Text analysis in translation

THE MINSTER (by the late Chancellor F. Harrison)

Beloved to Yorkshiremen, renowned the world over. This is true. Of great and noble churches in this country, probably three attract the greatest number of visitors. These three are Westminster Abbey, St. Paul's Cathedral and York Minster). (...) The east window deserves a note of its own. Seventy-six feet high and thirty-two feet broad, containing therefore more than two-thousand square feet of medieval glass - the great window at Gloucester Cathedral measuring seventy-two feet by thirty-eight feet, and containing more than two-thousand-three-hundred square feet of glass, but not wholly coloured - this great and grand window never ceases to excite admiration and wonder. The master-glazier, John Thornton, of Coventry, received for his work, in all, the sum of Ј 55 in three years, worth in modern currency - Ј 2,000? Who knows, even approximately? This was the pay of only one man. (From the brochure City and County of the City of York, Official Guide, 112 pages. I have left out the 12 pages on the history of the Minster.).

There are many small old churches, quaint and often glorious towers and the breathtaking spectacle of the Minster. It took two-and-a-half centuries, from 1220 to 1470, to complete this poem in stone. Inside, a kaleidoscope of light explodes from windows of medieval stained glass that are among the art treasures of the world. (Last of the three paragraphs on York, from the book AA Illustrated Guide to Britain, 544 pages)

York Minster is the largest of England's medieval cathedrals. The result of 250 years of building, it shows a variety of styles. The transepts are the earliest part of the present building, dating from 1220-1260; the nave, chapter house, and vestibule were built in 1291-1345 in Decorated style; the choir in 1361, the central tower in 1400-1423, and the western towers in 1433-1474 in early and late Perpendicular. The Minster contains some of the earliest glass and the biggest acreage of stained glass in Britain. The lancet lights of the "Five Sisters" in the north transept are a particularly fine example of 13th-century grisaille glass. (Paragraph on York Minster - under the heading "York" -from The New Caxton Encyclopedia, 18 vols.)

For translation-oriented text analysis, it is most important to elicit features typical of the medium, i.e. features of content and/or form, and to classify them as culture-specific or transcultural or even universal. This is particularly relevant in those cases where the target text is to be transmitted through a medium or channel different from that of the source text.

How to obtain information about the medium

If the source text is not available in its original medium, but only in a copy or typescript (which actually occurs fairly frequently in translation practice), the translator must insist on having detailed information about the medium, as it is rather difficult to identify the medium from intratextual analysis alone. There may be some clues in the dimensions of the sender and his/her intention or motive; time and place, too, sometimes narrow the field of possible media. In some cases, the choice of medium is determined by convention since there are favourite media for particular communicative purposes in every culture (e.g. posters or newspaper advertisements for product promotion, leaflets for tourist information, etc.).

Checklist

The following questions may help to find out the relevant information about the dimension of medium or channel:

Has the text been taken from a spoken or a written communication? By which medium was it transmitted?

Which medium is used to present the text to the target audience? Is there any extratextual information on the medium?

What clues as to medium or channel can be inferred from other situational factors (sender, intention, motive, function)?

What conclusions can be drawn from the data and clues obtained about the medium as regards

other extratextual dimensions, such as the addressees and their expectations, motive, and function, and

the intratextual features?

Place of communication

The dimension of space refers not only to the place of text production, i.e. the actual situation of the sender and the text producer, but also, at least in connection with certain media, to the place of text reception. It cannot be equated with the dimension of medium. The dimension of space is of particular importance where languages exist in various geographical varieties (such as the Spanish spoken in Spain as opposed to Latin America or even Peru, Mexico, Argentina etc., and the English spoken in Great Britain as opposed to the United States, Australia, India etc..

Example

The Portuguese version of the information brochure published by the Tourist Office of Munich was accepted unhesitatingly as being correct and appropriate by a group of Brazilian teachers in a seminar on translation, whereas their colleagues from Portugal classified the text as "more or less understandable, but unidiomatic and not conforming to normal usage". In this case, an analysis of the dimension of place could not throw any light on this problem because the text had been produced in Munich for "Portuguese"-speaking receivers. As the name of the translator was not specified in the text imprint, the participants in the seminar could only assume that the translator - whether he or she was a native speaker or not - had used the Brazilian variety of Portuguese. The sender/initiator (the Tourist Office) had probably not been aware of the problem. For the German version of this brochure, however, the dimension of place (of reception) would suggest that the text is written in the variety used in Germany (as opposed to Austria or Switzerland).

In addition to the linguistic aspects, the dimension of space can be important for the comprehension and interpretation of a text in that the place of text production may be regarded as the centre of a "relative geography". The distance or significance of other places must often be judged in relation to this centre. The translator has to take into account that the "relative geography" from the standpoint of TT production may be quite different from that of ST production.

Example

The difference in cultural or social level could be called "downgrade" or "upgrade", depending on whether it is seen from the lower or the higher level.

The distance between London and Liverpool is much "shorter" as perceived by a Texan than by an Englishman,

c) The names of places, areas and tribes listed in Act 2, 9-11, do not make sense as a description of the "horizon of the Jewish world" unless Syria is assumed to be the place of text production, and not Jerusalem, where the Pentecostal event is set.

What to find out about the dimension of space

In the dimension of space we have to consider not only linguistic aspects but also cultural and political conditions. A text published in a country where literature is censored must be read "in another light" than a text whose author has not been subject to any restrictions, since authors under censorship often write "between the lines".

In addition to the name of the state or country the text comes from, it may even be necessary to know the exact area or town of text production in order to be able to interpret the deictic elements correctly. This applies to the ST as well as to the TT, which would normally be read in the target cultural environment.

Example

In the case of newspaper articles, the place where the paper is published is normally taken to be the place of text production as well. Therefore, readers of the Sunday Times can assume that the information "Mortgage cut in sight" refers to Great Britain, while all articles on the first page of the international edition of the Herald Tribune have to indicate the place the article refers to: "U.S. Banks Lower Prime Interest Rate", "In Leipzig, Protesters Fear Resurgence of Communist Power", "Tamil Guerrilla Army Nears Goal in Sri Lanka", etc. If correspondents send their reports from somewhere else, the place of text production is usually specified together with the author's name ("By David Binder, New York Times Service, Bucharest") or at the beginning of the text ("LEIPZIG, East Germany"), so that the reader can interpret a sentence as "Now everything is quiet around here again" correctly. In a translation, too, the dimension of place has to be specified either externally (e.g. in an introduction) or internally (e.g. "Now everything is quiet around Leipzig again").

Information about the place of text production also gives an indication of the cultural affiliation of the sender and/or the addressees, the medium (in the case of culture-bound or culture-specific media), the motive (at least where combined with the dimension of time) arid the in-tratextual features (such as regional dialect or deictic expressions).

How to obtain information about the dimension of space

As a rule, information about the dimension of space can be found in the text environment in the form of the place of publication, the name of the publishing company, the first edition details, or newspaper headlines, or in the secondary literature. Sometimes, it is presupposed to be part of the receiver's general background knowledge (e.g. in the case of publications by international organizations or institutions or by world-famous writers). From the intratextual point of view, certain linguistic features may provide a clue as to where the text was written or intended to be read.

Other clues may be obtained from the information about the sender (e.g.: Where did s/he live, work, etc.?), the addressed audience (e.g.: What culture-specific information may be presupposed to be known by the receiver?), medium (e.g.: Is it bound to a certain culture?), or motive (e.g.: Is it a culture-specific motive?).

Checklist

The following questions may help to find out the relevant information about the place of communication:

Where was the text produced or transmitted? Is any information on the dimension of space to be found in the text environment? Is any information on space presupposed to be part of the receiver's general background knowledge?

What clues as to the dimension of space can be inferred from other situational factors (sender, receiver, medium, motive)?

What conclusions can be drawn from the data and clues obtained about the dimension of space as regards

other extratextual factors (sender, receiver, medium, motive) and

the intratextual features?

Lecture 3. The relevance of the dimension of time and text function

Time of communication

Every language is subject to constant change in its use and its norms. So the time of text production is, first and foremost, an important pre-signal for the historical state of linguistic development the text represents. This applies not only to language use as such (from the sender's point of view) but also to the historical comprehension of linguistic units (from the receiver's point of view), which is itself bound to a certain period or epoch, since linguistic changes are usually determined by socio-cultural changes.

Moreover, this process of change affects the area of text types. Certain genres are linked to a particular period (e.g. oracles and epic poems as opposed to weather reports and television plays), and, of course, genre conventions also undergo change. Depending on the age of the text, the receiver/translator may have totally different expectations as to the typical features of the text type in question. S/he may even expect obsolete forms that are not used any more.

Example

Being asked what they thought to be the typical syntactic feature of a German recipe, the majority of competent native speakers of German mention the subjunctive of the present tense: "Man nehme...", whereas modern German recipes are written exclusively in infinitive constructions. Today, the subjunctive is used only to give a recipe an old-fashioned touch, as if it was from Grandmother s Recipe Book.

In addition to the linguistic aspects, the dimension of time can throw some light on the communicative background of the sender and the addressed audience, and thus provide a clue to understanding the sender's intention. In the case of text types of topical interest, such as news items and news reports, political commentaries, election speeches, weather reports, etc., the dimension of time can be the decisive criterion as to whether there is any point in a text being translated at all, or, if there is, under which circumstances and with which skopos it may be worthwhile.

In connection with the dimension of space, deictic elements refer directly to the situation. Like spatial deixis, temporal deixis can only be interpreted correctly if the receiver knows the time of text production.

Example

In the International Herald Tribune of January 9, 1990, we find the following notice: "NEW YORK - The hopes entertained that the grippe was relaxing have been destroyed by the mortality returns of yesterday (Jan. 7), which show an increase of nearly 100 over the toll given three days ago, with 134 deaths traceable to the epidemic." No need to be alarmed: the notice is to be found under the heading "100, 75 and 50 years ago", and dates from 1890.

However, it may also be necessary to know the genre conventions in this respect, as the following example shows.

Example

In Madras, I was surprised to read in the morning paper lying on my breakfast table that "there was a train crash this afternoon". Of course, the text had probably been written late at night, and the author was quite right to say "this afternoon" - but in a German newspaper (and normally in British and American papers as well) the author would have written yesterday afternoon because it seems to be a convention here for newspaper writers to imagine themselves in the situation of the reader who receives the text the next morning, whereas obviously the Indian readers are expected to put themselves in the writer's shoes.

Sometimes it may be wise for the translator to check on the validity of the information given in the source text (if possible) or at least to point out to the initiator that some information in the text may not be up to date.

Example

In some tourist information leaflets, the information on opening hours, prices etc. or warnings such as "is being repaired" (cf. example 3.1.4./2a) are not up to date. For example, the latest (translated) published information on the famous Altamira caves in Northern Spain specifies that the caves can be visited by anybody "on request". When I went there to have a look at the prehistoric paintings, I found out that there was a pavilion with beautiful reproductions of the paintings - but the caves had not been open to the public for the past few years. Only persons presenting proof of a particular research project were allowed to enter.

The dimension of time influences directly or indirectly the dimensions of sender (e.g.: Is s/he a contemporary of the receiver/translator or not? What situational presuppositions can be made?), intention, audience (expectations, temporal distance between ST and TT addressees), medium (historical or modern forms of medium), motive (e.g. topicality), and, above all, intratextual features (e.g. presuppositions, historical language variety, deictic elements).

The traditions and conventions of translation

The dimension of time encompasses not only the time of ST production and reception but also that of TT production (= translation) and reception. The original communicative situation as well as the inter-cultural communicative situation are determined by their respective temporal contexts.

In connection with the dimension of time, we must therefore look at the traditional translations of classical texts and consider the problems involved in translating or re-translating old texts. Whether and how the dimension of time has to be taken into account for the translation of, say, Homer's Iliad, Shakespeare's King Lear, or Cervantes' Don Quixote depends on the translation skopos. Popovic ([1977]1981: 103f.) distinguishes between the "synchronous translation" of a contemporary author and modern translations of older texts, which in his opinion can be either "re-creative" (i.e. actualizing) or "conservative" (i.e. historicising).

Which approach is regarded as the "correct" one depends on the prevailing translation tradition or concept, which may be regarded as a kind of culture-specific convention.

How to obtain information on the dimension of time

Information on the dimension of time can sometimes be inferred from the date of publication of the text or other clues from the text environment, although this is not always reliable, as texts are often published years after they have been written. However, they cannot be published text type, it will be mainly the following intratextual features that are determined by the motive of communication: content (insofar as the motive is explicitly mentioned in the text), vocabulary and sentence structure (e.g. in a memorial address), suprasegmental features (memorial address vs. election speech), and non-verbal elements (e.g. black edging round a death announcement).

How to obtain information about the motive for communication

Although the motive for communication is closely linked with the dimension of time, the two factors must not be confused. While the dimension of time is part of the communicative situation (in the narrower sense), the dimension of motive relates the communicative situation and the participants to an event that is outside, or rather prior to, the situation.

It is not always easy therefore (and not always relevant to translation!) to find out which event has motivated a certain text. Sometimes the motive is referred to in the text or mentioned in the text environment (e.g. in the title: To Honor Roman Jakobson on the Occasion of his 70th Birthda); but there are communicative situations in which the motive is only an indirect reason for the author to deal with a loosely connected subject.

Example

On March 12th, 1984, the Spanish daily paper El Pais published a commentary under the title "El Dfa de la Mujer" (International Women's Day). It is the motive for text production this title alludes to and not the subject matter, because the text deals with the situation of working women in Spain in 1984. The newspaper reader was expected to be familiar with the occasion, International Women's Day, since it had been commented on quite frequently at the time. If the text is to be translated, it is the motive for translation (as well as the dimensions of time and place) that has to be taken into account. Only a few days later the date will have been pushed into the background by other events, and a title like "International Women's Day" will arouse specific expectations about the subject matter, which the text cannot meet.

As is illustrated by the example, the dimension of motive is of as much interest to the translator as that of time, because s/he has to contrast the motive for ST production with the motive for TT production and find out the impact this contrast has on the transfer decisions. While the motive for ST production is often to be found in the "environment" of the sender or text producer, the motive for TT production can be inferred from what is known about the transfer situation, i.e. the initiator and the translation brief. The effect of the motive on intra-textual features - as opposed to that of the dimension of time - is often merely an indirect one.

We can restate that the clues as to the motive or motive type are to be inferred from certain situational factors, such as medium (e.g. political section of a newspaper), place and time (in connection with the receiver's general background knowledge), and, of course, text function, if this is specified by unambiguous pre-signals, such as genre designations (e.g. "protocol") or text-type features (e.g. black edging). The information obtained on the sender and the intention usually permits only indirect conclusions as to the motive for communication.

Checklist

The following questions may help to find out the relevant information about the motive for communication:

Why was the text written or transmitted? Is there any information on the motive of communication to be found in the text environment? Is the ST receiver expected to be familiar with the motive?

Was the text written for a special occasion? Is the text intended to be read or heard more than once or regularly?

What clues as to the motive for communication can be inferred from other extratextual dimensions (sender, intention, receiver, medium, place, time, function)?

What conclusions can be drawn from the data and clues obtained about the motive for communication as regards

other extratextual factors (expectations of the receiver, sender and intention), and

the intratextual features?

5. What problems can arise from the difference between the motive for ST production and the motive for translation?

Text function

The relationship between text function and genre

Let me briefly restate that the notion of text function means the communicative function, or the combination of communicative functions, which a text fulfils in its concrete situation of production/reception. It is derived from the specific configuration of extratextual factors (sender/sender's role, intention, receiver/receiver's expectation, medium, place, time, and motive). The notion of text function is related to the situational aspect of communication, whereas the notion of genre is related to the structural aspect of the text-in-function. It is like looking at the two sides of a coin: they cannot be separated, but they are not the identical.

As was pointed out above, text can be classified on various levels of generalization. It is therefore not surprising that some authors specify text types as "newspaper reports", "sermons", or "resolutions", while others prefer a more general categorisation into "informative", "expressive", or "operative" texts.

Literariness as a text function

The notion of text function as a particular configuration of situational factors can be illustrated by the special function of literary texts. The senders of a literary text are usually individual authors who are also text producers and who in the literary context are known as "writers". Their intention is not to describe "reality", but to motivate personal insights about reality by describing an (alternative) fictitious world. Literary texts are primarily addressed to receivers who have a specific expectation determined by their literary experience, and a certain command of the literary code. As a rule, literary texts are transmitted in writing (= medium), although sometimes orally transmitted texts (such as fairy tales) are included in literature as well. The situational factors (place, time, motive) may not be of great significance in intracultural literary communication but they do play an important part in literary translation because they convey the culture-specific characteristics of both the source and the target situation.

The importance of ST function for translation

The basic principle of functionalism in translation is the orientation towards the (prospective) function of the target text. Since I have argued that a change of function is the normal case, and the preservation of function the special case in the process of intercultural text transfer.

If a translation is an offer of information about the source text, there can be two fundamental kinds of relationship between source and target text. Here again we find the two translation theories which have split translation scholars into two camps: the supporters of liberty and the adherents to fidelity. The target text can be (a) a document of a past communicative action in which an SC sender made an offer of information to an SC receiver by means of the source text, and (b) an instrument in a new TC communicative action, in which a TC receiver receives an offer of information for which the ST provides the material. Accordingly, we can distinguish between two translation "types": documentary and instrumental translation.

Documentary translations (such as word-for-word translation, literal translation) serve as a document of an SC communication between the author and the ST receiver, whereas the instrumental translation is a communicative instrument in its own right, conveying a message directly from the ST author to the TT receiver. An instrumental translation can have the same or a similar or analogous function as the ST.

In a documentary translation, certain aspects of the ST or the whole ST-in-situation are reproduced for the TT receivers, who is conscious of "observing" a communicative situation of which they are not a part. A documentary translation can focus on any of the features on each rank of the source text, pushing others into the background. In a word-for-word translation, for example, which aims to reproduce the features of the source language system, the focus is on the morphological, lexical, and syntactic structures presented in the source text, whereas textuality is bound to be neglected.

An instrumental translation, on the other hand, serves as an independent message-transmitting instrument in a new communicative action in TC, and is intended to fulfill its communicative purpose without the receiver being aware of reading or hearing a text which, in a different form, was used before in a different communicative action. This translation type comprises three forms. First, if the target text can fulfill the same function(s) as the source text, we speak of an "equi-functional" translation (used, for example, in the case of operating instructions or business correspondence). Second, if the ST functions cannot be realized as such by the TT receiver, they may be adapted by the translator, provided that the TT functions are compatible with the ST functions and do not offend against the sender's intention (e.g. the translation of Swift's Gulliver s Travels for children). This form is referred to as "heterofunctional translation". The third form is intended to achieve a similar effect by reproducing in the TC literary context the function the ST has in its own SC literary context. This form is often found in the translation of poetry.

How to obtain information about text function

The most important source of information is, again, the text environment, since designations like "operating instructions" or "anecdote" call on the receivers' reading experience of the text type in question and build up a specific expectation as to text function(s). It is obvious that these "labels" can be misleading if they are used inadequately by the author or sender (whether intentionally or unintentionally). On the other hand, it may be assumed that in normal communication such designations are in fact intended as a guideline for the receiver.

If there is no genre designation, the text function or functions have to be inferred from the configuration of the external factors. This is why text function should be analysed last when as much information as possible is available. As was illustrated by the example of literary texts, the intention of the sender and the expectations of the receiver are the crucial dimensions in this respect. However, other factors may also narrow the range of possible functions, such as sender (e.g. a candidate for presidency), medium and place (e.g. a public speech in the market place of a mountain village), time (e.g. shortly before the general elections), and motive (e.g. an election campaign).

The pragmatic relationships between sender, receiver, medium, and motive, provide the translator with a number of pre-signals announcing a particular function, which will be either confirmed or rejected by the subsequent analysis of the intratextual features. If the translator finds his or her expectations confirmed, s/he has reason to believe that s/he has elicited the correct function - if not, there are two possible explanations: either the author has intentionally violated the norms and conventions of the text type, or the translator has interpreted the pre-signals wrongly and therefore has to go through the process of eliciting the text function on the basis of pragmatic pre-signals again.

Checklist

The following questions may help to find out the relevant information about text function:

1. What is the text function intended by the sender? Are there any hints as to the intended function in the text environment, such as text-type designations?

2. What clues as to the function of the text can be inferred from other extratextual dimensions (motive, medium, receiver, intention)?

3. Are there any indications that the receiver may use the text in a function other than that intended by the sender?

4. What conclusions can be drawn from the data and clues obtained about text function as regards

other extratextual dimensions (sender, intention, receiver, medium, time, place, and motive), and

the intratextual features?

The interdependence of extratextual factors

The checklist questions suggested in connection with the extratextual factors illustrate the interdependence of the extratextual factors on the one hand, and of the extratextual and intratextual factors (which have so far not been specified), on the other. Data and clues about a single factor can be derived from the data and clues obtained about the other factors.

The most important principle, however, is that of recursiveness. This type of analysis is no one-way process, but contains any number of loops, in which expectations are built up, confirmed, or rejected, and where knowledge is gained and extended and understanding constantly modified. This applies not only to the analysis of the text as a whole and to the individual text factors but also, if the analysis and translation of microstructures leads incidentally to new discoveries requiring previous transfer decisions to be corrected, to the processing of smaller text units such as chapters or even paragraphs.

The interdependence of the extratextual factors is illustrated by a diagram (Figure 5), in which arrows are used to show the course of the analytical procedure. Those steps which yield reliable data are depicted by a continuous line, while the steps which merely lead to clues are represented by a dotted line.

2. “Intratextual Factors in Translation Text Analysis”

Lecture 1. Basic notions

It is the verbal elements (lexis, sentence structure and the suprasegmental features, i.e. the "tone" of the text) which are most important for conveying the message. In both written and spoken texts suprasegmental features serve to highlight or focus certain parts of the text and to push others into the background. All these elements have not only an informative (i.e. denotative), but also a stylistic (i.e. con-notative) function.

The intratextual features are influenced to a large extent by situational factors (e.g. the geographical origin of the sender, the special requirements of the chosen medium, the conditions of the time and place of text production, etc.), but they can also be determined by genre conventions or by the sender's specific communicative intention, which affects the choice of the intratextual means of communication. We also have to account for the fact that stylistic decisions are frequently interdependent. If, for example, the sender decides on a nominal style in the area of lexis, this will naturally affect the choice of sentence structure.

We distinguish eight intratextual factors: subject matter, content, presuppositions, composition, nonverbal elements, lexis, sentence structure, and suprasegmental features. In practical analysis it has proved effective to deal with the factors in the order in which they appear here. However, there is no real reason why this cannot be changed, since the principle of recursiveness again allows any feedback loops which may be deemed necessary.

In the practical application of the model it may not always be necessary to go through the whole process of intratextual analysis step by step. Some translation briefs will be such that merely a cursory glance at the intratextual features is sufficient (just to find out, for example, whether or not the framing of the text corresponds to genre conventions), whereas others may require a detailed analysis right down to the level of morphemes or phonemes.

Example

If a strongly conventionalized text, such as a weather report, has to be translated in such a form that the target text conforms to the target-culture conventions of the text type, there is no need to analyse all the intratextual details of the source text, once it has been stated that they are "conventional". Since the intratextual framing of the TT has to be adapted to TC conventions anyway, the intratextual framing of the ST may be regarded as irrelevant for translation.

When we analyse the linguistic features of a particular text, we soon realize that they all have to be evaluated in a different way, depending on the function they have in the text. There are features that depend on situational conditions which cannot be controlled or modified by the sender (e.g. pragmatics of time and space, geographical or socio-cultural background of the sender himself) or features that may have been determined by a decision taken prior to text production (e.g. choice of medium or addressee orientation). Then, there are other features which are dictated by social norms (e.g. text-type or genre conventions and so on). During the process of analysis, therefore, the translator constantly has to go back to factors which have already been analysed (= principle of recursiveness). Lastly, there is a type of feature which depends on the sender deciding on one out of several alternative means of expression, a decision determined by the intention to produce a certain effect on the receiver.

General considerations on the concept of style

In order to be able to understand a stylistic signal or sign, the receiver has to be equipped, like the sender, with a knowledge or command of stylistic patterns and of the functions that they are normally used for. This knowledge is part of text competence and will enable the receiver to infer the intentions or attitudes of the sender from the style presented in the text. It is based on the fact that most communicative actions are conventionalized and that text producers almost always proceed according to a given pattern. In ordinary communication an intuitive, unconscious, or "passive" knowledge of stylistic patterns will be more than sufficient to ensure the comprehension of the text. However, the receiver/translator cannot manage without an active command of such patterns of expression both in SL and TL, since it enables them to analyse the function of the stylistic elements used in the source text, and to decide which of these elements may be appropriate for achieving the target function and which have to be changed or adapted.

Subject matter

How to obtain information about the subject matter

As was mentioned above, the conventions of certain text types seem to dictate that the title or heading or the title context (comprising main title, subtitle(s) and the like) represent a kind of thematic programme. An example of this is the following title of a linguistic article: "Understanding what is meant from what is said: a study in conversationally conveyed requests" (Clark & Lucy 1975).

Where the information is not given by a thematic title like this, the subject matter of a text can be formulated in an introductory lead, as is very often the case, for example, in newspaper articles (cf. Liiger 1977: 49ff.) or in the first sentence or paragraph which can then be regarded as a kind of "topic sentence" paraphrasing the thematic essence of the text.

Example

The Soviet Disunion

UNITED IT STANDS ...DIVIDED IT FALLS

While 1989 was the year of eastern Europe, 1990 may be the year of the Soviet Union. Confronted by growing nationalist unrest and economic mayhem, the empire is beginning to come apart at the seams. James Blitz in Moscow reports on the crisis in the Kremlin (...). (The Sunday Times, 1 January 1990, p. A l.)

Example

Title: Ford Is Rebuffed By Mazda Sub-title: No Chance Seen For Larger Stake

TOKYO - Mazda Motor Corp. said Monday that it saw no opportunity for Ford Motor Co. to enlarge its stake in the Japanese company and that Mazda had no plans to raise funds by issuing new shares, warrant bonds or convertibles. (...) (InternationalHerald Tribune, 9 January, 1990, p. 9)

This applies not only to titles which are a shortened paraphrase of the text, but also to descriptive titles, e.g. of literary works.

Example

The original title El sigh de las luces ("The Age of Enlightenment") indicates the subject matter of the novel, while the titles of the English and the German translation (Explosion in A Cathedral Explosion in der Kathedrale) use the name of a picture that plays a symbolic part in the story. The reader, however, cannot recognize it as such and will probably interpret it as an indication of the subject matter or content. This may lead to a (wrong) classification of the book as a kind of thriller.

If the subject matter is not described in the title or title-context, it can be elicited by reducing the textual macro-structures to certain basic semantic propositions or information units, which constitute a kind of resume or "condensation" of the text. Occasionally, the translator is even asked to produce a short version of the text (i.e. a summary, abstract, or resume) in the target language. In translation teaching, the production of summaries can be used for checking text comprehension.

Condensing and summarizing, however, does not in all texts lead to an elicitation of the real subject matter, since in some cases this is obscured by a "false" subject occupying the foreground of the text. In these cases it is the analysis of other intratextual factors, mainly of lexis, which may lead to success.

The crucial concept in the analysis of the subject matter at the level of lexical items is that of isotopy. Isotopic features are semes shared by various lexical items in a text, thus interconnecting the lexical items and forming a kind of chain or line of isotopies throughout the text. The lexical items linked by isotopy are referred to as the "isotopic level", which may indicate the subject matter(s) of the text. There can be various isotopic levels in a text, either complementing each other or hierarchically subordinate to one another.

Checklist

The following questions may help to find out the relevant information about the subject matter of the text:

Is the source text a thematically coherent single text or a text combination?

What is the subject matter of the text (or of each component of the combination)? Is there a hierarchy of compatible subjects?

Does the subject matter elicited by internal analysis correspond to the expectation built up by external analysis?

Is the subject matter verbalized in the text (e.g. in a topic sentence at the beginning of the text) or in the text environment (title, heading, sub-title, introduction, etc.)?

Is the subject matter bound to a particular (SL, TL, or other) cultural context?

Do the TC conventions dictate that the subject matter of the text should be verbalized somewhere inside or outside the text?

Content

General considerations

Where the translator has a good command of the source language and is fully conversant with the rules and norms governing text production, s/he will usually have little or no difficulty in determining the content of a text. Even so, it would still be useful to have some means of checking this intuitive understanding. It would be even more useful, of course, to have some guidelines available in translator training, where competence in this area is still inadequate.

Paraphrase as a procedure for content analysis

By "content" we usually mean the reference of the text to objects and phenomena in an extralinguistic reality, which could as easily be a fictitious world as the real world. This reference is expressed mainly by the semantic information contained in the lexical and grammatical structures (e.g. words and phrases, sentence patterns, tense, mood, etc.) used in the text. These structures complement each other, reduce each other's ambiguity, and together form a coherent context.

Therefore, the starting point for the analysis of content has to be the information carried by the text elements linked on the surface of the text by the text-linguistic linking devices, such as logical connections, topic-comment relationships, functional sentence perspective, etc.

Since at this stage the external analysis of the communicative situation has been completed, the meaning of the text can be elicited, as it were, "through the filter" of extratextual knowledge.

Analysing the content of syntactically or semantically complicated texts can be made easier by a simplifying paraphrase of the information units, which can be formulated independently of the sentence structure. However, in so far as they are explicitly verbalized in the text, the logical relationships between these units should be noted. This procedure permits the translator to identify (and possibly compensate for) presuppositions, and even defects in coherence, which frequently occur in texts.

These paraphrases have to be treated with great caution, however. The paraphrased information units form a new text which is in no way identical to the original. Paraphrases can only be used in order to simplify text structures, making them more transparent. When paraphrasing lexical items we also have to take account of the connotative content, which has to be preserved, or at least marked, in the paraphrased text.

In any case, it must not be the simplified paraphrase which should be taken as a starting point for translation, but the original source text.

Connotations

The amount of information verbalized in a text includes not only denotative but also connotative (or "secondary") meaning, i.e. the information expressed by a language element by virtue of its affiliation to a certain linguistic code (stylistic levels, registers, functional style, regional and social dialects, etc.). By selecting one specific element in preference to another from a number of possible elements the author assigns a secondary meaning to the text. Since the connotative meaning can only be analysed in detail in connection with the stylistic values of lexis, sentence structure and suprasegmental features, I would recommend at this stage of the analysis provisionally marking those text elements which can be intuitively classified as "probably connotative". The extratextual category of text function often provides a certain expectation here.

Example

Kate Saunders in The Sunday Times, 7 January 1990: Career woman - or just the little woman?

Chic dinner tables are resounding with funereal orations over the twitching corpse of the women's movement - they come to bury it, certainly not to praise it. It was so selfish, so uncaring, so unnatural - surely home-building is nicer and more fulfilling than hacking through the professional jungle? The Eighties' ideal was the woman who ran a business, made breakfast appointments with her own husband, and spent 20 minutes' "quality time" a day with her children. But women are wondering now whether the effort of juggling home and career was worthwhile. All you got for your pains was nervous exhaustion, and kids who spoke Icelandic because they were brought up by the au pair. How much simpler to give up the struggle and devote yourself to stoking the home fires. Part of the problem seems to be that women are discovering the real snag about equality - work is a pain. Any man could have told them this. (...)

Certain connotations are a part of every speaker's communicative knowledge whether they speak the standard language or a particular regional and/or social dialect. They are linked so closely to a lexical item that they would be specified in the dictionary (e.g. kid is marked "slang" in OALD 1963, and "informal" in OALD 1989, whereas snag is marked "colloquial" in OALD 1963 and not marked at all in OALD 1989). Connotations such as these, even though they may change in the course of time, must be considered to be part of the "linguistic competence" of sender and receiver. Other connotations, however, are merely valid for certain persons, since they can only "work" if the participants know particular social, political, regional or cultural phenomena, e.g. career woman vs. the little woman or the allusion to Shakespeare's Julius Caesar in example 3.2.274. Such connotations belong to the "horizon" of sender and receiver.

In his famous book How to Be an Alien, George Mikes gives a humorous example.

Example

"You foreigners are so clever," said a lady to me some years ago. First, thinking of the great amount of foreign idiots and half-wits I had had the honour of meeting, I considered this remark exaggerated but complimentary. Since then I learnt that it was far from it. These few words expressed the lady's contempt and slight disgust for foreigners. If you look up the word clever in any English Dictionary, you will find that the dictionaries are out of date and mislead you on this point. According to the Pocket Oxford Dictionary, for instance, the word means quick and neat in movement, skilful, talented, ingenious. (...) All nice adjectives, expressing valuable and estimable characteristics. A modern Englishman, however, uses the word clever in the sense: shrewd, sly, furtive, surreptitious, treacherous, sneaking, crafty, un-English, un-Scottish, un-Welsh (Mikes 1984: 42).

The "internal situation"

The information in the text can be "factual", i.e. based on the facts of what is conventionally regarded as "reality" by sender and receiver, or "fictional", i.e. referring to a different, fictitious world imagined or invented by the author, which is quite separate from the "real world" in which the communicative action takes place. However, this distinction is not of immediate importance for content analysis. Fictionality is a pragmatic property which is assigned to a text by the participants in communicative interaction. Its definition depends on the notion of reality and the norms of textuality prevailing in the society in question. If the notion of reality changes, then a text which was intended to be factual might be read as fictional, or vice versa. If we look at Aldous Huxley's Brave New World or George Orwell's 1984 we might come to the conclusion that a fictional text describing a Utopian situation could even become factual if reality were to change accordingly. However, the question of fictionality or factuality really becomes relevant to translation when we consider presuppositions.

Nevertheless, an analysis of content will have to specify whether or not the internal situation of the text is identical with the external situation. If it is not, the internal situation will have to be analysed separately, using the same set of WH-questions applied in the external analysis. This is very often the case in fictional texts, and in factual texts of the complex text type which contain embedded texts of another category.

In an internal situation there might be an internal sender (speaker, narrator), who may adopt various attitudes or perspectives towards the narration (e.g. "author's perspective", or "camera-eye" perspective), or there might be an implicit reader or listener, and implicit conditions of time and place; there may also be hints as to the medium used, the motive for communication and the function assigned to the particular embedded text. The internal situation may even, like the famous Russian doll, contain further embedded situations.

The situational factors of an embedded text are normally mentioned explicitly in the frame text, whereas the internal situation of a fictional text (i.e. its "setting") can often only be inferred from hidden clues or indirect hints, such as proper names of persons and places, references to culture-specific realities, elements of regional dialect in a dialogue, etc.

However, there are cases where an analysis of the external situation yields information on the internal situation, as shown in the following example.

Example

In one of his short stories written in his French exile in Paris, the Argentinian author Julio Cortazar describes an urban environment which is not named explicitly, but hinted at by the information that from the window of his multistorey apartment block the auctorial narrator sees a sign saying Hotel de Belgique. The reference to the setting is not crucial to the interpretation of the story, which deals with the problem of daily routine and the hopelessness of life in modern society. The plot might equally well be set in any big city of the Western industrial world. But still, by describing (or pretending to describe) the view from his own window, the author gives a "personal touch" to the story, which makes it more authentic. This may be important for the translator when she has to decide whether to translate the description of a routine breakfast situation (tomamos cafe con leche) by "we drink our morning coffee" (neutral), "we have our coffee with milk" (non-specific strangeness) or "we have cafe au lait" (specific strangeness, explicitly referring to France as the setting of the story) or even "we have our ham and eggs" (receiver-oriented adaptation).