The role of discourse markers in an Afrikaans stage translation of The Merchant of Venice.

Summary

In drama texts the written and the spoken modes work together to communicate multiple and complex messages simultaneously. Consequently, the dramatic text has a dual role in both the literary and the theatrical systems of a particular culture. This duality has also influenced the translation of drama. If the drama is intended to appear in print only, the translator is likely to approach the translation as a literary text and will then produce a page translation. In contrast, if the main aim is staging the drama, the translator will create a stage translation that will appeal to contemporary theatre-goers. Both page and stage translations of drama texts are written for spoken delivery. In other words, the dialogue in such texts is usually designed to simulate real-life, face-to-face communication. This is also the case in Shakespeare dramas and their translations. When a recent stage translation of The Merchant of Venice in Afrikaans is compared to an older page translation it is clear that the stage translator has deliberately employed certain linguistic features to simulate participation or "involvement" between characters and make them sound more like real people in authentic situations (Kruger 2000). It is therefore no surprise that the stage translation exhibits more contractions than the page translation--this is a primary method in any language to indicate spoken speech. What is unusual, though, is the deliberate insertion of a far wider range of discourse markers in the stage translation, despite its being much shorter than the page translation. The only logical explanation for this particular finding is that the translator has actively attempted to influence the conversational coherence of the dramatic dialogue of the stage translation by foregrounding the interpersonal and text-building functions of discourse markers such as feedback words, interjections, exclamations, vocatives and courtesy adjuncts.

Opsomming

In dramatekste werk geskrewe en gesproke apekte saam om tegelykertyd veelvoudige en komplekse boodskappe te kommunikeer. Gevolglik speel die dramatiese teks 'n tweeledige rol in beide die literere en die teatersisteme van 'n besondere kultuur. Hierdie tweeledigheid beinvloed ook die vertaling van dramatekste: as die vertaling slegs in druk gaan verskyn, word die vertaling as 'n literere teks benader en 'n boekvertaling is die resultaat; andersins, as die vertaling opgevoer gaan word, word 'n verhoogvertaling gemaak wat eietydse teatergangers lok. Beide boek- en verhoog-vertalings word vir gesproke aflewering geskryf, met ander woorde, die dialoog word gewoonlik so geskep dat dit ware kommunikasie in die regte lewe naboots. Dit is ook die geval met Shakespearedramas en hul vertalings. Toe 'n onlangse verhoogvertaling van The Merchant of Venice in Afrikaans met 'n ouer boekvertaling vergelyk is, het dit geblyk dat die verhoogvertaler doelbewus sekere talige verskynsels ingespan het om deelname of "betrokkenheid" tussen die karakters na te boots en hulle te laat klink soos regte mense in ware kommunikasiesituasies (Kruger 2000). Dit is dus geen verrassing dat die verhoogvertaling meer sametrekkings as die boekvertaling vertoon nie--dit is immers 'n basiese metode waarop gesproke taal in enige taal nagemaak word. Wat wel ongewoon is, is die doelbewuste toevoeging van 'n wyer reeks diskoersmerkers in die verhoogvertaling, hoewel hierdie vertaling veel korter as die boekvertaling is. Die enigste logiese verklaring vir hierdie bevinding is dat die vertaler daadwerklik gepoog het om die gesprekskoherensie van die dramatiese dialoog in die verhoogvertaling te beinvloed deur die interpersoonlike en teksbouende funksies van diskoersmerkers soos terugvoerwoorde, tussenwerpsels, uitroepe, vokatiewe en hoflikheidsadjunkte op die voorgrond te stel.

1 Introduction

Since the first translation of Hamlet by L.I. Coertze in 1945, no fewer than 33 Afrikaans translations of Shakespeare plays have appeared in South Africa, with a concentration of efforts in the late 1960s and early 1970s to celebrate the quatercentenary of the Bard's birth (cf Kruger 2000). Most of these translations have been performed at some stage or another, with a greater or lesser degree of success. However, these translations have since been stigmatised as "museum theatre" and are not performed any longer. When the (now defunct) Performing Arts Council of the Transvaal (PACT) proposed the performance of an Afrikaans translation of The Merchant of Venice, because it was prescribed as a setwork for South African schools in 1991, they were confronted with exactly this problem. The existing translations of The Merchant of Venice, namely Die Koopman van Venesie by D.F. Malherbe (1949), a renowned Afrikaans novelist, poet and playwright, and Die Koopman van Venesie by Anna Neethling Pohl (Pohl 1969), doyenne of the Afrikaans theatre, director and professor of drama studies, were deemed too "archaic" and page-oriented and therefore unsuitable for performance, especially since most performances would be attended by school pupils.

Tjaart Potgieter, at that stage the literary advisor of PACT and an actor, was commissioned to produce a translation of The Merchant of Venice that would be stage-oriented. He subsequently produced Die Sakeman van Venesie (literally, "the businessman of Venice") for the PACT production in June 1991 (Potgieter 1991). According to various theatre critics, this particular Afrikaans production was well received at the time. Barry Ronge, a theatre critic, stated:

Ilse van Hemert's production of Die Sakeman van Venesie ... bristles with ideas and grapples with key issues to the play in a provocative and intelligent manner. She modernises it, working on the assumption that our society, so greedy, so bent on achieving wealth and status is no different from that Venetian community Shakespeare described so many years ago. She strips the play bare of period fancy-dress....

(Ronge 1990/1991: 94)

Barry Hough (1991), another critic, said the performance of Sakeman was "'n wenner" [a winner] and Darryl Accone (1991: 7) thought it was "absorbing" and "playful" and that the translation was "efficient". These remarks are, of course, impressionistic but they prompted my interest in the idiosyncratic and stylistic aspects of the language of this particular translation of The Merchant of Venice. Clearly, the production had an enthusiastic reception and the dialogue was found to be not only exciting but also efficient. However, can statements such as these be supported by textual evidence from the script? Has the translator also "stripped" the dialogue just as the producer has literally stripped the play "bare of period fancy-dress"? What sort of textual features did the translator employ to simulate involvement between characters and make them sound more like real people in authentic situations?

These questions led to the research conducted in Kruger (2000; cf also Kruger 2004) which aimed to explore linguistic features of involvement in different registers of Shakespeare translation in Afrikaans. "Features of involvement" are those linguistic features which are typical of face-to-face communication (e.g. private verbs, contractions, first- and second-person pronouns, analytic negation, demonstratives, emphatics, discourse markers, causative subordination, amplifiers, questions, time- and place-adverbials). Such features are usually much less prominent in written texts. Both types of translation examined by Kruger (2000, 2004) are drama texts written for spoken delivery; the research question was whether, despite production constraints, the dialogue of a Shakespeare stage translation indeed exhibited more features of involvement than a page translation.

The overall finding was one of a statistically highly significant difference between the two registers, with the stage translation displaying more features of involvement than the page translation (Kruger 2000, 2004). (1) The fact that the stage translation exhibits more contractions than the page translation is no surprise--this is a primary method in any language to indicate spoken speech. What is unusual, though, is the deliberate insertion of a far wider range of discourse markers in the stage translation, despite its being much shorter than the page translation (cf Appendix A). It is the aim of this article to revisit earlier findings in this regard and to establish in more detail what role discourse markers play in the Afrikaans stage translation of The Merchant of Venice.

What then, does involvement in dramatic dialogue entail and what role do discourse markers play in this particular translation? In order to answer this question and to demonstrate how the research was conducted, it is necessary to discuss issues such as the nature of the drama text and dramatic dialogue, concepts such as performability, playability, speakability and modes of drama translation.

2 The Nature of the Drama Text

Ideally, every drama is written, and translated, to be performed. The dramatic text, as written text, addresses a context of performance which requires a change in the mode of discourse--the transformation and transmutation of the written lines into the dynamics of spoken speech, which involves more than the recitation of the lines of the text by actors (Herman 1995: 13). In other words, a dimension beyond the solely linguistic is involved, "for a play is much more than a literary text, it is a combination of language and gesture brought together in a harmonious frame of timing" (Bassnett-McGuire 1978: 161). This combination of the written and the spoken medium gives the drama its typical dual nature. As a result, it is possible to regard a drama as a written text and treat it as a literary text only, thereby ignoring its performance potential. Alternatively, it is also possible to treat a drama as a theatrical performance which can only be properly understood and evaluated in performance.

Separating the dramatic text from the performance makes the written text subject to literary scrutiny as it then becomes the domain of literary criticism, whereas the theatre text becomes the "happy hunting ground for producers, reviewers, reminiscing actors and theatrical theorists" (Elam 1978: 140). In this regard Link observes that

dramatic art has so far been considered as a mixed art, considering the dramatic text as literature and the production as a performing art. It is, of course, possible to consider the dramatic text as literature only. Actually this is the way it has predominantly been done in the past. It is meaningful if the literary qualities of the text are the main object of interest.

(Link 1980: 49)

Link (1980) is correct: the dramatic text as entity can be, and frequently is, analysed without any account being taken at all of performance--as a glance through the vast collection of writing on, for instance, Shakespeare's plays, will reveal. In doing so, literary critics usually implicitly or explicitly assume the priority of the dramatic text over the performance; in other words, the written text becomes a "blueprint" (Bassnett 1990: 72) for an eventual production. In the case of Shakespeare texts it has become almost impossible for an English director "to be freed from the tyranny of the written ... text which becomes a straight-jacket preventing mobility" (Bassnett-McGuire 1985: 88).

However, although it is useful to separate the two types of textual material--that produced for the theatre and that produced in the theatre--it is important to note that it has been pointed out that neither text precedes the other nor is more important as an object of study (Elam 1978: 140). The relationship between the two texts can best be described in Kristevan terms as intertextual because "we are dealing with different kinds of text which have an intimate relationship with each other" (Elam 1978: 157). The fact that this intertextual relationship is problematic is reflected in the different approaches to the study of drama texts. Bassnett (1990: 72) says that it is one of the obstacles to research on drama that it is studied separately in two different camps and that the literary and theatre scholars usually ignore each other's work. According to Mouton (1988: 7), a broad distinction can be made between the so-called "traditional" drama theorists and "modern-day" drama and theatre semioticians. Mouton (1988: 4, 37) contends that, although traditional theorists acknowledge the performance orientation of the dramatic text and thus also study the relation between the text and the performance, it is safe to assume that the dramatic text (and thus literary aspects) enjoy preference in their studies. The semioticians, on the other hand, are specifically interested in the communication process which is carried out in the dramatic text and the performance; in other words, they study the creation, transfer and reception of dramatic and theatrical signs, sign systems and codes. They have actually gone further by refining insights regarding aspects such as the dramatis personae, time, space, fictionality, language, action and the structuring of the dramatic world. Mouton's (1988) own research makes a valuable contribution towards linking the dramatic text with the theatre text by means of the notion of the performance orientation of a play and reconciling previously neglected aspects such as the role of the implicit spectator and the function of didascalies (cf also Keuris 1996).

It is a fact that the dramatic text does constrain the performance, the mise en scene, in obvious ways, not only in determining what the actors say and in establishing the structure of the action, but also by indicating movement, setting, props, and so on. And it is equally true that the performance constrains the written text by its performability (Elam 1980: 209). Performability, the "implicit, undefined and undefinable quality of a theatre text" (Bassnett-McGuire 1985: 101), involves the non-verbal, cultural and staging aspects of a drama, that is, the "encoded gestural subtext" (Bassnett 1991: 102), and stems from the requirement that "a play written for a performance must be actable and speakable" (cf Zuber 1980).

2.1 Performability, Playability, Speakability

According to Bassnett-McGuire (1985: 90-91), from a textual viewpoint, the concept of performability is usually applied to the fluency with which actors can utter the dialogue in a translated text; in this sense performability is often equated with "speakability" or "breathability". From the viewpoint of the performance, the concept is applied to a range of translation strategies aimed at adapting the "Other", such as replacing dialect in the source language by dialect in the target language or omitting passages deemed to be too rooted in the linguistic context of the source culture. Thus, "from the viewpoint of theatrical practice, playability, or actability are used as synonyms for performability" (Espasa 2000: 50). Because the term "performability" often appears in translators' prefaces, it seems to suggest that one particular translation is more congenial to eventual performance than other translations which do not adhere to this criterion. For Bassnett (1991: 102), this term is used to excuse the practice of handing over a so-called literal or draft translation to a playwright who then adapts it to suit his or her own purpose. It seems to be used to justify substantial modifications to the translation and to describe the supposedly concealed performance text within the written.

However, the realisation that an abstract notion of performance should not be put before textual considerations has resulted in drama scholars such as Bassnett (1991: 102, 111) reversing their opinion and arguing the case "against performability". She claims that "performability" as a criterion has "no credibility because it is resistant to any form of definition" (Bassnett 1998: 95). The time has come to focus more closely on the linguistic structures of the text itself for, after all, "it is only within the written that the performable can be encoded and there are infinite performance decodings possible in any playtext" (Bassnett-McGuire 1985: 102). The written text, incomplete though it may be, is "the raw material on which the translator has to work and it is with the written text, rather than a hypothetical performance, that the translator must begin" (Bassnett-McGuire 1985: 102). It is a fact that the principal problems facing the translator involve close engagement with the text on the page and the need to find solutions for a series of problems that are primarily linguistic ones, such as differences in register involving age, gender and social position. Therefore, Bassnett (1991: 111) argues that an investigation into the linguistic structuring of drama texts should take precedence over "an abstract, highly individualistic notion of performability" and that "the satisfactory solution of such textual difficulties" will result in the creation of a translated text that can be submitted "to the pre-performance readings" of those who will be involved in the performance.

"Speakability" is another vague term that concerns producers, actors and translators alike. Pavis (1992: 143), who regards speakability as meaning "easy pronunciation", warns against the danger of banality which "is lurking under cover of the text that speaks well". In contrast, Schultze (1990: 268) claims that speakability is an important instrument for producing literary and theatrical meaning, but that it should not be confused with convenient pronunciation. What matters is the type of speakability and its function in the process of generating theatrical meaning. According to Aaltonen (2000: 43), defining speakability in terms of simplicity is not an accurate way of characterising theatre texts, as such texts "do not need to be simple and easy to speak". Theatre texts may, and often do, differ from texts in the literary system.

The frequent use of terms such as "performability", "playability" and "speakability" can be regarded as "generalised descriptions of translation strategies in the theatre" (Aaltonen 2000: 43), which seem to set them apart from dominant views in translation studies of how translations relate to their source texts. If the norm is "faithful" translation, then a "deviant" approach to the source text, a "free" translation, can be justified in some way by saying that theatre requirements in terms of performablility and speakability have provided such justification. In effect, the use of these terms seems to generate two different types of translation: one, more related to the text or the page, and the other, to the specific style of presentation of the translator and/or the production company, that is, the stage.

2.2 Shakespeare Page and Stage Translations

According to Bassnett (1991: 105), two principal modes of theatre translation seem to have existed side by side since the seventeenth century, which in effect reflects the two-fold nature of theatrical discourse: "'aesthetic" and "commercial" translation. The first mode of translation perceives the drama as essentially a literary text--to be read on the page and translated as a literary text. Bassnett (1991: 105) maintains that the history of such translations is to be found in the history of the translation of poetry, for the debate on the quality of the translation centred on the re-creation of suitable verse forms in the target language. The principal criteria for the translators were "the power of the verse form and the status of the written text" as it would be judged according to aesthetic criteria for literature. According to Bassnett (1991: 106), at the same time, however, the commercialisation of theatre (particularly in northern Europe) led to a rapid turnover in "translations that could be adapted for performance in the new theatres by the emergent companies". Because the eventual performance was crucial, texts were anything but sacred and were reshaped according to very basic needs--audience expectations, size of company, repertoire of performers, limitations of time and space.

Van den Broeck (1993: 105) has also reduced the possible approaches to drama translation to a distinction between two basic modes. One mode, called "retrospective" translation, "treats the original in the manner of a 'poetic' or 'literary' text whose translation is to appear in print for the benefit of a potential readership". The other mode, called "prospective" translation, "strives for an overall reconstruction of the theatrical text in its broader sense of a 'performance' text(ure), which is to be understood as the complex network of signs belonging to dramatic, theatrical and cultural codes". According to Van den Broeck (1993: 105) "retrospective" can be replaced by page-oriented (or poetry-centred) and "prospective" by stage-oriented translation (cf also Habicht 1993: 49); notions which roughly correspond to Bassnett's (1991) aesthetic and commercial translation. These terms are, however, cumbersome and therefore only the terms "page translation" and "stage translation" will be used. In principle, practitioners of either mode are governed by Toury's (1980: 54-55) well-known "initial norm", that is, that the translator decides beforehand whether her approach will be more source-text-oriented or more target-culture-oriented. Van den Broeck (1993: 105) claims that adherence to the source text almost inevitably results in plays that do not acculturate the foreignness of the original, that is, page translations, whereas concern for the target audience will normally produce translations "which are living theatre texts capable of appealing to contemporary theatregoers", that is, plays that acculturate the original, or stage translations. The merit of page translations "depends on the degree to which the translator succeeds in reconstructing the linguistic, stylistic and textual properties of the foreign text for its new users"; while, on the other hand, the success of stage translations is likely "to coincide with the degree to which the translator manages to bring the foreign play under the theatrical home rule, for instance by adapting it to the demands and expectations of a target audience" (Van den Broeck 1993: 105-106). However, this does not mean that stage translations inevitably result in drama texts that observe home conventions and submit to the predominant aesthetic, cultural and theatrical norms--translators may also want to modify existing theatrical standards, but their chances of success depend on whether the target system is prepared to adopt foreign models.

Van den Broeck (1993: 240) stresses that his two concepts are purely theoretical and that in practice they are actually never fully manifested. He points out that page translations are problematic in that it is hard to imagine that a translator would make a translation of a drama text only meant for reading. However, as pointed out by Short (1998: 18), there are actually plays that were never meant to be performed, so this is not too far-fetched an idea. Snell-Hornby (1997: 195) maintains that a page translation may be adequate "for the armchair reader", but not for a performance on stage; in this case, "a new dramatic 'score' [is needed] for a performance that is coherent and acceptable within the target culture". Page translations are more literal than stage translations, which are "more rhythmical and hence more speakable ... giving the actors more scope for expression".

According to Delabastita and D'Hulst (1993), the high status of Shakespeare texts in particular has led many translation critics to compartmentalise Shakespeare translations into page and stage translations. Van den Broeck (1993: 105) for instance, states that when it comes to translating such highly canonical texts as Shakespeare's, "one should never forget that paying tribute to their 'literariness' is only one thing, whereas doing justice to their 'performability' is quite another thing". The first group of critics are those who feel themselves called upon to "defend" the Bard against disrespectful translators who dare to go beyond a mere linguistic transcoding to include cultural and theatrical transpositions as well. These are the critics who usually complain about "unfaithful", unpoetic and non-metrical translations. The second group of critics are those who reject every form of "slavish" source-oriented copying in favour of a more "creative" treatment of the source text which is geared to the taste and conventions of modern audiences. In their view, faithful translation merely results in stilted museum theatre.

Habicht (1993:241) criticises Van den Broeck's page-stage distinction for being "somewhat too timeless". He points out that what are considered today as the highly archaic Schlegel-Tieck translations of Shakespeare in Germany in the nineteenth century were then considered the most suitable way in which to present Shakespeare productions. This is also what happened in 1947 when the first Afrikaans translation of Hamlet was staged. According to a report in Die Burger, it was estimated that 20,000 people attended the 17 performances in Johannesburg alone and that surely is some kind of gauge of how well this page translation was received in South Africa then (Kruger 2000: 50). At that time, a stage translation of Hamlet would not have been tolerated at all!

Schmidt (in Delabastita & D'hulst 1993: 241) is of the opinion that a body of "faithful" translations of Shakespeare is very valuable in any literature because

directors then can still do what they like with them, and this is in fact what you see happening nowadays. They quite often stick to Shakespeare's words but it is the staging or the costumes, the decor and the gestures that produce the new interpretation.

(Schmidt in Delabastita & D'Hulst 1993: 241)

A Shakespeare page translation usually contains only minor additions or omissions above sentence level. It is generally faithful, literal and metrical because the translator's primary aim is to retain as much of the blank verse, word-play, imagery and style levels of the original as possible. Such translations are "written to be read as if heard"--the reader acts out the play in his imagination (House 1981: 172). Page translations are therefore not as suitable for performance as stage translations, which is not to say that such translations have not been, and sometimes still are, performed.

In contrast, a Shakespeare stage translation is a play that is translated exclusively for performance, that is, "written to be spoken as if not written" because the dialogue is to be spoken by actors during a performance. There is thus a subtle difference, albeit not too clear-cut. It is usually only available as the script from which the producer and the actors work and is not published. In many cases the translator receives a very specific brief from the initiator of the translation, the production company, as regards budgetary constraints and audience expectations. In order to save costs, the stage translator will sometimes cut minor characters out or use the same actors for more than one role and avoid intricate decor, props and period costumes by opting for a contemporary setting. The dialogue is shortened and scenes are cut or switched to avoid scene changes. Stage translations of Shakespeare plays that are prescribed as setworks for schools, in particular, are fairly free and idiomatic, with instances of linguistic and cultural adaptation of foreign elements to make the play more accessible to the target audience. This, according to reviews, is exactly what happened in the case of Potgieter's (1991) stage translation of Die Sakeman van Venesie:

Karakters van die hoogheilige Shakespeare war kougom kou, tekkies dra, rook, koerant lees en sonbrandolie aansmeer? O, Nee! O ja, in Ilse van Hemert se aangrypende Die Sakeman van Venesie vir Truk werk dit so. En dit werk soos 'n bom. Veral vir die skoolkinders wie se voorgeskrewe werk dit is.

(Hough 1991)

[Shakespeare characters who chew bubblegum, wear tackies, smoke, read the paper and put on suntan oil? Oh no! Yes, this is what happens in Ilse van Hemert's fascinating Die Sakeman van Venesie. And it works well--especially for the school pupils for whom this is a prescribed book.]

It seems therefore as if page and stage translations of Shakespeare plays are similar in many respects, but that their functions as acts of communication in generating theatrical meaning differ. We can therefore assume that they are different types or registers of translated drama and because they have different functions they will present information differently: the dialogue of a stage translation, in particular, might be expected to be more "speakable" than that of a page translation precisely because it comprises more spoken language features. In Kruger (2000) it was not my intention to examine whether dramatic dialogue is seen to mirror real-life conversation or not. It was my intention, however, to show that the dialogue of the stage translation sounds more like ordinary speech than that of the page translation.

In order to do so, the assumption was made that the "rules" underlying the orderly and meaningful exchange of speech in everyday contexts are the resource that dramatists use to construct dialogue in plays and that fabricated speech in plays is under no necessity to mimic some given original, except as a specific dramatic strategy. Even then, it is the illusion of real-life conversation that characterises consummate art (Herman 1995: 6). Simpson (1998: 41) echoes this view:

It is a truism to say that drama dialogue differs from everyday speech. Drama dialogue clearly is fabricated interaction between fictional characters, mediated and controlled by playwrights in the first instance, and, in the case of dramatic performances, by directors and actors in the second. This is not to say, however, that principles of social interaction ... cannot be brought to bear in the interpretation of dramatic dialogue. In fact, drama dialogue can only be accessed through its relationship to the social context outside the play-text.

(Simpson 1998: 41)

In what ways then does dramatic dialogue on the one hand differ from, and on the other resemble, real conversation?

2.3 Dramatic Dialogue and Real Conversation: Differences

The most obvious difference between the dialogue in a drama and naturally occurring conversation is that dramatic characters are simply not real people and that the channel of communication is more complex. Not only are we dealing with two layers of discourse, playwright [right arrow] audience/reader and character [right arrow] character, but also with a combination of the written and the spoken medium in one text (cf Elam: 1980: 135; Short 1996: 169ff.). I am primarily concerned with the second layer of discourse: the "character talk" that is embedded within the "overarching level of discourse ... between the playwright and the audience" (Short 1996:169).

A second major difference between dramatic dialogue and ordinary conversation is so-called normal non-fluency (Short 1996: 176). Normal non-fluency features are interactive communication management features (Allwood 1999) such as pauses, false starts, interruptions, digressions, overlaps, voiced fillers (of course, you know), repetitions and mispronunciations which are usually not noticed in conversation, unless they occur with unusual frequency. In contrast, normal non-fluency does not occur in dramatic dialogue, precisely because it is written (even though it is written to be spoken)--readers will not tolerate such features in written language; also, the time constraints within which a drama operates do not allow such features. If features that are normally associated with non-fluency do occur in dramatic dialogue, they are perceived by readers and audience "as having a meaningful function precisely because we know that the dramatist must have included them on purpose" (Short 1996: 177). The relative orderliness of dramatic dialogue accounts for its repeatability, something that is very difficult in ordinary conversation because of its spontaneous and fragmentary nature.

A third major difference between dramatic dialogue and ordinary conversation relates to what is called "feedback" (Short 1996: 179; Allwood 1999). According to McCarthy (1991: 127), in face-to-face conversation there are specific linguistic devices used in turn-taking for framing transactions that vary greatly in level of formality and appropriacy to different situations (e.g. If I may, Mr Chairman versus Shut up! I want to speak). "Back-channel responses" are available when listeners do not take the turn but simply want to make it clear to the speaker that they are attending to the message: non-verbally by for instance nodding the head, or verbally by means of words such as umm or yeah. Quirk et al. (1985: 444) divide such linguistic devices into two small classes, viz. "reaction signals" (e.g. no, yes, yeah, yep, m, hm, umm) and "initiators" (e.g. well, oh, ah, oh well, well then, why). In following Allwood (1995: 25), these devices are called feedback words, so as to show that they have four basic pragmatic functions in communication: contact, perception, understanding and reaction to the main evocative intention. Feedback also occurs in dramatic dialogue, and sometimes may be indicated in the script. But it definitely does not occur as regularly as in real conversation. Dramaturgical conventions sometimes have the listener stand motionless so that the audience can concentrate on the speaker and not get distracted by other characters nodding or giving feedback noises. Feedback words and other discourse markers that indicate participation between speaker and listener that were identified in the page and the stage translation receive attention below.

2.4 Dramatic Dialogue and Real Conversation: Similarities

As stated above, dramatic discourse does not mirror everyday conversation. As Herman (1998: 24) aptly puts it, dramatic speech is "tidied up speech" which "follows certain of the constitutive and regulative rules of extra-dramatic conversation" (Elam 1980:178). Some of the similarities discussed below are turn and floor management, speech acts, (im)politeness and implicature, and involvement.

The order of speech in a drama is organised to project the order of turns to be taken by the characters and is controlled by the dramatist (Herman 1998: 24). The one-speaker-speaks-at-a-time kind of floor and the turn-management strategies that construct it are the dominant mode of organising speech in drama. But within this overall mode, the use of turn-lapses, pauses, hesitations, gaps, interruptions and overlaps also occur. These elements are investigated in detail by Herman (1995: 76-163; cf also Herman 1998: 24; Bennison 1998), who shows that they assign meaning and function to what is said by a character. For instance, a character who is consistently interrupted and who is denied the opportunity to speak, can be interpreted as less powerful than another. Also, if a character fails to take the floor such an option can dramatise his ineffectuality. Herman (1998: 25) lists the following variables that can influence the dramatic situation: who speaks to whom; who is not spoken to; who listens or does not listen; whether listeners are responsive in turn, or not; whether those who respond are those targeted by the speaker or not; length of speeches; linguistic style and texture of a character's speech; how changeovers are effected; the uses of silences, either intra- or inter-turn. According to Herman:

Situation, event and character thus emerge, develop, in the "here and now" of speech as speech alternation is blocked or progresses in troubled or untroubled fashion.

(Herman 1998: 25)

Speech acts, like other acts, have effects on people and, in turn, make them do things. The same potential gap between what words mean and what they do is found in dramatic dialogue as in ordinary conversation (cf Short 1996: 193-204). Elam (1980: 181) says that dramatic dialogue is "illocutionarily purer", in other words, it is far better structured than "real-life" exchanges as it is essential to the development of the action. Lowe (1998: 140) says that the choice of different speech acts tells us something about the integrity of a character: for instance, commands and questions can indicate a relatively powerful character and tell us about relationships between characters. Short (1996: 204) points out that the speech-act value of an utterance might be different for different characters on stage, or for the characters on the one hand and the audience on the other, leading to dramatic irony.

According to Culpeper (1998: 83-84), politeness is "about the strategic manipulation of language, about expediting our conversational goals by saying what is socially appropriate" and therefore saving face. In contrast to the social harmony created by politeness, impoliteness is a face-threatening act. It is a type of aggression which, although socially outlawed, has been a source of entertainment for years. In drama (e.g. twentieth-century courtroom films), it generates the disharmony and conflict between characters which causes character development, and often moves the plot forward. Both speech acts and (im)politeness therefore serve very specific functions in dramatic dialogue.

As in real conversation, dramatic characters often say one thing but mean another. The audience immediately infers that a dramatic character flaunting a conversational rule, for example by lying, is doing so on purpose and that the dramatist is telling them something about the character and the relations between the characters (cf Short 1996: 240ff.; Cooper 1998).

Simpson (1998: 41) was quoted above as saying that dramatic dialogue is clearly "fabricated interaction between fictional characters". However, as I have shown, it draws on aspects of real speech and therefore shows certain similarities--it "represents face-to-face conversation and thus is also characterised by linguistic features of interaction and involvement" (Biber & Finegan 1992: 691). Following Chafe (1982), Biber (1988: 43) defines "involvement" as those linguistic features which reflect the fact that speaker and listener typically interact with one another in conversation, while writer and reader typically do not. Owing to this interaction, speakers often refer directly to the listener by means of, for instance, second-person pronouns and questions and they express their own thoughts and feelings by means of first-person pronouns, affective forms such as emphatics and amplifiers, and cognitive verbs such as think and feel. As a result, real speech often has a distinctly non-informational and imprecise character. One of the major features of normal non-fluency that assists in establishing interpersonal involvement in real conversation, but which is used only to a certain extent in dramatic dialogue, is repetition (cf Tannen 1987; 1992). In a drama text, although in a stylised manner, repetition in the dialogue will obviously also attempt to simulate the interaction of real-life conversation.

As the manifestation of features of involvement in drama translation formed part of the research question examined in Kruger (2000), the term "involved production features" was adopted from Biber (1988) for those linguistic features that signal involvement and interaction between speaker and addressee in the dialogue of drama texts and their translations. The concept of involved production is used in the domain of register variation studies. Following Biber, who uses the terms "register" and "register variation" as cover terms "for the full range of language varieties associated with differences in communicative situation" (Finegan & Biber 1994: 316), the page and stage translations of Shakespeare's The Merchant of Venice are also regarded as different registers of drama translation. Therefore, in order to analyse involvement in these translations, the necessary analytical tool was accessed from register variation studies. As will be explained below, register variation studies compare the linguistic characteristics of two or more registers, usually using quantitative techniques to isolate linguistic similarities and differences among registers. In this regard, Biber's (1988) multidimensional analysis of huge corpora of spoken and written texts was found extremely useful for analysing the texts in my corpus, that is, the original Shakespeare text and two translations in Afrikaans.

3 Framework and Procedures

The multidimensional approach to register variation, referred to here simply as the MD approach, was developed by Douglas Biber of Northern Arizona University (USA) in 1988. It was originally developed for comparative analyses of spoken and written registers in English (Biber 1988, 1995a, 1995b) and has subsequently been refined and used for cross-linguistic analyses of universal tendencies of register variation in non-western languages (cf Biber & Hared 1992; also Biber & Finegan 1994). To the best of my knowledge it has not yet been applied to translated texts. Methodologically, the approach uses computer-readable text corpora as well as computational tools to identify linguistic features in texts, and multivariate statistical techniques to analyse the co-occurrence relations among linguistic features, thereby identifying underlying dimensions of variation in a language. The primary aim of the MD approach is to provide comprehensive descriptions of patterns of register variation, that is, identifying underlying linguistic dimensions of variation and specifying linguistic similarities and differences among registers with respect to those dimensions (Biber 1995a: 19-20).

Biber (1988) found several functional differences between typical speech and writing which are associated with the typical situational characteristics of the two modes: for instance, some linguistic features are used to elaborate information in typical writing, while others are used to mark interaction or to express personal feelings in typical speech. However, no absolute distinctions were found between speech and writing. In terms of its situational characteristics, Biber (1988: 37) found that typical speech is interactive and involved, and depends on shared space, time and background knowledge; typical writing has the opposite characteristics. In terms of its linguistic characteristics, typical speech is structurally simple, fragmented, concrete and depends on exophoric reference; again, typical writing has the opposite characteristics. In line with an earlier study by Chafe (1982), Biber (1988: 21, 43) also found that linguistic variation must be analysed in terms of sets of co-occurring dimensions because they work together to mark some common underlying function. This is why he calls his approach "multi-dimensional". Each dimension comprises an independent group of co-occurring linguistic features, and each co-occurrence pattern can be interpreted in functional terms such as "involved", "informational" and so on.

As my corpus was not tagged, the MD approach could not be followed in all respects. For my purposes, I also did not need to analyse every one of the 67 linguistic features of the MD approach and therefore selected only certain linguistic features from some of the dimensions that comprise the MD approach (Kruger 2000). I was mainly interested in linguistic features typically found in spoken language, so most of the features analysed were taken from Dimension 1, labelled "Informational versus Involved Production" and defined by Biber (1988: 115) as "discourse with highly informational purposes which is carefully crafted and highly edited" as opposed to "discourse with interactional, affective, involved purposes, associated with real-time production and comprehension constraints". In a diachronic study of two speech-like genres written for spoken delivery (i.e., dialogue taken from plays and from fiction), Biber and Finegan (1992: 689) themselves used only three dimensions of the MD approach.

In Kruger (2000) the MD approach was subsequently adapted as follows to allow for the identification and analysis of linguistic features signalling involvement in drama texts. From Dimension 1 (Involved vs. Informational Production) the following features were chosen: private verbs, contractions, second-person pronouns, analytic negation, demonstratives, emphatics, first-person pronouns, causative subordination, discourse particles, amplifiers and questions. These features were chosen because they can be characterised as verbal, interactional, affective, fragmented, reduced in form and generalised in content (Biber 1988:104)--exactly the kind of features found in dramatic dialogue. Type-token ratio was also chosen from this dimension to allow for the measurement of lexical variation or diversity in the corpus. From Dimension 3 (Elaborated vs. Situation-dependent Reference) time- and place-adverbials were chosen as the dialogue in drama texts in particular features aspects of the here and now of a particular dramatic situation.

The first of the research procedures involved preparing the source text (The Merchant of Venice) and the translations by Malherbe (1947) and Potgieter (1991) for automatic and semi-automatic analysis by means of WordSmith Tools, a Windows-based suite of text retrieval programs used for the lexical analysis of texts. It was developed by Mike Scott of Liverpool University, UK (Scott 1999). The names of the actors as well as the stage directions were manually deleted from each text so that only the actual dialogue remained. Thereafter, by means of the WordList Tool, alphabetically arranged word lists (and corresponding frequencies) for each of the texts were generated. The list of items of each linguistic feature supplied by Biber (1988: 221-245) was checked against the lists in Quirk et al. (1985)--the source to which Biber himself refers. This procedure assisted me in identifying those items that appear in The Merchant of Venice and in recording their frequencies. Translation equivalents and synonyms appearing in similar contexts in the Afrikaans translations were subsequently identified manually from the frequency lists and their frequencies were also recorded. The recorded number of frequencies for each feature (e.g. private verbs) was added to obtain total counts of a particular linguistic feature per text.

Raw frequency counts cannot be used for comparison across texts because texts differ in length and such counts will give an inaccurate assessment of the frequency distribution of features. That is, long texts will tend to have higher frequencies simply because there is more opportunity for a linguistic feature to occur; in these cases, the higher frequency count does not indicate a more frequent use of the linguistic feature. Comparing the frequency per 100 or 1000 words eliminates this bias (Biber 1988: 14; 75-76; Biber, Conrad & Reppen 1998: 263-266). All frequency scores were therefore "normed" or converted to densities, that is, to a basis per 1000 words. (2)

3.1 Analysis of Discourse Markers in the Corpus

According to Biber (1988: 241), discourse particles are used to maintain "conversational coherence" and to monitor the information flow in involved discourse. I find Biber's (1988) view of discourse particles restrictive because I take both the grammatical and the pragmatic function of different markers or signals that link discourse into account, and therefore, in agreement with Carstens (1997: 304-308) and Schiffrin (1987), who examines oh, well, and, but, or, so, because, now, then, as discourse markers, have chosen to use the term "discourse markers" as an umbrella term. According to McCarthy and Carter (1994: 68, 85, 206, 207), discourse markers are generalised "interactive markers" used to signal new segments of information in conversation--and certainly also in drama dialogue. These markers have "important interpersonal and text-building functions" by taking, keeping and yielding the turn through speech actions, by empathising or communicating with the listener; and by structuring the message (cf also Carstens 1997:304-308 as regards pragmatic functions and semantic classification of discourse markers). Of the discourse markers listed by Biber (1988:241), (3) only well and now occur in The Merchant of Venice, and it was therefore necessary to identify discourse markers that occur in The Merchant of Venice and the two Afrikaans translations, to categorise these as (i) feedback words, (ii) exclamations/interjections/ formulae, (iii) vocatives and (iv) courtesy adjuncts and to record them (cf Appendix A). These categories are explored in more detail below.

Feedback Words

Feedback words have four basic pragmatic functions in communication: contact, perception, understanding and reaction to the main evocative intention. Allwood (1999) maintains that different forms of yes function to signal acknowledgement, agreement, affirmation and acceptance in a conversation, and therefore all such occurrences in The Merchant of Venice and the translations were recorded. Similarly, all occurrences of nay as response signals were also recorded. McCarthy and Carter (1994: 85) state that forms such as O, yea, no, and so are "not answers to polar questions but serve more as discourse markers". Concordances of so were obtained, and only instances of so as feedback words in the translations were recorded, as there were no such occurrences in the source text. The following feedback words were identified in The Merchant of Venice and the Afrikaans translations by means of frequency and concordance lists (cf Appendix A):

   Well [Wel], O [O], Ay, Yea, Yes [Ja, Ja-wel, Nou-ja, Goed, Top],
   Now, [Nou], Why, Nay, No [Nee], [So]

Exclamations/Interjections/Formulae

Exclamations, interjections and formulae all add to the emotive value of an utterance (also in dramatic dialogue). Through these words a speaker reveals emotions such as pain, anger, surprise, disbelief, embarrassment, fear, refusal, despair, contempt, appeal, amazement, anxiety and sympathy. They are therefore included in this category (cf Quirk et al. 1985: 833-834, 852-853; Van Schoor 1983: 60, 329). Exclamations, interjections and formulae such as the following occur in The Merchant of Venice and the translations and were recorded (cf Appendix A):

   What! [Wat!], How! [Hoe!], Alas, Fie, fie! [Nee! Foei!], Good morrow
   [Goeie more], Fare ye well [Totsiens/ Vaarwel]

Vocatives

According to Quirk et el. (1985: 773), a vocative is an optional element, usually a noun phrase. One obvious function of a vocative is to seek the attention of the addressee, and especially to single him out from others who may be within hearing. A second function, less obvious but certainly no less important, is to express the attitude of the speaker towards the addressee. Vocatives are generally used as a positive mark of attitude, to signal either respectful distance or familiarity (varying from mild friendliness to intimacy) and may take different forms. Combrink (1982), Kotze (1982), Wybenga (1982) and Ponelis (1979:36-40) investigate different types of vocatives, their uses in Afrikaans and how such forms are influenced by the participants of the communicative situation and their relationship, the context and different aspects of the situation, as well as the topic. Following these authors, the following types of vocatives were identified in The Merchant of Venice and the translations. Concordances of the individual items were obtained so as to determine whether such an item was indeed used as a vocative. For example, of the 49 entries for Antonio obtained by means of WordList, only 15 are vocatives in the source text; of the 48 entries in the page translation only 14 are vocatives; and of the 44 entries in the stage translation only 15 are vocatives.

These were subsequently recorded (cf Appendix A).

* Names: Antonio, Bassanio, Balthazar, Gobbo [Gopse], Gratiano, Jessica, Nerissa, Portia, Launcelot [Langeraad], Leonardo, Lorenzo, Shylock, Stephano, Tubal 

* Standard appellatives: Terms for family relationships: father [vader, pal Titles of respect: (your) Grace/Honour [Hoogedele, (U) Hoogheid/Edele]; (my) Lord [(my) heer, meneer]; wise/noble/upright judge [wyse/edele/regverdige regter] Markers of status: doctor [dokter]; sir [meneer/signior]; madam [mevrou/ juffrou/signiora/ signiorina] Terms for occupations: goaler [sipier/bewaarder/konstabel] Epithets expressing an evaluation: Terms of endearment: sweet [harteliefie/vroutjie/vrou/liefste/liefling] Impolite/offensive terms: Jew, infidel, old man [Jood/heiden/ongelowige/ou man]

* Any of the above may be expanded by the addition of modifiers or appositive elements of various kinds, for example my friend Stephano, lord Bassanio, Signior Antonio, learned judge, most rightful judge, upright judge, good sir, young man, sweet doctor, sweet lady, fair ladies. The vocatives were only counted once, even in cases where a particular vocative comprises more than one lexical item, for instance, of the 40 entries for Bassanio in the source text, it occurs 7 times as a vocative, good Bassanio occurs once, (my) lord Bassanio occurs 5 times, sweet Bassanio occurs twice and Signior Bassanio occurs twice, and so a total of 17 vocatives was recorded in the source text (cf Appendix A). A similar procedure was followed as regards the translations.

Courtesy Adjuncts

Courtesy adjuncts have an appellative function. They are used when a speaker wants to influence an addressee and are generally used to convince an addressee to accept a certain viewpoint. Please is a courtesy adjunct that "is generally confined to imperatives or to sentences constituting a request or containing a reported one" (Quirk et al. 1985: 571). All occurrences of please in The Merchant of Venice are verbs; however, in the translations the Afrikaans courtesy marker, asseblief occurs quite frequently and that is why I recorded all instances of asseblief (cf Appendix A).

Let us now turn to the findings yielded by the research.

3.2 Findings and Interpretation

The stage translator seems to have consciously created opportunities for interpersonal interaction and involvement between interlocutors by means of discourse markers, which made taking, keeping and yielding tunas possible through speech actions, by empathising or communicating with the listener, and by structuring the dialogue. The results in Appendix A show a density count of 30,2/1000 for the stage translation, 24,3/1000 for the page translation and 26,1/1000 for the original. These density counts are all much higher than Biber's (1988: 265) highest finding for discourse particles--6,6/1000 for telephone conversations--but it is important to mention that I identified and counted a far greater range of discourse markers than he did.

For example, Appendix A shows that the stage translator inserted far more feedback words than the page translator in an attempt to make the dialogue sound more like real speech:

* Ja wel, daar's nou vir jou 'n jong vulletjie. (Yes well, there's a young foal.)

* Nou ja, is Antonio hier? (Now then, is Antonio here?)

* Goed, bieg dan. (Good, then confess.)

* Nee, ons sluip weg terwyl die ete aan die gang is. (No, we'll slip off during dinner.)

* So, totsiens, tot ons weer ontmoet. (So goodbye, till we meet again.)

One interjection that occurs consistently in the stage translation and which is more or less absent from the page translation is Toe!, meaning more or less "come on", "please", "all right", a typical feature of spoken Afrikaans used to encourage someone else to say or do something, for example Toe, praat (Come on, speak); Toe, pa, kom (Come on, dad); Toe, weg is jy (Come on, be gone; Toe, menere (Come on, gentlemen); Toe, gaan jy vooruit, meneertjie (Come on young man, you go ahead.)

The page translator in general used more formal, but dated, exclamations and interjections such as Mag God my bewaar van hulle (May God protect me from them) which fit in with his attempt to raise the status of Afrikaans as a literary language back in 1949 (Kruger 1996). In contrast, the stage translator used ordinary, everyday expressions such as Liewe hemel, skroef op jou kop (Good heavens, screw on your head; i.e., come to your senses) and Deksels, maar ek het haar lief (Gee, I really love her). The hmm at the end of the following line will be spoken with raised intonation, eliciting an answer, making the dialogue more interactive between the characters: Wat se daai gek van Hagar se gebroedsel, hmm? (What does that fool say about Hagar's brood, hmm?)

Typically, vocatives in dramatic dialogue raise the level of personal involvement between interlocutors. Potgieter (1991) inserted extra vocatives in the stage translation to this effect as regards the following characters: Antonio, Bassanio, Nerissa and Portia. In the stage translation there are double the number of occurrences where Portia functions as a vocative compared to the source text and the page translation (cf Appendix A), e.g.:

(1)   ST         BASSANIO   "Madam, you have bereft me of all words"
                            (II.ii.176).
      Page tr    BASSANIO   "O jonkvrou, jy het my die spraak ontneem."
                            (Oh young lady, you have taken away my
                            words.)
      Stage tr   BASSANIO   "Portia, jy laat my sonder woorde."
                            (Portia, you leave me without words.)

The stage translator obviously attempted to retain the Italian flavour of the original play by consistently using signior, signiora and signiorina as forms of address in his translation. The page translator does not use this term of address at all, and the word signior(s) only occurs three times in the source text itself(cf Appendix A). By using the Italian loan words, Potgieter has not only 2avoided the more archaic forms of address used for male characters (kerel(s)/jongeheer/jonge = chap/young man/boy) that occur in the page translation, but again attempts to remind the audience of the Venetian setting in which the play takes place. Also, by using the Italian terms of address (signiora, signiorina) and more ordinary terms of endearment for the female characters in the play (liefling = my love), he has avoided the more "poetic" and "literary" terms such as my skone (my beauty), skoonste heldin (beautiful heroine) and edele vrou/jonkvrou (honourable lady/young lady = madam) which appear in the page translation. In contrast to the formal term of address, vader (father), used in the page translation, Potgieter (1991) uses the informal pa (dad), typical of spoken Afrikaans.

Most importantly, by using the term sakeman (businessman) as opposed to the more archaic koopman (merchant) in the stage translation, he has managed to give the play a thoroughly modern ring which is absent from the older translation. Even though "merchant" can be literally translated as both koopman and sakeman in Afrikaans, these two terms differ in evoked meaning. Sakeman in modern Afrikaans means "businessman"; whereas koopman, a loan word from Dutch, is archaic. (4)

The frequent occurrence of the courtesy adjunct asseblief(please), cleverly combined with a vocative, is a prime example of how the translator attempted to influence the visual by means of the verbal. This combination makes a request such as the following more personal and more involved. One can only imagine that the actor playing Antonio will use facial expressions, gestures and body language to bring the verbal message across to Bassanio that he should carefully reconsider before getting into an argument with Shylock:

(2)   ST         ANTONIO   "I pray you, think, you question with the
                           Jew" (IV.i.71).
      Page tr    ANTONIO   "Onthou tog dat jy met die Jood staan twis."
                           (Remember that you are quarelling with the
                           Jew.)
      Stage tr   ANTONIO   "Asseblief, Bassanio, jy argumenteer met die
                           Jood." (Please Bassanio, you are arguing
                           with the Jew.)

In Kruger (2000) the second statistically highly significant finding concerned discourse markers (cf table in Note 1). The high density count and wide range of discourse markers displayed by the stage translation (Appendix A) revealed that this is an easy and efficient way to signal involvement and interaction in dramatic dialogue. The insertion of ordinary, everyday expressions ensured that the characters speaking on stage will sound more authentic. The insertion of vocatives in combination with the courtesy adjunct asseblief pointed up informal relations in the text and affective stance. They also assisted in maintaining textual coherence where omissions could cause incoherence. The different forms of address used by the different translators reveal why the page translation is branded as "museum theatre"--we just do not address a young man as jongeheer or a young lady as juffer any more.

4 In Conclusion

It was the aim of this article to examine the role that discourse markers play in an Afrikaans stage translation of The Merchant of Venice. As stated above, concepts such as "performability" and "speakability" are vague and not easy to define. What does emerge, though, is that stage translations, in particular, seem to require dialogue that is somehow more "speakable" than that of page translations. And in order to be more "speakable" the stage translator of the Afrikaans translation of The Merchant of Venice has ingeniously applied a variety of discourse markers to bridge the gap between real conversation and the fabricated, "tidied up speech" (Herman 1998: 24) usually found in dramatic dialogue. The feedback words, exclamations and interjections all have a thoroughly modern ring and the vocatives, combined with courtesy adjuncts, present the Shakespeare characters in this play as more real and more human--in fact, more like us!

APPENDIX A

FREQUENCY DISTRIBUTION OF DISCOURSE MARKERS
BETWEEN SOURCE TEXT AND TRANSLATIONS

English Source Text
No. of Tokens: 20 908

(i) Feedback Words

Well (then)             32
O (heavens/hell)        37
Ay                      14
Yea                      3
Yes                      8
Why                      1
Now/(How) now            8
Nay                     12
No                      17
--

Subtotal               132

(ii) Exclamations/ Interjections/ Formulae

What(!)                  3
How!                     5
Alas                     4
Fie                      6
Ha                       6
                         4
Ho                       8
Sola                     1
Wo
Alack (the day)          3
By my hood               1
Lord                     2
Gramercy                 1
O rare fortune!          1
God bless ...           10
Lord worshipp'd
might he be              1
Good morrow              1
Fare ... well            2
Adieu                    6
Fare ... well           14
Rest you fair            1
First                    5
To be brief              2
Lastly                   1

Subtotal                88

(iii) Vocatives

(Signior) Antonio       15
(my lord/good/
Signior/sweet)
Bassanio                17
Balthazar                1
(gentle) Gratiano        9
Gobbo                    6
Jessica                 14
(good/honest)
Launcelot                9
Leonardo                 1
(good) Lorenzo           8
Morocco                  1
Nerissa                 12
(sweet) Portia           5
(good/old)
Shylock                 11
Stephano                 1
(good) Tubal             7
(wise young/
upright) judge           9
Jew                     13
(your) Grace             7
(your) honour            1
(my) Lord(s)            19
Sir                     38
signior(s)               3
gentleman                2
gentlemen                3
madam                   15
lady                    16
ladies                  i
mistress                 2
(good) sweet             2
(sweet) soul             1
love                     5
doctor                   2
(young/old) man         10
(good)youth              1
goaler                   5
merchant                 1
friends                  1
infidel                  1
old carrion              1
father                  10
Clerk                    1

Subtotal               287

(iv) Courtesy Adjunct

(I) pray (you)          38

Subtotal                38

TOTAL                  545

Density/1000         26.10

Malherbe (1949)
No. of Tokens: 20 688

(i) Feedback Words

Wel                    27
O (aarde)              23
Ja (wel)/(Nou) ja      27
Goed                    5
Top                     1
Nou (goed/mooi)         8
(Kyk/kom) nou           9
Nee                    22
--

Subtotal              122

(ii) Exclamations/ Interjections/ Formulae

Wat,/Wat?             711
Hoe,
Toe,
Ai                      2
Ag                      6
Foei                    3
Ha                      4
Hei!                    2
Ho                      2
Hoha                    4
Woha                    1
Helaas                  6
By my kool              1
By my siel              2
Liewe tyd               1
(Liefste) hemel         5
Wrintig                 1
Wonderlike geluk        1
Mag God ...             9
Aarde                   2
Goeiendag/more          1
Totsiens                1
Vaarwel                10
Tot weersiens          10
Alle heil vir jou       1
Eerste dan              1
Om dit kort to se       3
Kortom                  2
Dank/Dankie             4

Subtotal               94

(iii) Vocatives

(meneer/my goeie)      14
Antonio
(my goeie/heer/
meneer) Bassanio       15
Balthazar               1
(my goeie)
Gratiano                8
Gobbo                   6
Jessica                15
(eerlike/vriend)
Launcelot               8
--
Lorenzo                 9
Morokko                 1
Nerissa                12
(liewe/beminlike)
Portia                  5
(goeie/ou) Shylock     11
Stephano                1
(goeie) Tubal           7
(wyse/jeugdige)
regter                  9
Jood                   13
(soetste) Jodin         1
(U) Edele              12
Hoogedele               5
meneer                 34
(my) heer/here         22
mevrou                  7
juffer                  2
juffrou                 4
(edele) koningin        1
(edele) jonkvrou        5
(edel/e) vrou           3
dames                   1
liefste                 6
(skoonste) heldin       1
(my) skone              4
doktor                  1
man                     2
kerel/kerels            9
jongeheer               1
jongman                 2
(my) jonge              1
sipier                  3
koopman                 1
vriend/vriende          6
ongelowige              1
vader                  11
Klerk                   1

Subtotal              282

(iv) Courtesy Adjunct

asseblief               5

Subtotal                5

TOTAL                 503

Density/1000         24.3

Potgieter (1991)
No. of Tokens: 18 273

(i) Feedback Words

Wel                    17
O (hemel/
afskuwelik)            24
Ja (wel)/(Nou) ja      42
Goed                   12
Nou (goed)              6
(En/kyk) nou            5
Absoluut                1
Nee                    26
So                      4

Subtotal              137

(ii) Exclamations/ Interjections/ Formulae

Wat,/Wat?               7
Toe,                   22
A/Aa/Ai                 3
Ag (tog)                9
Ha                      4
Haai                    2
Hallo                   1
Ho                      2
Hola                    8
Wooo                    1
Hmm                     1
Sjjjj                   1
Verbeel jou             1
So by my kool           1
So by my eer            1
Liewe Vader             1
Genade                  1
(Liewe) hemel           5
Deksels                 1
Die Here seen ...       5
Prys die Heer           1
God rus .../            4
Om Godswil              1
Aarde                   1
(Goeie) dag             2
Totsiens               13
Vaarwel                 8
Dit gaan u goed         1
Nommer een              1
En volgende             1
Dankie (tog)            1
Kortom                  1
Dankie                  3

Subtotal              116

(iii) Vocatives

(Signior) Antonio      15

(liewe/Signior)
Bassanio               17
Balthazar               1
(my liefste)
Gratiano                8
Gobbo                   6
Jessica                12
(liewe/ordentlike)
Launcelot               8
Leonardo                1
Lorenzo                 9
Marokko                 1
Nerissa                13
(liewe) Portia         12
(liewe) Shylock        11
Stephano                1
(liewe) Tubal           7
(wyse jong/
geleerde) regter        9
Jood                   13
(U) Edele               7
(U) Hoogheid            2
Meneer/menere          29
meneertjie              1
Signior/-tjie          24
(my) heer               5
signiora               15
signiorina              1
juffrou                 2
(liewe) siel            2
liefling                1
liefste                 6
soetste                 2
(beeldskone) heldin     1
doktor                  2
man                     6
(signior) jongman       1
meester                 2
konstabel               3
sakeman                 1
vriende                 1
heiden                  1
ou aasvleis             1
vader                   2
pa                      9
omie                    3
Klerk                   1

Subtotal              275

(iv) Courtesy Adjunct

asseblief              24

Subtotal               24

TOTAL                 552

Density/1000         30.2

Notes

(1.) Frequency distribution of Involved Production features between the source text (ST) and two drama translation registers (cf Kruger 2000--the entries for discourse markers have been revised).

                   Source Text     Page Translation
 Linguistic       20 908 words       20 688 words
  Features

                Actual   Density   Actual   Density
                Scores    /1000    Scores    /1000

Contractions       302      14.4       69       3.3
Discourse          545      26.1      503      24.3
Markers
First-person      1496      71.6     1357      65.6
Pronouns
Analytic           220      10.5      205       9.9
Negation
Emphatics          274      13.1      356      17.2
Second-            873      41.8      753      36.4
person
Pronouns
Private Verbs      472      22.5      491      23.7
Demonstra-         152       7.3       88       4.3
tives
Causative            6       0.3       72       3.5
Subordination
Questions          206       9.9      203       9.8
Time- &            272      13.2      380      18.4
Place-
Adverbials
Amplifiers         114       5.4       80       3.9
Totals            4932     235.8     4557     220.3

                Stage Translation
 Linguistic        18 273 words
  Features                              Com-
                                       puted
                Actual   Density    [chi square]
                Scores    /1000       Results

Contractions       151       8.3      41.97 **
Discourse          552      30.9      15.64 **
Markers
First-person      1349      73.8      10.17 **
Pronouns
Analytic           234      12.8       7.31 **
Negation
Emphatics          257      14.1       6.19 **
Second-            743      40.7       4.78 *
person
Pronouns
Private Verbs      489      26.8       3.63 *
Demonstra-          98       5.4       2.51
tives
Causative           51       2.8       1.46
Subordination
Questions          198      10.8       1
Time- &            350      19.2       3.3
Place-
Adverbials
Amplifiers          76       4.2       2.1
Totals            4548     249.8      46.39 **

As these results show, the overall difference in frequency of involved production features between a page and a stage translation is statistically highly significant ** (p [less than or equal to] .01) and the null hypothesis (i.e., There is no difference in the density of involved production features between a page and a stage translation) was therefore rejected. This finding lends support to the fact that translation register, that is, page or stage translation (the independent variables), does influence the use of linguistic features of involvement in drama translation.

(2.) The raw frequency count is divided by the number of words in the text and then multiplied by 1000 as follows: 552 discourse markers in Stage Translation/ 18273 x 1,000 = 30,2 discourse markers per 1,000 words.

(3.) Biber (1988: 241) lists the following "discourse particles": well, now, anyway, anyhow, anyways.

(4.) This fact is borne out by the relatively high frequency with which sakeman has occurred in the texts comprising Pharos Dictionaries' main corpus, the text archives, as well as in issues of Die Burger (a provincial newspaper) between 1988 and 1998, compared to the occurrences of koopman (Luther 1998):

            Pharos: Main Corpus   Pharos: Text Archives   Pharos: Die
             (3 599 765 words)     (12 996 939 words)        Burger
                                                          (1988-1998)

sakeman       55   0,002%            131   0,001%              7157
sakemanne     33                      67                       3524
sakemense      7                      20                        342
sakelui       31                      64                       8514

Total        126                     282                      19537

koopman        8   0,0002%            16   0,0001%             2945
koopmanne      1                       8                          7
koopmans       2                       4                        176
kooplui        0                       1                         13

Total         11                      29                       3141

References

Aaltonen, S. 1996 Acculturation of the Other: Irish Milieux in Finnish Drama Translation. Joenssuu, Finland: Joensuu University Press.

2000 Time-sharing on Stage: Drama Translation in Theatre and Society. Clevedon: Multilingual Matters.

Accone, D. 1991 An Absorbing, Playful "Sakeman". The Star Tonight, 20 June.

Allwood, J. 1995 An Activity-based Approach to Pragmatics. Gothenburg Papers in Theo retical Linguistics. Department of Linguistics, Gothenburg University.

1999 Activity-based Analysis of Meaning and Interaction. Paper read at the Annual Conference of the Linguistic Society of Southern Africa ("Putting Linguistics into Practice"), 5-7 July, held in Pretoria.

Bassnett, S. 1990 Translating for the Theatre: Textual Complexities. Essays in Poetics 15(1): 71-83.

1991 Translating for the Theatre: The Case Against Performability. TTR: Traduction, Terminologie, Redaction: Studies in the Text and its Transformations (Special Issue: Languages and Cultures in Translation Theories) 4(1): 99-111.

1998 Still Trapped in the Labyrinth: Further Reflections on Translation and Theatre. In: Bassnett, S. & Lefevere, A. (eds) Constructing Cultures: Essays on Literary Translation. Clevedon: Multilingual Matters, pp. 90-108.

Bassnett-McGuire, S. 1978 Translating Spatial Poetry: An Examination of Theatre Texts in Performance. In: Holmes, J., Lambert, J. & Van den Broeck, R. Literature and Translation: New Perspectives in Literary Studies. Leuven: Acco, pp. 161-176.

1980 An Introduction to Theatre Semiotics. Theatre Quarterly 10(38): 47-53.

1985 Ways Through the Labyrinth: Strategies and Methods for Translating Theatre Texts. In: Hermans, T. (ed.) The Manipulation of Literature: Studies in Literary Translation. London: Croom Helm, pp. 87-102.

Bennison, N. 1998 Accessing Character Through Conversation: Tom Stoppard's Professional Foul. In: Culpeper, J., Short, M. & Verdonk, P. (eds) Exploring the Language of Drama. London: Routledge, pp. 67-82.

Biber, D. 1988 Variation Across Speech and Writing. Cambridge: Cambridge University Press.

1995a Dimensions of Register Variation. Cambridge: Cambridge University Press.

1995b On the Role of Computational, Statistical, and Interpretive Techniques in Multi-dimensional Analyses of Register Variation: A Reply to Watson. Text 15(3): 341-370.

Biber, D., Conrad, S. & Reppen, R. 1998 Corpus Linguistics: Investigating Language Structure and Use. Cambridge: Cambridge University Press.

Biber, D. & Finegan, E. 1992 The Linguistic Evolution of Five Written and Speech-based English Genres from the 17th to the 20th Centuries. In: Rissanen, M., Ihalainen, O., Nevalainen, T. & Taavitsainen, I. (eds) History of Englishes: New Methods and Interpretations in Historical Linguistics. New York: Mouton de Gruyter, pp. 688-704.

1994 Sociolinguistic Perspectives on Register. Oxford: Oxford University Press.

Biber, D. & Hared, M. 1992 Dimensions in Register Variation in Somali. Language Variation and Change 4: 41-75.

Brown, J. R. (ed.) 1985 The Merchant of Venice (New Arden Shakespeare.) London: Methuen.

Carstens, W.A.M. 1989 Norme vir Afrikaans: Enkele riglyne by die gebruik van Afrikaans. Pretoria: Academica.

1997 Afrikaanse tekslinguistiek: 'n Inleiding. Pretoria: Van Schaik.

Chafe, W.L. 1982 Integration and Involvement in Speaking, Writing, and Oral Literature. In: Tannen, D. (ed.) Spoken and Written Language: Exploring Orality and Literacy. New Jersey: ABLEX Publishing Corporation, pp. 35-54.

Cooper, M.M. 1998 Implicature, Convention and The Taming of the Shrew. In: Culpeper, J., Short, M. & Verdonk, P. (eds) Exploring the Language of Drama. London: Routledge, pp. 54-66.

Combrink, J.G.H. 1982 Noem hom op sy naam: Die nut van die vokatief. In: Van Jaarsveld, G.J. (red.) Wat se jy?: Studies oor taalhandelinge in Afrikaans. Johannesburg: McGraw-Hill, pp. 1-24.

Culpeper, J. 1998 (Im)politeness in Dramatic Dialogue. In: Culpeper, J., Short, M. & Verdonk, P. (eds) Exploring the Language of Drama. London: Routledge, pp. 83-95.

Delabastita, D. & D'Hulst, L. 1993 European Shakespeares: Translating Shakespeare in the Romantic Age. Amsterdam: John Benjamins.

Elam, K. 1980 The Semiotics of Theatre and Drama. London: Methuen.

Espasa, E. 2000 Performability in Translation: Speakability? Playability? Or just Saleability? In: Upton, C-A. (ed.) Moving Target: Theatre Translation and Cultural Relocation. Manchester: St Jerome, pp. 49-62.

Finegan, E. & Biber, D. 1994 Register and Social Dialect Variation: An Integrated Approach. In: Biber, D. & Finegan, E. (eds) Sociolinguistic Perspectives on Register. Oxford: Oxford University Press, pp. 315-349.

Habicht, W. 1993 The Romanticism of the Schlegel-Tieck Shakespeare and the History of the Nineteenth-century German Shakespeare Translation. In: Delabastita, D. & D'Hulst, L. European Shakespeares: Translating Shakespeare in the Romantic Age. Amsterdam: John Benjamins, pp. 45-54.

Herman, V. 1995 Dramatic Dialogue: Dialogue as Interaction in Plays. London: Routledge.

1998 Turn Management in Drama. In: Culpeper, J., Short, M. & Verdonk, P. (eds) Exploring the Language of Drama. London: Routledge, pp. 19-33.

Hough, B. 1991 Shakespeare in stoombad: Sakeman-karakters faks tot liefdesbriewe. Rapport, 26 Mei.

House, J. 1981 A Model for Translation Quality Assessment. Tubingen: Gunter Narr.

Keuris, M. 1996 The Play: A Manual. Pretoria: Van Schaik.

Kotze, E.F. 1982 Djy kan nie vir my djy nie, djong! In: Van Jaarsveld, G.J. (red.) Wat se jy?: Studies oor taalhandelinge in Afrikaans. Johannesburg: McGraw-Hill, pp. 151-60.

Kruger, A. 1996 "Bless Thee, Bottom, Bless Thee! Thou art Translated!" The Shakespeare Phase in South Africa: A Socio-cultural Perspective. Journal of Literary Studies 12(4): 408-428.

1998 Shakespeare Translations in South Africa: A History. In: Beylard-Ozeroff, A., Kralova, J. & Moser-Mercer, B. (eds) Translators' Strategies and Creativity. Amsterdam: John Benjamins, pp. 107-115.

2000 Lexical Cohesion and Register Variation in Translation: The Merchant of Venice in Afrikaans. Urdversity of South Africa, Pretoria.

2004 Shakespeare in Afrikaans: A Corpus-based Study of Involvement in Different Registers of Drama Translation. Language Matters 35(1): 282-301.

Link, F.H. 1980 Translation, Adaptation and Interpretation of Dramatic Texts. In: Zuber, O. The Languages of Theatre: Problems in the Translation and Transposition of Drama. Oxford: Pergamon, pp. 24-50.

Lowe, V. 1998 "Unhappy" Confessions in The Crucible: A Pragmatic Explanation. In: Culpeper, J., Short, M. & Verdonk, P. (eds) Exploring the Language of Drama. London: Routledge, pp. 128-141.

Luther, Jana 1998 Personal Communication and Correspondence. Senior Editor (Afrikaans) of Pharos Dictionaries, a Division of Nasionale Pers-Boekhandel in Cape Town.

McCarthy, M. 1991 Discourse Analysis for Language Teachers. Cambridge: Cambridge University Press.

McCarthy, M. & Carter, R. 1994 Language as Discourse: Perspectives for Language Teaching. New York: Longman.

Malherbe, D.F. 1949 Die Koopman van Venesie. Johannesburg: Afrikaanse Pers-Boekhandel.

Mouton, M. 1988 Dramateorie vandag: Die bydrae van die drama- en teatersemiotiek. Potchefstroom: PU vir CHO, Departement Sentrale Publikasies.

Pavis, P. 1992 Languages of the Stage: Essays in the Semiology of the Theatre. New York: Performing Arts Journal Publications.

Pohl, A.S. 1969 Die koopman van Venesie. Johannesburg: DALRO.

Ponelis, F.A. 1979 Afrikaanse sintaksis. Pretoria: Van Schaik.

Potgieter, T. 1991 Die sakeman van Venesie. Unpublished manuscript.

Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. 1985 A Comprehensive Grammar of the English Language. London: Longman.

Ronge, B. 1990/91 Yippee for Ilse's Venesie yuppies. Shakespeare in Southern Africa 4: 94-96.

Schiffrin, D. 1994 Approaches to Discourse. Oxford: Blackwell.

Schultze, B. 1990 In Search of a Theory of Drama Translation: Problems of Translating Literature (Reading) and Theatre (Implied Performance). In: Associacao Portuguese de Literatura Comparada: Os Estudos Literarios: (Entre) Ciencia e Hermeneutica. Proceedings of the 1st Congress of the PCLA in 1989, pp. 267-274.

Scott, M. 1999 WordSmith Tools Version 3.0. Oxford: Oxford University Press.

Short, M. 1996 Exploring the Language of Poems, Plays and Prose. London: Longman.

1998 From Dramatic Text to Dramatic Performance. In: Culpeper, J., Short, M. & Verdonk, P. Exploring the Language of Drama. London: Routledge, pp. 6-18.

Simpson, P. 1998 Odd Talk: Studying Discourses of Incongruity. In: Culpeper, J., Short, M. & Verdonk, P. (eds) Exploring the Language of Drama. London: Routledge, pp. 34-53.

Snell-Hornby, M. 1997 "Is this a Dagger Which I See Before Me?": The Non-verbal Language of Drama. In: Poyatos, F. (ed.) Nonverbal Communication and Translation: New Perspectives and Challenges in Literature, Interpretation and the Media. Amsterdam: John Benjamins, pp. 187-201.

Toury, G. 1980 The Adequate Translation as an Intermediating Construct: A Model for the Comparison of a Literary Text and its Translation. In: Toury, G. In Search of a Theory of Translation. Tel Aviv: The Porter Institute for Poetics and Semiotics, Tel Aviv University, pp. 112-121.

Van den Broeck, R. 1993 Report. In: Delabastita, D. & D'Hulst, L. (eds) European Shakespeares: Translating Shakespeare in the Romantic Age. Amsterdam: John Benjamins, pp. 105-109.

Van Schoor, J.L. 1983 Die grammatika van standaard-Afrikaans. Kaapstad: Lex Patria.

Wybenga, D. 1982 Wie sal die wa smeer? In: Van Jaarsveld, G.J. (red.) Wat se jy?: Studies oor taalhandelinge in Afrikaans. Johannesburg: McGraw-Hill, pp. 339-354.

Zuber, O. (ed.) 1980 The Languages of Theatre: Problems in the Translation and Transposition of Drama. Oxford: Pergamon.