ENGLISH FOR SPECIFIC PURPOSES: WHAT IS IT AND WHERE IS IT TAKING US?

English for Specific Purposes (ESP) is teaching with the aim of assisting learners’ study or research in the particular variety of English they may need. It has emerged from over 50 years of research and classroom practice and has become a major influence in university and workplace classrooms in many parts of the world. The basic idea behind ESP is that learners’ needs differ enormously according to future academic or occupational goals, and this is why ESP has become so influential in universities around the world in recent years. There is a growing awareness that students have to take on new roles and engage with knowledge in new ways when they enter university and, eventually the workplace. They find that they need to write and read unfamiliar genres, and that communication practices are not uniform across the subjects they encounter. Simply, the English they learnt at school rarely prepares them for that which they need in Higher Education and in the world of work. In this paper I sketch some of the major ideas and practices that have shaped contemporary ESP and look at the main effects it is having on language teaching.


Introduction
English for Specific Purposes distinguishes itself from more general language study through a focus on particular, purposeful uses of language, or what Cummins (1982) refers to as 'context-reduced' language. This tends to be generally more abstract and less dependent on the immediate setting for its coherence than everyday language use. A commitment to language instruction that attends to students' specific purposes for learning English has given ESP a unique place in the development of both theory and innovative practice in language instruction since the term first emerged in the 1960s. With countless students and professionals around the world now required to gain fluency in the conventions of their particular communicative domain of English to steer their learning and promote their careers, ESP has consolidated and expanded its role. It is now a major player in both research and pedagogy in applied linguistics, with a large and growing contribution from researchers around the world.
ESP has been widely adopted in many countries to better address the communicative needs of learners as students increasingly find themselves having to read, and often write, their subject papers in English. This presents challenges to both teachers and students. For students, they encounter a variety of English very different to that which they are familiar with from school, home or social media, while teachers recognise that they have to go beyond teaching grammar to assist students towards new professional or workplace literacies. ESP addresses these issues by drawing from a variety of foundations and a commitment to research-based language education. It takes the most useful, successful and relevant ideas from other theories and practices and combines them into a coherent approach to language education. In so doing it helps reveal the constraints of social contexts on language use and provides ways for learners to gain control over these.
In this paper I want to try and give an overview of ESP to help us understand it a little better. To do this I first sketch some of the ideas that have influenced it, focusing on needs analysis, communicative teaching, ethnography, social constructionism, and discourse analysis. I then go on to look at some of the effects ESP is having on what we do in classrooms, arguing that it has encouraged teachers 3 to highlight discourse rather than language, to adopt a research orientation to their work, to employ collaborative pedagogies and to be aware of discourse variation.

What are the main characteristics of ESP?
ESP emerged in the early 1960s as a response to the increasing globalisation of world markets and the growth of English as a commercial lingua franca to facilitate this (e.g. Hutchinson & Waters, 1987). Early in its history, Peter Strevens (1977) distinguished ESP in terms of: the primacy it gave to language-using purposes, the need to align curricular content with learner goals, and the use of appropriate teaching methods. Language teachers found themselves teaching technical English to non-native students and needing information about their discourses to do so. ESP thus grew out of text-based counts of grammar features in written technical documents, which quickly gave way to more explanatory models which sought to connect technical lexico-grammar and authors' rhetorical purposes. Since then, we have seen a strong interest in different research and teaching perspectives and a need to closely combine research and practice (Anthony 2018;Hyland, 2006;Johns, 2013).
We have also witnessed, under the broad umbrella of ESP, an increasing diversification of practice, and acronyms, so that the original Academic Purposes and Occupational Purposes labels no longer accurately represent the field. This is the natural outcome of following specificity, and Belcher points out that: There are, and no doubt will be, as many types of ESP as there are specific learner needs and target communities that learners wish to thrive in. (Belcher, 2009: 2)  Reviewers of the field have attempted to identify the key areas of ESP (e.g. Belcher, 2009;Basturkmen, 2021), with needs analysis, genre, corpus studies, and specialised language skills and lexis all figuring prominently. Handbooks add themes such as intercultural rhetoric, English as a Lingua Franca and critical perspectives to these (Hyland & Shaw, 2016;Paltridge & Starfield, 2013). There have also been studies of papers in the two flagship journals of the field, English for Specific Purposes (ESPJ) and Journal of English for Academic Purposes (JEAP) which show a trend toward the analysis of written texts (Gollin-Kies, 2014;Swales and Leeder,2012). More recently, these surveys have been supplemented by quantitative studies using bibliometric techniques. Hyland & Jiang (2021), for example, tracked changes in ESP research through an analysis of all 3,500 papers on the Social Science Citation Index since 1990 dealing with ESP topics. The results indicate that classroom practices remain central to the discipline and that there has been a consistent interest in specialised texts, particularly written texts, and in higher education and business English, with a massive increase in attention devoted to identity and to academic and workplace discourses.

What are the main influences on ESP?
ESP, in contrast to many approaches, can be characterised by its openness to the methods and insights of other fields. Most centrally it depends on a better understanding of what students' target texts are like, so it is part of applied linguistics, and particularly discourse analysis. ESP, then, can be seen as English language teaching with a stronger descriptive foundation for pedagogic materials.
In the classroom it incorporates elements from Communicative Language Teaching, Task-Based Language Teaching, Project-Based Learning (Richards & Rodgers, 2014) and, more recently, corpus-oriented and text analytic methods (Hyland, 2012;Reppen, 2013). Here, however, I want to briefly introduce five of the most salient aspects of ESP: (i) needs analysis, (ii) genre analysis, (iii) communicative teaching methods, (iv) ethnography, and (v) social constructionism. This is perhaps an idiosyncratic list, but they are core ideas which define what ESP is, assisting teachers to interpret how aspects of the real communicative world work and to translate these understandings into practical classroom applications.

i Needs Analysis
While not unique to ESP, nor the sole driver of ESP research, needs analysis is a defining element of its practices (e.g. Basturkman, 2021; Upton, 2012). It is conducted to establish the 'what' and 'how' of a course and is the first step in ESP course design. Investigating the specific sets of skills, texts and language a particular group of learners must acquire is central to ESP. It informs its curricula and materials and is a crucial link between perception and practice, helping ESP to keep its feet on the ground by softening any excesses of theory-building with practical applications. Hyland (2006: 73) defined it like this: Needs analysis refers to the techniques for collecting and assessing information relevant to course design: it is the means of establishing the how and what of a course. It is a continuous process since we modify our teaching as we come to learn more about our students, and in this way it actually shades into evaluationthe means of establishing the effectiveness of a course. Needs is actually an umbrella term that embraces many aspects, incorporating learners' goals and backgrounds, their language proficiencies, their reasons for taking the course, their teaching and learning preferences, and the situations they will need to communicate in. Needs can involve what learners know, don't know or want to know, and can be collected and analysed in a variety of ways.
How we understand what must be analyzed and the frameworks we use to describe it have both changed over time. Early needs analyses focused on the lexical and syntactic features of scientific and technical English texts. Interest then moved to the rhetorical macro-structure of specialist texts to describe scientific writing as patterns of functional units (Trimble, 1985). In Europe this approach was informed by functional-notional syllabuses and attempts to specify the competence levels students needed to perform particular tasks (Munby, 1978). This interest in seeing texts as part of their social contexts has continued through to the present. The use of genre analysis pioneered by Swales (1990) and Bhatia (1993), for instance, has provided a 6 useful tool for understanding how language is used in particular academic and professional communities as required by learners.
Conducting a needs analysis is a complex process and Bocanegra-Valle (2016:563) identifies six types of sub-analyses: • Teaching context analysis: resources, time, teacher skills and attitudes, etc.
• Task analysis: identification of tasks required in target context.
These analyses have become more diverse and, simultaneously, the concept of need has been expanded beyond the linguistic skills and knowledge required to perform in a target situation. On one hand, it has moved to include learner needs, or what the learner must do in order to learn, incorporating both the learner's starting point and how they see their own needs (Hutchison & Waters, 1987). Most recently, the question of 'whose needs?' has been asked, raising political questions about target goals and the interests they serve. Do large corporations benefit more than the individual student by focusing on target needs? Is accommodating to big business or academic disciplines in the best interests of the student? The term rights analysis has been introduced to refer to a framework for studying power relations in classrooms and institutions and for helping teachers to reflect on their role to bring about greater equality (Benesch, 2009). Clearly however, the imperative of need, to understand learners, target contexts, discourses, and contexts, means that the starting point for any ESP activity must be a strong research base.

ii Genre Analysis
Genre analysis is probably the most important item in the ESP toolbox. The importance of genre is underpinned by the fact that few people have explicit knowledge of the rhetorical and formal features of the texts they use every day. 7 Genre analysis seeks to "make genre knowledge available to those outside the circle of expert producers of the texts" (Shaw, 2016: 243).
Genres, most simply, are abstract, socially recognized ways of using language that we use to respond to repeated situations. In ESP, a fruitful line of research has been to explore and identify the lexico-grammatical features and rhetorical patterns which help characterise particular genres. This has helped to reveal how texts are typically constructed and how they relate to their contexts of use through specific social purposes, as well as providing valuable input for classroom teaching. Genre analysis also helps show how texts are related to other texts, how they borrow from and respond to other texts in a situation. Analyses of genres are therefore informed by function and situation This idea draws on the concept of intertextuality (Bakhtin, 1986). Intertextuality suggests that any instance of discourse is partly created from previous discourses, and this helps us to see how texts cluster together to form sets, and how they come to form particular social and cultural practices. Texts and their related activities may be linked one after the other, as in a formal job application: anapplication -interview offeracceptance sequence, or more loosely as a repertoire of options, say in the choice of a press advertisement, TV campaign, or social media posts to announce a new product. Researchers and teachers have been greatly assisted in recent years by being able to analyse text corpora to collect and study representative samples of texts from a given context. Counting frequencies shows what language and vocabulary features are important in a given genre while collocational analyses show how writers in different professions or disciplines use words in regular patterns. In this way more specific and accurate descriptions of target texts can be made.
Genre analysis in ESP has been influenced by the pioneering work of Swales (1990) and by Systemic Functional Linguistics (e.g. Halliday & Matthiessen, 2014).
Both see language as a system of choices which allow users to most effectively express their intended meanings. This, of course, fits neatly with ESP's aims to explore and explain the academic and professional genres that will enhance learners' career opportunities. Genre analysis has thus become the principal means 8 by which ESP practitioners identify the features that distinguish the texts most relevant to students (Cheng, 2021;Hyon, 2018;Tardy, 2017).
In the last few years, academic activity and communication are increasingly mediated by digital technologies, which enable scholars to engage in new social practices but with different affordances and challenges (Luzon & Pérez-Llantada, 2022). As a result, studies of blogs, 3-Minute theses, wikis and other Web 2.0 applications are emerging which both describe these genres and how they are being taught in classrooms (Nakamaru, 2012;Pérez-Llantada, 2021;Zou & Hyland, 2022).

iii Ethnography
In addition to close analyses of texts, a more recent research influence on ESP moves away from an exclusive focus on texts and studies the activities that surround their use (e.g. Guillén-Galve & Bocanegra-Valle, 2021). Ethnography is a type of research that explores contexts and tries to appreciate the participants' perspectives on writing, reading and using texts, drawing on the understandings of insiders themselvesan approach known as an emic perspective. Members of discourse communities and the physical settings in which they work become the main focus of study, with detailed observations of behaviours together with interviews and the analysis of texts (Paltridge, Starfield & Tardy, 2016). Together, these methods provide a fuller picture of what is happening, helping us to "understand our students and our students to understand the nature of the University and of EAP" (Collins & Holliday, 2022). This approach lends itself well to ESP research as it provides insights into educational and workplace practices, offering descriptions from actual investigations of people using texts.
Ethnography has been important in ESP in three main ways. First, it has begun to provide valuable insights into target contexts, helping to identify what happens in the production, distribution, and consumption of texts (Paltridge, Starfield & Tardy, 2016). So, for example, this approach was used by Gollin (1999) to analyze a collaborative writing project in a professional Australian workplace, and by Na and Hyland (2019) in their study of a Chinese scholar who spoke little English but had a 9 successful career publishing in international journals. Second, ethnographic techniques have also been useful in exploring student practices, revealing how they participate in their learning, engage with their teachers, and experience their classrooms. One example is Starfield's (2015) research into the experience of black undergraduates in a formally whites-only university in South Africa. Third, ethnography has been used to argue for appropriate pedagogic methods in contexts where overseas students study in Anglo countries or where Anglo teachers and curricula are employed in overseas settings. Holliday's (1994)  iv Communicative teaching practices ESP recognises that the communicative demands on students in universities and workplaces go far beyond control of linguistic error or 'language proficiency' (e.g Dudley-Evans & St John, 1998). There is now a considerable body of research and experience which emphasises the heightened, complex, and highly diverse nature of communicative demands in these contexts (e.g. Bazerman & Paradis, 1991;Manchon, 2011). Students find that they need to write and read unfamiliar genres and that communication practices reflect different, disciplinary or professionallyoriented, ways of constructing knowledge and engaging in study (Nesi & Gardner, 2012). In other words, ESP does not see students' writing difficulties as a linguistic deficit which can be improved by remediation in a few language classes, but as their attempts to acquire a new literacy and, more specifically, new discourse practices. In the classroom, this shifts language teaching away from isolated written or spoken texts towards contextualised communicative genres and an 10 increasing preoccupation with identifying strategies suitable for both native and non-native speakers of the target language (Anthony, 2018;Hyland, 2006).

So ESP is driven by a stimulus similar to that behind Communicative Language
Teaching back in the 1970s: to make the language purposeful by relating it to credible, real-world outcomes. As a result, it often relies on communicative methods which use tasks involving the negotiation of meaning, which employ portfolios, which use consciousness raising activities (such as comparison exercises) and those which ask students to reflect on text choices. Stoller (2016: 578-82) identifies several broad areas relevant to classroom materials and tasks in EAP classes: • Authenticity: the use of materials not designed for the classroom vs those adapted for student abilities.
• Motivating tasks which supplement textbooks and engage students.
• Materials and tasks that work together to scaffold students to achieve course goals • Relevant vocabulary for students' needs and vocabulary-learning strategies.
Genre approaches are widely used, and teachers seek to exploit relevant and authentic texts through tasks which attempt to help students increase their awareness of the purpose and linguistic features of these. More generally, providing students with an explicit knowledge of target genres is seen as a means of helping them gain access to valued genres, jobs and careers. The public and free availability of online corpora make teacher-student collaborations around relevant genres feasible and there are several sources which help guide students in their use (e.g. Reppen 2013;Hyland, 2004). Genre approaches, in fact, also seem to offer the most effective means for learners to both see relationships between texts and the contexts in which they are commonly used, and to critique those contexts (Hyland, 2018).
By providing students with a rhetorical understanding of texts and a metalanguage to analyze them, students can see that texts can be questioned, compared, and deconstructed, so revealing the assumptions and ideologies that underlie them. 11 Teaching, therefore, involves a commitment to real communication, to learner centeredness, and, where it is possible, a close connection with specialist subjects.
There has, as a result, been a focus on inductive, discovery-based learning, authentic materials and an emphasis on a guided, analytical approach to teaching (e.g. Anthony, 2018;Bell, 2022). Despite this, however, Bell (2022) has recently argued that classroom methods remain peripheral to discussions in ESP and deserve greater prominence than they are currently given in the literature. Hyland (2018) has also made similar comments and Hyland & Jiang's (2021) analysis of the ESP literature largely supports this view. While currirulum and assessment papers have increased significantly since 1990, discussions of classroom practices seem to have actually declined.

v Social Constructionist Theory
Social constructivism is a theory which suggests that knowledge and social reality are created through daily interactions between people, and particularly through their routine discourse. Originating in the symbolic interactionism of Mead (1934Mead ( /2015 and developed within social psychology, it is now perhaps the mainstream theoretical perspective in ESP today (Bazerman & Paradis, 1991;Latour & Woolgar, 1986). Although not an explicit framework for shaping and changing practice like, say, Legitimate Code Theory or Critical Realism (Ding & Evans, 2022), social constructionism provides a theory of knowledge-building for ESP. It underpins how the field understands discourse variation and its role in recontextualizing and reproducing knowledge (Hyland, 2004).
Social constructivism takes a critical stance towards taken-for-granted knowledge and, in opposition to positivism and empiricism in traditional science, questions the idea of an objective reality. It says that everything we see and believe is actually filtered through our theories and our language, sustained by social processes, which are culturally and historically specific. We see and talk about the world in different ways at different times and in different cultures and communities. Discourse is therefore central to relationships, knowledge, and scientific facts as all of these are rhetorically constructed by individuals acting as members of social communities.
The goal of ESP is therefore to discover how people use discourse to create, 12 sustain, and change these communities; how they signal their membership of them; how they persuade others to accept their ideas; and so on. Stubbs succinctly combines these issues into a single question: The major intellectual puzzle in the social sciences is the relation between the micro and the macro. How is it that routine everyday behavior, from moment to moment, can create and maintain social institutions over long periods of time? (Stubbs, 1996: 21).
Social construction, together with situated learning theory (Lave & Wenger, 1991) has thus become a central idea for many who work in ESP (e.g. Hyland, 2015a;Johns, 2019). It sets a research agenda focused on revealing the genres and communicative conventions that display membership of academic and professional communities, and which create those communities. From this, ESP practitioners set a pedagogic agenda focused on employing this awareness of communicative conventions to best help learners participate in such communities. The fact that social constructionism makes truth relative to the discourses of social groups sometimes draws criticism from those in the physical sciences, who prefer to see the world as a tangible and observable thing which is knowable independently of the language used to talk about it. This can sometimes make collaboration with the sciences difficult. Barron (1992), for example, found that the ontological superiority of science lecturers at Hong Kong University made them rigid when negotiating learning tasks and assignments with ESP teachers and Hyland (2013b) found that lecturers in science and engineering fields often treated student writing as peripheral to knowing 'facts'.
Nor do constructionists agree on precisely what the term community means, despite its importance in this approach. Harris (1989), for example, argues we should restrict the term to specific local groups to avoid the risk of representing abstract groups (such as professions or disciplines) as static, abstract, and deterministic. Discourse communities, however, are not monolithic and unitary structures but the result of interactions between individuals with diverse experiences, commitments, and influence. As a result, Porter (1992) understands a community in terms of its forums or approved channels of discourse, and Swales (1998) sees them as groups 13 constituted by their typical genres, of how they get things done, rather than existing through physical membership. For the most part, recent research has sought to capture the explanatory authority of the concept by replacing the idea of an overarching force that determines behavior with that of systems in which multiple beliefs and practices overlap and intersect.

What are the main impacts of ESP on teaching?
I now turn from some influences on ESP to offer a brief consideration of how ESP itself influences classroom practices: where these influences have taken us.
Basically, ESP centres around a general acceptance that institutional practices and understandings strongly influence the language and communicative behaviors of individuals. It also stresses that it is important to identify these factors in designing teaching tasks and materials to give students access to valued discourses and the means to see them critically. I want to draw attention to four aspects of this characterization: (i) the study of discourse rather than language, (ii) the role of teacher as researcher, (iii) the importance of collaborative pedagogies and (iv) the centrality of language variation.

i The Study of Discourse not (only) Language
In the past ESP materials were often based solely on the lexical and grammatical characteristics of scientific and business discourses in isolation from their social contexts. Today materials are more likely to acknowledge wider contexts, where language and tasks are more closely related to the situations in which they are used.
These might include the use of English to negotiate problems on an international building site (Handford & Matous, 2015), understand university tutorials (Coxhead & Dang, 2019), or express a stance in academic blogs and three-minute theses (Zou & Hyland, 2022). ESP practitioners now tend to address wider communicative skills in their teaching. Central to ESP, then, is a focus on discourse rather than just language and how communication is embedded in social practices and disciplinary epistemologies. 14 To understand language and the functions it performs for people, we have to appreciate how it is used within particular situations, so that identifying the participants involved and the purposes they have in using the language are integral to the construction of particular writing processes and written products. We need, for instance, to understand the interpersonal conventions a sales manager might observe when giving a client presentation or the knowledge a chemist assumes of his or her audience when writing up a lab report. In the classroom, these concerns translate into finding ways of preparing students to participate in a range of activities and to see ESP as concerned with communicative practices rather than more narrowly with specific aspects of language.

ii The Teacher as Researcher
ESP is, most centrally, research-based language education; a pedagogy for learners with identifiable professional, academic, and occupational communicative needs.
This means that teachers can rarely be just consumers of the materials provided by textbook publishers. The imperatives of specific English mean they must consider the relevance of the studies they read in journals or the activities they find in set textbooks to their own learners and, often, conduct their own research. Exploring the texts or the target situations relevant to their students.
While ESP textbooks and so-called "English for General Academic Purposes" or "English for General Business Purposes" courses are widespread, and may be useful in some situations, there is a growing awareness that many of the skills, language forms, and discourse structures these materials include are not easy to transfer across situations (Hyland, 2016). In addition, many teachers are not only becoming researchers of the genres and practices of target situations, but also of their classrooms. Teachers have used qualitative techniques such as observations and interviews to discover students' reactions to assignments, the ways they learn, and content instructors' reactions to learners' participation and performance (e.g. Hyland, 2013a;Li, Y. & Casanave, 2012). This information then feeds back into the design of ESP courses in the materials, tasks, and problems that are employed in the classroom.

iii Collaborative Pedagogies
A third major impact is the distinctive methodological approach that ESP has developed as a result of its view of specificity. ESP teachers must often work in tandem with specialists in those fields it seeks to describe, explain, and teach.
English teachers bring an expertise in communicative practices to the subject skills and knowledge of those working in particular target areas. As I have discussed above, the idea that professional communities possess their own distinguishing discoursal practices, genres, and communicative conventions is central to ESP.
Learners need to acquire a specific literacy competence together with the knowledge and tradecraft of their professions, so subject knowledge becomes the context for learning language. The topics, content, and practices of the profession thus act as vehicles for teaching particular discourses and communicative skills.
The fact that the ESP practitioner is generally a novice in these areas means that collaboration with both students and subject specialists is desirable, if not essential.
Students bring to their ESP classes some knowledge of their specialist fields and the kinds of communication that go on in them, and this implicit communication knowledge is important in a number of ways. First, it means that ESP teachers should try and make use of the specialist expertise of their students to engage them in relevant communicative activities. An imperative of ESP has always been a reliance on tasks and materials that display authenticity, mimicking real-world texts and purposes as far as possible, and learners themselves are among the best judges of whether these are appropriate. Second, teachers can use the specialist knowledge of their students in class as a learning resource. ESP tends to be strongly focused on the idea of rhetorical consciousness-raising, helping students to become more aware of the language and communicative practices in their fields. This means the teacher seeks to assist learners to activate their implicit understandings and to build on these, harnessing the methods of their fields to explore the ways that communicative intentions are expressed.
Teachers also often need to collaborate with subject experts, and there are a number of ways this can be done. First, the specialist can assist as an informant, providing teachers, or students, with background and insights into the kinds of practices that experts engage in and their understandings of the texts they use (Johns, 1997).
Alternatively, such collaboration can involve the specialist acting as a consultant, assisting the ESP teacher to select authentic texts and tasks. Finally, and more centrally, subject specialists sometimes collaborate directly with ESP teachers, either in a team-teaching relationship or through a linked course which runs parallel with the ESP course. This involves the ESP course supporting the content course with the two teachers jointly planning tasks and coordinating instruction.
The literature reports mixed experiences of this kind of collaboration, with some teachers describing ESP and subject teacher alliances as unrewarding. As noted above, faculty teachers may tend to treat the English teachers as subservient with the ESP course merely supporting the content course rather than being of equal importance to it (e.g. Barron 1992;Turner, 2004). Others, however, report more positive relationships (e.g. Arno-Macia & Mancho-Barés, 2015). Hyland (2015b), for instance, discusses how various degrees of cooperation with different faculties, including co-teaching and co-assessment, helped invigorate the English curriculum at Hong Kong University as well as providing valuable professional development opportunities and gaining the teachers greater respect for their work.

iv The Importance of Discourse Variation
Finally, as I have emphasised, ESP research strongly suggests that professional and academic discourses represent a variety of specific literacies. While there may be a "common core" of generic skills and linguistic forms which are transferable across different settings and professions, this is likely to be very limited (Hyland, 2016).
The distinct practices, genres, and communicative conventions of each community are directly related to the different purposes they have and their different ways of seeing the world. As a result, investigating and teaching these inevitably takes us to greater specificity in our classrooms. The idea of linguistic variation has been central to ESP since its beginning and owes its origins to Michael Halliday's work on register in the 1970s, but it has gathered momentum as a result of a number of factors.
One reason has been a growing awareness of the complexities of community literacies and the training that leads to professional membership. A large body of survey research in the 1980s and early 1990s, for example, revealed the considerable variation of discourses across the university (e.g. Horowitz, 1986). This work shows that not only do different disciplines employ different genres but that the structure of common genres, such as the experimental lab report, differed completely across disciplines (Braine, 1995).
At the most obvious level, of variation is lexis, with disciplines having completely different ways of talking about the world, so that students in different subjects have to learn completely different vocabularies. Less obviously, a study of an academic corpus of 4 million words showed that the so-called universal items in Coxhead's (2000) Academic Word List, actually have widely different frequencies and preferred meanings in different fields (Hyland & Tse, 2007). So that • "consist" means 'stay the same' in social sciences and 'composed of' in the sciences.
• 'volume' means book in applied ling and 'quantity' in biology.
• 'Abstract' means 'remove' in engineering and 'theoretical' in social sciences.
So words which appear to be the same to students, can have widely different meanings across fields. Similarly, Ha and Hyland's (2017) study of a 6-million-word corpus from economics and finance identified 837 words which had a meaning specific to those fields, although most of them also had a different general meaning too.
More generally, we know that different disciplines value different kinds of argument and set different writing tasks, so that analysing and synthesising multiple sources are important in the humanities and social sciences while more activity-based skills such as describing procedures, defining objects, and planning solutions are required in science and technology fields. It is also the case that different fields make use of different genres, so that in their large-scale corpus study of 30 disciplines in UK universities, Nesi & Gardner (2012) found 13 different "genre families", ranging from case studies through empathy writing to essays and reports. These differ considerably in social purpose, genre structure and the networks they form with other 18 genres. Equally, in the workplace, the ability to communicate as an insider is increasingly recognized as a marker of professional expertise. The professional competency statements of nursing, law, and accountancy, for instance, all refer to communicative abilities as central to these jobs, while caregivers, therapists, doctors, and other professionals are also often judged by their ability to gather and give information effectively.
This idea of different literacies is not just found in the genres professionals and academics use or the tasks they perform but is supported by close textual analyses of those genres. Successful communication depends on the projection of a shared context, showing others that you are like them and can understand their communicative needs and expectations. Communication, then, is effective only if writers and speakers can draw on knowledge of prior texts to frame messages in ways that readers and hearers recognize, expect and find persuasive. Their messages must appeal to appropriate cultural and institutional relationships. This directs us to the ways professional texts vary not only in their content but also in different appeals to background knowledge, different means of persuasion, and different ways of engaging with readers.
In sum, this research challenges the view that professional discourses are differentiated only by specialist topics and vocabularies. It also undermines the idea that there is a single 'English' that can be taught as a set of grammar rules and technical skills usable across all situations of use. This helps teachers to see that if students are having difficulties with the tasks they are asked to do at university, these difficulties may not be due to proficiency or laziness. Their frustrations cannot always be regarded as weaknesses easily corrected by additional grammar classes.
Instead, it encourages ESP teachers to find ways of integrating the teaching and learning of language with the teaching and learning of disciplines and professions.

Conclusion
This overview has been necessarily selective, as limitations of space prevent a fuller coverage of the theories that have influenced the growth of ESP and of the contributions it has itself made to applied linguistics and language teaching. 19 There are, however, two clear ideas that emerge from this survey and which might be seen as representing two basic principles of the field: • First is the fact that ESP is founded on the idea that we use language as members of social groups. This in turn means that it is concerned with communication rather than language and with the ways texts are created and used, rejecting an autonomous view of literacy to look at the practices of real people communicating in real contexts.
• The second point is that ESP is unashamedly applied. The term applied, however, does not mean lacking a theory. It means gathering strength by drawing on those disciplines and ideas that offer the most for understanding language use and classroom practice.
Not only is there an interdisciplinary research base at the heart of ESP, but this results in a clear theoretical stance that distils down to three main commitments: to linguistic analysis to the principle of contextual relevance, and to the classroom replication of community-specific communicative events.
It is clear that the same concerns which initially encouraged the pioneers to turn to specialised English language teaching remain central to the field. An interest in research-informed language instruction based on an understanding of specialised discourses and the demands these make on users. But nothing remains static, and ESP continually requires us to step into new domains and face new challenges.
Among these are finding ways to adequately marry textual and experiential methods which allow us to better understand new domains of practice and explore unfamiliar communicative worlds. In particular, the affordances of the internet, online teaching, digital genres and automated feedback on learning will require our attention, as will the growing demand for ESP by professional, technical, migrant and blue-collar occupations. This will almost certainly require adding to our existing toolkit of theories, methods and approaches, but there is no reason to suppose it will mean abandoning those that have proven so useful in helping us thus far in building plausible theories, detailed descriptions, relevant curricula and useful pedagogic tasks.