could be interpreted as to mean that the official policy goes relatively unquestioned and has strong support.The reverse would mean that the official policy is formulated by language planners without support from the elite, and would pose a problem for the official policy.Of course one could argue that if those belonging to the lower strata support the official language policy to a higher degree than those representing the higher strata, this is a sign of the official policy being in accord with the general population.The problem of this argument is, if one accepts a Bourdieuian model of society, that the support of the language planners is not among their "own kind", viz. the elite, but among, as it were, socially opposing groups.
For this study, as mentioned above, "class" is operationalised by dividing the populations according to the degree of formal education.This is done on the one hand because "degree of formal education" is probably the most objective criterion of social class in a broader sense, and on the other, because educational data are easily accessible from the survey material.
One could criticise correlating formal education with attitude towards English by claiming that those with a longer education will of course be the more positive towards English.They have more qualified, and supposedly more international, jobs, and therefore use English more and are more positive towards English.This may be a fair criticism with regard to the attitudes towards English as such.However, as specified in Section 2.2, the questions asked in this study are specifically about English influx in the respective languages, not about English influence more broadly speaking.If there is any explicit correlation between high exposure (and competence) in English and language purism regarding the national language, it would supposedly be in terms of a norm of not mixing the two; a kind of double purism Qørgensen 1998:142).However this is speculative.What is important for the present is that there is no direct logical correlation between exposure to English and national language purism.

Method
The survey data are the same as the ones used in the analysis above.The difference lies only in the analysing of the data as a variable dependent on the independent variable "education".
The study distinguishes only between "short" and "long" formal education.This rather crude distinction is used on the one hand because the data are gathered using different measures for education (e.g. the Norwegian material uses length of education in years, the Danish material uses the type or name of the education), on the other hand because simplifying the data to merely two groups, "high" vs. "low" education, makes the data more transparent.The restriction to only two variables is therefore not exclusively a weakness of the data.
The dividing line between high and low formal education for the Danish, Swedish, Icelandic, and Finnish material is placed so that the upper secondary level (gymnasium) is counted as a long education.For the remaining countries the division is set between 3 and 4 years of education beyond primary school (Danish gymnasium usually equals 3 years).Of course one could argue that 3 years is not a long education.The counterargument is that 3 years of formal education at a gymnasium is the crucial social dividing point between "the educated" and "the non-educated". 5he school systems are not immediately comparable, so in order to judge the usefulness of this division, the percentages of the population in the different educational groups are included.As one can see from Tables 3a and 3b the split between high arid low education is roughly the same across all societies and ranges from a 50%-50% split in Iceland to a 40% to 60% split in Finland 6 .

Results
As in study 1, the results are presented as arithmetical means.The significance testing is done in SPSS ver.10.1 using the Mann-Whitney test which assumes the data to be ordinal. 7One general trend and a few exceptions are apparent: In most countries, the recurrent pattern for both questions is that higher formal education correlates with more positive attitudes towards the English influx.Thus, the hypothesis that one would see the official policy as an emblem for the more educated rather than the less educated does not hold true.One could claim that the hypothesis is confirmed for the more laissez faire countries (Denmark and Sweden).However, a more reasonable interpretation seems to be that attitudes towards English influx in the Nordic countries represent a more fundamental trend than the policies of the individual countries.My claim is that the results can best be interpreted as a sign that a laissez faire attitude towards English influx is generally used as a symbolic emblem of belonging to the more educated classes.I will return to possible implications of this general pattern in the conclusion.
One undeniable exception to this general pattern is Iceland with regard to question 4b (the agenda question) where the pattern is the opposite of the general pattern that higher education correlates with more purist attitudes.A pro-neologism discourse seems to be part of the elite discourse in Iceland as opposed to most of the other Nordic countries.Apparently Iceland's much admired and criticised purist policy reflects and/or constitutes a truly unique linguistic environment.
The other exceptions to the general pattern of correlation are the two Finnish societies, Finnish-speaking Finland and Swedish-speaking Finland (Tables 3a and 3b), question 4a for Norway (Table 3a) and 4b for Sweden (Table 3b), which are all statistically non-significant.One should of course always hesitate to interpret statistical non-significance as a sign of a particular tendency.A few speculations, however, are in order.One could claim that the questions posed are not a part of any discourse that marks social class affiliation; they have no emblematic function.The questions are either not value-laden at all, or they are connected with national rather than class identity.Alternatively, one could speculate that the lack of differences may be interpreted as a methodological shortcoming, since "attitude towards English influx" may, as sketched above, be used to mark group affiliation in a complex way that is hidden using this design.It may be that one would find differences between e.g. business executives and intellectuals educated within the humanities, but that these differences disappear when the two are grouped together.If this is the case one would claim that "attitudes towards English influx" marks the difference between "material" and "symbolic" capital, thus leaving those without capital somewhere in the middle, rather than marking the difference between the capital "have's" and "have-not's".To justify such a claim, further studies are required.

Conclusion
This paper has presented two studies, or rather two analyses, of data from a survey regarding Nordic peoples' attitude towards English influx on their languages and on linguistic purism.The first study offered empirical support for the stereotypical images of the Nordic linguistic communities with few exceptions.The other presented a somewhat more fragmented result when we tried to correlate attitude towards English influx with social status.The general pattern was for those with a high status to be more positive to English influx than those with a low status.However, significant exceptions were also apparent.To conclude I wish to offer some speculations about the significance of the correlation between level of education and attitude towards English influx.
The most significant finding is that the official language policy does not play an important role.It is remarkable that both Denmark and the Faroes (at opposite ends of the "linguistic consciousness" spectrum) show the same pattern irrespectively of being each other's opposites with regard to policy and their overall attitudes (as expressed in the results of the first study).Apparently, official language planning has little impact on the general tendency for the elite to have positive attitudes towards English influx and/or negative attitudes towards purism.This could be interpreted to mean that a purist discourse, in those countries that follow the general tendency, is associated with traditionalism and nationalism; two "-isms" in sharp opposition to the educated elite's image of itself.
The long-term, but also very uncertain, consequences of these findings could be that the purist language policies in the Faroes and Norway are under threat.If the elites in these countries do not support a purist policy, or, stated differently, if purist discourse is associated with low status, it would be hard to imagine a long-term future for it.
On the other hand, the Danish and Swedish laissez faire policy seems to be under no threat from the educated elite, which may be a sign that the current influence of English on the Nordic languages is seen as an inevitable development.It is my hope that others will propose their interpretations of the data.Especially the interpretation of the dubious correlation between education and language attitudes would gain from being discussed by members of the different linguistic communities.I believe the results presented here warrant an analysis of attitudes as an emblem of social group membership that can give new insights regarding the Nordic linguistic communities.But the empirical data offer no simple, unambiguous interpretation.Future interpretations should perhaps look more into differences between societies, and attempt a plausible intrasocietal interpretation, rather than the unified interpretation for all societies that I have attempted in this paper.

Table 2a .
Attitudes towards English influx.Question 4a.Higher values signify more positive attitudes towards English influx.

Table 2b
Attitudes towards creation of new words.Question 4b.Higher values signify more positive attitudes towards English influx.

Table 3a .
Formal education and attitude towards English influx.Question 4a.Higher values signify more positive attitudes towards English influx.

Table 3b .
Formal education and attitude towards English influx.Question 4b.Higher values signify more positive attitudes towards English influx.