How advanced is advanced when it comes to speaking?

How advanced is advanced when it comes to speaking? Tomáš Gráf ABSTRACT: The huge achievement of becoming advanced in alanguage is often undermined by the feeling shared by the learners and their teachers that whilst there still is considerable progress to be made it is difficult to see what concrete steps to take. The study explores different ways of describing advancedness— using CEFR descriptors, the AdLs’ self-perceptions, the views of their teachers and, finally, through asample analysis of spoken production of advanced learners. The study argues that in order to prevent the advanced learner from remaining stranded on aplateau an active approach is to be taken by both the learners and their teachers using analyses of learner and native-speaker language and challenging the learners to use complex and idiomatic language well beyond the boundaries of their habitual language use. KEY WORDS: advanced learners, CEFR, learner language analysis, speaking 0. INTRODUCTION Defining levels of language proficiency is ataxing task, and defining ‘advancedness’ is no exception. Advancedness may be seen as afunction of time (as in “aman of advanced years”), but this is problematic in our context— some learners will attain the highest levels of proficiency in almost no time whilst others will never reach the heights even after many years of toil and moil. Perhaps amore fitting view is one which puts levels of proficiency on acline, where advancedness is at the extreme end or at the very top. But how is the learner to know he has reached this stage, or, indeed any other? Are there any common, typical features? What is the advanced learner (henceforth AdL) able and unable to do? How advanced can one become? These are just some of the questions we may legitimately ask. Knowing the answers to them could provide valuable guidance to both teachers who are often somewhat at aloss as to how to teach the most advanced, and also to the AdL himself, telling him which steps to take and in which direction. The following lines offer different perspectives of the AdL from the point of view of the CEFR and also considering the experience of AdLs, students of English philology, and their teachers at the Faculty of Arts, Charles University in Prague. The prime focus of the study is on spoken production. 66 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1. THE CEFR AND THE ADVANCED LEVEL (C2) Ortega and Byrnes (2008, p. 7) list four different approaches to defining advancedness: 1. taking into account the learner’sinstitutional status (e.g. students of philology at university level can be expected to be advanced); 2. using information from standardized tests; 3. evaluating which advanced features the learner has acquired; 4. evaluating the level of language sophistication. The learners and their teachers themselves are currently most likely to use descriptions from the Common European Framework of Reference (CEFR, 2001). Aiming to provide “acommon basis for the elaboration of language syllabuses, curriculum guidelines, examinations, textbooks, etc. across Europe” (CEFR, 2001, p. 1), it uses “can do” statements in an attempt to provide idealized descriptions of possible learning outcomes. It aims for maximum brevity in its descriptions, and it practically dismisses the model of native or near-native competence as apoint of reference. Let us explore to what extent the CEFR succeeds in defining clear, achievable goals for the AdL. 1.1 CEfR’SGLOBAL SCALE The CEFR’s(2001, p. 24) ‘Global Scale’— possibly its most widely quoted part— aims to provide asuccinct summary of key competencies. Its description of the C2 level reads as follows: Can understand with ease virtually everything heard or read. This descriptor is not without problems. What does “with ease” mean? And what is “virtually everything”? Can just about anybody aspiring to be aC2 user be expected to understand with ease, for example, alecture delivered in afairly simple language on the topic of quantum physics or generative grammar? As for “everything heard or read”, how far can we expect the learner to cope with all varieties of English? Is this actually achievable and desirable? Surely what plays arole here is the actual complexity of the task itself, and not only the inherent linguistic features. The Global Scale carries on: Can summarize information from different spoken and written sources, reconstructing arguments and accounts in acoherent presentation. Can express him/herself spontaneously, very fluently and precisely, differentiating finer shades of meaning even in more complex situations. As we see throughout the CEFR, there is agreat deal of non-specificity and subjectivity (Alderson, 2007; Hulstijn, 2007; Weir, 2005). What are the “different sources” mentioned here, and what is meant by “coherent presentation”? Besides, the concept of the advanced here seems to have been replaced by academic in amisconstrued confusion between linguistic and cognitive skills. Many highly fluent and coherent native speakers would not get past this requirement, as without relevant academic training their linguistic skills would not add up to the task. Reading the second quote TOMáš GRáf 67 we cannot help wondering what “spontaneity” is, how we measure “fluency”, what “precisely” means, and what we are to imagine under “finer shades” or “more complex situations”. The very existence of the global scale presupposes that learners might have the same proficiency across all of the skills, which is afallacy, as many alearner has experienced that whilst he may be ahighly proficient reader, his oral skills might lag seriously behind, and he might not even be able to understand spoken English very well at all. 1.2 CEfR— SPEAKING As the Global Scale does not seem to offer much help, let us see if the actual description of speaking could provide more guidance. This is the description of spoken interaction in the CEFR’s(2001, p. 27) self-assessment grid: Ican take part effortlessly in any conversation or discussion and have agood familiarity with idiomatic expressions and colloquialisms. Ican express myself fluently and convey finer shades of meaning precisely. If Ido have aproblem Ican backtrack and restructure around the difficulty so smoothly that other people are hardly aware of it. As above, there is adegree of non-specificity here (e.g. effortlessly, any conversation, good familiarity, finer shades of meaning etc.). Interestingly idiomaticity and colloquiality are mentioned. These are two features which we can take for granted with native proficiency and yet no direct comparison with the native speaker is attempted here. The description then switches from linguistic to academic again: Ican present aclear, smoothly-flowing description or argument in astyle appropriate to the context and with an effective logical structure which helps the recipient to notice and remember significant points. This has more to do with the skill of planning aspeech rather than speaking spontaneously. The speaker is supposed to provide alogical structure for the recipient’sease of orientation in the text, which can hardly be classified as purely alinguistic skill.1 1.3 CEfR— CONCLUSION It seems the CEFR can hardly provide the advanced learner with sufficient guidance as to which steps he is to take in order to achieve ahigher level of proficiency. The CEFR provides avague description of the AdL— whilst it tries to dispense with the native-speaker model it provides aportrait of alanguage learner who has many of his features (Abrahamsson— Hyltenstam, 2009; Birdsong, 2005; Coppieters, 1987; 1 Similarly the focus on academic skills is apparent in the description of writing, and the C2 learner is expected to be able to write summaries and reviews of professional or literary works. 68 STUDIE ZAPLIKOVANÉ LINGVISTIKY Piller, 2002). Moreover, it is very unspecific and does not cover aspects which really describe the learners’ level and concrete performance. It does not present aconcrete model for the AdL, neither does it point the teacher in aclear enough direction. As it is not language specific it cannot show which concrete aspects of the language in question the learner is to acquire and to what end. The AdL thus has to rely on the presumed expertise of publishers of language textbooks, who, however, provide adearth of information on the criteria which determine the selection of language elements for inclusion in their publications. Neither has the AdL any guarantee that these materials will map the most efficient way for him to follow (Granger, 1999). 2. HOW DO ADVANCED LEARNERS SEE THEMSELVES? In order to gain abetter understanding of the AdL, asurvey has been carried out amongst the students of English philology at the Department of English Language and ELT Methodology, Faculty of Arts, Charles University in Prague (Ledvinka, 2013; Gráf, 2012). The aim of the survey was to explore the students’ evaluation of their own English proficiency, their perceived strengths, weaknesses, problems and goals. The total of 103 students participated in the two surveys, and two students were interviewed. The students evaluated their own level of global proficiency2 as C1− (10%), C1 (51%), C1+ (27%), C2 (10%), bilingual/native-like (2%). When asked to explain why they have reached such ahigh level of proficiency the students most frequently reported it was owing to the length or intensity of exposure. This was explained by the following reasons: an early starting age; intensive contact with English through films, TV shows, games, and communication on the internet (Skype); intensive reading; long-term stay abroad or studying at an international school. Another significant factor seemed to be their opportunities, such as regular contact with native speakers, the ability to attend language courses on aregular basis, having good teachers, having the possibility to travel to English-speaking countries, and to quote one of the respondents “awealth of opportunities to use English”. Yet, amongst the most advanced were also students who had never had any of these opportunities and benefited solely form their dedication, hard work and regular contact with English through television or the internet. Anumber of respondents spoke of their love of English, wanting to become excellent and having atalent for languages. Others also mentioned that they are generally good at learning, whilst amere two students acknowledged using elaborate learning strategies. As for their strong and weak points, listening appeared to be the least difficult skill for them, whilst speaking the hardest. They also expressed their dissatisfaction with the level of their vocabulary knowledge, and admitted their grammar still needed some attention. Having been asked about what prevents them from becoming even better, the students gave the following reasons: lack of systematicity, unstrategic approach to 2 Whilst self-evaluation may not be entirely reliable, ahigh proportion of students at the department takes atest of general English proficiency either prior to or during their studies. TOMáš GRáf 69 studying English; complacency (i.e. being satisfied with what they have achieved already); not making aconscious effort to become more advanced; laziness. Aselection of some of their opinions is perhaps the best conclusion to this part: Ihave English relatives so I’ve been learning English since childhood. What really helped me though was watching movies or series in English. Idon’t go any further with my level of English because Ikeep on using aquantity of words (that are also used in those movies or series) and often don’t feel the need of learning new ones— Imanage very well with what Ihave. Ithink the main reason is just using English everyday for along time (translating the music lyrics, movies, communication with people, computers, books etc). As for not getting further, Ifeel like Ihaven’t made much progress for the last two years because Ididn’t have an opportunity to use more advanced language (for example most of my English speaking friends would never use some of the words in our compulsory vocab, even though they’re native speakers), so Iwas basically just practicing the stuff Iknow and not learning anything new. However, Ifeel like I’ve been making progress again since the school started. Why did Ibecome so good? In my opinion, because Irealized the immense interface between will and talent. When Istarted to like English, Itried to set myself in an English environment as much as possible. Istarted to read books in English. Istarted to watch films in English with English or no subtitles. Ifound some native speakers to chat with. Istarted to listen to BBC and RTÉ radio. My sister likes English as well, so we often speak English together. Istarted to talk to myself in English :-). We had aBritish teacher at grammar school, who influenced me alot. My pronunciation got better thanks to her. Iwent to Ireland for 3 months to work as an au-pair there. Istill feel my English needs to be improved, especially my speaking skills. It was better when Iwas in Ireland, but it got worse after Ireturned. I’m quite afraid of speaking again. 3. HOW ARE ADVANCED LEARNERS SEEN BY THEIR TEACHERS? When speaking to the most advanced, one cannot help feeling impressed by how much they have achieved. And yet, almost with all of them something always appears to be missing, and we can nearly always tell that they are not native speakers however close to the native standard they might be (Byrnes, 2012). What are the general impressions and most striking features? Perhaps the most significant fact is that we almost always know what they are attempting to say— the need to negotiate meaning rarely arises. Advanced learners manage to get their messages across, and they generally succeed in completing their thoughts. This happens at arate of speech of c. 160 words per minute which was 70 STUDIE ZAPLIKOVANÉ LINGVISTIKY found to be lower than in the parallel native-speaker corpus with the native speakers’ mean of 218 words per minute (Götz, 2013). Compared to native speakers they use more filled and unfilled pauses and frequently insert these within semantic units rather than between them (Götz, 2013). Serious errors (i.e. such which would obscure the meaning) are absent, and minor errors occur only infrequently. There is anegligible number of mispronounced words (word stress being the most common problem), which are however fully intelligible. The speakers make use of native-like selections (Pawley— Syder, 1983) but compared to native speakers at alower frequency (Götz, 2013, p. 102). All of this comes at some cost. In order to make themselves understood at arelatively high rate of speech the learners seem to adopt various avoidance and compensatory strategies3 especially with regard to complex structures and vocabulary. In this way they have become the masters of camouflage, skillfully hiding whatever they do not know or cannot do in the language. They overuse certain expressions which consequently become devoid of their actual meaning and which are used as almost contentless function words (e.g. the use of modifiers and intensifiers such as great, amazing(ly), incredible, incredibly etc.). The resulting production is correct and fluent but sometimes not particularly sophisticated. The learners seem to find it difficult to strike the right balance with idiomaticity: some have clearly not acquired or internalized asufficient range of native-like expressions whilst others try so hard to be idiomatic that they may sound unnatural. Many of the speakers seem to be overly concerned with the image they create of themselves, they want to be seen as advanced, native-like speakers or even pass for anative speaker (Piller, 2002) and when they are not sure about the appropriacy of achosen word, phrase or structure they want to use, they might try speaking extra fast or intentionally unclearly to mask any possible mistakes or gaps in their knowledge. Others are obsessed with accuracy and carefully monitor their speech so that they do not make the slightest slip or so that they can instantly correct it if they do. Many of them seem to be at aloss as to how to make further progress and why. Complacency is but one part of the problem but perhaps the most important one. Once they feel they can do almost whatever they desire with the language there is very little urgency to do any more work. By becoming so proficient at using what they have got they do not challenge themselves to use more advanced language, using avoidance and compensatory strategies means they do not have to experiment as much and test new language. In this light they have become too good for their own good. But there is yet another facet to it. At this stage, progress is hard to measure and hard to feel which can be demotivating both to the AdL and to his teachers. Without aguided approach taking into account all areas of development, the AdL is at ahuge risk of getting stranded on aplateau, where without sufficient practice and challenge he can start slipping away from the heights he has reached so far. 3 These are evidenced by frequent pausing, self-corrections and false starts. TOMáš GRáf 71 4. THE ADVANCED LEARNER OF ENGLISH— AN ANALYTICAL VIEW As has been stated above, the AdLs’ error rate is relatively low. In the analyzed sample of 16 speakers the error rate was 1.3 errors per hundred words. Some of the errors they make are fairly characteristic and Ihave attempted to analyze and classify them. For this purpose Ihave used recordings and transcriptions Ihave made for the Louvain International Database of Spoken English Interlanguage (LINDSEI).4 The subjects were all thirdor fourth-year students of English philology at the Faculty of Arts, Charles University in Prague. Their first language is Czech. The students were interviewed by anative English speaker following aset design. Each interview was preceded by abrief period of time during which the interviewees prepared for Task 1. In Task 1, the students spoke on achosen topic. Task 2 was adialogue covering common personal topics. In Task 3, the student had to reconstruct astory from aset of four pictures. Subsequently the interviews were transcribed using the LINDSEI transcription guidelines. To date 53 speakers have been recorded, 30 transcriptions have been finalized (i.e. transcribed and checked), and 20 have been error tagged. However, for the analysis below, only the first 25 interviews have been used. Figure 1 provides an overview of the metadata. Population Length of Aand B turns Length of Bturns Average length of study of English prior to university Average length of study of English at university Average length of stay in an English- -speaking country N = 25 64649 tokens 51335 tokens 9years 3.5years 5months4males 21females figure 1: Metadata. As asecond step, an error analysis and tagging were completed following the Louvain Error Tagging Manual Version 1.3. (Dagneaux et al., 2008). Figure 2 provides an overview of the types of most frequently occurring errors. Lexis Articles and determiners Tenses Idiomaticity Form Word order Complementation Other Total 143 (32.6%) 92 (21%) 49 (11.7%) 33 (7.5%) 31 (7%) 25 (5.7%) 9 (2%) 55 (12.5%) 437 (100%) figure 2: Asummary of frequency of the most frequently occurring types of errors. 4 LINDSEI is the spoken counterpart to The International Corpus of Learner English (ICLE). It is currently the largest spoken multi-national learner corpus, containing interviews with advanced learners of English (students of English at university level). To date, 12 countries have completed their contributions, giving LINDSEI 1.1 million words, and work is currently being carried out by afurther 9 universities. 72 STUDIE ZAPLIKOVANÉ LINGVISTIKY The overview presented in Figure 2 shows that the most frequent type of error is of lexical nature (10% of the total number of errors were owing to the wrong use of prepositions). This is in conflict with the results of Llach’sstudy (2006) which reports acorrelation between proficiency and the frequency of lexical errors, but in agreement with Lennon (1991). The second most frequent type of mistake is with articles and determiners. The third group contains mistakes with tenses (48% perfect tenses, 30% tense agreements, 22% other). Using Corder’s(1967) classification of errors, we can classify most of the errors as postsystematic as they are largely the result of an inconsistent use of arule. Corder distinguishes between errors (which the speaker cannot correct because he has not yet acquired the appropriate rule) and mistakes which are simply slips the student would be able to identify and correct. Interestingly, in our recordings we find examples of systematically occurring mistakes. These are instances of mistakes which the student appears to be making regularly whilst he would actually be able to identify and correct them. Even at this high level of proficiency, errors caused by L1 transfer— in this particular case the influence of Czech— appear to be the most common type (cf. Dušková, 1969; Granger, 1999): (1) but they didn’t find us on the beach and if they *did Ithink that they *would kill us… This should read “and if they had done … they would have killed us…” and can be explained by reference to Czech, where the present conditional is commonly used in place of the past conditional, which native English speakers do not do. (2) so much *of literature This should read “so much literature”, and is probably the influence of the genitive valency of Czech quantifiers. (3) especially on *Sunday This is how Czech refers to aregular action repeated on aparticular day. English needs the plural here, thus “especially on Sundays”. (4) we *should have read Ithink two or three books aweek The students occasionally do not distinguish between “should” and “be supposed to”, which is probably explainable by the fact that in Czech both concepts are expressed by asingle verb (‘mít’). The speaker meant to say ‘we were supposed to read’. (5) *on the picture This is one of the most common errors (30 occurrences in the analyzed sample), and aword for word translation of the Czech ‘na obrázku’. The correct form is ‘in the picture’. TOMáš GRáf 73 (6) we have just sports clothes … and *it’sused very much Pluralia tantum cause uncertainty about the agreement. The Czech equivalent of “clothes” is the singular “oblečení”, thus the correct form should have been “they are used very much”. (7) he told us *it’sraining everywhere Even at this level, tense agreements in reported speech are not fully automatic, but the students would be able to identify and correct the mistake. The correct form is “it was raining everywhere”, and the mistake can be explained by the use of the present tense in Czech in the same situation. (8) I’m in the *half of the book Here the student did not use the correct equivalent of the Czech “polovina”, which can be both “half” or “middle” with the latter being the only option here, thus correctly “I’m in the middle of the book”, or perhaps even more idiomatically “I’m half way through the book”. (9) we spent *much time; there were also *many people there Spoken English does not use much and many in positive declaratives. Whilst this might be the influence of Czech, it might also be explained as ateaching induced error. Amongst the most difficult aspects of grammar for the advanced feature articles (especially when expressing generic reference, and reference to abstract concepts), and the present perfect. In the following list Igive the erroneous example first, and the more appropriate option second, after the symbol →. (10) in *the society → in society (11) in *the recent years → in recent years (12) *the translation has always been adream of mine → translation has always been adream of mine (13) we’ve never done that ever again → we never did that again This erroneous use of the present perfect is fairly common in sentences with never. Granger (1999, p. 200) points out that teachers overemphasize the role of formal markers (e.g. ever, never, since and for) when teaching English tenses. Consequently, the error in question is probably an induced one— the use of the adverb triggers off the inappropriate use of the present perfect. Example (14) is of similar nature, in teaching the present perfect much emphasis is placed on expressing experience, and here the student confuses past experience relevant at present with an event anchored in the past. (14) it’sbeen great → it was great