2. The Council of Europe's Modern Languages Programme

The Council of Europe was set up in 1949 to strengthen pluralist parliamentary democracy, then still in a fragile state in a number of countries, to protect and extend human rights, to develop mutual understanding and respect between peoples and to promote cooperation among its member states in tackling common social issues. It was not until after the signing of the Convention on Cultural Cooperation in 1954 that the Council concerned itself with language policies. The Second World War and its aftermath had seriously weakened language learning. The relations between peoples had been totally disrupted for ten years and were still marked by mutual antagonism, distrust and ignorance. Transport facilities were rudimentary and international travel was still subject to heavy political and financial restrictions. International conferences, congresses and even journals were few and far between. For all the efforts of the reformers, language teaching had for the most part been confirmed in its traditional mode, concentrating on formal accuracy and the study of literary texts, often from an earlier period. Listening and speaking were of marginal importance – there was after all no great likelihood of meeting someone to hold a conversation with. By the middle of the 1950s, however, immediate post-war reconstruction was largely completed, travel restrictions were lowered and communication facilities of all kinds were rapidly developing. Young people in all countries were keen to see more of the world and to meet their counterparts in other countries.

The Council's first initiatives were taken at the instigation of the French government, which was anxious to protect and restore the French language to its pre-war position as the main medium of international communication in diplomacy and cultural affairs. It was impressed by the work of the Center for Applied Linguistics in Washington and saw the potential of developing technologies. As a result, a major ten year project was launched in 1964, covering all sectors of education, with the aim of modernising language teaching at all levels. The project strongly supported the development of applied linguistics by launching the Association Internationale de Linguistique Appliquée (AILA), and promoted the use of audiovisual methods and materials in schools. In adult education it was thought that the concept of unités capitalisables, allowing students to build up qualifications in a modular fashion over a period of time as opportunity allowed, might be applicable to language learning, and in 1971 a working group was convened to investigate the feasibility of setting up a European system of this kind.

3. Objectives for Language Learning and Teaching

It was clear to the group, bearing in mind Meillet's dictum: ‘Chaque langue forme un système où tout se tient’, that it was not possible to organise a language into discrete modules in the way proposed for other curricular subjects such as ecology. Nor was it thought advisable to separate from each other the so-called four skills, or the different domains of language use. Instead, attention was concentrated on the basic principles for the construction of a language learning system workable on a European scale but flexible enough to respond to the widely varying needs, interests, characteristics and situations of adult learners. First, however, the group had to make some fundamental decisions regarding its basic aims, which should be closely geared to the general educational and political aims of the Council of Europe. They were agreed to be:

• • To facilitate the free movement of people, information and ideas in Europe with access for all and to encourage closer cooperation by providing the linguistic means of direct interpersonal communication in speech and writing, both face-to-face and at a distance.

• • To build up mutual understanding and acceptance of cultural and linguistic diversity in a multilingual and multicultural Europe, with respect for individual, local, regional and national identities, developing a common European intercultural identity by unforced mutual influence.

• • To promote the personal development of the individual, with growing self-awareness, self-confidence and independence of thought and action combined with social responsibility as an active agent in a participatory, pluralist democratic society.

• • To make the process of learning itself more democratic by providing the conceptual tools for the planning, construction, conduct and evaluation of courses closely geared to the needs, motivations and characteristics of learners and enabling them so far as possible to steer and control their own progress.

• • To provide a framework for close and effective interaction and cooperation among all the individuals, institutions and authorities engaged in the organisation of language learning.

Over the years these aims have gained wide acceptance in the language teaching profession. In pursuing them, successive projects have consistently advocated and promoted the following principles:

• • A systems development approach should be adopted, interrelating aims, objectives, methods, materials assessment and evaluation.

• • Objectives should be appropriate, desirable and feasible:

  • ○ appropriate in the light of the characteristics of the learners

  • ○ desirable in the light of the needs, (vocational, recreational, cultural and personal) and motivations of learners and the interests of society

  • ○ feasible in the light of the human and material resources which can be brought to bear.

• • Methods and materials should be selected and/or developed which are appropriate to learners, teachers and the learning situation and used so as to achieve the agreed objectives.

• • Methods of assessment and evaluation should be employed and developed which are directly related to learning objectives and provide accurate and relevant information to learners, teachers and other interested parties.

• • The approach to language learning and teaching should be learner-centred rather than subject- or teacher-centred.

• • The function of teachers and other partners is to facilitate appropriate and effective learning.

• • There should be consultation and agreement on objectives and methods among all the partners for learning.

• • A permanent education perspective is needed, as developing individual learners are enabled, in a life-long process, to understand and to communicate in diverse ways for diverse purposes in a multilingual and multicultural Europe.

• • All educational programmes should involve preparation for future independent learning.

The group saw the specification of worthwhile, appropriate and feasible objectives to be central to any system constructed in accordance with these guiding principles. Ideally, since all individuals are unique, every learner should be free to follow objectives and methods uniquely suited to him- or herself. However, the practicalities of language teaching on a social scale place strict limits on individualisation. It is necessary, when dealing with large numbers of learners, to identify and follow objectives which embody their common needs and interests. Given agreement on a shared objective, the various agents, often acting independently, who are responsible for the many aspects of the process can then work together harmoniously. These include bodies ranging from ministries setting out curricular guidelines and controlling qualifications, through multimedia course designers, textbook publishers, teaching institutions and examining bodies to the classroom teachers and, ultimately, the learners themselves. Such collaborative working is important even when these groups are in competition with each other or working in different countries under different educational systems – an increasingly important consideration as increasing educational and vocational mobility across the continent made the portability of qualifications of ever greater significance. The group placed the development of a model for specifying objectives at the centre of its concerns.

One problem immediately arose. The process of language learning is lengthy and continuous. Was it possible to identify natural breaks which would justify our dividing the process into a series of levels, providing a succession of definable objectives to satisfy our criteria, the attainment of which could be recognised as capitalisable in a European unit or credit scheme? If not, any system of levels would be arbitrary, or the result of other more practical considerations, such as the organisation of the academic year, the viability of class sizes, the scope of published textbooks, etc. On inspection we found that the practice of different institutions in the field of adult education varied widely. At that time, UCLES regarded its two levels as the only ones of public interest, whilst Trinity College, by analogy with its popular music grade tests, recognised many more, largely for motivational purposes. We felt that the traditional annual division of classes into beginner, intermediate and advanced, each further divisible into upper and lower, had the force of common sense and pointed towards a six-level scheme. However, beyond proposing brief level descriptors, we temporarily shelved the decision. There was, however, one objective we thought might be considered a natural level. In the earliest stage of language learning, a student learns to do specific things, such as greeting (hello, goodbye), identifying and naming objects (this is a book). Lexical and grammatical resources are patchy, with large areas still unexplored. It is not possible to speak of a level of communicative ability until these bits and pieces of language cohere into a competence that can be called upon across the more common situations of everyday life. We felt that at this point there was a threshold effect and so termed this level the ‘Threshold Level’. We thought at the time that this would be the lowest level in the system. We decided to concentrate our efforts on developing a model for the specification of this level.

4. The Threshold Level

The first possible model to consider was that of le français fondamental, consisting of a basic vocabulary selected on word frequency criteria and a set of basic grammatical structures. This had been used as the specification for the successful audio-visual course Voix et Images de la France and the model had been adopted by the International Certificate Commission (ICC) set up by the Volkshochschulverein in Germany, Austria and Switzerland for its suite of certificates in a range of languages, including English. Our Dutch colleague Jan van Ek, a pupil of Zandvoort with long experience and expertise as a grammarian of English, had worked on this project and drew upon it in a first draft definition of the Threshold Level (Trim et al. Reference Trim, Richterich, van Ek and Wilkins1980: 101–28). However, the group's thinking moved in a different direction, partly because of scepticism with regard to a lexicostatistical basis for the selection of vocabulary, partly because we wanted to escape from the formalism of structure drilling and the memorisation of constructed dialogues, which characterised the audio-lingual and audio-visual methodologies of the time. We wished to displace language forms from the centre of attention, not that they were unimportant, but because we saw their mastery not as an end in itself but as a means to an end, namely the satisfaction of the universal human need to communicate feelings and ideas and to get things done.

This pragmatic orientation was very much in the British tradition, particularly as promoted by H. E. Palmer and A. S. Hornby. In a seminal paper (Trim et al. Reference Trim, Richterich, van Ek and Wilkins1980: 129–46), David Wilkins stated ‘since the threshold level is, by definition, a limited competence, its content will be determined by the minimal set of notions that will permit communication with native speakers in a typically European environment’. He then set out a scheme dividing notions into ‘semantico-grammatical categories’ and ‘categories of communicative function’. On the basis of these studies and drawing on a wide range of sources from Austin's How to Do Things with Words to Palmer and Blandford's Everyday Sentences in Spoken English, also using the scheme of Sprechintentionen developed by Anthony Peck for the Nuffield Foundation's course Vorwärts and a similar scheme drawn up by Svante Hjelmström for the Swedish distance learning scheme In the Air Footnote ¹ , extensive lists of notions and functions were drawn up and distributed on a restricted scale by the Council of Europe. Over the following months this material was consolidated and reorganised by Jan van Ek and published by the Council of Europe in 1974 as The threshold Level, followed in 1976 by The Threshold Level for Modern Language Learning in Schools. This work had an immediate impact on the teaching of English well beyond the expectations of the group. Though intended as just one exemplification of a minimal level specification for one defined audience, in other words ‘people who want to prepare themselves to communicate socially with people from other countries, exchanging information and opinions on everyday matters in a relatively straightforward way and to conduct the business of everyday living when abroad with a reasonable degree of independence’ (van Ek & Trim Reference van Ek and Trim1991a: 1), it provided a common point of reference for educational authorities, examination bodies, textbook authors and teacher trainers in a language that was comprehensible and relevant to teachers and learners, who felt that their needs were being directly addressed.

The Threshold level model was widely followed. The functional and notional categorisation became an accepted part of mainstream planning and the ability to communicate ideas, emotions and intentions in everyday situations gained acceptance as the overarching aim of language learning, teaching and assessment, except in more traditional university departments. There were, however, a number of misconceptions. One was that the Threshold level objective was the sole objective for all learners up to that level, crowding out the diversity of provision in response to the diversity of learner needs, motivations and situations advocated by the project. Our French colleagues were so anxious to avoid that danger that they produced a counter model in Un Niveau-seuil (Coste et al., Reference Coste, Courtillon, Ferenczi, Martins-Baltar and Papo1976 ), giving an extensive survey of the resources of French for communication at a modest level, which was too rich to be taken as an early learning objective, obliging teachers, course designers, examiners, education authorities, etc. to make choices appropriate to the learners concerned, whereas in the model exemplified for English choices were made for the defined audience. Users were then invited to make any changes they felt to be appropriate. Un Niveau-seuil was thus more a framework than the specification of an objective. In the event, most users preferred to adopt and adapt the Threshold level model. Copyright law and the pressures of commercial competition ensured variety and innovation in the English language teaching industry. Other misconceptions were that it was minimalist, that it abandoned grammar in favour of a glorified phrase-book approach, that it had no cultural content and limited itself to transactional exchanges. The apparent neglect of grammar arose from the fact that many users took up with enthusiasm the categories of functions and specific notions, which they found new, concrete and appealing, but left aside the chapter on general motions, which was more abstract, highly grammatical and difficult to exploit directly. In fact none of the lexical and grammatical content of van Ek's first draft was lost. It provided many of the exponents of the functional-notional categories. The phrasebook impression was due to the fact, emphasised by Palmer, that many communicative functions in everyday life are indeed performed by fixed formulae rather than synthesised ad hoc from general lexical and grammatical resources. The absence of explicit cultural objectives was a consequence of the defined target audience. It attracted criticism particularly from the surviving adherents of the older elitist order of language teaching based on the aims and methods of the teaching of Latin and Greek, still strong at university level, as well as from those who believed Landeskunde, the study of the geography and the various institutions of the country concerned, to be an important aspect of language teaching. This criticism was stronger with regard to languages more closely identified with one country than with English, which is polycentric and learnt more as a medium for global communication. It was as important for learners to express their own culture in English as to know, let alone adopt, one or more of the many cultures and subcultures of its native speakers. In fact, the cultural neutrality of the Threshold level may well have been one of the reasons that the model could be followed with only minor adaptations in drawing up similar specifications for some 25 other European languages.

5. The Waystage and Vantage Levels

As to minimalism, though at the time the Threshold level was considered as the earliest point in the language learning process at which it would be possible to speak of a general level of linguistic competence, experience soon showed that this was not the case and that the objective as specified was far from minimal. When it was used as the basis for courses in the Viennese Volkshochschulen, it was found that after a one-year course under normal conditions average adult learners had acquired some 70% of its content receptively, but no more than 33% productively. When it was proposed to adopt the Threshold level as the objective for the Anglo-German multimedia coproduced English course Follow Me, it was felt to be necessary to set a first-year objective with about half the content. This work was undertaken by Jan van Ek together with the Longman's textbook author Louis Alexander, in association with Tony Fitzpatrick of ICC (the International Certificate Commission). As it was not thought to qualify as a ‘level’ but rather as ‘one of many possible intermediate objectives on the way to threshold level’ (van Ek & Trim Reference van Ek and Trim1991b: 1) it was termed Waystage. In fact, the authors found it possible to maintain a wide functional, notional coverage by extracting ‘what were considered to be the most basic categories within each of its parameters – the most essential situations, topics and functions, inescapable general notions and their simplest and most basic lexical and grammatical exponents’. Their task was made easier because in composing the Threshold level the desire to give as full an account as possible of notional and functional categories had outweighed the desire to produce a minimal learning objective and much economy could be produced by eliminating or reducing alternative exponents.

For some years, attention was focused on the use of the Threshold level and the concepts and methods associated with it to modernise the teaching of foreign languages, especially English, in schools across Europe and indeed more widely. However, in the course of the 1980s pressure grew for a more complete system. Following a series of preliminary studies (Trim et al., Reference Trim, Holec, Coste and Porcher1984), van Ek wrote a summative work Objectives for Foreign Language Learning dealing with their scope and level (van Ek Reference van Ek1985–86). In that work he presented a framework for comprehensive foreign language learning objectives, distinguishing the following components:

• • Communicative ability

  • ○ linguistic competence

  • ○ sociolinguistic competence

  • ○ discourse competence

  • ○ strategic competence

  • ○ socio-cultural competence

  • ○ social competence

• • Optimal development of personality

  • ○ cognitive development

  • ○ affective development

In further response to pressure from the field, van Ek and I were commissioned to review and if need be revise the two existing levels and to add a third, to be as far above Threshold as Waystage was below it. In the revisions, the treatment of functions and notions – general and specific – were little changed, the main innovation being the addition of two new categories of language functions: ‘structuring discourse’ and ‘communication repair’, which had been introduced in Kontakte, the German version of T-level, and the introduction of several open-ended items in the list of specific notions, specifying thematic fields rather than a particular selection of words from those fields. The intention was to encourage diversity in accordance with the particular needs and interests of individual learners. The principal innovations were ‘that now those components of communicative ability which particularly allow it to be related to a wider educational context are identified and explicitly incorporated into the objective’ (van Ek & Trim Reference van Ek and Trim1991a: 7). They included discourse strategies, a sociocultural component, compensation strategies and a ‘learning to learn’ component. The grammatical content was systematically set out in a 30 page appendix. In another, objectives for pronunciation and intonation, a previously neglected field, were introduced.

The same model of specification was used for the third level. This was given the title Vantage, as representing a more elevated viewpoint conferring an advantage. Learners were ‘not so much called upon to do entirely new things in the language, as to meet the challenge of daily living in a more adequate and satisfying way, less restricted by limited resources’ and with ‘a more fluent and accurate control over the communication process’ (van Ek & Trim Reference van Ek and Trim2001: 2). Accordingly, Vantage went beyond Threshold particularly in the following respects:

• • the refinement of functional and general notional categories, especially with regard to the expression of emotions and the conduct of discussion

• • enlargement of concrete vocabulary in wider thematic areas according to the individual learner's needs and interests

• • recognition and limited control of important register varieties

• • understanding and production of longer and more complex utterances

• • increased range and control of conversational strategies

• • greater sociocultural and sociolinguistic competence

• • improved and diversified reading skills applied to a wider variety of texts

• • a higher level of skill in carrying out the processes of language reception and production.

With the publication of Vantage a system was now in place which would ‘offer to all practitioners a description of the language needed to assure a learner's ability to deal effectively with the challenges presented by everyday life, presented at three levels rising from a minimal equipment to deal with the highest priority needs, through the minimum needed to deal with the full range of requirements for a visitor or temporary resident, to an enriched equipment adequate to deal effectively with the complexities of daily living’. The use of a single model for the successive levels allowed for – and encouraged – considerable flexibility in the use of the descriptive apparatus to define appropriate objectives for a particular set of learners, supplementing, replacing or eliminating categories and elements as necessary. The material could be divided up into more or fewer levels or stages of learning as the organisation of a particular institution or educational system indicated. The concept of ‘level’ was loose. The series had developed one by one in response to pressure from the field rather than as an organised overall system of levels and since the aim had been to set out in some detail needs-related objectives, questions of comparability and standards of attainment and qualification did not arise.

6. The Common European Framework of Reference for Languages

It was such considerations that led the Swiss Government in 1991 to request the Council of Europe to undertake the construction of a common European framework for languages, covering their learning, teaching and assessment. Switzerland has no federal ministry of education, which is controlled by Cantonal authorities. Problems had arisen in relating qualifications obtained in one Canton to those awarded in another. This was a microcosm of a more general European situation produced by accelerating educational and vocational mobility. The need was felt for a commonly accepted point of reference against which the diverse qualifications awarded by institutions of all kinds could be calibrated. There was, of course, no desire to limit the diversity of provision, rather to encourage it by enabling all to find an appropriate niche in the overall scheme. In line with the aims and values of the Council of Europe's language policies, awareness and independence of thought and action of language professionals and of the learners themselves should be strengthened by giving them a clear and comprehensive account of what competent language users are called upon to do in order to communicate effectively with other users of the language and of the competences, i.e. the knowledge and skills, that enable them to do so. It would also be helpful to learners and professionals to have a clearer picture of what learners could be expected to do at successive stages in the learning process. Those involved in teaching, assessment and curricular planning would also benefit from a survey of the options open to them. This apparatus as a whole would enable practitioners to reflect on their current practice, setting it in a wider context and, hopefully, improving communication and cooperation amongst them in an international context.

The Common European Framework of Reference for Languages: Learning, teaching, assessment (Council of Europe, 2001) covered all the above aspects: a taxonomy of the parameters of language use and competences, descriptors of language proficiency, activities and competences at six levels and options for learning, teaching, assessment and curricular planning. Its publication was very warmly received by the field, but while the account of language use and competences was well taken up in the area of teacher training, attention was otherwise focused on the system of six levels of language proficiency. It rapidly became clear that it was meeting a real need for a simple, robust and easily intelligible descriptive scheme for calibrating language proficiency, within which publishers, examining bodies and so on could situate their products. The choice of six levels, or rather bands, of proficiency was sufficient to distinguish the major stages of language learning without being so fine as to cause difficult problems if relatively minor differences of interpretation led to inconsistency or even disputes over boundary points. It allowed a fair margin of flexibility to accommodate relatively minor differences within the same band. The descriptors, particularly those for the global scales which were generally used, were brief and selected from the stock of existing attempts at specification by a process that ensured that they were readily intelligible to practising teachers and commanded their agreement as to the ordering of the levels as described in the learning process. The descriptors were broad and flexibly interpretable within commonsense limits, which made them widely acceptable and avoided the disputes that any attempt to impose a more rigid, finely differentiated and detailed logical system would certainly have evoked.

No attempt was made to define lexical, grammatical or functional inventories for any of the levels. Instead, a ‘can do’ principle was adopted, both for the global scales and those set up for particular activities and competences. This followed a trend, established in the level descriptions developed by ALTE and used also in the DIALANG scheme, as well as already to be seen in the 1990 revisions of Threshold and Waystage, and particularly in Vantage, in which the chapter on specific notions was replaced by one on topic-related tasks and lexicon, consisting only of can-do specifications, lexical exponents being relegated to an appendix. The level descriptors in the Framework were therefore lacking in linguistic content, leaving greater freedom and responsibility to the user. However, a link with the T-level series was established by giving the names Waystage, Threshold and Vantage to levels A2, B1 and B2, to which they roughly corresponded. Although never stated explicitly, the nomenclature encouraged those looking for a concrete exemplification of those levels, which were those with which most courses leading to qualifications were concerned, to look at the T-level series. In fact, of course, the processes by which the successive specifications in the T-level series were arrived at were unconnected to those by which the Framework descriptors were selected. There was no guarantee that the two would exactly correspond. The Vantage specification, for instance, pioneered the way forward from the Threshold level, but in the absence of any higher level specification its upper limit was unspecified and it appeared to many that it contained some features which in the framework system might appropriately be placed at C1.

7. Profile deutsch and English Profile

In the case of languages for which only a Threshold level description had been developed, neither upper nor lower limits had been established. This was felt to be the case, for example, with the German Kontakte (Baldegger et al., Reference Baldegger, Müller, Schneider and Zusammenarbeit mit Näf1980), the authors of which had drawn upon both the Threshold level and Un Niveau-seuil as well as introducing some new categories. The authorities in Germany, Austria and Switzerland considered that the specification of learning objectives for German as a foreign language should be reorganised in the light of the Common European Framework, with the content of Kontakte distributed over levels A1 to B2, supplemented as necessary and new level descriptions developed for levels C1 and C2. A trinational team was set up for this purpose and the results of its work were published as Profile deutsch (Glaboniat et al., Reference Glaboniat, Müller, Rusch, Schmitz and Wertenschlag2005). This project was innovative in a number of ways, including its publication partly in hard copy, partly as a computer program enabling the various lexical, grammatical, functional and notional components of the specification to be searched and combined, selected and augmented, in a more flexible manner. One consequence of the flexibility introduced by the computer program as opposed to hard copy publication, one to which the team attached particular importance, was interactivity between user and disc. Individual learners as well as course designers could amend their disc so as to build up their own inventories according to their developing experience. In this way, each user could construct his or her own unique profile (hence the title of the work).

The team developing Profile deutsch found the definition of levels C1 and C2, for which no specifications had been produced in the T-level series, a formidable undertaking. ‘It is clear, that the language needs for learners at a high level are hard to define ‑ the higher the language level, the more complex the actions involved, the more specific the needs of the individual learner and so the less linguistic means and structures can be laid down as a whole’ (Glaboniat et al., Reference Glaboniat, Müller, Rusch, Schmitz and Wertenschlag2005: 6). The team therefore decided to confine themselves to setting out can-do objectives at these levels. As with other levels, the can-do statements were set out for the various categories of activity according to their treatment in CEFR: interaction, reception, production and mediation, in each case, distinguishing spoken and written and divided into ‘global’ and ‘detailed’. The latter were accompanied by specific examples. The scheme is very thorough and logical, but can in places appear long-winded and redundant, as some statements can occur in a number of places with or even without minor changes of wording.

The decision of the Profile deutsch team not to attempt any selection of functional/notional, lexical, grammatical, idiomatic or phonetic/prosodic elements or structures beyond level B2 is certainly understandable. It is not that their specification at level B2 is exhaustive. Experience of all kinds, in addition to any formal instruction, will have brought a great expansion of vocabulary (words, idioms, regular collocations) at C1 and again at C2, which itself cannot represent an exhaustive, closed system. The total lexicon of English, even if one limits oneself to the content of a major dictionary, such as Chambers' or The Shorter Oxford English Dictionary, is beyond the competence of any native speaker. It is very difficult to see how any formal limit could be set, especially if the higher levels are characterised, as is claimed, by increasing specialisation. But is that in fact the case? Specialisation is less a matter of language than of vocational and leisure interests. Professionals whose general language ability is at level B1 may be required to read documents or even give presentations in their field involving technical language which no-one would consider specifying in a general description of that level. A general specification at the C1 or C2 level would also have to abstract away from individual specialisation and attempt to identify what might reasonably be expected of any language user able to satisfy the general can-do criteria for that level. One possibility might be to base oneself on for example the language not of headwords but rather of definitions in a major unilingual dictionary since these are chosen to be known to any well-educated user. Whether that would be feasible is by no means clear, but should it be excluded a priori? Level C1 in CEFR is not envisaged as an unattainable ideal (as tends to be the case in some schemes) but as what may be expected on a social scale as the final objective of a general educational system, beyond which the organisation of language courses and qualifying examinations is no longer practicable. It is, of course, in that way the apex of a pyramid and the numbers reaching that level are likely to be relatively limited, but the level as defined should meet the criterion as stated. It is then open to any organisation wishing to reach higher for a small number of highly trained specialists to propose and define a D level. It was in order to leave that possibility open that A1 was made the lowest rather than the highest in the system.

English Profile is a major, many-faceted project, offering a coherent framework within which knowledge, understanding and experience of many kinds can be brought together and further developed in an interactive process. It is a logical application and an important step forward from the projects described above, developing over the past 40 years, just as they built on foundations laid by Jespersen, Palmer and others. Innovation is perhaps most valuable when it develops, rather than rejects, the achievements of the past, providing its own distinctive contribution to a continuous, living tradition.