The role of attitudes in the development of Russian as a foreign language : A retrospective study

The article reports the findings of a retrospective study which looked at Hungarian learners’ attitudes towards Russian people, the Russian language and teachers of Russian. Mixed-methods sequential explanatory design (Ivankova, Creswell, & Stick, 2006) was applied which combines the collection and analysis of quantitative and qualitative data in two consecutive phases. First person accounts turned out to be a useful and relevant resource for exploring individual differences in proficiency in Russian. Differences in attitudes appear to play a role in the developmental process, but to what extent this refers to attrition or non-acquisition is unclear.


Introduction
The current literature (Szilágyiné, 2006; see also Dörnyei, Csizér, & Németh, 2006) suggests that Russian lessons were not effective in Hungarian schools, where it was taught as an obligatory subject from 1945 to 1988, mainly due to attitudinal factors towards the occupants.However, the period of occupation can be divided into three phases (Szilágyiné, 2006), and during the final phase, the students were probably less affected by the negative attitudes towards the Russian language and people.This study examines the attitude of the learners in the final phase by means of a motivational survey and in-depth interviews.Russian language proficiency was ascertained by means of a lexical test (recognition and recall) tapping the number of words the participants knew from a set of words in Russian.The assumption was that many factors play a role in the lexical development (i.e., acquisition and attrition) of words.By means of a mixed-methods sequential explanatory design (Ivankova, Creswell, & Stick, 2006), the study explored the extent to which a negative attitude towards the speakers of the language may have played a role in the recall of Russian words in Russian as a foreign language.The main goals of this study were to measure the impact of attitude(s) on proficiency in Russian and to identify different factors contributing to low or high language retention by analysing first person accounts.

Teaching and learning Russian in Hungary
Russian teaching in Hungary between 1945 and 1989 is considered by many people, including linguists (Dörnyei et al., 2006;Enyedi & Medgyes, 1998) and former learners and teachers of Russian, to have been a failure.It is thought that one of the main causes of this failure was the general negative attitude towards learning Russian; this was also the main assumption of this study.I supposed that among the measured variables, negative attitudes towards the language and the speakers of the target language (at that time the Russian "invaders") would have the strongest impact on the degree of language retention.Between 1944 and 1945 the Red Army invaded Hungary and drove out German soldiers and their allies from the country, an act referred to officially as liberation for 44 years.Due to several political events and agreements, Russian soldiers remained in Hungary until 1990, their number remaining at around 100,000 (Molnár, 1996).The relationship the citizens had with the soldiers changed from hatred to indifference during this period.The historical memory of Hungary is still processing the horrors of the first years of the invasion (1944/45) (see Pető, 2014), but as Ungváry (2002) notes, on an individual level there were those for whom the appearance of Soviet soldiers meant life, and others for whom it meant captivity, rape and uncertainty.For the majority of people the soldiers became almost invisible as they lived in their barracks functioning as a "state in the state" (Molnár, 1996).Thus, people had no or limited contact with the Russians in Hungary during the greater part of those 40 years.
Russian was initially introduced to Hungarian schools as a non-compulsory subject and only at some levels of the education system.At that time there was no history of Russian teaching in Hungary; therefore, basic materials and other essentials were lacking.There were no trained teachers, no textbooks, no teaching aids and no Cyrillic letters available in the printing houses.Lendvai (2005) claims that Russian language teaching and learning failed due to the lack of appropriate methodology, real opportunity for practice and the resulting lack of interest and/or positive attitude towards the language, as evident from the following quote from Bognár (2008) exemplifying the Hungarian prejudice against Russian: Between the end of the WWII and 1989, Russian was the compulsory foreign language in schools in Hungary.Almost all students started learning it at the age of 10 in the fourth or the fifth grade of primary education and studied it for at least four years in primary school and another four years in secondary schools.Thus, Russian was the number one foreign language.For obvious reasons, teaching Russian to Hungarians became synonymous with total failure in language education: after eight years of study of the language, the vast majority of Hungarians could hardly survive in Russian.A fine example of counter-motivation in education.(p. 6) This quote makes two basic assumptions that may not be valid: (a) that the teaching of Russian was a failure and (b) that this was due to a negative attitude.However, the compulsory nature of Russian may not have affected each period, school type, and level equally.Moreover, this assumption ignores the fact that some students were actually successful in acquiring the language to the extent that they were able to attend Soviet Union schools/universities, for both long and short term educational periods (Lendvai, 2005).However, there is little systematic research on Russian language teaching and learning in Hungary.One exception is Szilágyiné Hodossy Zsófia (2006), who compiled a comprehensive book on foreign language teaching in primary schools between 1945 and 1995 in Hungary.She divides the Russian period into three phases with different characteristics: • Pluralism (1944Pluralism ( -1949) ) was characterized by the fact that besides Russian other languages were taught in schools.The goal of Russian teaching was to teach the culture of the neighbouring nations.The period was characterised by a lack of basic conditions (teachers, books, dictionaries, teaching aids) for teaching the language.• Centralisation (1949)(1950)(1951)(1952)(1953)(1954)(1955)(1956) was characterized by the fact that no other foreign languages were taught in schools apart from Russian.The teacher training problem was seemingly solved by 7.5-month Russian training programs for teachers of other foreign languages, where, besides learning the language, they also had to gather knowledge about the culture, history, politics and geography of the Soviet Union.• Gradual liberalisation  was characterized by the fact that following the 1956 revolution restoration to previous times was impossible and gradually other foreign languages began to re-emerge in the educational system.In Russian teaching, ideology became less important, and there was an attempt to encourage students to use the language in real-life situations (Szilágyiné, 2006).As a result most of the students had pen pals in the Soviet Union.
The failure of 40 years of Russian teaching becomes clear from Vágó's (2000) discussion of the main tendencies of foreign language teaching in Hungary during the 1980's and 1990's.As she claims, the declaration of free language choice in Hungary happened in a European context, where, in most countries, it became clear that the main task of institutional education was to give answers to the global, regional and local challenges amongst which the development of communication competencies was of high priority.Foreign language knowledge, especially the knowledge of western languages, began to be valued in the Hungarian labour market, which created a need for its institutionalisation voiced by the parents (Vágó, 2000).This period was also characterised by the low level proficiency of the adult (older than 14 years) population; for example, in 1994 only 11.8% of adult Hungarians claimed to speak a foreign language, and 3.6% spoke two foreign languages besides the mother tongue (which is/was the European norm) (Terestyéni, 1995).According to Vágó (2000), these low results stem from a complete lack of motivation due to the long isolation of the country and the disinterest in the compulsory learning of Russian.She also claims that positive attitudes towards foreign language learning were formed in the 1980's and stabilised in the 1990's.
In summary, some authors (e.g., Szilágyiné, 2006) stress the lack of appropriate teaching conditions, such as the lack of teaching materials, as the main causes of failure.Others (e.g., Vágó, 2000) blame the low proficiency in Russian speaking among Hungarian citizens on the inefficiency of its teaching, the obligatory nature of the subject, a lack of interest, a lack of real-life situations in which the language might be used and the negative attitudes towards the speakers due to past experience.However, it is important to note that the claims are based on impressions, not on empirical findings.
The present study tried to establish empirically what the attitudes of students learning Russian between 1958 and 1988 were and indirectly measure the impact of these attitudes on the development of lexical skills in Russian.

Attitudes and motivation in FL learning
Attitudes and motivations are very complex, multidimensional constructs, which implies difficulties in their operationalisation.As Dörnyei et al. (2006) suggest, motivation is "intended to explain nothing less than the reasons for human behaviour.Because of this ambitious aim, there is no general consensus on the definition of the notion . .." (p.9).The concepts of attitude and motivation in the study of L2 learning often overlap and are in interaction with each other, given the fact that, for example, attitude towards the L2 community is a component of motivation.The idea of studying L2 attitudes and motivations by breaking them down into components, that is, attitudinal/motivational factors, comes from social psychology.Gardner (1985) developed a questionnaire, the Attitude and Motivational Test Battery (AMTB), which proved to be a powerful instrument for measuring attitudes and motivation of L2 and foreign language learners.
A difference between the original study by Gardner and Dörnyei et al.'s research is the presence/lack of the target language community in the vicinity of learners, that is, second language and foreign language settings.Gardner's general idea with the AMTB was to study the situation where the L2 community (French Canadians) was present in the country; thus, it was not a typical "foreign language learning" situation.While in Hungary, as Dörnyei notes, a "foreign language learning situation was faced where students learnt languages as school subjects with very little (if any) direct contact with members of the L2 communities" (p.10).Based on this discussion, Dörnyei et al.'s Language Disposition Questionnaire would appear to be the right choice to investigate Russian as a FL in Hungary; however, it measures real-time opinions and is therefore not suitable for a highly retrospective study.Furthermore, it focuses on generalizable motives rather than on situation-specific motives (e.g., attitudes towards the L2 teacher).Therefore, Gardner's AMTB may be a better instrument, especially when supplemented with personal interviews.Csillagh's (2015) analysis of the Swiss context highlights important motivational factors in language learning, pointing out that "theoretical developments have gradually moved away from the notion of a clear-cut division between internal and external factors in favour of a more complex representation of L2 motivation" (p.433).She found that different contextual elements play an important role in university students' motivation and attitude.Language learning does not happen in a vacuum, and the context is multi-layered including a linguistic, social, economic and political sphere.Dörnyei et al. (2006) refer to Russian teaching and learning in Hungary as dramatically ineffective: In 1979-1982 not more than 2.9% of the Hungarian adult population spoke Russian, which decreased by 1994 to 1% (Terestyéni, as cited in Dörnyei et al., 2006).They claim that Hungarians were reluctant to learn Russian "because it represented the oppressive power" (p.4) and they also lacked positive language attitudes.However, as previously mentioned, there is no empirical evidence on language attitudes and motivation from the period.

Measuring attitudes retrospectively
There is little research on the accuracy of retrospective attitudinal data.Gutek (1978) presents data that shows that with the passing of years the accuracy of attitudes does not necessarily deteriorate.Her research shows that "the assumption that attitudes in the past are recalled less accurately over time is unwarranted" (p.399).A review of the literature on the reliability of recall data (Dex, 1995) also shows that many types of retrospective data collection are actually sufficiently reliable, though of course for validity purposes very specific data with clear points of reference are needed.

Measuring effectiveness of FL teaching retrospectively
Foreign language learning through instruction requires a large amount of time, effort and devotion invested by both the students and teachers.Measuring the effectiveness of this effort retrospectively is not really possible, but it is possible to measure how much of the foreign language is retained.The assumption is that the better the language is taught, the more entrenched its forms are, and the better it is retained; but it is also a well-known phenomenon that if foreign language knowledge is not maintained (due to disuse or reduced input), it will start to decline.This phenomenon is called foreign language attrition, the non-pathological decrease of language skills in a language that had previously been learnt through instruction by an individual, a definition adapted from Köpke and Schmid (2004).
In this study, we cannot be sure how much had been learned to begin with, but we will assume that common nouns that occurred in all the textbooks were at least seen by the participants and the more words are remembered at the time of the study, the fewer words were forgotten and the higher the proficiency is.In the remainder of this article, the term attrition will be used, but it should be interpreted as the combination of non-acquisition and attrition.
The research question is whether there is a link between words remembered and attitude towards the Russian language or its people.

Participants
In this study 39 participants (29 females and 10 males) who had learnt Russian between 1958 and 1988, that is, the period called "gradual liberalisation," were tested.Participants were recruited and selected by the snowball approach (Goodman, 1961).To help prevent the occurrence of any age effect, the age range of participants was limited to 40-56.Participants were divided into three age categories (40-45: 9 participants, 46-50: 14 participants, and 51-56: 16 participants).The main criterion for selection was having had Russian as a mandatory subject before the dissolution of the Soviet Union (before the 1990s; see Bátyi, 2015).

Instruments and procedure
Several data collection instruments were used in the study to collect both qualitative and quantitative data: the AMTB, a short interview and a lexical test.

AMTB
The participants were asked to fill in an adapted online version of the Attitude and Motivation Test Battery (AMTB).In the questionnaire, the participants were asked to indicate on a 7-point Likert scale to which extent they agreed with the statements.There were seven randomized clusters: 1. Attitude towards learning foreign languages (ATLFL) (e.g., "If I were visiting a foreign country, I would like to be able to speak the language of the people"), 2. Attitudes towards Russian people (ATRP) (e.g., "Russians are a very sociable, warm-hearted and creative people"), 3. Integrative orientation (INT.O) (e.g., "Studying Russian was important for me because it enabled me to better understand and appreciate Russian art and literature"), 4. Instrumental orientation (INST.O) (e.g., "Studying Russian was important to me because I thought it would someday be useful in getting a good job"), 5. Attitudes towards learning Russian (ATLR) (positively worded items and negatively worded items, e.g., "Russian was an important part of the school programme"; "Learning Russian was a waste of time").6. Anxiety in Russian class (Anx) (e.g., "I always felt that the other students spoke Russian better than I did").7. Attitudes towards the teacher (ATT), which included positive and negative features (e.g., "friendly-unfriendly," "organised-disorganised"); unlike other parts of the questionnaire, participants had to indicate on a 6point Likert scale to what extent the positive or negative adjective characterised their teacher.
Table 1 shows Cronbach's a and the mean for each cluster.

The interview
The interview, as a complementary instrument, helped to "put flesh on the quantitative bones" (Singleton & Pfenninger, 2015, p. 12).While in the questionnaire attitude was operationalized by closed statements, the interview with the same questions elicited more detailed answers, which provided a degree of triangulation.

The lexical test
One hundred words were chosen to test participants' lexical knowledge.The stimulus words were selected from books that were used at the time of the language learning (e.g., Hlavács & Rédey, 1969;Kecskés, 1986;Szilágyi & Oszipova, 1989).It was important that the selected words appear in the word lists of most of these books.Only nouns were selected, and they were categorised into the following groups: people (e.g., boy, grandmother), places (e.g., house, shop), school-related (e.g., pen, desk), animals (e.g., dog, elephant), food and drink (e.g., milk, bread), body (e.g., hair, head), nature (e.g., air, tree), abstract (e.g., peace, song).To avoid a trigger effect, each level (recall and recognition) contained the same number from each category in a randomised order (Bátyi, 2017).Both on the recall and recognition level, 50 nouns were used as stimuli.
On the recall level, participants were asked to translate a given word from Hungarian to Russian, and on the recognition level, from Russian to Hungarian.

Design and analyses
The results of the lexical test served as the dependent variable.For such a complex construct as foreign language non-acquisition or attrition, it is important to be careful when drawing conclusions as predefined variables may not always fully explain variations.For this reason, I opted for a mixed-methods sequential explanatory design, which implies "collecting and analysing quantitative and then qualitative data in two consecutive phases within one study" (Ivankova et al., 2006).After the group level analysis, I looked at the individual data to find patterns in the quantitative and qualitative data.As a result of this approach, it was possible to put the participants into four categories or learner profiles.
As for the quantitative analyses an independent t test was used to compare the number of successfully retrieved words on the recall and the recognition level.A multivariate regression analysis was conducted to reveal the predictor variables, that is, to what extent the independent variables predict language attrition and/or retention.From the dependent variables (recall and recognition) a composite variable was created (their correlation was very high: r = .955,p < .001).For this part of the analyses, the results of the AMTB were used as independent variables.Traditionally, contact with and use of the language since the onset of attrition is taken into account; however, in this case, only a few participants mentioned the fact that once or twice they had given directions in Russian, had had a short conversation in the language or had read Russian on signs, which is why this dimension could not be included in the analyses.The following variables (extralinguistic aspects) were chosen: age (3 levels, see Section 4.1), sex, language exposure (the number of years of studying Russian), attitude towards Russian people (ATRP), attitude towards Russian language (ATLR) and attitude towards the teacher (ATT).In the case of ATLR and ATT, I wanted to create composite variables as both had positively worded and negatively worded items.The correlation of the positive and negative components of the ATLR (r = -781, p < .001,N = 39) enabled me to create the composite variable (after item reversal).For the teacher questionnaire the correlation between the two components was low (r = -.386,p < .005,N = 39), so I decided to include them separately: negative attitude towards the teacher (ATTneg) and positive attitude towards the teacher (ATTpos).

Quantitative analyses
Previous research (de Bot & Stoessel, 2000) suggests that participants recognise more words than they recall due to the fact that recognition and recall include different processes.The present results confirmed these findings: There was a significant difference between the two lexical test scores presented in Table 2: t(8) = -9.664,p < .001It is also important to note that the Min-Max-value range is very wide.In order to better visualize the spread of individual results, the z-scores were calculated for each participant and are presented in Figure 2.

Figure 2
The spread of individual scores on the lexical test Table 3 shows the variables which have predictive power on the level of retention.61.3% of the variance is explained by the attitude towards learning Russian (ATLR), number of years of studying, and sex.

Learners' profiles
While the multiple regression analysis works best with reduced numbers of variables, the pattern-seeking analysis on the qualitative side can include all of the relevant variables.Learner profiles were created to enable detection of commonalties among them in the interview data.
To reduce the number of variables from the AMTB, only clusters that showed clear differences among participants were used.The findings demonstrated few differences among the participants on the following factors, which were almost all positive: attitude towards Russian people, attitude towards the teacher, and attitude towards learning foreign languages.Therefore, these factors were not included in determining the profiles, and four measures remained: attitude towards learning Russian (ATLR; r = -781, N = 39, p < .001),integrative and instrumental orientation (r = .725,p < .005,N = 39), anxiety, and the score on the lexical test.For each participant, a mean for each factor was calculated on the independent variables (e.g., ATLR) and categorised as high (3.6-7) or low (0-3.5).Also, participants were split into two groups according to their score on the lexical test: high (51-100) and low (0-50).
Table 4 shows the distribution of the participants in the different learner profiles.Most participants were categorised according to these measures, although some of them (marked with asterisk) showed some "deviations" from the pre-set categories.* The participant fits the category except for anxiety; ** the participant fits the category except for ATLR; *** the participant fits the category except for anxiety; **** the participant fits the category except for orientation.Participant 67 is exceptional as he/she does not fit any of the categories.
Table 5 presents the means and standard deviations for each profile.The maximum score was 100 (50 points each for recall and recognition).Participants in Profile 4 scored higher than the participants of the other profiles.In the next section the four learner profiles are further analysed with reference to the participants' recollections and links and references to the (Russian) teaching and learning context of the period.

Profile 1
Overall, learners in this group lacked the motivation to learn Russian, which was fuelled by different negative components of the teaching and learning context.Most of them could not take the subject seriously due to the ineffective teaching methods.
#37 (at the college) in the first lesson our reading proficiency was assessed, and the results were catastrophic, even we, students could give this diagnoses . . .Our teacher then gave us the task to take the text and decide word by word if it was a noun or a verb . . .so all you had to do was saying: one, two, one, one, etc.
Participant #37 explained that the class was very heterogeneous in the sense that students came with very different language proficiency backgrounds, and the above detailed strategy of the teacher to cope with the situation was insufficient and ineffective, resulting in low motivation in the class.
Besides low motivation, low requirements can lead to a negative attitude, as was confirmed by another learner from the group, who claimed to have had several Russian teachers, none of whom put high demands on the class.
#38 There were no high requirements.We had a vocabulary copy book with 10-20 words in it and all you had to do was to learn the words by heart and get your 5 . . .It (Russian) was never as serious as History or Math.[5 is the highest grade in the Hungarian school system.]There was no real interpersonal contact between citizens and Russians in the period, which could have hindered the formation of a positive attitude towards the target language.A good requirement in the schools was to have penfriends from the target language community, although it was not obligatory in all schools.The success of this strategy to motivate learners was dependent on how it was implemented by the teacher.One of the learners (#1) from Profile 1 noted that in their school it was compulsory to have a penfriend, but the letters they sent were the same for the whole class and included sample sentences such as "It is raining here" or 'I am fine."The response was translated by the teacher, which led to disinterest and ignorance among the students.
The lack of instrumental motivation is exemplified by Participant #26, who claimed that he had no language aptitude and that this explains the low proficiency he had achieved in Russian, but, nevertheless, he learnt English when he lived abroad because he "needed" the language for communication.
Another component which contributed to the formation of negative attitudes was the compulsory nature of the subject: #37 Well . . . it was compulsory, there was nothing to love or not to love in it.
#1 The political system was such that we were forced to learn it and it showed in our Russian learning . . .we did not learn it for fun at all! #64 Because it was compulsory the only goal was to survive.#16 We were not in a good relationship, I considered it as necessary evil.
Generally, in the interviews there were no indications that the attitudes towards the language would have been shaped by the attitude towards its speakers, but in this extract a learner refers to negative feelings: #51 At high school, where my mind was just opening up I learned what damage the Red Army caused in the country and since then I had negative feelings.I wasn't adult enough to handle it separately from the language and the people.At the university it changed totally because I was in love with Russian literature and wasn't angry with the language any more.However, I had no motivation to learn it.

Profile 2
In this learner profile participants also scored low on the lexical test (< 50); however, their attitude towards learning Russian was positive as well as their integrative and instrumental orientation (> 3.5).Based on the questionnaire results, the only factor that could have contributed to the low performance are the high values on the anxiety measure.The interviews were analysed to find the sources of the anxiety of these learners.
Anxiety is a complex and multi-element construct (see, e.g., the anxiety scale in Horwitz, Horwitz, & Cope, 1986), and reconstructing it from the recollections of the participants is only one way to gather information; however, the interviews revealed multiple factors.Participant #5, for example, mentioned the teachers with whom she had no particular problems, the tasks she partly liked, her attitude towards Russians which, eventually, had no negative effect on her language learning (#5), and, finally, she briefly referred to the class as a community which, in her opinion, was somehow not good.Unfortunately, she did not elaborate on this remark.
Another participant (#55) reported on the traumatic experience when at secondary school she had become part of a class in which the teacher was not willing to say a word in her mother tongue, and only Russian was used.By the end of the first year, she almost failed but finally managed to survive the secondary school class with an acceptable level in Russian.
Participant #32 talked about the assignments and recitations in each class and punishment from the strict teacher, although her attitude towards the teacher was positive.Others (Participants #32, #44 and #2) referred to obligatory Russian learning as a "compulsion" and "duty." It is generally assumed that due to the Russian invasion the attitude towards learning the Russian language was negative.Apart from the quote from Participant #51, none of the interviewees referred to negative attitudes; all seemed to be rather neutral in this respect: #5 It was two different things what we learned at school and what we learned from our parents.In the school they tried to teach us the Russian culture, e.g., the Tretyakov Gallery, the monuments and famous buildings of Moscow, generally the culture.What I heard at home, that when they trooped into the town my parents found their behaviour very strange and unusual, and their thinking was so different from ours.So, for me it was hard to compare these two things, you know, what I heard from my parents and their experience and at school . . .There was this dichotomy.It was because at that time it was hard-core socialism.And it was strengthened in the education too, the big Soviet brothers, their culture, and . . .that we should be like them and follow them.That was in fashion to say so . . .Interestingly it did not influence my learning of Russian.I was interested in what I learnt at school [emphasis added].

Profile 3
This profile has only one distinctive feature in comparison to Profile 4, which is the score on the lexical test: Regardless of the high attitude measures and the low anxiety, the participants in this group performed low on the vocabulary test.
An interesting pattern emerging from the interview data of this group is that throughout the school years, participants lost their interest and motivation in learning Russian and the teacher had a key role in shaping their negative attitude.The following extracts describe a decrease in motivation and attitude.#3 I liked our primary school teacher very much, she had even no diploma, but she was very enthusiastic.I think I chose teaching English as a profession because I felt I had language aptitude . . .Our secondary school teacher was less good . . .I set myself to a minimal program.The methods of the teacher were not inspiring either, we had to learn rules by heart in Hungarian, and the words, no communication method could be traced.
#56 I had very nice Russian teachers and I loved the lessons too.But later, at high school, I could see we were not improving.The teacher had a great taste of humour . . .but we were not forced to study.Then I started learning English four years later and it was much more intensive.I think not because of the difference between the languages but probably the methods were more effective.
#29 My favourite was the primary school teacher, she could make me study.The second was the high-school teacher . . .At the university we basically repeated the highschool material.Effectiveness and methodology were far away from each other.
#36 I had a teacher in primary school, but in the second year we got our formmaster as a Russian teacher, who was a good teacher, but usually we put Russian aside and talked about class-related issues.At the university the teacher was just terrible.
In addition to negative feelings towards the teacher, participants also mentioned the ineffective teaching methods that had hardly any practical relevance: #36 It was very lexical without any practical knowledge.We wouldn't be able to ask a glass of water.If they had dropped us in Moscow we would have been able to find our way without any maps because we knew by heart where everything was.And we also knew the content and characters of Swan Lake by heart, although I think I would never need that knowledge.

Profile 4
Most of the participants (15) were put into this category due to the high scores on the lexical test, which was accompanied with high instrumental and integrative orientation, a positive attitude towards learning Russian and low anxiety.With only a cursory look at the interview data, it becomes evident why these participants performed well on both the recall and recognition test: Many of them took Russian as a major at university or took part in the 7.5 month Russian training programs for teachers of other foreign languages.But it is not clear why they decided to continue their Russian studies and what factors played a role in forming their positive attitude towards learning the language.
First of all, as most participants in the previous profiles, they did not identify the language with politics, as evidenced by Participant #54: "Many (older) people identified the language with the system, because both were mandatory, but I looked at them as separate things."In the following extract the participant refers to age-related issues in connection with the attitude towards Russian people saying that schoolchildren are not interested in political issues: #25 In my opinion we were a lucky generation.We didn't have to be scared that our parents would be taken from us if their way of thinking was different or they were not members of the Party.My parents were not members and I know there were some problems but we, children, knew nothing about it.They really protected us from any information . . .The truth is that I wasn't interested.I wasn't concerned or affected . . .I was a kid and all I cared for was that I could go and ride my bicycle, play tennis or play with my doll.
Another participant, who spent some time in Russia during her school years, also talked about Russians positively regardless of the fact that her father was in Russian captivity: #21 I can only talk about Russians positively . . .They have a very deep spiritual life despite the fact that they had to suffer through the years of communism . . .Both, the educated and ordinary people are likeable, hospitable and warm-hearted.My father was in Russian captivity but he also described them very positively.
Another factor which is revealed by the interviews is the role of the teacher in forming positive attitudes in the classroom towards learning the foreign language: #42 I loved my primary school teacher very much and I think I learned everything from her which is more or less stable since then.
#23 I loved it very much.Our teacher taught in a playful way, she used word-cards and authentic material.

Discussion and conclusion
This paper investigated a commonly held assumption that learning Russioan was ineffective in Hungary because Hungarians did not want to identify with its occupants.Thirty-nine Hungarians who learnt Russian as an obligatory foreign language between 1958 and 1988 were tested.After the AMTB was administered, the participants were interviewed and then tested with a lexical test.Two levels of lexical knowledge were tested: recall and recognition.In line with de Bot and Stoessel's (2000) study, participants recognised more words than they recalled.Regression analysis showed that the extent of retention is predicted by the attitude towards learning Russian (ATLR) and the length of language exposure.The explanatory power of time spent on learning Russian is not surprising, but contrary to expectations, attitudes towards the speakers of Russian (i.e.Russian people) have no impact on the number of retained word.If the context of language learning at that time is considered, these results are easily accounted for.Students in the period in question had had very few or no opportunities to have contact with the target language community (not only in Hungary but elsewhere in Europe too; see Littlewood, 1984), which means that their attitude towards learning the language could only be shaped in school by the teacher.This fact is reflected in the present findings, which show that attitude towards learning Russian has the strongest predictive power.Because the learning experience was limited to the classroom, the teacher played a key role in shaping the students' attitudes and in keeping their interest alive by using engaging methods and in setting requirements to the right level.
Based on the results of the lexical test (scores) and the attitude and motivation questionnaire, four learner profiles were created to explore further how the learners felt about the Russian people and the language.The interviews confirmed the finding that the learners did not have a negative view of the Russian occupants, and that teaching methods were highly ineffective, which was probably related to the low ATLR scores.The group with higher ATLR scores and medium lexical test scores were on the whole more positive about their teachers and lessons, especially those in elementary school.The one group with high scores were highly motivated with low anxiety, but most importantly they had continued exposure to Russian.It is not clear whether the motivation played a role in exposure, but having used more Russian was clearly related to the high scores.
From both the regression analysis and interviews, it may be concluded that rather than the attitude towards the Russian people, the attitude towards learning Russian was a strong predictor of retaining lexical items.The other strong predictor was the amount of exposure.
Thus contrary to the main assumption, negative attitudes towards Russian speakers and the extent of retention are not related.Authors (Dörnyei et al., 2006;Vágó 2000) often refer to the Russians as invaders and as such the main sources of negative attitudes towards learning their language by Hungarians.Our participants obviously had no negative attitudes towards Russian people as a whole as they did not identify the language with politics; they claimed they were simply too young to have any opinion on the topic.Results do show that teaching and learning Russian was not very effective.Participants complained of spending too much time on learning rules while getting little or no communication practice (see Lendvai, 2005;Szilágyiné, 2006;Vágó, 2000).
This study clearly demonstrated that the assumed connection between positive attitudes and language retention is more complicated than often assumed: The classroom environment and the role of the teacher have more impact on the extent of retention than, for example, the political environment or attitudes towards the speakers with whom students have no or limited contact.Although foreign language teaching methods and opportunities have changed tremendously since the dissolution of the Soviet Union, in many countries, including Hungary, it is still often the case that students have no real-life opportunities to use the language with the speakers of the target language (and form their attitudes), which often limits language use and language input almost exclusively to the class.Future studies on foreign language attitudes and motivation may take this into account.

Table 1
Reliability measures and means for the clusters

Table 2
Summary of scores on the lexical test

Table 4
Learners' profiles (the numbers are the codes of participants)

Table 5
Summary of the lexical test scores of the learner profiles