Discourse Analysis

Section A: Introduction: Key topics in the study of discourse analysis

A1: Why people don't say what they mean and mean what they say

---

One of the most important reasons why it is useful to learn about discourse analysis is the fact that people don’t always say what they mean, and people don’t always mean what they say. In fact, people quite frequently don’t say what they mean or mean what they say, and so social interaction is a complex process of ‘guessing’ about other people’s meanings and intentions.

For example, if I say to you, ‘Let’s get together for lunch sometime’, I might be inviting you to lunch, or I might be just trying to end the conversation in a polite way. This invitation might be serious or casual. It might even be a command to have lunch with me. ‘Sometime’ might mean next week or some unspecified time in the distant future. It might mean that I have something important to discuss with you and having lunch is just an excuse to do this. I might be inviting you to lunch because I harbour some hidden romantic feelings for you, or I might be inviting you out of a sense of obligation.

The way you interpret this utterance depends a lot on our relationship, the context in which it is uttered, and what was said in our conversation before it was uttered.

But why, you might ask, must we suffer through all of this ambiguity? If only we were clear and precise in our language we would avoid the need to do all of this guesswork and also avoid the need to study discourse analysis. And so, I should only say ‘Let’s get together for lunch sometime’ when I really want to have lunch with you, and if I want to do something else, I should say so, like ‘let’s get together to discuss discourse analysis.’

There are several good reasons why we don’t say what we mean and mean what we say. The first is that would be much more difficult than being ambiguous. Imagine having to be exact and precise about everything thing you said, for example, having to say something like ‘Let’s get together to have a meal which will serve as an opportunity to discuss our relationship and the fact that I fancy you. This meal will occur at midday and on some date in the near future which I prefer not to specify since I don’t want you to think me too pushy.’ The fact is that not saying what we mean is much more efficient, and we can often depend on the person we are talking to to be able to guess what we mean with minimal input.

Another reason we don’t always say what we mean or mean what we say is that it's dangerous. We may risk hurting someone else’s feelings or risking embarrassment. Not saying what we really mean allows us to be polite, and it also gives us some ‘wiggle room’ in potentially difficult situations. If I fancy you, and I say, ‘I fancy you—please go on a date with me’, there is always the possibility that you will reject me, which will be embarrassing for both of us. By just saying, ‘let’s get together for lunch sometime,’ I am able to express that I fancy you in an indirect way, and I avoid the possibility of being rejected. It is unlikely that you will reject ever having lunch with me, and even if you do, I can always rationalize it by saying that it's not me that you object to but something else – maybe you are on a diet and make a habit of skipping lunch.

But the most important reason that people don’t say what they mean and mean what they say from the point of view of a discourse analyst is that meaning is not the most important thing about saying. When we say things, what we are trying to do (or get other people to do) is usually much more important than what we ‘mean’. In fact, very often, the best way to get something done is not to say what you mean. If I want to get you interested in going out with me, maybe the best way to do that is not to tell you how much I fancy you (which might scare you off), but to treat the whole thing in a rather casual way.

Finally, one of the main reasons that we don’t say what we mean or mean what we say is that life is more fun that way. One of the things that gives richness to human relationships is the space we create in our discourse for multiple meanings and multiple interpretations. Ambiguity allows us to joke and flirt and bargain and do many of the other things that make life worth living.

Useful Links

‘Words Don't Mean What They Mean’ by Steven Pinker (Book Excerpt: Time Magazine, Sept. 6, 2007) www.time.com/time/magazine/article/0,9171,1659772,00.html

‘What is pragmatics?’ by Samia Adnan www.echoingmemories.wordpress.com/2007/06/20/pragmatics/

‘Analysing Verbal Data’ by Jay Lemke academic.brooklyn.cuny.edu/education/jlemke/papers/handbook.htm

Back to top

A2: Other Perspectives on what makes a text a text

---

In Unit A2 I explained that the three main things that make a text a text are cohesion (the fact that different parts of the text are joined together in certain meaningful relationships), coherence (the fact that the text fits into some overall pattern that is recognizable to readers, and intertextuality (the fact that the text is related in some way to other texts).

Although these three ingredients are generally thought of as being fairly essential for ‘texture’, other scholars have pointed out other defining qualities that texts have. The text linguists Robert de Beaugrande and Wolfgang Dressler, for example, point out seven qualities that they believe to be essential if a collection of words and sentences is to be considered a text. Three of these, cohesion, coherence and intertextuality, we have already covered. The additional four are: intentionality, acceptability, informativity and situationality. All of these aspects of texts are also extremely important to discourse analysts and are covered in great detail in later units of the book.

Intentionality has to do with the attitude or purpose on the part of a writer. Most people believe that some kind of intention to make a text is essential for something to be regarded as such. In other words, ‘accidental’ utterances like coughing fits cannot be considered texts. A focus on the intention of speakers is a very important part of speech act theory, which will be introduced in Unit B5. Others, however, might take issue with placing too much focus on the intentions of the text producer, pointing out that text receivers (hearers and readers) play a big role in inputting intentionality onto a text. You may have ‘not meant anything’ by your fit of coughing, but if I interpret this as an evasive technique to avoid answering an embarrassing question, then it becomes a text of sorts.

Acceptability, which might also be called ‘appropriateness’ has to do with how a collection of words or sentences fits into what speakers and hearers or readers and writers deem useful or relevant in particular situations. The text on a stop sign sitting in a warehouse somewhere may be a kind of a text, but since it is not useful or relevant to those outside of the context of the warehouse, they do not really consider it a text (see below). At the same time, most public places are full of discourse that most people who inhabit those places do not regard as texts—the writing on the underside of furniture or serial numbers on electrical equipment. Of course, whether you consider these pieces of text to be texts depends a lot upon who you are and what you are doing, a matter I will explore in some detail when we consider mediated discourse analysis in Unit A8.

Informativity refers to the amount or quality of new information that a group of words or sentences delivers to a hearer or reader. While we might not go so far as to say that a newspaper article containing ‘old news’ is no longer a text, we may not regard it as a text with as much value as one capable of informing us of something we do not already know. That's why we use yesterday's paper for non-communicative functions like wrapping fish. There is, however, a danger in thinking about texts only in terms of information. As we will see in Units A4, A6, B4 and B6, conveying information is only part of what texts do. Texts also play an important role in establishing and maintaining relationships among people. In fact, many of the most important texts in our daily lives, like greeting sequences, contain very little new information, but are very important in holding together our web of social relationships.

Situationality is probably the most relevant of de Beaugrande and Dressler’s categories for discourse analysts. As I discussed in Unit A1, discourse is always in some way situated, and much of its meaning and its effect on the world is a consequence of how words and sentences interact with situations. A recurring theme in this book is that ‘discourse’ results from the connections that are formed between language and the various elements that make up social situations (people, places, actions, objects, times of day, etc.).

Texts and situations

When we talk about texture being the result of connections being made between different elements in a text (such as words, sentences or pictures), we are implying that it is impossible to have a text which only has one element. Some people might point out that there are, in fact, a lot of texts that appear to only have one element. A stop sign, for example, only has one word: ‘STOP’, and it undoubtedly has meaning and so is undoubtedly a text.

IMAGE - STOP SIGN

Although there is only one word in this text, it does not mean that there is only one element. There are other features of this text that must be connected to the word ‘STOP’ in order for it to have the meaning: ‘stop your vehicle here’. One element is the colour red, and the other is the octogonal shape. And so what makes a stop sign mean ‘stop’ is the relationship among the word, the colour and the shape of the sign.

There are other relationships that are also crucial if this text is going to have the meaning ‘stop your vehicle here’. The most important is where the text is situated. If this text is on the wall of your bedroom or in the warehouse of the Highway Department, it is still a stop sign, but it does not have the meaning ‘stop your vehicle here’. Another crucial ingredient is a person inside a vehicle to read the sign. In the absence of a vehicle and a driver, the meaning ‘stop your vehicle here’ cannot be made.

The point I’m trying to make is that it is impossible for just one word in the absence of any connection to other words, symbols, situations or people, to be considered a text. On the most basic level, at least two elements need to be present to create meaning: a thing (such as an object, a person, a place or an idea) and an action. The word ‘milk’ does not make sense by itself. It is only when combined with an action (e.g. ‘We need to buy milk’) that it is meaningful. By the same token, ‘stop’ does not make any sense unless you have some thing or some body to do the stopping (a vehicle and driver)

Useful Links

The Prison Door: What makes a text a text? (PowerPoint presentation)
www.didattica.spbo.unibo.it/adon/files/rilm_cohesion1.ppt

What is a text? by Joel Gilberthorpe (a more critical approach)
www.arts.mq.edu.au/current_students/new_and_current_hdr_candidates/hdr_journals/hdr_journal/articles/2009/articles/what_is_a_text

Back to top

A3: Discourse communities, speech communities and communities of practice

---

The idea of 'community' is very important in helping us understand the relationship between discourse and the social groups that use it. The way the word community is used, however, is different for different schools of discourse analysis, and these different definitions of community help to illuminate different aspects of social organization and the role of discourse in it.

As was discussed in Unit A3, genre analysts speak of 'discourse communities', groups of people who are united by their use of specific types of texts to reach shared goals. For genre analysts, genres are like ‘membership cards’ to discourse communities, ways of showing that you belong. Doctors use medical charts and prescriptions to do the work of curing people and to show other doctors and their patients that they are ‘qualified’ to do this. Solicitors use contracts and legal briefs to defend people’s rights and to show judges and clients that they are capable of doing this. These different genres, then, not only help the people in these groups get certain things done; they also help to define these groups, to keep out people who do not belong in them, and to regulate the relationships between the people who do belong.

You might also come across discourse analysts using the term 'speech community'. This term is much more associated with sociolinguists and ethnographers of communication (see Units A7 and B7). The purely linguistic definition of ‘speech community’ is all the people who speak a particular language or dialect. The problem with this definition is that it shifts our focus away from people and makes the definition of a language (or dialect) the same as the definition of a community. This definition is taken to its extreme in the work of linguist Noam Chomsky who proposed a ‘completely homogeneous speech community’ (1965, pp. 3-4) as one basis for his study of language. Of course, the ‘completely homogeneous speech community’ doesn’t really exist. It is an abstraction, a theoretical construct designed to facilitate a theoretical approach to language. Discourse analysts, on the other hand, prefer their ‘speech communities’ to exist in the real world.

Definitions of ‘speech community’ that are more relevant for discourse analysts focus not just on language but also on how people use language to conduct their day to day affairs. This is the approach taken by Labov, who emphasizes not the shared language of members of a ‘speech community’, but their shared norms for using language. He writes (1972 pp. 120-1):

The speech community is not defined by any marked agreement in the use of language elements, so much as by participation in a set of shared norms; these norms may be observed in overt types of evaluative behavior and by the uniformity of abstract patterns of variation which are invariant in respect to particular levels of usage.

This definition shifts the emphasis away from linguistic criteria to things that individuals themselves feel make them members of the same community.

A similar emphasis can be seen in the definition of ‘speech community’ promoted by Dell Hymes and used by those applying his model of the ethnography of communication. For Hymes, a ‘speech community’ is a group of people who share both a 'variety' of language and a set of expectations about how that variety will be used in day to day social life. Hymes defines 'speech community' as a group of people who share ‘rules’ for when and how to speak (p. 54). In this formulation, people who speak American English would probably not be considered a ‘speech community’ since different people who speak American English may have very different ideas about how to speak it in different situations. A more likely candidate for the label of ‘speech community’ might be a group a skateboarders who not only share a common language but also share a common way of using that language to organize their skateboarding activities (see Unit C7).

Still, other discourse analysts use the term ‘communities of practice’ when they refer to the groups that they study. The idea of ‘community of practice’ was developed by educational psychologists Jean Lave and Etienne Wenger (1991) to describe the groups within which people learn. Their argument was that learning is not just about learning particular skills or knowledge, but learning how to become a part of a group of people who do things together.

This concept of ‘communities of practice’ was later taken up by sociolinguists Penny Eckert and Sally McConnell-Ginet (1992), especially in their work on gender in language. They define ‘community of practice’ as:

An aggregate of people who come together around mutual engagements in some common endeavor. Ways of doing things, ways of talking, beliefs, values, power relations - in short, practices - emerge in the course of their joint activity around the endeavor. (1992:87).

This particular definition of community highlights the situated nature of discourse mentioned in Unit A1. It also reminds us that discourse, power and social identity are all products of concrete actions that people take in the real world.

The idea of ‘communities of practice’ is particularly useful for discourse analysts interested in the relationship between language and what people do. Consequently, it is a concept that appears prominently in the work of Ron Scollon (1998) and others working in mediated discourse analysis.

References

Chomsky, N. (1965) Aspects of the theory of syntax, Cambridge, MA: MIT Press.

Eckert, P. and McConnell-Ginet, S. (1992) ‘Communities of practice: Where language, gender, and power all live’, in K. Hall, M. Bucholtz and B. Moonwomon eds., Locating Power, Proceedings of the 1992 Berkeley Women and Language Conference, Berkeley: Berkeley Women and Language Group, 89-99. Available online at: www.stanford.edu/~eckert/PDF/Communitiesof.pdf

Hymes, D. (1972) ‘Models of the interaction of language and social life’, in J. Gumperz & D. Hymes (eds.), Directions in sociolinguistics: The ethnography of communication. New York: Holt, Rinehart, Winston.

Lave, J. and Wenger, E. (1991) Situated learning: Legitimate peripheral participation, Cambridge: Cambridge University Press.

Scollon, R. (1998) Mediated discourse as social interaction, London: Longman

Back to top

A4: Discourses and Cultural Models

---

In Unit A1 I introduced James Paul Gee’s ideas of ‘Discourses’ (with a capital D) and ‘cultural models’. Discourses are combinations of ways of saying, ways of acting and ways of thinking that express certain ideologies or ‘versions of reality’.

Discourses include sets of (often unstated) rules and assumptions about what is good or bad, right or wrong, normal or abnormal as well as how people ought to treat other people in different kinds of situations. Since these rules are often unstated, they can have a powerful effect on us – rather than regarding them as ‘rules of a language’ we regard them as ‘rules of nature’. We forget that these rules were invented by human beings in societies and come to think of them as ‘absolutely true.’ Gee calls these sets of rules ‘cultural models’.

Cultural models provide us with frameworks for how the world is ‘supposed to be’. When things or people we encounter fit into these models, we think of them as ‘good’ or ‘normal’. When they don’t fit into these models, we think of them as ‘bad’ or ‘abnormal’.

One of the biggest problems with ideologies and the models they provide is that they limit the way we look at reality. Another big problem is that they marginalize certain people or groups of people who do not fit into our models (marginalize means to push someone or something to the ‘margins’, to exclude them).

One way to make sure that we are controlling our models rather than the models controlling us is to try to make them explicit, that is, to try to clearly state the unstated rules and then to analyze them.

Gee gives as an example of cultural models the sets of rules and expectations that university aged women on many American campuses use when they talk about and judge males, especially potential romantic partners. This cultural model, he says, includes the following rules:

• In the prototypical relationship, the two parties are equally attractive and equally attracted to one another.
• If the discrepancy in relative attractiveness is not great, adjustments are possible.
• A relatively unattractive male can compensate for his lesser standing by making extraordinary efforts to treat the female well and make her happy.
• A relatively unattractive female can compensate by scaling down her expectations of good treatment.

Gee goes on to observe that women often label males in terms of this model. For example, males who are attractive or popular with women and who take advantage of their attractiveness to gain affection and intimacy without having to give back the friendship or caring that would normally follow mutual attraction are called by such names as ‘Don Juans,’ ‘jocks,’ ‘chauvinists,’ ‘hunks,’ ‘studs,’ and ‘playboys,’ terms often, though not always, used as insults.

In contrast, males who are not attractive to females and who are not adept at pleasing them and pursue a female who is more attractive than they are, sometimes hanging around and becoming an embarrassment to the woman they are attracted to, lowering her prestige and forcing her to treat them badly are labelled by such terms as ‘jerks,’ ‘turkeys,’ ‘nerds,’ and ‘wimps’.

Cultural models in Chinese Opera

In order to further explore the idea of cultural models, have a look at the short video clip at the link below:

This clip is from a Cantonese movie called 虎度門 (Fu Dou Mun) starring the famous Hong Kong actress Siu Fong Fong. In the clip, an old opera singer is having an argument with the director of the opera during a rehearsal. They are arguing about the fact that the director had punished another opera singer, named Bing, who is the old opera singer’s disciple. The old opera singer, Uncle Tin, doesn’t think the director has the right to punish his disciple.

One of the main sources of misunderstanding in this argument (and in most arguments) is that the participants are applying different cultural models to the situation. These different cultural models come from different Discourses.

In the film, Uncle Tin operates within what we might call ‘The Discourse of Traditional Chinese Opera’. This Discourse is quite complex. It contains many different elements like a special register to talk about opera and the various kinds of equipment and techniques used in its performance, particular dialects in which opera is sung, certain visual codes like the colours and shapes of the make-up and clothing (which show the different roles the actors are playing) and various gestures used both on stage and off stage. Along with these tools, the Discourse also involves certain cultural models about how people in communities of opera singers should treat one another. This cultural model of how Chinese opera singers should treat their colleagues includes the following rules:

• Respect is based on age and experience. The oldest and most experienced opera singer in the troupe should be given the most respect.
• People become full members of a community of opera singers by serving for a particular time as a disciple to a more experienced member.
• The relationship between the master and his disciple is almost like that between a father and a son.
• The disciple is responsible for following the orders of the master, and the master is responsible for teaching, taking care of and protecting the disciple.
• The main purpose of communication within the troupe is to confirm these social relationships, and so language tends to contain a lot of signals for politeness and respect.

In this way, a Chinese opera troupe is very much like a Chinese family, and success in performance is seen as the result of everyone in the troupe fulfilling their social roles (disciples obeying their masters and masters taking care of their disciples). Since the disciple is like a son, nobody else has the right to punish or discipline him/her except the master. For an outsider to criticize the disciple is the same as criticizing the master.

The director, on the other hand, is operating within a different Discourse, which we might call the ‘Discourse of Commercial Theatre’, a Discourse with different cultural models. The cultural model for treating colleagues in this Discourse contains the following rules:

• Respect is based on the job/position held by the individual, and both age and experience are largely immaterial.
• The most powerful person in the troupe is the director and everybody must do what he says.
• People become full members of the community of theatrical performers by performing in shows and accumulating ‘credits’ which they list on a resume that they use when they want to get a job performing in a new show.
• They may have learned their trade by doing shows, or they may have taken some courses in an educational institution. They might also do things like join unions or professional organizations.
• The purpose of language is to convey information. If one of the actors is not doing a good job, the director will tell him so directly. If the problem continues, the director might fire him. The director has the same power over all of the actors, regardless of their relationships with one another.

In this model, a group of theatrical performers is not like a family, but more like employees in a company. Success is dependent on ‘teamwork’.

Even though the director and Uncle Tin are both ‘Chinese’, that is to say, they both belong to the same ‘culture’, since they operate within different Discourses and apply different cultural models to their job situation, conflicts such as that shown in this clip are practically inevitable.

Useful Links

Discourse and Sociocultural Studies in Reading, James Paul Gee
www.readingonline.org/articles/handbook/gee/

Cultural Models in Video Games: A summary of a book chapter by James Gee on Eduiwiki
www.edutech.csun.edu/eduwiki/index.php/Gee_Chapter_6_-_Cultural_Models

Critical Discourse Analysis
www.users.utu.fi/bredelli/cda.html

Jackie Chan drinks Mountain Dew
www.wisc.edu/english/zuengler/pdf/JackieChan.pdf

Back to top

A5: Pragmatics and Conversation Analysis

---

As discussed in Unit A5, pragmatics and conversation analysis represent two different approaches to how we solve the problem of understanding what other people mean in conversations. These two different approaches are based on two fundamentally different ideas about what is most important in this process of sense making.

Scholars in the field of pragmatics, including speech act theorists, focus primarily on how the resources speakers have for expressing certain meanings and intentions interact with the social conditions in which the resources are used and the social identities of those who use them. For them, conversation is about ‘logic’. People make sense of other people by taking into account various factors (how certain meanings are usually made, what the probable intentions of speakers are, what is possible in particular situations) and coming to a conclusion about what they probably mean.

Conversation analysts on the other hand, are more interested in order than in logic. They focus more on how people make and interpret meanings by exploiting the orderly structure of conversation, especially its sequential structure.

These two different approaches to conversation mean that discourse analysts in these different camps have different ideas about how to go about studying conversation and what constitutes appropriate data. While both use ‘naturally occurring’ conversations as data, those approaching the problem from the point of view of pragmatics might supplement that data with other things like interviews with the people involved, questionnaires, experimental or elicited conversations, and even conversations from literature or films. Conversation analysts, on the other hand, are very strict about what they consider valid data. They will only accept transcripts from naturally occurring conversations.

'Doing being ordinary': Conversation Analysis and its ethnomethodological roots

'Conversation Analysis' (CA) has its roots in a branch of sociology founded by a scholar named Harold Garfinkle called ethnomethodology. Perhaps the best way to describe ethnomethodology is that it is the study of how people act normal – in other words, how people use shared commonsense knowledge and reasoning to conduct their everyday affairs. We might think acting normal is not such a big deal, but it is actually a very complicated thing, and something that varies a lot from culture to culture. Understanding how people 'do being normal' (as Grafinkle put it) is an important basis for understanding how and why social interaction works (or does not work). Studying being normal is also a good way to remind yourself that 'being normal' is not something that you are, but something that you do, and something that you have to learn to do. Children are not as good at 'doing being normal' as adults.

Like ethnomethodology, the goal of CA is to determine the unspoken, shared understandings and methods of commonsense reasoning that guide and orient participants' actions in a given context. Its focus however is much narrower, emphasizing in particular the mechanics or procedural rules of everyday conversation.

Developed collaboratively by Harvey Sacks, Emanuel Schegloff and Gail Jefferson, CA starts from the premise that conversation is not random, but has an underlying order. In fact, conversation analysts believe that the people use the rules of conversation to jointly construct an orderly world.

The main principles of CA are as follows:

• Conversation is a kind of activity with which people bring order to the world.
• This activity is made up of a sequence of orderly actions (utterances are ‘doings’ in which certain things are accomplished like greeting, asking, leave-taking, etc.).
• These actions are governed by rules or sets of expectations which people share with one another.
• These rules determine things like how we begin and end conversations, who gets to talk about what and when, and how we know when it is our turn to talk and when it is not.

As I said in Unit A5, the core of conversation analysis is the exploration of the sequential structures of social action which shape the world, turn by turn. According to Sacks, Schegloff and Jefferson social interaction is basically arranged in pairs of utterances – what one person says basically determines what the next person can say.

But conversation analysts are not just interested in adjacency pairs. They are also interested in other aspects of the mechanics of conversation. Two of the most difficult problems we encounter when we study conversations (and when we have them) is understanding how people manage to successfully start conversations and how they manage to successfully end them.

Openings

The most important thing about any conversation is beginning it. According to Schegloff, openings of conversations consist of three distinct moves:

• the opening of the 'channel' (consisting of a summons-answer sequence)
• an identification sequence
• a topic negotiation sequence.

For example, a typical phone conversation might go something like this:

Summons-Answer Sequence	Ring Hello?
Identification Sequence	Hi Rodney, this is Iris. Oh, hi Iris...
Topic Negotiation Sequence	What can I do for you? I'm calling to ask some questions about the assignment.

Of course, often conversational openings are much more complicated than this. They must always however, contain these three sequences in some form or another.

One thing that often complicates openings is that people often insert some kind of 'facework' in between the identification sequence and the introduction of the topic. This is especially true in situations where people are friends, one person has power over another person or the topic involves some kind of face threatening act.

Different forms of communication tend to come with different rules about who speaks first, how the identification sequence is managed and who is supposed to introduce the topic. In a phone conversation like that above, for example, the person who is called speaks first, but the caller is generally expected to identify themselves and to introduce the topic. These rules might be different for different forms of conversation (like mobile phones, in which the identification sequence is often made unnecessary by caller-display and replaced by a location sequence – 'where are you?', or ICQ where the person who is summoned is just as likely to introduce the topic as the person who issues the summons). In face to face conversation between friends, the identification sequence is often very subtle, consisting of 'recognizing' the other person and 'ratifying' the relationship.

Closings

One of the hardest things about having a conversation is knowing how to end it. This is because ending a conversation works against the 'pairwise' organization of talk (see Unit D5). Different kinds of conversations end in different ways. Conversations that are instrumental in nature (such as asking a teacher some questions about an assignment) are easier to end – they end naturally when the purpose of the conversation is fulfilled. Casual conversation between friends is much harder to end, as you probably know from you own experience. People usually start ending the conversation well before the actual closing, offering what we call 'pre-closings’ – signals to the person they are talking to that they want the conversation to end. This is sometimes done rather directly ('Well, I've got a lot of homework to do') or in rather subtle ways with things like minimal responses (not elaborating too much or raising new topics), speed and intonation (slower, lower) and use of discourse markers ('so', 'well'). One of the most common ways to end a conversation 'politely' is by making it seem like the other person wants to end it ('I know you're busy, so I won't bother you anymore'). Ending a conversation abruptly creates the implication that you are angry at the other person or do not feel very close to them (which is okay if you really are not very close to them).

Useful Links

Pragmatics (from Stanford Encyclopedia of Philosophy)
www.plato.stanford.edu/entries/pragmatics/

Routledge Encyclopedia of Philosophy entry on Pragmatics, Kent Bach
www.online.sfsu.edu/~kbach/spchacts.html

Speech Acts
www.online.sfsu.edu/~kbach/spchacts.html

An Introductory Tutorial in Conversation Analysis, Charles Antaki
www-staff.lboro.ac.uk/ssca1/intro1.htm

Methodological Issues in Conversation Analysis, Paul ten Have
www.paultenhave.nl/mica.htm

Back to top

A6: Strategic interaction and the presentation of the self: Erving Goffman's contribution to discourse analysis

---

As I said in Unit A6, the two most important analytical constructs used in interactional sociolinguistics – face and frames – come from the work of the American sociologist Erving Goffman. Goffman however, contributed much more than these two concepts to the study of interaction. In his famous book The Presentation of Self in Everyday Life he introduced a whole new way of looking at social interaction. Social interaction is, he argued, primarily a performance. Here I would like to give a short summary of some of the concepts Goffman discussed in that book in order to give you a wider framework in which to understand the ideas of face and frames described in Units A6 and B6.

In order perform successfully, Goffman said, we need to have a stage, scenery, costumes and a collection of voices, facial expressions and movements. These are what Goffman calls our 'front'. He says the front is the expressive equipment  we use in performances, in other words, the equipment we use (including our clothes, our make-up, our laughs, smiles, frowns, our mobile phones, our restaurants, bars, karaoke boxes and classrooms) to create the kind of 'impression' (or 'line') that we want.

Goffman divides front into three parts. The first is the setting. Setting is all the things around you that you use to play the role. My university with its classrooms, offices and lecture theaters provides me with the equipment I need to play the role of a teacher and the equipment my students need to play the role of students. A restaurant is a setting that allows people to play the roles of waiters, cooks and customers.

The second part of front is appearance. Appearance includes the clothes you wear, your make-up and hairstyle. Of course, you put on different costumes to play different roles. When I play the role of a lecturer I wear a tie. But I have a very different appearance when I am shopping on Saturday afternoon. Waiters, police officers and secondary school students have their uniforms, but when you think about it we all have uniforms. If you look around you, you can see that there is a certain type of uniform that you wear when you are being a student, which is of course very different from the uniform you will wear when you get married.

The last part of front is manner. While setting and appearance are more constant, manner can change according to the situation. There are however certain set ways of acting when you perform certain roles, certain tones of voice, gestures and facial expressions. Notice the manner that a police officer has when he or she is directing traffic, when a teacher teaches, when a doctor has a consultation with a patient. You will notice that there are certain things these people are supposed to do when they perform these roles. These are what we call manner.

In a way, you can think of front as all of the 'cultural tools' you have available to you to play certain roles, the things (objects, places), ways of talking, ways of looking and ways of acting you use to project the impression that you want, to say, 'Look, now I am being a teacher’ (or a father, or a shopkeeper, or a pop star, or a businessperson, etc.).

If you don't have the proper expressive equipment available to you, then it is very difficult for you to play the role. A policeman without a badge is not a policeman. A teacher without a classroom would have a very difficult time getting people to listen to him or her. A businessman without an office would find it difficult to do much business.

Routines

Goffman says that we employ 'fronts' to perform 'routines'. Routines are specific performances or kinds of performance. As we have said, a 'performance' is any activity that you do which takes place in front of an 'audience'. So performance is a more 'generic' (general) term. What we perform are routines. They are 'strips of action' that usually have some kind of internal logic or structure. It is usually pretty easy to tell when a routine begins and ends.

Most of the routines that we perform are routine, that is they follow rather consistent and predictable patterns. You could even give names to the different routines that you perform: the 'lecture routine', the 'riding the bus routine', the 'shopping routine', the 'eating at McDonald's routine'. That is not to say that we always act in exactly the same way whenever we perform a particular routine. But, every routine has its own sets of expectations about how you should perform and limitations regarding how you should not perform. Routines are not something that comes naturally – we have to learn routines just as actors have to study their lines and if you are unfamiliar with a routine, you probably won't be able to perform it very well. It is not uncommon for people who are about to perform an unfamiliar or particularly important routine to rehearse before they perform. Someone who is about to go out on their first date, for example might spend some time in front of a mirror practicing their facial expressions and someone going for a job interview might prepare themselves by thinking about the questions that might be asked.

Different routines also have different Fronts associated with them. I employ one Front when I am performing the 'teaching routine', for example, and a different front when I am performing the 'taking my dog for a walk routine'. Of course, it is possible also to employ the same front for different routines. I use more or less the same front when I am teaching and when I am having staff meetings with my colleagues.

So, if your life is like a play, your Routines are like the different scenes you play.

'What's Your Line?'

In Goffman’s dramaturgical approach to interaction, we use fronts to perform roles. When we perform these roles, we try to maintain and promote a certain 'line'. Our line is basically ‘our  version of reality’, our idea about what’s going on.

It is easy to confuse the notion of role and the notion of line. Role is basically the general ‘type’ of character we are playing (such as a teacher, a soldier, a student or a pop singer). Line is much more specific. It includes the attitude we take up towards ourselves and our performance of this role, the attitude we take towards other people, and the attitude we take towards the situation.
The idea that we want to ‘promote’ our line is very important.

So we can say that the role that Barak Obama plays is that of the President of the United States. He uses his office, his airplane, his suit, his smile etc. as his front in order to play this role. The line that he is promoting includes what he believes is good or bad for the United States and the world, his policies, his ‘agenda’. As a politician his main job is trying to sell his line to the people who elected him.

A similar situation can be seen with your boyfriend. He is playing the general role of a boyfriend. He may use flowers, candy, smiles, kisses and romantic words as his Front. His Line is basically the kind of relationship he is trying to promote. Some boyfriends may want to promote a very serious relationship with you and others might want to promote a more casual one. The problem of course comes when your line is different from his.

So to sum up, your line is the version of who you are, who the other person or people are, and what is going on that you try to ‘sell’ to the other people. If they ‘buy’ (accept) your line then things go smoothly. But if they challenge your line, then you might have to either defend your line or change your line. 

Regions

‘Regions’ are the places where we perform and where we prepare for our performances. The idea of ‘regions’ is easy to understand if you imagine a theatre. In a theatre you have the ‘stage’ where the performance takes place and you have ‘backstage’ where the actors put on their costumes and make-up and  rehearse their routines. In the same way, in real life you can divide places into those where you are ‘performing’ and those where you preparing for your performance.

‘Onstage’ regions include public places like classrooms, restaurants, trains, shops etc. Certain regions in your flat can also be onstage (typically the living room), if people are visiting your home. Backstage places are more private, like toilets, your bedroom etc. They are places where you let down your front and act more ‘natural’, where you get dressed, look at yourself in the mirror and check your appearance, put on your make-up and so forth.

What makes a region an onstage or a backstage region is not so much the place itself but whether or not an audience is present. Even a very public place can be like backstage if you think nobody is watching you. Similarly a private place can become onstage if an ‘audience’ enters it. It is also important to remember that not everybody counts as an ‘audience’. We sometimes perform in ‘teams’ and so a ‘team member’ is not really the same as an audience. A ‘team member’ is really a fellow actor in your performance who is cooperating with you to promote your line. So it is typical to invite other team members into our backstage regions, and we treat our team members differently when we are with them backstage. For example, we may criticize our team members backstage in ways that we might not when we are on stage.

Disturbances

According to Goffman, when we are ‘performing’ in social interaction the goal is to maintain our ‘line’, to help others to maintain their ‘lines’ and to ensure that the interaction goes smoothly. When this happens the ‘face’ of the people involved is preserved. When something goes wrong, that is, when the performance is ‘disrupted’ in some way as to cause the ‘loss of face’ for participants, this is called an incident. Incidents involve the discrediting of one or more of the participants (a challenge to his or her ‘line’ and/or ‘face’). Some incidents are intentional—that is, one or more participants try to challenge another’s ‘line’ or take away their ‘face’ on purpose. This can be called a scene, as in the well known expression ‘making a scene’. A scene is when the performance is disrupted intentionally.

Often however, the performance is disrupted unintentionally either by an unintentional action or an intentional action or statement whose significance is not appreciated by the person who says or does it. The first kind of disruption might occur when somebody, for example, reveals a secret ‘by accident’ or loses control of what they are saying or doing. The second kind of disruption, known a faux pas, is when you say or do something on purpose but you do not realize that what you are saying or doing will cause a disruption to the performance. Faux pas is a French word which literally means ‘false step’ and that’s a good way to remember what it means: a kind of mis-step in social interaction. Goffman calls a faux pas an incident in which ‘a performer unthinkingly makes a contribution which destroys his own team’s image’. So faux pas specifically have to do with mistakes we make which jeopardize our own face or the face of our team members. When we mistakenly make a comment that jeopardizes the image of the other team, this can be referred to by the English idiom ‘putting your foot in your mouth’.

Useful Links

The Presentation of Self in Everyday Life (Introduction), Erving Goffman
http://www.clockwatching.net/~jimmy/eng101/articles/goffman_intro.pdf

The Erving Goffman Archives
http://www.unlv.edu/centers/cdclv/ega/

Erving Goffman, Dramaturgy, and On-Line Relationships, Nikki Sannicolas
http://www.cybersociology.com/files/1_2_sannicolas.html

Jonathan Franzen: Read Some Erving Goffman please, Savage Minds
http://savageminds.org/2011/05/31/jonathan-franzen-read-some-erving-goffman-please/

Back to top

A7: The Text/Context problem

---

Ever since Malinowski’s famous proclamation in 1947 that ‘utterance and situation are bound up inextricably with each other and the context of situation is indispensable for the understanding of the words’, scholars of language have been struggling to understand exactly how utterance and situation are bound up with each other. What is the relationship between text and context? This question of course brings us to an even more fundamental issue of how text and context are to be defined. Where does text end and context begin?

As I said in Unit A1, different schools of discourse analysis define context differently. I mentioned Halliday’s tripartite definition of context (field, tenor and mode), but this is just part of Halliday’s overall theory, describing what he referred to as the ‘context of situation’. Along with this he also talked about the larger ‘context of culture’, which includes the broader set of social relationships, beliefs, customs, ideologies and Discourses within which situations occur.

Other analysts take a much more limited view of context. Conversation analysts for example consider only the immediate context of a conversation, in particular the context that participants create and manage through the sequential organization of talk (Schegloff 1991, 1997), to be relevant. Issues that are not attended to in participants’ talk are not considered part of the context.

Whether we are considering the ‘context of culture’ or the more ‘micro-level’, ‘proximate’ context that is of interest to conversation analysts, one of the most important aspects of context that we must not loose sight of is that of dynamic, not static. Context is not like an unchanging theatre set within which people interact. It is created, defined and altered through the interaction itself.

This is a point that Goffman made very strongly with his notion of framing (see Units A6 and B6). For him, context is made up of multiple and complex layers of reality and deceit through which people actively manage what they are doing and the alignment they are taking up to others.

One of the problems with approaches like Halliday’s tripartite model of field, mode and tenor and Hymes’s SPEAKING model is that by encouraging analysts to define different aspects of context they also encourage the illusion that context is stable and definable rather than fluid, contingent and negotiated.

Another problem with such approaches is that they tend only to consider things in the external, physical world and ignore the cognitive dimension of context. This is a concern that is taken up by the discourse analyst Tuen van Dijk in his book Discourse and Context (2008). For van Dijk, context is not limited to the physical reality surrounding the text, but also includes the ‘models’ that people build up in their minds about the kinds of situations they find themselves in, models which they use to make predictions about the kinds of meanings that are likely to be important.

Context in the Digital Age

The discussion in Unit A1 and above illustrates that the problem of context is still a matter of considerable debate and contention among discourse analysts. With increased use of digital media, this problem only promises to become more complex and contentious.

In a paper I wrote in 2004 called ‘The problem of context in computer mediated communication’ I discussed how digital media are altering the way we think about context. When we operate personal computers for example, all of the windows we have open, all of the applications, all of the pages open in our web browser, all of the people on our instant messaging list, create multiple contexts for our activities. As we work on documents, chat with friends and check our Facebook page, we move rapidly from context to context, so that a piece of discourse which at one moment might be considered ‘text’, at another moment might become part of the context that surrounds some other text. These ‘virtual’ contexts are also linked to multiple physical contexts in which the different people we are interacting with are sitting in front of their computer screens.

Nowadays with advances in mobile technology, we are able to carry this complex collection of contexts around with us. So when we are sitting in a coffee shop with our friend, the context of our interaction extends far beyond the walls of the coffee shop. With our mobile phones at our disposal many other spaces, conversations and relationships can become part of the context of our cup of coffee. We can text or email our friends and include them in our conversation, or check out the details about a movie we’re discussing on our portable web browser, or take a picture of ourselves and our companion and upload it to Facebook, so that not only does the context of our cup of coffee expand across time and space, but it also becomes part of the context of other people’s activities.

From the point of view of computer mediated communication, most conventional models of context I have discussed simply don’t work because they begin from the basic assumption that communication takes place in the form of focused social interactions which occur in particular physical spaces and involve particular participant’s assumptions that do not hold true in much digitally mediated interaction. In the ‘digital surround’ created by new communications technologies, communication is more polyfocal, forcing us to rethink the traditional separation between text and context.

As I will discuss in Unit A8, mediated discourse analysis attempts to address this problem by beginning not with texts and contexts but with people’s actions and experiences around texts.

Useful Links

The problem of context in computer mediated communication
www.personal.cityu.edu.hk/~enrodney/Research/ContextCMC.pdf

What is wrong with modern accounts of context in linguistics?
Roman Kopytko, Poznań
www.univie.ac.at/Anglistik/views/03_1/KOP_SGLE.PDF

Back to top

A8: What is a social practice?

---

Many contemporary approaches to discourse, including critical discourse analysis (see Unit A4) are concerned with the relationship between discourse and practice. Mediated discourse analysts however, view the idea of ‘social practice’ a bit differently from other analysts.

As I wrote in Unit A8, mediated discourse analysts are interested in the relationship between discourse and action. But the actions that we do are not random and rarely unique. We tend to do the same actions or combinations of actions over and over again in our lives, appropriating the same or similar cultural tools to do these actions. For example, ‘having a cup of coffee with friends’ might be something that we do quite regularly. Whether we go to Starbucks or a small boutique coffee shop, the actions we engage in and the tools we appropriate to perform this practice will be the same.

So, mediated discourse analysis defines practices as actions or chains of actions which are recognized by a group of people as doing ‘a certain kind of thing’ that usually has some label like ‘having a cup of coffee’, ‘going out on a date’, or ‘attending a lecture’. Mediated discourse analysts are very interested in such practices because they help to define societies and, to some extent, exert control over the people in those societies.

But what keeps social practices going? What ensures that practices like ‘going shopping’ and ‘waiting in a queue’ are performed in the same way over and over again by the people in a particular ‘community of practice’?

The first thing is the community itself. People in communities of practice tend to maintain and enforce certain ways of doing things; they reward people who do things in the conventional way and often punish in some way people who do things differently. ‘Old timers’ in communities usually teach newcomers how to perform the practices that are important to the community and learning how to perform these practices is a sign of successful integration into the community.

The second thing is the individual. As people learn how to perform practices, these practices gradually become ‘habitual’, that is, they become second nature to people to the extent that they can hardly imagine doing things any other way. The sociologist Pierre Bourdieu (1977) used the term habitus to describe how social practices become part of the way people think, behave, even the way they move and experience their bodies.

Perhaps the most important thing that sustains social practices from the point of view of a discourse analyst however, is discourse.  Discourse helps to maintain social practices in a number of ways. One way is through some kind of explicit description or set of rules governing the practice. Most practices however, are not described explicitly in statutes or rulebooks. In this case they may be described or represented in more indirect ways.

In his book Discourse and Practice, Theo van Leeuwen argues that nearly all discourse is in some way a representation of practice. For him and other critical discourse analysts the best way to understand social practices is to examine how they are represented, for example how certain participants and actions are made more or less prominent in discourse.

Mediated discourse analysts are less interested in the way practices are represented and more interested in the role of discourse in how they are actually performed. So for example, certain forms of discourse or genres may help to sustain a social practice through making certain kinds of actions easier to perform and other kinds of actions harder to perform.

One example of this is the ‘Like’ button in Facebook. This piece of discourse helps to define and maintain the practices of sharing and commenting on this social networking site in a profound way. Since ‘liking’ something is the easiest way to comment, it makes expressions of ‘liking’ much more common than other kinds of expressions (e.g. expressions of ‘concern’, ‘interest’, ‘ambivalence’ and ‘dislike’). Consequently, this also affects the kinds of things people habitually share, making it more likely that they will share things that will elicit an expression of ‘liking’ (pictures of my birthday party) as opposed to things more likely to elicit a different kind of response (e.g. information about the war in Afghanistan).

References

Bourdieu, P. (1977). Outline of a theory of practice, Cambridge: Cambridge University Press.

van Leeuwen, T. (2008) Discourse and practice: New tools for critical discourse analysis, Oxford: Oxford University Press.

Useful Links

Mash-up of Discourses: A Critical Analysis of the Videotext, "Dream America Movie", Nancy Fox
www.infohost.nmt.edu/~xchanges/index.php?option=com_content&view=article&id=79&Itemid=74

Communities of practice, a brief introduction, Etienne Wenger
www.ewenger.com/theory/

Habitus and field, Robert O. Keel
www.umsl.edu/~keelr/3210/3210_lectures/habitus_field.html

Bibliography of research on social networking sites
www.danah.org/researchBibs/sns.php

Back to top

A9: Multimodal cohesion

---

In Unit A2 I talked about what makes a text a text. One of the things I pointed out is that texts have cohesion. One of the main things that give a text ‘texture’ is the way different parts of the text are joined together through devices like repetition, reference, conjunction etc. But what about texts that are not verbal such as images, films and musical compositions? How do they achieve texture? Do they also have cohesion?

In his book Introduction to Social Semiotics (2004), Theo van Leeuwen discusses the different ways that cohesion is achieved in multimodal texts. Two of the most important ways multimodal texts achieve cohesion are rhythm and layout. Rhythm is more associated with texts that are organized in time, and layout is associated with texts that are organized in space.

Rhythm

Rhythm provides coherence and meaningful structure to events unfolding over time. It plays a crucial role in everyday interaction as well as in time-based modes such as film, music, and dance. Rhythm divides the flow of time in such texts into discernable units (measures) and organizes measures into phrases that are often repeated throughout the text in the same way words or phrases might be repeated in a written text. In this way, the rhythm of a measure of music for example, points backwards to previous measures and foreshadows subsequent measures.

Rhythm also helps to link up different modes in multimodal texts – the rhythm of the music on a skateboarding video for example, synchronized to the movement of the skaters depicted in the video.

Van Leeuwen argues that the reason rhythm has such a powerful cohesive effect has to do with the underlying rhythms of the human body and of nature. Just as rhythm makes multimodal texts cohesive, the beating of our hearts, the rising and setting of the sun and the regular changing of the seasons make our lives cohesive.

Layout

Whereas rhythm gives cohesion to texts that are organized in time, layout gives cohesion to texts that are organized in space. In Unit B9 I discuss at length how the arrangement of elements in images and the use of 'vectors' helps to join elements together into a meaningful whole.

In spatial texts, cohesion is created primarily though balance. The elements arranged in the different parts of a picture or a webpages, for example are often balanced in terms of their 'visual weight' (based on their size, colours, shapes and other aspects contributing to their prominence).

In two dimensional images this balance is achieved along the three axes of left and right, top and bottom, and centre and margin. In three dimensional texts like architectural structures, sculptures, parks and other built environments, the axes of front/back and top/bottom are also important.

Just as rhythm is related to the body, so is balance. Human beings are themselves symmetrical and we rely on a sense of balance to stand, move and navigate our way through the world.

Of course, some kinds of texts make use of both rhythm and layout for cohesion. In films and videos for example, the composition of shots and the arrangement of sets is organized spatially, based on principles of layout, whereas the movements of characters, the dialogue, and the musical soundtrack unfold according to rhythmic principles.

Useful Links

Transmedia transversals, Jay Lemke
www-personal.umich.edu/~jaylemke/papers/transmedia_traversals.htm

Speech, music, sound, Theo van Leeuwen
www.us.macmillan.com/speechmusicsound

Back to top

A10: Using corpora in discourse analysis

---

Using corpora in discourse analysis

In Unit A10 I considered the contribution tools from corpus linguistics could make to the analysis of discourse. The analysis of corpora is particularly useful for:

• Providing a ‘big picture’ perspective on discourse by revealing patterns in language across a range of texts
• Determining what grammatical patterns, lexical choices or generic structures are typical and unusual in given circumstances
• Analyzing how particular topics are ‘talked about’ in terms of grammatical patterns and lexical choices
• Analyzing the lexical and grammatical features of particular genres
• Mapping the occurrences of words or grammatical features in different parts of texts

Ideology and corpora

Perhaps the most valuable contribution corpus assisted analysis can make to the study of discourse is in the area of ideology. As I said in Unit A4, ideology is expressed through the words they use and how they combine those words. These include things like grammatical ‘agency’ (which participants are portrayed as performing actions and which are portrayed as having actions performed to or for them), modality (how possibility and obligation are expressed) and collocation (the kinds of words that are associated with other words). These lexical and grammatical choices combine to create certain ‘versions of reality’ in texts.

Just noticing these choices in one text however, does not really tell us much about how widespread or pervasive this ‘version of reality’ is. This is where corpus assisted analysis can be particularly useful. By examining these kinds of choices across a wider range of texts, we can understand better if and how much certain ideologies are being circulated through society.

Many analysts have employed corpora to do just that. One notable study was that of Stubbs (1996), who used corpus based techniques to analyze speeches to the Boy Scouts and Girl Guides in terms of grammatical structures and word collocations, showing how these two influential institutions promote particular ideologies regarding men and women's roles in society. Another example is Stubbs’s (2001) analysis of keywords associated with the education policy of the UK Conservative Party during the 1980s and 1990s (for example, words like ‘grammar’ and ‘standards’). Yet another example is Tuebert’s (2000) analysis of the language of the euro-skeptic debate.

As I note in Unit C4, another important way ideology is expressed in texts is through the ways the ‘voices’ of different people are represented (through, for example, direct quotation, paraphrases, presupposition or different kinds of reporting verbs). Corpus based analysis can also be useful in understanding this aspect of discourse. Garretson and Ädel (2008) for example, investigated a corpus of newspaper reports relating to the 2004 US presidential election, examining how speech was reported with respect to the use of direct versus indirect speech, the explicitness of source identification and the choice of reporting verbs.

Pragmatics and corpora

Corpora based analysis can also be useful in the study of spoken language and there are a number of corpora of spoken language available, including the London-Lund Corpus of Spoken English, which Aijmer (1996) used to examine conversational routines, such as thanking, apologizing and making requests. Similarly, Deutschmann (2006) used the British National Corpus to investigate apologizing in British English. Using corpora to study pragmatics, of course poses many methodological problems for the analyst since, as I discussed in Units A5 and B5, people often express speech acts indirectly, and their interpretation can be highly context dependant, whereas corpus based analysis is usually based on searching for clearly defined sets of lexical forms or grammatical patterns.

References

Aijmer, K. (1996) Conversational routines in English. Convention and creativity, London and New York: Longman.

Deutschmann, M. (2006) ‘Social variation in the use of apology formulae in the British National Corpus’, in A. Renouf and A. Kehoe (eds) The changing face of corpus linguistics, Amsterdam: Rodopi, pp. 205–22.

Stubbs, M. (1996) Text and corpus analysis, Oxford: Blackwell.

Stubbs, M. (2001) Words and phrases: Corpus studies of lexical semantics, Oxford: Blackwell.

Garretson, G., and Ädel, A. (2008) ‘Who’s speaking?: Evidentiality in US newspapers during the 2004 Presidential Campaign’, in A. Ädel and R. Reppen (eds) Corpora and discourse: The challenges of different settings, Amsterdam: John Benjamins.

Teubert, W. (2001) ‘A province of a federal superstate, ruled by an unelected bureaucracy. Keywords of the Euro-sceptic discourse in Britain’, in Musolff, A., Good, C., Points, P., and Wittlinger, R. (eds), Attitudes towards Europe: Language in the unification process, Aldershot: Ashgate, pp 45–88.

Useful Links

Annotated bibliography of corpus linguistics
www.courses.washington.edu/englhtml/engl560/corpusannbib.html

Discourse, news representations and corpus linguistics, Paul Baker
www.ling.lancs.ac.uk/staff/paulb/bham2011.pptx

What can a corpus tell us about pragmatics, Christoph Rühlemann
www.slideshare.net/perezparedes/corpus-linguistics-and-pragmatics

Back to top