English and German in Swedish Classrooms: Writing in the two languages compared

Swedish learners of English are bombarded daily with authentic communication. Many of them spend hours a day with computer games where they both receive and produce English in an interactive process. The situation for the other major European languages is not quite so favourable. German, for instance, is seldom heard or seen in computer or other media contexts nor is it so generally recognised as important to speak German. This can be assumed to have an effect on both teachers’ and learners’ concepts of the goals for language learning in schools, as well as on the development of language proficiency in the classroom. We have therefore chosen to study an aspect of language proficiency, writing, and compare the results of learners in school and in the first term of university studies in English and German. We focus on the vocabulary in written free production, in order to analyse some of the differences and similarities between Swedish pupils’ proficiency in the two languages. Teachers’ attitudes to and methods of teaching the two languages may well differ and this is also a question we raise in the study.


Introduction and aims
Swedish learners of English are bombarded daily with authentic communication.Many of them spend hours a day with computer games where they both receive and produce English in an interactive process.The situation for the other major European languages is not quite so favourable.German, for instance, is seldom heard or seen in computer or other media contexts nor is it so generally recognised as important to speak German.This can be assumed to have an effect on both teachers' and learners' concepts of the goals for language learning in schools, as well as on the development of language proficiency in the classroom.
We have therefore chosen to study an aspect of language proficiency, writing, and compare the results of learners in school and in the first term of university studies in English and German.We focus on the vocabulary in written free production, in order to analyse some of the differences and similarities between Swedish pupils' proficiency in the two languages.Teachers' attitudes to and methods of teaching the two languages may well differ and this is also a question we raise in the study.
The aims are as follows: • To make a comparison of the level of proficiency in some aspects of vocabulary in written free production in English and German for pupils in class 9 of the compulsory school, class 2 of the senior secondary school and the first term of university studies.A comparison will also be made between class 9 in English and class 2 in the senior secondary school in German, as they are at the same stage, step 4, in the school system for languages.• To analyse the progression in the same aspects of vocabulary proficiency in written free production between class 9 and the first term of university studies within each language.
• To study teachers' attitudes to the teaching of English and German and the methods they use in teaching respective language.

English and German in Sweden
It is considered completely natural that English should be used in studies at the tertiary level in Sweden.Lectures and literature are frequently in English.It is often the language used for reports and written work at university level and from class 3 or 4 English is compulsory in Swedish schools.Most Swedes consider that it is important to speak the language.This is also reflected in the enormous amount of access to English television programmes, films and computer games.In this case English refers to the entire spectrum of English-speaking countries.The US is the dominant influence in Sweden but primarily Great Britain and to lesser extent Australia are also prominent in Swedish media.According to Svenska Dagbladet, the Swedish daily paper (27/2 2008: 4 f), 58% of programmes on Swedish television are in English.In fact on the channels with the largest percentage of viewers, i.e.SVT 1, SVT 2 and TV 4, programmes in Swedish dominate with over 80% on the state run channels SVT 1 and SVT 2, and nearly 70% on TV 4. On the majority of commercial channels, however, the US is dominant, with 72% on TV 4 comedy and 84.9% on TV 6.The situation is very different for German.Until 1945 German was the main foreign language in Swedish schools.After that English took its place.German was the second foreign language between 1946 and 1962.That year a new school form, 'Grundskolan', was introduced and pupils could then choose German or French in class 7.
At present pupils can choose to study German from class 6 or 7 in the compulsory school.They can later continue German in the senior secondary school and study for a further three years.The majority, however, choose to limit their studies of a second foreign language at that level to one or two years.At most a pupil today can study German for six or seven years.It is also possible to choose to start German as a beginner in senior secondary school and study it for one to three years.
There is very little input from German culture in Swedish media and what there is, is often negative.You very seldom see a positive picture of German culture and the major amount of input refers to the Second World War and the Hitler era.As German has decreased in popularity in Sweden, Spanish has become more and more popular, both in schools and universities.At present the majority of Swedish school learners who study a second foreign language over and above English choose Spanish.German is in second place and French in third place.Many small universities are being forced to close the German department and put resources into Spanish instead.
This raises the question as to how the imbalance in the perceived usefulness and access to English and German affect the way the two languages are taught in schools and the level of proficiency of the learners.In the following section we will refer to some previous research on writing and vocabulary in English and German as second languages.

Previous research into writing
We will here mention a few studies of writing with direct relevance for the present investigation.After a brief summary of the results of an earlier investigation of Swedish learners' writing in English, French, German and Spanish, we will continue with research into writing in English and end with research into German as a foreign language in Sweden.Linnarud et al (2005) studied the writing of school learners of English, French, German and Spanish.The learners were at the same level, step 4 in the Swedish school system, which means that they were from class 9 (age 16) in the compulsory school in the case of English and class 2 at senior secondary school (age 18) in the case of the other three languages.The results showed the greatest difference between English and German.The vocabulary in the German compositions was much poorer than in those in English and the descriptions of the setting and main characters were very basic in German.In contrast the English compositions were very creative and contained a wide range of descriptive terms.According to the Swedish curriculum the learners, who were at the same stage in their foreign language education, should have the same ability to express themselves in English and German and the goals for the lessons are the same.The results of the composition analysis showed that in reality the texts in German were of a much lower standard than those in English.
In an early study of the vocabulary of foreign learners of English in writing Linnarud (1986) compared the writing of Swedish learners with native speakers of the same age, writing on the same topic.Here we will confine ourselves to a description of her results for measures used in the present study, in order to allow a comparison.
The first measure that is common to the two studies is the number of words in the compositions (W/C).The Swedish learners wrote on average significantly shorter compositions than the native speakers in the same amount of time.There was also more variation in W/C among the Swedish learners than among the native speakers.
The second measure that is common to the two studies is Lexical density (LD).This is the percentage of lexical words in the total number of words in a text.According to Ure (1971) written texts have an LD of 40% or more.Informal written texts not surprisingly have the lowest density and formal texts the highest.In Linnarud's (1986:58f) study of learners in class 2 at the senior secondary school it was shown that most learners wrote within a native speaker-like range for LD but there were a few learners with significantly lower scores.The average LD for a control group of native speakers of the same age writing on the same subject was higher on average than that of the Swedish learners.The compositions were graded by three different groups of evaluators, Swedish teachers of English, native speaker teachers of English in Sweden and native speakers without teaching experience.For all three groups the compositions with low lexical density received low grades.This would suggest that lexical density is a measure of a native-speaker like norm.It is not a case of the higher the better but rather that an appropriate level must be reached.
The third measure is the amount of variation in vocabulary in a text (Lexical Variation, LV).Linnarud (1986:57f) found that non-native speakers of English repeated themselves more frequently than native speakers when writing on the same topic.Again there was a correlation between a low grade and a large amount of repetition.The native speakers also varied their vocabulary irrespective of the number of words they wrote.The correlation between W/C and LV was statistically significant for the Swedish learners and not for the native speakers.
The measure of Lexical Variation can be problematic in other ways than the number of words in the text.As Granger & Wynne (2000) point out, the figures for LV can be affected by the amount of error in the words included.In a comparison of Dutch, French, Polish and Spanish learners the Spanish learners came out on top.After adjusting for error the Spanish compositions ended up at the bottom of the groups.A native speaker corpus ended up in the middle but the fact that these compositions were longer is only mentioned briefly.This adds uncertainty to their results.
Kjellén-Simes (forthcoming) has focused on the vocabulary of learners in the International Baccaloreate programme.She used a measure of low frequency vocabulary in order to chart the progress in English of a group of learners in an IB class, where English is used to teach other subjects, compared to a group in a national programme where English is only taught as a subject.The two groups had a similar starting point measured in terms of a diagnostic test at the beginning of the study.Kjellén-Simes shows that those with the lowest results in the diagnostic test, improve most in the immersion group over a time period of almost two years.The control group students achieved lower results over time and ended with a lower result than at the start of the study.This would suggest that a communicative use of the language has beneficial effects on vocabulary learning.It would also seem to confirm that vocabulary is a fruitful area in which to make comparisons between groups of learners.
German as a foreign language has also been studied fairly extensively in Sweden.The importance of vocabulary is stressed by Krohn (1992).He lists a basic vocabulary of 1450 lexemes from different corpora that all learners should know actively as they form the basis for the learner's communicative competence.Kjär (2002) emphasises the need for learning strategies, which can provide the learners with the means to learn vocabulary more efficiently.He stresses the need for consciousness raising and the importance of context.
In a study of texts in German produced by Swedish learners at different levels Rosén (2006) found that the information structure differs greatly from the target language and that direct transfer from Swedish is very frequent.Other aspects than the actual structure of the text, such as grammar and morphology, are emphasised in teaching German.Rosén means that there is a difference in how texts produced by learners in English and German are evaluated.As Archibald (1994:222) says, the emphasis in second language writing in English is normally on the text as a product.This is not always the case in German.This is supported by the results of Wernersson (2008) who interviewed teachers of German.One teacher says specifically that he feels forced to work considerably with grammar although it takes a lot of time and the learners think it is difficult (Wernersson 2008:17).He means that it is a matter of priorities.The time spent on grammar means, of course, that there is less room for other activities such as learning words.
Some insights into the problems involved in the teaching of German are provided by Dentler (2003).Dentler poses the question why Swedish students tend to drop out of language courses, more specifically German.The results of her study suggest that learners are dissatisfied with the learning outcomes and the way the languages are taught at school.In other words it is not just the unpopularity of German in general that creates a lack of interest in the subject but that, in fact, a new approach to teaching might improve interest.
In the Nödinge Immersion project, late partial immersion was tested as a possible way of reducing drop-outs (Dentler 2003:154ff).The results showed a clear success for the immersion programme compared to a control group.The learners' verb morphology, syntax, vocabulary and text production improved markedly and the group was more homogeneous than the control group.The immersion group was more strongly motivated than the control group but did not show any substantial difference in attitude to Germany and Germans compared to the control group.They both had negative responses.Dentler concludes that three years in an immersion class was not enough to change negative attitudes that the pupils have been exposed to throughout their lives.Immersion in the language being taught has been an often-tried method of improving language learning in the classroom.
In Sweden English is the second language used most frequently to teach other subjects.In fact German is used in a small number of cases, 10% of the total for SPRINT (Content and language integrated learning and teaching in Sweden) in the compulsory school and 6% for upper secondary level.The corresponding figures for English are 68% and 82% (Nixon 2000: 23).The results of Dentler's study would suggest that it is in fact possible to teach and learn German in a more communicative fashion and that immersion programmes could be more widely spread in other languages than English.
We will now go on to describe the material collected for this study and the methods of analysis.

Material
The empirical material consists of compositions written in English and German on the same topic by learners at different stages in the education system.The topic of the composition is a series of pictures showing a bald man who sees a girl with lovely long hair.They talk and eventually he asks her to marry him.After the wedding he finds her long hair in the food and looks disgusted.What happens next is up to the writer.The topic gives some room for creativity but also provides a common framework in order to allow for comparison between writers.
The writing task was given to learners in class 9 of the compulsory school, aged 15 -16, class 2 of the senior secondary school, aged 18, and the first term of university studies, aged from 19 upwards, in respective language.Twenty compositions were picked at random from each group with the exception of the university group in English where there were only 19.
From now on class 9 will be referred to as Gr (from 'Grundskolan', compulsory school in Sweden).Class 2 in the senior secondary school will be referred to as Gy (from the Swedish 'gymnasium') and the university group will be referred to as Un.EGr being class 9 in English and GGr being class 9 in German and so on.
The writers were given 60 minutes to complete the task and were not allowed the use of dictionaries.The instructions given to all were the same, that they should use their imagination and write as much as possible in the time allotted.
The pupils who participated in our investigation in class 9 were in their third or fourth year of German (step 2), compared to their sixth or seventh year for English (step 4).Those in the second year of senior secondary school were in their fifth or sixth year of German, compared to their eighth or ninth year of English.Class 9 in English and Class 2 in the senior secondary school in German are at the same stage, step 4, in the Swedish school system.The students in the A-course, the first term of university studies of the subject had between three and six years of German before starting at university and nine or ten years of English.
There is a large difference in the amount of preparation that Swedish learners have had before starting university studies in English and German but the demands made on them are very similar.They are expected to be able to read literary texts and use the language for written and spoken communication and both groups are expected to be able to teach English and German in schools after completing their university studies in teacher education.It is therefore important that the difference in the number of years of language learning prior to tertiary education does not lead to a low level of proficiency in German.

Methods
The investigation focuses on the vocabulary of the learners.This does not mean to imply that correct grammar is unimportant.In the study of compositions in English, French, German and Spanish (Linnarud et al 2005) there was a clear correlation between the amount of correct language in the text and the evaluation given by teachers.Spelling errors were very common in the English and French texts.It must be remembered that the English texts were written by younger learners although they were at the same stage, step 4, as the older learners of the other languages.
Instead of measuring what has been done wrong we decided to measure the positive aspects of the text in terms of the richness of the vocabulary aimed at although the spelling may not be correct.
The compositions were analysed as follows: 1) Number of words (W/C) A word is defined here as an orthographic word.All words have been counted.Contracted forms have been counted as two words, e.g.'don't' as 'do not.' 2) Lexical density of the texts (LD) The LD is counted as the percentage of lexical words, in this case restricted to nouns, lexical verbs and adjectives in the total number of words in the text.
3) Lexical variation measured as a type/ token ratio of nouns, lexical verbs and adjectives (LV) As the length of the text affects the amount of repetition the compositions were divided into groups according to number of words in the text.The lowest group was -100 words and the highest 601-700 words.
Comparisons were made as follows: The results for all groups in English and German were compared.The results for English and German at the different levels, Gr, Gy and Un were compared.Specific emphasis was placed on a comparison between EGr and GGY.In spite of the age difference the learners are at the same stage step 4 according to the curriculum and the goals for learning are almost identical.
There was also a comparison of the results for the various levels, Gr, Gy and Un in English to see any possible progression and a similar comparison of the different levels in German.
Questionnaires were administered to teachers who teach both English and German.We received eight answers from teachers at the senior secondary level and six from teachers of classes six to nine.Two interviews were carried out to follow up points brought out in the questionnaires.The questions included the amount of time spent on writing, the type of tasks, how correction is carried out, how they work with vocabulary learning, how the learners' contacts with the language outside school are used in the classroom and if there is any difference in the goals for English and German in the classroom.

Number of Words (W/C)
The results of the quantitative analysis of the English and German compositions show that the greatest difference between the two languages is the number of words in the text (see figure 1).The English texts are significantly longer on average than the German ones (see table 1).
Table 1: W/C English and German all groups collated -mean and standard deviation W/C mean sd English (all) 307.1 147.7 German (all) 145.5 66.2 p = 0.000 The differences are also significant for each level, Gr, Gy and Un (see table 2).
Table 2: W/C English and German Gr, Gy and Un -mean and standard deviation The English texts have a steady increase from 248 in EGr to 378 in EUn and the difference between EGr and EUn is significant at the 1% level (see table 3).The English Gr group has a higher W/C than the German Gy group, who are the same stage, step 4.
Table 3: W/C EGr, EGy and EUn compared Gr-Gy: p = 0.244; Gy-Un: p = 0.107; Gr-Un: p = 0.005 In contrast, the German groups reach a peak in class 2 at the senior secondary school and regress at university level.The difference is statistically significant at a 1% level (see table 4).If the ability to produce more words in the same amount of time is to be looked on as a sign of development in language proficiency then the English groups show more progress than the German groups.There are no statistically significant differences between the English and German texts or between the different levels in each language.As can be seen in table 5 there is, however, after an initial advantage for the German texts a higher figure for English in the Gy and Un groups.In GGy there is a decrease in LD compared to GGr, although there is progression between GGr and GUn for both English and German.The increase is greater for English, 5%, as against German, 2.2%.Regarding LD there is an appropriate level.It is not a case of trying to achieve as high a density as possible, but rather a sufficiently high level for the text to appear to be normal written English.It is impossible to compare the figures for English in this study exactly with the results for Ure (1971) and Linnarud (1986) as a slightly different definition of lexical words has been chosen.It is, however, apparent that the EUn group most closely approaches the normal figure for written texts and that there is a small tendency through time to achieve higher results.The lexical density for the two groups at step 4 in the school system, EGr and GGy, is almost identical.

Lexical Variation (LV)
Variation in vocabulary in the texts has been measured by modifying the type/token ratio (T/T ratio).This is the percentage of nouns, lexical verbs and adjectives that are used only once in the total number of lexical words in the text.Thus a high figure for T/T ratio means that there is little repetition.The T/T ratio is sensitive to the length of the text as it is easier to avoid repetition in a short text than in a long one.The length of the texts is highly relevant in this study where the learners write much shorter texts in German.In Linnarud's study (1986) of Swedish learners and native speakers writing in English it was found that the native speakers varied their vocabulary almost regardless of the length of the text.The results for LV in this investigation were correlated with W/C in order to see if this was the case for the most advanced learners of English (see figure 2).The figures for W/C are 10% of the actual number in order to present them in the same figure as LV. Figure 2 shows clearly that the tendency is for a high LV to correlate with a low W/C.In other words the most advanced learners of English, EUn, do not show the same trend as the native speakers in Linnarud's study, but vary their vocabulary less with increasing length.There are, of course, individual differences, but the results support the need to take into consideration the W/C in a text before making any pronouncements about variation.A frequently used method of allowing for differences in length is to count the T/T ratio for the first 100, 150 or 200 words.As there is such a wide spread in the W/C in this material we chose instead to divide into groups according to the W/C, with a span of a hundred words in each.The smallest group was 0-100 and the largest 700-800.A comparison of the English and German texts in each W/C group shows that those in English have a higher T/T ratio throughout than the ones in German.There are no German texts above 400 W/C and only small numbers in the English groups (see figure 3).As can be seen in table 6 (numbers in bold mark where there is one individual only) the average LV is higher for English than German at all levels apart from university in spite of the significantly higher number of words in each English group.The numbers are too small to draw any definite conclusions but where there is more than one person in a level and W/C group there is an increase in LV over time in English while the LV in German show little or no change.The EGr group has a consistently higher figure for LV than the GGy group.
Table 6: Lexical Variation for English and German Gr, Gy and Un per W/C group In order to further clarify the differences between English and German in Lexical Variation the results for each W/C group were divided into quartiles.Quartile1 is the group with highest LV and quartile 4 the group with the lowest LV.The number of compositions in English and German in each quartile has been counted as a percentage of the total number of compositions in that language in that W/C group.An example is that in quartile 1, W/C group 101-200, the quartile with the highest result for LV contains 6 compositions in English and 5 in German.Counted as a percentage of the number of compositions in each language, 15 in English and 29 in German, this means that 40% of the compositions in English and 17.2 % of the German compositions are in the highest group.
In W/C group 0-100 there are only two English compositions, one in quartile 2 and one in quartile 3 and in W/C group 301-400 there are only two German compositions, one in quartile 2 and one in quartile 4. For the rest of the higher W/C groups there are no compositions in German.Division into quartiles has therefore only been carried out in W/C groups 101-200 and 201-300.
The results for English and German for W/C group 101-200 are given in table 7. It can be seen that 73.3% of the English compositions are in quartiles 1 and 2, and only 26.7 in quartiles 3 and 4. The situation for the German compositions is the opposite.The majority, 51.7%, are in quartiles 3 and 4. The results of the LV analysis show that when the W/C factor has been accounted for the compositions in English have consistently higher figures than the compositions in German.This would suggest that the learners use a more varied vocabulary in English.

Summary
To summarize, the results of the quantitative analysis in general show differences between German and English both in the comparison between the two languages and the comparison between levels within each language.The most significant difference between the languages is the number of words the learners write in the same amount of time.
There is also a steady increase in W/C between proficiency levels in English whereas German increases between Gr and Gy and decreases between Gy and Un.Lexical Density is higher in the English compositions with the exception of Gr.The difference is not statistically significant.Lexical Variation is greater in the English compositions than in the German ones.In all this would suggest that the learners have a greater command of vocabulary in English and are able to use it in written free production.
In order to illustrate what lies behind the statistics one composition at university level in each language has been chosen and is presented here in its entirety.Both texts are close to the mean W/C for that level.The texts were based on pictures of a bald man who falls in love with a woman with long hair.The pictures end with the man finding a long hair in his food and it is left to the writer's imagination to supply the rest of the story.The English text treats the topic in a humorous fashion, adding original touches about a furry dog and the man coughing up hairballs.EUn W/C: 378 LV: 68.4 Mr Baldy changes his mind Mr Baldy considered himself a quite attractive man.He though that he had nice chiseld facial features, a good strong and firm body, glistening eyes … But there was one thing about his apperance Mr Baldy didn't like, he had no hair.Because of his lack of hair, Mr Baldy considered everything hairy extremely beautiful and attractive.Mr Baldy had a large collection of long-haired wigs, the furriest dog you could imagine, and lots of pictures on his walls of people with long flowing hair.Now don't think that Mr Baldy hadn't tried to get his hair to grow.He had tried every product in the world but nothing worked.So after having spent thousands of pounds on hair-gaining products Mr Baldy started to collect hairy things instead, such as his dog.
One day when Mr Baldy was out looking for new hairy things to buy, a beautiful woman with long blond hair caught his eye.Mr Baldy fell in love on the spot just by the sight of her.So he started following her around, admiring her long hair, until one day when the woman who couldn't help but noticing Mr Baldy invited him to sit down beside her as she was resting on a bench.She introduced herself as Wendy Warey.The two soon fell in love, Mr Baldy more in Wendy's hair than in her person.
After a year or two Mr Baldy asked Wendy to marry him.After a lot of thought Wendy said yes and they hurried to the chapel to be husband and wife.Life for Wendy living with Mr Baldy wasn't easy.All he wanted to do was to stroke her hair, comb it and make fascinating braids.Wendy soon got tired of her husbands fixation of her hair and started to work on a plan to get her husband to stop liking hair that much.
Her plan was to put a couple strays of hair in Mr Baldy's food everytime they ate together.After having coughed up his 311 hairball Mr. Baldy's fascination for hair ceased and he threw away all his hairy things, got rid of the dog and pleaded Wendy to get a hair cut.And Wendy was pleased Mr Baldy had changed his mind.The German text is much shorter than the English one although they are both the mean length for their level and language.The story follows the pictures and provides an adequate description of the events.There are, however, not many original touches.The two compositions are typical in that the writers in English seem to treat the task mainly as a chance to show their imagination and to produce an interesting text.

GUn
The next section will deal with the questionnaire distributed to teachers of English and German.

The questionnaire
As was shown in the study by Linnarud et al (2005) there was a substantial difference between learners ability to communicate creatively in written free production in English and German although the goals for the courses are almost identical.In order to find out teachers' attitudes to and methods of teaching English and German we distributed questionnaires to teachers who taught both languages, 14 in all.Of these, six taught at classes 6 or 7 to 9 in the compulsory school and eight at senior secondary level.Two teachers with long experience from senior secondary level were interviewed.
The results of the questionnaire show that learners work considerably more with free written production in English than in German and that the tasks are more advanced.As the teachers say 'they know a lot more English.'The amount of time spent on writing in English varies, however, from once a week to once a term at an examination, with a couple of short practice sessions before the exam.We also received the answer 'Not often enough!' In German in classes 6-9 the tasks tend to be more closely related to the textbooks while in English the learners are encouraged to write diaries and short compositions.This tendency continues in class 2 in the senior secondary school.At both levels the majority allow the learners access to dictionaries and grammar books while they are working with writing.
The question as to how the teacher corrects the pupils' work gives varied answers.Underlining errors often followed by letting the pupils correct their texts seems to be used in both English and German.In a few replies, the teacher emphasises the importance of positive feedback and that the amount of correction can vary according to the learner's competence.Better pupils get more direct correction.There is, however, a tendency among teachers to correct morphological and syntactical errors more in German than in English.
Not surprisingly the question 'How much do you think the pupils come in contact with English and German respectively in their spare time, via e.g.computers, film, TV and music?' receives the answer that there is much more contact with English.One comment illustrates this: 'English -I am convinced that the teacher fills a very small function and that the computer, film etc. are of great importance.German -The teacher's role is important.A very small percentage see any of the above in German in their spare time.' The final question is 'What do you see as being the goals for teaching in English and German?Is there any difference between the two languages in what you are trying to achieve and, if so, what?' The teachers' responses show that they do not really think that there should be a difference in the goals for the English and German classroom.In both cases the learners should be able to use the language in different situations.The reality in school, however, differs from this picture.Many teachers find it very much easier to motivate pupils to study English in school than German.The learners are aware that they will need English for future studies and in their future professions.Many pupils find it difficult to see any point in studying German at school.Contrary to the facts they believe that English is enough.One teacher suggests that this lack of motivation on the learners' part might lead to lower motivation for the teacher to teach German than English.This is, of course, not a general observation.The fact that German is under threat might instead lead to greater enthusiasm among teachers.
One thing that is striking in the teachers' responses is that a processoriented approach to writing, with response from the peer group is not practised.One teacher mentions the possibility of letting other pupils respond if there is time, but it is almost exclusively the case that the teacher is in total control of the process.
The conclusions that can be drawn from the questionnaire are that although teachers would like to work with English and German in similar ways this does in fact not happen.It appears that it is tradition to put an emphasis on grammar in the German classroom and thus focus more on grammatical correctness than in English.In English more emphasis is placed on learning vocabulary.As one teacher referring to German expressed it, "We don't have time to learn vocabulary because grammar takes up so much time."Access to authentic material in English in schools is much greater in English than in German thus making it easier for the teacher to find texts for the learners to read and discuss.
The teachers' responses pinpoint the difficulties of teaching German in Sweden where English has such an unassailable position as the primary and, in some opinions, only foreign language necessary for international communication.Regarding writing and vocabulary there appears to be an acceptance of their importance in both English and German.The results of the composition analyses show, however, that the goals achieved by the learners differ considerably in the two languages despite the teachers' feeling that they should be the same.

Conclusions
In spite of the aims for teaching English and German being in all essentials similar there is a marked difference, not least in the ability of learners to use the languages in written free production.
The results of the quantitative analysis show that learners given the same written task write significantly more in English than in German in the same amount of time.There is also an increase in W/C for English from Gr through Gy to Un.This suggests steady progress.For German the largest W/C was found in Gy.The university students had a lower result.Any explanation for this must be pure speculation.
The measure of Lexical Density showed an increase in English between Gr, Gy and Un.After initially having higher results than for English, the German texts reached their lowest point in Gy and increased again at Un level.The differences at all levels are small and can only be regarded as tendencies.
The results for Lexical Variation based on a type/token ratio adjusted for the number of words in the text showed that English had a higher amount of variation than German when all levels were collated.There is more progress between levels in English than in German.The English compositions had a consistently higher percentage in the upper quartiles for results in LV when placed in groups according to the number of words in the text.The opposite was true for the German compositions where the majority were in quartiles 3 and 4.There is obviously more that can be done to refine the measure of LV.Not least a larger corpus would have given a more reliable result.The tendency is, however, clear that the lexis in the English texts was more varied than that in the German texts.
A comparison of the two groups that are at the same stage in the school system in spite of age differences, EGr and GGy, showed that EGr achieved higher results on all counts than GGy.The effects of the age difference between the two groups can only be a matter for speculation.It must be assumed that increased maturity should affect the results in some way, negative or positive.The fact remains that in spite of the official goals being identical the young learners of English produce more text with a higher degree of variation than the older learners of German.
The comparison of university students of English and German shows major differences between the two.The English learners have studied the language for a greater number of years than the German learners but both are faced with similar expectations as to language proficiency when they start university studies.Authentic texts are used in English and German and the students are expected to produce written communication at a high level of proficiency.This task is made difficult by the low starting point in German when entering university studies.
The questionnaires given to teachers of both English and German showed that writing is less practised in German than in English and that there is more focus on grammar and morphology in German than in English.In English communication is the main focus and use is made of more authentic texts.
The question remains whether it is possible to have a more communicative approach to the teaching of German and thus encourage more learners to choose to study what is after all a major means of communication in Europe.The Nödinge project would suggest that this is in fact possible and that it does lead to positive results.There are, of course, reasons why vocabulary receives less attention in German than in English.German morphology is highly complex.The three genders and four cases cause major difficulties for learners.An example is the definite article.In German the learner must keep track of 16 different alternatives to reach the correct result.The tradition in German has been to emphasise translation of German texts grammatically correctly.If that is the goal then learning demands an extremely high amount of time and practice, which can lead to vocabulary learning being neglected.Added to this the fact that the learner is exposed to very little German outside the classroom compared to English means that all learning has to take place in a couple of hours a week.
The situation for English is very favourable as there is easy access to authentic communication and an awareness of the usefulness of the language in the world outside the classroom.The aims for teaching German are, however, in all essentials the same as those for teaching English and it is to be hoped that both German and English can be taught in a way that stimulates interest and encourages the learners to use both grammar and vocabulary in order to seek information and communicate with the world surrounding them.

Figure 1 :
Figure 1: Mean W/C -English and German Gr, Gy and Un

Figure 2 :
Figure 2: LV and W/C correlated -EUn group

Figure 3 :
Figure 3: LV English and German -W/C groups compared

Table 4 :
W/C GGr, GGy and GUn compared