¶ … language-in-use, whether it is presented as text or speech. The meaning of the term is very heterogeneous and covers more than one approach to this subject. These approaches are very different with regard to their focus, purpose and techniques. As far as focus is concerned, discourse analysis may concentrate on the conclusions of the discourse itself or on the social processes and structure in accordance to which the discourse is constructed. Systemic linguistics approaches are appropriate for the first category, as there is always a very well defined boundary between language and society, with emphasis on the former. On the other hand, the common discourse analysis in sociology and social psychology has a broader focus and usually rejects the artificial distinction between discursive and social actions -- since "all discourse is action and all action is discursive."

The differences in purpose are not specific to discourse analysis but to social sciences in general, as researchers try to determine whether the exclusive goal is to produce knowledge or to concentrate on a social or political objective. Conversation Analysis and Critical Discourse Analysis are very different in this respect, as the former uses a strict neutral approach, while the latter supports the idea that serving political purposes cannot be avoided.

Discourse analysis may be applied upon a text in order to draw conclusions about the way in which the production and effects of the texts integrate themselves and further influence the social context to which they belong. In other occasions, discourse analysis is used to develop theories on the different types of discursive mechanisms in general. However, there is not always a clear distinction between these two approaches and it often seems that they are applied simultaneously. Critical Discourse Analysis studies the use of discursive devices for serving ideological purposes, while Conversation Analysis provides a theoretical foundation for abstract types of discursive strategy.

One very debated issue in the field of discourse analysis is how to relate the text to the context; various versions of discourse analysis have different study objects -- depending on the discourse feature on which they concentrate -- and different technical approaches. One major difference is the degree to which linguistic resources are employed: one problem is to determine the level at which analysis operates -- word, sentence, paragraph, the entire text etc. Another problem concerns the form of analysis employed. The particular type of analysis used may be modeled on a version of linguistics analysis. For instance "linguistic organization above the level of the sentence" was used a study object for J. McH Sinclair's and R.M. Coulthard's 1975 work on classroom discourse.

Conversational Analysis (CA) focuses on the use on one particular word and does not base its conclusion on linguistic research. CA is preoccupied with the way in which language is used in context and not on its structure. A very different approach was provided by twentieth century linguistics analysis, which has made a clear distinction between the structure and the actual use of language (or language and speech, linguistics competence and performance). Linguistic approaches, including systemic linguistics, try to find a relation between structure and function of speech, therefore reconciling language and its use.

Conversational Analysis concentrates on the way in which language is used in context, rather than on its structure. However, a very different approach is used in the works of linguists, who also use the term "discourse analysis." For instance, Zellig S. Harris states than the phenomenon of discourse analysis may be approached from two angles, as a result of the issues that lead to its study: one is the fact that descriptive linguistics usually works with one sentence at a time, which does not suit the needs of a more complex text analysis. The second issue at hand is the correlation of language and culture (briefly defined as the total amount of material and spiritual achievements of the population from a certain territory).

The first problem is caused by the simple and ineluctable fact that sentences are the normal "stretch" of words with meaning - in almost all languages, including English. Therefore, descriptive linguistics links any element in a stretch of speech...

However, due to the fact that grammar and logic combined offer a reasonable solution to speech construction, this was not viewed as a very important limitation.
The second problem refers to the fact that the scope of linguistics does not include any connection between social demeanor and language. Descriptive linguistics has not had the appropriate tools to engage in a an analysis of this issue, as it is not preoccupied with the meaning of the words, and can only offer the frequency rate of one morpheme among the others.

From this particular and very technical point-of-view, discourse analysis performs a series of operation upon a text of connected sentences. The first step is the collection of the elements or sequences of elements which have identical or equivalent environments of other elements within a sentence and considers that all these are equivalent to each other. Should materials that do not belong to any equivalence class appear, they would be associated with the class they are more closely tied to. The sentences are then divided into intervals, each one being a succession of classes, so that each interval is as similar as possible, regarding class composition, to the other intervals of the text. This succession of intervals is then subjected to analysis in order to establish a pattern of class occurrence.

These operations do not take into consideration the meaning of the words or the intent or conditions of the author of the text. According to Zellig. S. Harris, discourse analysis "yields considerable information about the structure of a text or type of text, and about the role that each element plays in such a structure." The role each element plays in the structure of a sentence is analyzed by descriptive linguistics.

Discourse analysis is also useful to determine the way in which a discourse should be constructed in order to meet certain specifications. Also, unlike descriptive linguistics, it provides information about stretches of speech which are longer that a single sentence. Therefore, relation existing between sentences are brought to light, although they are not immediately visible in the structure of the sentence.

Another approach relates to rhetoric and coherence. Rhetorical composition may be defined as the arrangement and sequencing of sentences, and rhetoric is the art of bounding those sentences together in a coherent way. Each declarative sentence contains "given" information (that which the speaker supposes that the listener is already aware of) and "new" information (the fact that are unknown to the listener and that he/she should be told about). There is a difference between the way the two kinds of information are presented: the given information is expressed by unstressed intonation or sentence position. Although this aspect is controversial, Goodwin and Perkins state that the most appropriate solution is that the unstressed position is that of the logical (and not necessarily grammatical) subject, and the stressed position is that of the logical predicate.

As for sentence construction, Clark and Haviland make the following suggestion: "Try to construct your utterance such that the listener has one and only one direct antecedent for any given information and that is the intended antecedent." By antecedent, the two authors mean the information presented to the listener / reader in the preceding sentences and that needs to be taken into account for a proper interpretation of the current sentence. The point of this technical construction is the detection of incoherence. All sentences that Clark and Haviland declare (by applying this method) as unacceptable or awkward are also incoherent.

I have chosen Lleyton Hewitt as the personality of choice, because he is not only one of the most appreciated people in Australia but because he has also lately captured the media's attention by breaking-up with his fiancee -- tennis star Kim Clijsters

The first article comes from "The Sydney Morning Herald" and is written by Alexandra Smith, who does not specify what her position in the newspaper is. The title of the article reminds the reader that the two stars have met on the tennis court, which was the most powerful thing that brought them together. "Hewitt and Clijsters call game over on romance" is a sentence conveying sadness and the impression that the joy of the game has ended.

The tile is interesting in itself. Although some may interpret it as being cynical (after all, a game is not something very serious), the noun "game" does not have a frivolous connotation. It reminds the reader of happy moments and time spent with friends, of competition and team spirit. Since the affair of the two tennis players was centered around tennis, the tile is more than appropriate.

The opening paragraph has the same structure as the title: first positive moods and then the news of the break-up. "They fell in love court-side and planned two weddings in each of their home countries to keep both families happy, but tennis stars Lleyton Hewitt…