A psycholinguistic approach to theatre translation (original) (raw)

To cite this article:

Tarantini, Angela Tiziana. “A Psycholinguistic Approach to Theatre Translation.” The AALITRA Review: A Journal of Literary Translation 11 (May 2016): 60-77.

A Psycholinguistic Approach to Theatre Translation

ANGELA TIZIANA TARANTINI

Monash University

Abstract

Ever since the publication of the first edition of Venuti’s The Translator’s Invisibility, the dichotomy between the foreignizing and the domesticating approaches to translation has been an object of debate in translation studies. In this article I do not wish to discuss the political implications of this opposition, which have already been widely discussed (Robinson, Bassnett and Trivedi, Tymoczko, among others). Rather, I wish to demonstrate that when translating for the theatre, a higher degree of domestication might be necessary because of the medium involved. I show that the translator not only has to take into account the spoken nature of the dialogue, but also the aural nature of its reception. The aural nature of the reception of a theatre text also implies its impermanence and irreversibility. These features do not allow for re-examination of the linguistic input provided. Drawing from studies in psycholinguistics and cognitive psychology, I show how certain foreignizing strategies successfully applied to the translation of the written page might prove problematic when applied to stage translation. For that purpose, I use a single case study, an ongoing drama translation project: the translation of Convincing Ground (Mence) into Italian. My purpose is to demonstrate that a foreignizing strategy may not only hinder the audience’s lexical decision response, or prevent it altogether within the given time of utterance, but may also result in failure to convey the characterization of the people depicted in the fictional world, as well as the relationship they have with one another, shaped and negotiated through language.

1. Translating for the listener

In this article I do not wish to discuss the political implications of the opposition foreignization vs domestication, where the two items should be considered extremes of a continuum rather than an actual dichotomy (as Baker has cautioned us). Rather, I wish to demonstrate that when translating for the theatre, a higher degree of domestication might be necessary, at least on the lexical and syntactic levels.

Many scholars have focused on the spoken nature of the theatrical discourse, and on how the translation has to be “speakable” or “performable” on stage (see, for example, Espasa; Morgan; Bassnett-McGuire, Nikolarea); others have drawn attention to the fact that the translation of texts meant to be spoken and/or performed can exploit the possibilities of expression of the human voice and body (Pavis, Serpieri, among others). 1{ }^{1} However, to date, no scholar has analysed the psycholinguistic and the cognitive implications of the mode of production of the translated text on the listener/viewer. Working across disciplines, I apply some of the findings of the vast body of research in psycholinguistics to translation for the

[1]

1. 1{ }^{1} The literature on theatre translation has flourished in recent years; from the mid 1990s onwards, an increasing number of practitioners have engaged in research on the translation for the stage (for a detailed overview, see Serón Ordóñez). Here I have cited only a few researchers who have specifically addressed the issue of speakability and performability. ↩︎

stage. 2{ }^{2} I aim to show that some of the foreignizing strategies that may be successfully applied to translation for the page, in stage translation may hinder audience understanding altogether. I will focus in particular on foreignizing strategies such as maintaining the foreign syntax in the translated playtext, and retaining certain culture-specific items in the translation. I am aware that some translators and practitioners may not be familiar with psycholinguistics. I hope that this article will make some of the research in the field accessible to theatre translators, insofar as it could be relevant for a more effective translation for the stage, and for increased awareness on the translator’s part. In section 1.1 I focus on aspects of spoken language such as prosody and segmentation, and how those influence the listener’s process of word recognition. In section 1.2 I show how the complexity of the message impacts the time it takes for both the reader and the listener to decode a message. In section 2, I analyse the possible implications of some of these findings for the theatre translator. In section 3, I look at the “affective environment” (Johnston 18) of the audience by analysing the emotional impact of certain lexical items on the spectators. All the theory introduced in the first three sections comes together at the end of the article, and in section 4 I draw my conclusion.

1.1 Prosody, segmentation, and word recognition

If we ask a lay person if it is easier to process written or spoken language, the answer is very likely to be: spoken. As surprising as this may seem to most people, the opposite is true: the process of decoding written language is easier than that of decoding spoken discourse (Cacciari). Some data on language processing time are necessary in order to lay the groundwork for my advocacy of domestication in theatre translation.

One of the main differences between written and spoken language processing lies in the organ designated for such aim. The ear is a serial device, i.e. it processes information spread over time, while the eye is a parallel device, i.e. it can process much more information at the same time - information related to three-four words (Foss and Hakes). In this article, I will refer to experiments on both spoken and written language processing, to highlight how lexical frequency, ambiguity, and structural complexity affect language processing time.

As Dąbrowska states: “In informal conversation people produce, on average, about 150 words per minute” (13), whereas while reading, a person reads somewhere between 200 and 400 words per minute; that is the commonly accepted figure for newspaper reading (Gibson and Levin, cited in Foss and Hakes 327). Just to give a concrete example, a conference paper consists of anywhere between 2,500 and 3,000 words, and is read (and consequently listened to and decoded) in approximately 20 minutes. This means that the audience will process the linguistic information at a rate of 125-150 words per minute (this is in line with Cacciari’s and Dąbrowska’s data). For the same audience, the processing time for reading the same material would oscillate between two thirds and one third of the time, that is, between 12.5 and 6.25 minutes. So, even if spoken language is acquired while the ability to read and write is taught, the written input is processed faster than the spoken input. Moreover, there are factors which can make spoken language processing more challenging: firstly, “the conditions of noise in the environment where we often speak and hear people speak” (Cacciari 111). 3{ }^{3} That is even more the case in a public place like the theatre. If someone in the audience speaks or whispers during a performance, it might interfere with the listening process of the other people present. When a reader engages with a written text, instead, what usually happens is that s/he automatically shuts out other stimuli and focuses on the referential value of the information provided. Another difference, and a fundamental one, is the possibility of re-examining the linguistic

[1]

1. 2{ }^{2} This work is not meant to be experimental research, but rather an interdisciplinary approach to theatre translation.
   3 “le condizioni di rumore ambientale in cui spesso parliamo e sentiamo parlare” (all translations from Italian are mine unless otherwise indicated). ↩︎

input. When reading, one can go back and re-read a sentence, if necessary, while “a listener instead has to elaborate the spoken discourse at the pace set by the speaker” (Cacciari 111-12). 4{ }^{4} Most of all, segmenting spoken discourse into the sound units it is made of is more difficult than isolating single words or sentences in written language. “Segmentation is almost absent in fluent spoken language, and is a by-product of the process of word recognition” (Cacciari 114). 5{ }^{5} Where single words are not isolated, the process of word recognition is a lot more difficult,
asAltmanndemonstratedinhisbooktheascentofBabelbywritingasentencewithoutsignallingtheen dofthewordsandthusmakingacaseinpoint (Altmann). Linguist John L.M. Trim, with due acknowledgement to cartoonist Peter Kneebone, shows us that it is the identification of words that allows for segmentation of spoken discourse, which consequently enables the interlocutor to decode the message, as the examples in Figure 1 reveal.

Figure 1. The role of segmentation in decoding (from Trim 76-77)
If we look at any spectrogram of spoken sentences, we will see how boundaries between words are not acoustically marked, and that the only interruption we can actually notice, is that of the air flux when certain consonants (such as stops) are articulated. Word division is the result of a cognitive, 6{ }^{6} not a phonological process.

An important feature of spoken language is prosody. Research carried out in the mid1980s by Jacque Mehler showed that infants as young as four days old can distinguish the prosody of their own language from that of a foreign language; this ability wears off at about nine months of age (Altmann; see also the more recent research by Kuhl). Cacciari observes that it is reasonable to assume that “people develop implicit strategies of segmentation of their own language based on its typical rhythm” (Cacciari 115). 7{ }^{7} To put it in Cutler’s words, “listening itself is language specific. It is always native listening” (Cutler 72). She maintains

[1]

1. 4 “chi ascolta qualcuno che parla deve elaborare, invece, il parlato alla velocità decisa dal parlante”
   5 “La segmentazione è quasi assente nel linguaggio parlato fluente ed è un prodotto collaterale del processo di riconoscimento delle parole.”
   6{ }^{6} Here I use the word cognitive in the sense of intellectual, not as used in cognitive linguistics, where it takes on a completely different meaning.
   7 "che le persone sviluppino delle strategie implicite di segmentazione della propria lingua basate sul suo ritmo tipico […][\ldots] " ↩︎

that “listeners from different language backgrounds develop different ways of listening, propelled by differences in the native vocabulary structure” (Cutler 72). As a listener hears a sentence in his/her own native language, s/s / he automatically applies the segmentation strategy of that particular language. Therefore, recognizing a foreign word in a language which applies a different segmentation strategy from that of one’s own native language within the context of the latter can be more challenging than it may appear, even if the person is a fluent speaker of the foreign language in question.

Applying Marslen-Wilson’s revised Cohort Model, a model of recognition of spoken words (“Functional Parallelism”), to culture-specific items will enable me to demonstrate how some foreignizing translation strategies applied to the translation of the written page might not be suitable for the translation for the stage. The cohort model is based on the assumption that “when we hear a word we simultaneously build a cohort of possible items that share the initial part of the word (more or less the first syllable)” (Cacciari 122). 8{ }^{8} As the speaker progresses, such cohort will contain fewer and fewer items, until the word is finally recognized by the listener, as Table 1 shows.

Table 1. Illustration of the Cohort Model for the word

It takes the listener somewhere between 30 and 300 milliseconds (ms) to identify words in isolation, and as little as 200 ms when words are inserted in the context of a sentence (Dąbrowska 13); some even maintain that it can take as little as 125 ms (Cacciari 107), which means before the speaker has finished uttering them, i.e. before their acoustic offset. 9{ }^{9} These data refer to optimal conditions, but there are some factors that can hinder understanding. Cacciari identifies five distinct factors that can influence the language processing time:

1. Structural complexity (i.e. syntax, but also word frequency);
2. Lexical (or syntactic) ambiguities;

[1]

1. 8 “quando sentiamo una parola, costruiamo contemporaneamente una coorte di possibili item che condividono una parte iniziale (grossomodo la prima sillaba).”
   9{ }^{9} Luce instead maintains that most words cannot be recognized until at or after their end (cited in Cutler). ↩︎

2. Degree of cohesion;

3. Length;

4. Time pressure.

For the purpose of my study, I will only focus on structural complexity, lexical ambiguities and time pressure.

A series of experiments by means of the probe latency technique carried out by Caplan (1972), and Walker, Gough and Wall (1968) demonstrated that the listener’s reaction time is indeed affected by the clausal structure of a sentence (cited in Foss and Hakes). In probe latency studies carried out by Caplan, listeners heard a sentence, and were asked immediately afterwards to decide as quickly as possible whether a probe word had been presented in a sentence. Caplan found that the average time to answer the yes/no question was longer in sentences where the structure was more complex. So, the more complex the sentence structure, the longer the listener’s reaction time. Maintaining a foreign syntax for a spoken dialogue puts extra pressure on the listeners, who might not have enough time to work their way through a complex syntactical structure during a performance. This is not the case in novel reading, for example, when readers can take their time to work their way through a foreign or foreignizing syntax, putting in a little extra work (and time). Whereas in novels certain foreignizing strategies can be stylistic (or political) choices, and can be very effective and pleasant, in theatre translation they may impede understanding altogether.

1.2 Complexity of the message: written vs spoken

In an experiment on perception of spoken words carried out by Vitevitch and Luce (1998), subjects were asked to repeat either a word or a nonword (a sequence of consonants and vowels). Figure 2 shows the reaction times in milliseconds for the words and nonwords for each probability and density condition. 10{ }^{10}

Figure 2. Reaction times for repetition of words and nonwords (from Vitevitch and Luce 327)

[1]

1. 10{ }^{10} The probability refers to the frequency of a phonotactic structure. For example, in English the vowel-consonant combination CCVCC is very common, while in Italian it is less common. Words in dense similarity neighbourhood, instead, are words that share with other words some phonological features (e.g. minimal pairs). It is important to point out that the experiment was carried out on single words in isolation, not words in the context of a sentence. ↩︎

Experiments on eye fixation times in reading carried out by Just and Carpenter, and by Rayner and Duffy demonstrate how the complexity of the message influences the processing time of the message itself. Rayner and Duffy analysed the effects of word frequency, verb complexity and lexical ambiguity on written language processing time. For the purpose of my study, I will take into account only word frequency and lexical ambiguity. 11{ }^{11} The experiments carried out by both groups of researchers distinguish between fixation and gaze, gazes being the “[c]onsecutive fixations on the same word” (Just and Carpenter 329). Both experiments show that “longer fixations are attributed to longer processing time caused by the word’s infrequency and its thematic importance” (Just and Carpenter 330). In their experiment, Rayner and Duffy measured the fixation and gaze durations on a target word, and on what they defined the “disambiguating region” (i.e. the word preceding the target word, and the one following). The experiment confirmed what was expected, which is that "subjects spent significantly longer on both the first fixation on the infrequent word […] and the gaze on the infrequent word"12 (Rayner and Duffy 195). When a target word was infrequent, the mean gaze duration was also longer on the word immediately following. 13{ }^{13} A second experiment by Rayner and Duffy on equibiased and non-equibiased ambiguous words confirmed what was predicted, i.e. that “subjects spent extra time looking at the ambiguous word when two meanings for the ambiguous item were fairly equally likely. This was not the case for ambiguous words for which one meaning was highly likely” (Rayner and Duffy 197).

2. Implications for the translator

Translators of written texts can count on the readers to follow their own rate of information input. In the case of lexical ambiguity, for example, readers can spend time on the challenging word, and the words immediately preceding and following. This operation will take anywhere between 1,423 and 1,923 milliseconds (Rayner and Duffy), without considering the time to read a footnote or an entry in the glossary, if any. But during the disambiguation and/or interpretation process, readers will not be subjected to further inputs, whereas listeners will.

Another aspect to take into account is that in many theatrical performances the audience is not the main addressee of the utterance; to put it differently, the communication system differs from that of naturally occurring conversation (and from that of the novel as well). Since the communication between playwright and audience is “embedded” within that between characters (Short), the audience cannot interfere with the communication between the characters; so the audience cannot halt the speaker and ask him/her to clarify what s/\mathrm{s} / he means. At a challenging point, the listener will apprehend listening in order to process the linguistic information. That is what happens when the reader fixates a word on the written page, spends time on the disambiguating region (the words preceding and following the challenging word); re-reads the information provided (i.e. gazes on the disambiguating region); or spends the time s/\mathrm{s} / he needs in order to decipher the complexity of the foreign syntax. The reader will move on only once the linguistic information has been processed. The listener does not have the same possibility, because while s/he\mathrm{s} / \mathrm{he} is trying to process the linguistic input provided the actors will keep speaking, since actors usually speak at natural speed (unless otherwise required for a specific dramatic effect). The consequence is that the listener will be able to process neither the challenging input, nor the next incoming input. This is consistent with Cacciari’s hypothesis

[1]

1. 11{ }^{11} The results of the experiment on verb complexity for causative, factive, and negative verbs offer no evidence that verb complexity affects processing time.
   12{ }^{12} The mean first fixation duration on the target word was 557 milliseconds, while the mean gaze duration was 1,492 milliseconds (Rayner and Duffy 195).
   13{ }^{13} 1,443 milliseconds (Rayner and Duffy 195). ↩︎

that time pressure is a crucial factor in the understanding process. As semiotician Cesare Segre points out, theatre is:
a secondary modelling system totally different from the narrative text. It is a system which resorts to the physicality of actors, to their voices and gestures, to their costumes; to the physicality of the stage […]; to the physicality of the duration itself, because what the audience witnesses […] takes place in the very time of the utterances of which it is made up, a time which is irreversible, similar to lived time.
(Segre 5) 14{ }^{14}
It is precisely this irreversibility of the time of the utterances that may prevent the audience from elaborating the complexity of a foreign syntax, the lexical ambiguity, and the decoding of infrequent words at their own rate of input, and may ultimately preclude the audience’s understanding. As early as 1976, Italian linguist and lexicographer Giovanni Nencioni claimed that
[…] in theatre, and in every type of theatre, the recipient is more important than in any other type of literary communication. He [sic] is physically present and can count on two perceptual organs, sight and hearing, but applied to a passing and irreversible reality; the author, the director, the actors have to adapt the text and the acting to the average perceptual and mnemonic abilities of the listeners, and keep in mind the paraphrastic consequences, if they do not want to repulse them. The audience then influences all those who participate in the realization of the show.
(Nencioni 45, my emphasis) 15{ }^{15}
Nencioni did not mention the theatre translator, since in the mid-70s very few scholars had addressed the issue of translating for the theatre. 16{ }^{16} However, Nencioni’s vision most certainly applies to theatre translators.

A common foreignizing translation strategy consists in leaving culture-specific items unchanged in the translated text, adding a footnote, an entry in a glossary, or simply trusting that the reader will infer the meaning from the context. A footnote or a glossary is not applicable to theatre translation for obvious reasons. The only viable option would be to leave the culture-specific item untranslated in the target text. While that strategy could have a certain effect in the translation of the written page, in theatre translation it may result in the audience not understanding the spoken message, as the following example shows (for clarity I will provide the whole exchange):

[1]

1. 14 “un sistema modellizzante secondario del tutto diverso dal testo narrativo. È un sistema che ricorre alla fisicità degli attori, delle loro voci e gesti, dei loro costumi; alla fisicità del palcoscenico […]; alla fisicità stessa della durata, perché ciò a cui il pubblico assiste […] si svolge nel tempo stesso degli enunciati che lo compongono, tempo non reversibile, analogo a quello vissuto.”
   15 “[…] nel teatro, e in ogni tipo di teatro, il destinatario ha maggior peso che in qualsiasi altra comunicazione letteraria. Egli è presente fisicamente e può fare assegnamento su due organi di percezione, la vista e l’udito, ma applicati a una realtà trascorrente e irreversibile; l’autore, il regista, gli attori devono commisurare il testo e la recitazione alle medie capacità percettive e memorizzatrici degli ascoltatori, e interessarsi delle conseguenze parafrastiche, se proprio non voglioni ributtarli. Il pubblico dunque condiziona profondamente tutti coloro che concorrono a realizzare lo spettacolo […]”
   16{ }^{16} Ibsen’s translator Michael Meyer, and Jiří Levý were among the first to discuss theatre translation. The prolific author and researcher Susan Bassnett started writing on the topic in 1978. ↩︎

The tone of the passage is playful, and the two characters here are sharing fond memories of their life together. The reason I did not translate Renanghi’s swear word into Italian is precisely the frisky mood of the whole passage. My choice will become clearer in section 3, where I analyse the impact of taboo words on the listener. 17{ }^{17} For the scope of my argument, I wish to focus only on the word “darkie”. In this passage, the word “darkie” is used by Dutton as a kind of endearment. In Italian, the English loanword “dark” is an entry in the dictionary, and is associated with Gothic music and fashion style. If we apply the Cohort Model to the word darkie for an Italian audience, the result would be as follows:

Table 2. Cohort Model for the word in Italian

By applying the Cohort Model to the word “darkie”, we can see that an Italian audience will end up with a nonword of low density (i.e. not many words have a similar sound in Italian) and high probability (i.e. the phonotactic sequence CVCCV is very common), 19{ }^{19} so the audience’s processing time is likely to be over 1,000 milliseconds, but in the end the message

[1]

1. 17{ }^{17} Moreover, the connotations of marked female language are stronger (see Lakoff). Also, swearing in Australian culture is a lot more acceptable than in Italian culture, particularly for women (see also Berruto).
   18{ }^{18} I have chosen to transcribe phonetically the word dark in the way the average Italian speaker would pronounce it, i.e. with the rolling <r> and the Italian vowel /a/. BrE and AusE: /'da:k/, AmE: /'da:®k/
   19{ }^{19} I specifically refer to the phonotactic sequence; the graphemic sequence is CVCCVV, and in Italian it would be a low-probability sequence. ↩︎

will probably not be decoded. Because of the average speaking rate, while the audience is still processing this linguistic input, the actor will have uttered other 2-2.5 words. The audience would be unable to decode the word darkie and would probably also miss those immediately after it, since during the utterance they would still be engaged with the target word. Retaining the word “darkie” in Italian, therefore, does not seem viable. In my translation I have opted for the word “negretta” which contains the root “negr-”, which is politically incorrect and offensive, but also the suffix “-etta” which is a modification to express endearment (it could also be used to belittle and diminish someone, but that is not the case in the example provided). Elsewhere in the play, where Dutton uses the word “darkie” as a derogatory term, I translated it as “negra” (the feminine singular of “negro”). If I had to translate the same term in a novel, I would probably leave the word darkie and enter it in a glossary, or just let the reader infer the meaning from the context, given the possibility for readers to re-examine the linguistic unit and to process the linguistic input at their own pace. Different texts may require different translation strategies, as the following example shows:

The word gin was used to signify an indigenous woman. It was common practice for sealers and whalers to keep an Aboriginal wife, a woman to satisfy their sexual urges, but also to help them survive on the harsh Australian frontier (Taylor). The term is derogatory and offensive. Consulting the Collins Dictionary online, we find (among others) the following definitions:
gin 20^{20}
noun

1. an alcoholic drink obtained by distillation and rectification of the grain of malted barley, rye, or maize, flavoured with juniper berries
2. any of various grain spirits flavoured with other fruit or aromatic essences

⇒\Rightarrow sloe gin
3. an alcoholic drink made from any rectified spirit
noun
4. a primitive engine in which a vertical shaft is turned by horses driving a horizontal beam or yoke in a circle
5. Also called: cotton gin a machine of this type used for separating seeds from raw cotton
6. a trap for catching small mammals, consisting of a noose of thin strong wire
7. a hand-operated hoist that consists of a drum winder turned by a crank
noun
8. (Australian, offensive, slang) an Aboriginal woman
[1]

1. 20http://www.collinsdictionary.com/dictionary/english/gin?showCookiePolicy\={ }^{20} \mathrm{http}: / / w w w . c o l l i n s d i c t i o n a r y . c o m / d i c t i o n a r y / e n g l i s h / g i n ? s h o w C o o k i e P o l i c y= true (accessed January 2015) ↩︎

When presented with the word gin, an Australian audience will have to decode a nonequibised ambiguous word. However, as Marslen-Wilson demonstrated, “the frequency of an item and the frequency of its close competitors should interact to determine the timing of lexical choice” (Marslen-Wilson, “Activation” 150). This means that, since the word gin is more frequently used in relation to the alcoholic drink, the processing time for an Australian audience would be longer than for a high-frequency word. Here is the cohort model for the word gin for an Australian native speaking audience:

Table 3. Cohort Model for the word in Australian English

An Australian audience will have to disambiguate the word, exclude the more or less frequent non-equibised lexical items, and finally the process of recognition will take place. There are two distinct factors which will foster the lexical decision response of an Australian audience. Firstly, the recognition process will be facilitated by cross-modal priming: the visual presentation of a target word (the indigenous woman on stage) is presented concurrently with the auditory presentation of the related word in the context of a sentence (Zwitserlood 30). This facilitates the disambiguation of the term and the final lexical decision response. However, that is possible only for an Australian audience; it is the very presence of the lexical item “gin = (Australian, offensive, slang) an Aboriginal woman” in the mental lexicon of the Australian audience that allows for the activation of the lexical item in question, and for the final lexical decision response. The multimediality of the theatre allows for cross-modal priming, which could not take place on the written page (with the exclusion of illustrated books). Secondly, as Hill and Kemp-Wheeler note, “[c]ompared to neutral words, aversive words are easier to identify as words in a lexical decision task” (cited in Harris, Ayçiçeği and Gleason 562-63).

Let us now look at the cohort model for the same lexical item in Italian:

Table 4. Cohort Model for the word in Italian

In Italian, the word gin only indicates the liquor; it is quite unlikely that an Italian audience would make the connection between the lexical item gin and an indigenous Australian woman.

3. The “affective environment” of the spectator

As theatre translator and practitioner David Johnston observes, "[t]ranslation, and especially translation for the theatre, is a process that in this way engineers two-way movement - a traffic

between the narratives, concepts and structures of life embodied in foreign texts, and the affective and cognitive environment of the spectator" (Johnston 18, my emphasis). So far I have dealt with the cognitive environment of the spectator; let us now look at the affective environment. As a theatre translator, I see myself as a translator of spectacle, performance. The immediate impact that the translated text will have on the audience is therefore of vital importance. To put it differently, the translator should not weaken “the force the text has in performance […] what counts is not the degree of distance from an ontological original but the effect that the reconfigured text (as performance) has on the receiving culture and its networks of transmission and reception” (Marinetti 311, my emphasis). The emphasis placed on reception involving “a reconceptualization of the role played by spectators as well as a rethinking of more general notions of reception” (Marinetti 311, original emphasis) opens the way to new research paradigms with the spectators at their centre. To some extent, the “performative force” of an utterance (Worthen 2003 9-13, cited in Marinetti 311) can be measured through the impact it has on the audience. The idea of impact (or effect) actually rests on psychological and physiological grounds. Cognitive psychologist Steven Pinker gives us insight into the way we use language to negotiate relationships, but also to impose negative emotions on our interlocutor, as when speakers use swear words, for example (Pinker “Stuff of Thought” lecture). According to Allan and Burridge, language is both “a shield and a weapon” (3). It is used as a shield to avoid being offensive (as in the case of euphemisms), but as a weapon when the speaker deliberately uses dysphemisms to be offensive or abusive. In fictional dialogue, swear words can be a good indicator of how the author characterizes the people inhabiting the world of the play, but also of what s/he\mathrm{s} / \mathrm{he} wants the audience to feel during the performance. Indeed, the dialogue between the characters in a play has two functions: one in the fictional world of the performance, and one in the real world. In the fictional world, dramatic dialogue creates the fictional world and shapes the relationship between the characters. In the real world, it is a message from the playwright to the audience. 21{ }^{21} My analysis will focus on the translation of taboo words also because swear words are often culturespecific, and therefore belong to a debated category in translation studies. Some of the scholars in favour of a foreignizing strategy (first and foremost Venuti) may recommend preserving culture-specific items untranslated in the target text.

Swearing differs across cultures, but taboo words belong to the same five semantic areas in all cultures: religion and the supernatural; bodily secretions; death and illnesses; sexuality; race and minority (i.e. disfavoured) groups (Allan and Burridge; Pinker). Two generalizations need to be made at this point. First, that “taboo words activate brain areas associated with negative emotions” (Pinker Stuff of Thought), such as the right hemisphere of the brain, and the amygdala. Second, that taboo words are processed involuntarily; this means that when someone reads or hears a taboo word, not only can s/he not ignore its meaning, but s/he can also not help perceiving the negative emotion associated with it. 22{ }^{22} In order to prove that something is processed involuntarily by the brain, psychologists use the stroop test: people are asked to name the colour in which a word is printed, and ignore what the word spells out. 23{ }^{23} When the font colour does not match what the word spells out (e.g. the word “black” is written in red ink), subjects take longer to perform the task (Pinker Stuff of Thought). That is because as literate people we simply cannot treat a word as a cluster of sounds, or scribbles on a page;

[1]

1. 21{ }^{21} See Short’s and Segre’s theatrical communication models.
   22{ }^{22} In the light of the research on taboo words, and the negative feelings that they impose on the listener, I have chosen not to translate the word “fucken” in the example in section 2 . The playfulness of the exchange did not seem to suggest the intention on Renanghi’s part to inflict negative emotions on Dutton. Additional factors, such as rhythm, influenced my translation choice.
   23{ }^{23} The stroop test got its name from John Ridley Stroop, who first published on the stroop effect in 1935. ↩︎

we automatically process the written or spoken word. The same thing happens cross-modally: if people have to name colour patches, but a voice recites a different sequence of colours unrelated to the patches shown, the people involved in the experiment get confused. Psychologist Don MacKay introduced a variant of this test, where he asked subjects to just name the font colour and to ignore what the word spelt out, but this time swear words were among the words presented. Not only did MacKay find that subjects were slowed down almost as much as in the standard stroop test; he found that “presenting taboo words enhances skin conductance, an unconscious index of sympathetic nervous system activity and emotional arousal” (Mackay et al. 475). Harris et al. refer to experiments carried out by means of electromodal monitoring 24{ }^{24} to measure the autonomic arousal when subjects hear a taboo word in their first or in their second language. These experiments have proven that words in one’s native language have “greater emotional resonance” than words in a second language (Harris, Ayçiçeği and Gleason 563). If a theatre translator wishes not to weaken the “performative force” of an utterance, awareness of this neurophysiological aspect of language processing could be of help. The impact of a written utterance and that of a spoken utterance differs also on an emotional level. As Harris et al. state
[s]poken language is acquired before visual language (for L1 acquisition). To the extent that linguistic representations that are learned early become connected with emotional regulation systems (Bloom & Beckwith, 1989), auditory language may be more closely tied to emotional arousal than visual language.
(Harris, Ayçiçeği and Gleason 565)
As theatre translator, I feel that failure to engender a strong psychophysiological reaction in my audience would be the worst form of “betrayal” of the source text, when such reaction was originally envisaged by the author (and such intention is evident in the use of a certain kind of language).

Let us look at the implications of the taboo stroop effect on the translation of the culture specific item gin examined above. As we have seen, the word gin is both misogynist and racist, and therefore belongs to one of the five semantic areas that activate the areas of the brain associated with negative emotions. As Pinker observes, “[t]hanks to the automatic nature of speech perception, a taboo word kidnaps our attention and forces us to consider its unpleasant connotations” (Pinker Stuff of Thought 339). The use of taboo words fulfils two functions in the playtext: on the level of the fictional world, it characterizes the protagonist uttering the swear word as someone willing to impose negative emotions on the other character/s. On the level of the communication between the playwright and the audience, by using a taboo word the playwright imposes that same negative emotion on the audience. When hearing the word gin, the right hemisphere and the amygdala of an Australian audience will activate. Failure to create the same effect on the target audience in the name of a politically correct foreignizing approach would have two negative consequences. Firstly, the characterization of the protagonist uttering the sentence may not be the one originally envisaged by the author. Secondly, keeping the foreign word within the text will not cause that emotionally charged response from the audience. The third, obvious consequence - that is, the impossibility for the audience to process the foreign word - has already been dealt with from a psycholinguistic point of view (section 2).

Let us now look at the socio-anthropological implications of the same exchange. Anthropologist Alan Fiske maintains that there are three types of relationships throughout all

[1]

1. 24{ }^{24} A psychophysiological technique that records skin conductance responses. ↩︎

human cultures, namely Communal Sharing, or communality, Authority Ranking, or dominance, and Equality Matching, or reciprocity (Fiske; see also Pinker, Stuff of Thought). 25{ }^{25} Communality relationships are based on the assumption of equality between the members. Usually kin relationships, or the relationship with one’s spouse, are based on communality. Authority Ranking relationships are based on the dominance of one subject over the other(s). That is usually the case of relationships within a working environment, or in colonial situations. Equality Matching relationships are business-like relationships, where there is an exchange between the parties, for the benefit of both groups/members involved. What is appropriate in one type of relationship might not be appropriate in another. If we apply Fiske’s relationship grid to the exchange examined above, we could “translate” the dialogue as follows:

This crucial aspect of the different perception that the characters have about their relationship would get lost if I chose to leave the culture-specific item unchanged.

Elsewhere I translated the same word differently, as in the following example:

Here I translated the word “gin” as “aborigena” (lit. Aboriginal woman); as already mentioned, it was common practice for the settlers to have intercourse with indigenous women (Taylor), and here Dutton describes a place where it was easy for a man to find an indigenous woman to sleep with. Here I have chosen to favour the racial elements for two reasons: firstly, because of the negative emotions attached to words referring to disfavoured groups, 26{ }^{26} and secondly, because I wanted to underline the dominance relationship between the “blokes” and indigenous women, treated as sexual objects. I made a more radical choice to underline Dutton’s racist language earlier in the play, when translating “blackfella”, an Aboriginal English word. It is used by the indigenous population to refer to the people in their own community. However, if used by white people it takes on a pejorative connotation (Arthur) and is considered offensive, 27{ }^{27} as in the following example:

[1]

1. 25{ }^{25} There is a fourth relationship type, called Market Pricing, which applies to the whole system of modern market economies and therefore is far from universal, not applicable to the present study, and not relevant to this discussion.
   26{ }^{26} In today’s English, the most offensive word is the word “nigger” (Pinker Stuff of Thought).
   27{ }^{27} Farzad Sharifian, personal communication, 22 June 2015. ↩︎

This exchange takes place at the beginning of the play, and I wanted to render Dutton’s use of offensive language from the start. Elsewhere, I translated the word differently, also according to its use and the user, as in the following example:

Here it is Renanghi, the indigenous young woman, who uses the term blackfella, and uses it with a positive connotation, hence my translation choice.

If I were translating a different type of fiction, such as a novel, I would not translate the same culture-specific item in different ways in different parts of the text, as I did with the words gin and blackfella. Rather, I would apply the same translation strategy consistently throughout the novel (most likely by adding a glossary and preserving the lexical item in the source text). Given the difference between reading, and watching a play, it is not surprising that in theatre the stress is often placed on the immediate impact of the dialogue on the audience.

4. Conclusion

Venuti claims that “[f]luency assumes a theory of language as communication that, in practice, manifests itself as a stress on immediate intelligibility […]” (60). But, as we have seen, this can be central to theatre translation. As already mentioned, the different communication model could influence the audience’s processing time; it is my belief that a theatre translator who is aware of such mental processing is more likely to produce an effective translation that will work on stage. The question remains the same: what is the effect the translator strives for? If the translator wants to enable the audience to process the spoken message during the performance time, and does not wish to weaken the performative force of the utterance for the benefit of the audience, then a higher degree of domestication might be necessary in theatre.

As I hope to have demonstrated, this higher degree of domestication is justified by psycholinguistic research on spoken language processing, and by psychophysiological studies on the effect of words pertaining to certain semantic areas. Not only is the spoken mode of delivery crucial, as theatre translators and practitioners have explored (Morgan; Espasa; Pavis, among others), but also the aural aspects of its reception. Theatre translators know that the playtext they translate will be part of “a structural system [which] exists only when received and reconstructed by a spectator from the production” (Pavis 25, my emphasis). This

[1]

1. 28{ }^{28} In Italian, there is no derogatory term to refer to indigenous Australians. Since in Italy it is common knowledge that Australia was inhabited by indigenous Australians prior to the settlement, the audience will not be confused by the term “negro” (lit. nigger) and will not think it might refer to a different ethnic group, for example African Americans. The term “negro” is the most offensive word in contemporary Italian to refer to black people, and that is the reason why I have opted for this solution. ↩︎

reconstruction operated by the spectator through the affective, cognitive and psycholinguistic processes involved in the decoding of the spoken message, however, has often been overlooked in translation studies, despite its importance in the creation of meaning.

Bibliography

Allan, Keith, and Kate Burridge. Euphemism and Dysphemism: Language Used as Shield and Weapon. New York: Oxford University Press, 1991.

Altmann, Gerry T. M. The Ascent of Babel: An Exploration of Language, Mind, and Understanding. Oxford and New York: Oxford University Press, 1997.

Arthur, J. M. Aboriginal English: A Cultural Study. Melbourne: Oxford University Press, 1996.
Baker, Mona. “Reframing Conflict in Translation.” Critical Readings in Translation Studies. Ed. Mona Baker. London and New York: Routledge, 2010. 113-29.

Bassnett, Susan. “Translating Spatial Poetry: An Examination of Theatre Texts in Performance.” Literature and Translation. New Perspectives in Literary Studies. Eds. James Holmes, José Lambert and Raymond Van den Broek. Louvain: Acco, 1978. 171-76.

Bassnett, Susan, and Harish Trivedi, eds. Post-Colonial Translation: Theory and Practice. London and New York: Routledge, 1999.

Bassnett-McGuire, Susan. Translation Studies. London: Methuen, 1980.
Berruto, Gaetano. Sociolinguistica dell’italiano contemporaneo. Rome: La Nuova Italia scientifica, 1987.

Cacciari, Cristina Psicologia del linguaggio. Bologna: Il Mulino, 2001.
Cutler, Anne. Native Listening: Language Experience and the Recognition of Spoken Words. Cambridge, MA: MIT Press, 2012.

Dąbrowska, Ewa. Language, Mind and Brain: Some Psychological and Neurological Contraints on Theories of Grammar. Edinburgh: Edinburgh University Press, 2004.

Espasa, Eva. “‘Performability’ in Translation: Speakability? Playability? Or Just Saleability?” Moving Target: Theatre Translation and Cultural Relocation. Ed. Carole-Anne Upton. Manchester, UK and Northampton, Mass.: Northampton, Mass.: St Jerome, 2000.

Fiske, Alan Page. “The Four Elementary Forms of Sociality: Framework for a Unified Theory of Social Relations.” Psychological Review 99.4 (1992): 689-723.

Foss, Donald J., and David T. Hakes. Psycholinguistics: An Introduction to the Psychology of Language. Englewood Cliffs, N.J.: Prentice-Hall, 1978.

Harris, Cathrine L, Ayşe Ayçiçeği, and Jean Berko Gleason. “Taboo Words and Reprimands Elicit Greater Autonomic Reactivity in a First Language Than in a Second Language.” Applied Psycholinguistics 24 (2003): 561-79.

Johnston, David. “Metaphor and Metonymy: The Translator-Practitioner’s Visibility.” Staging and Performing Translation: Text and Theatre Practice. Eds. Roger W. Baines, Cristina Marinetti and Manuela Perteghella. Basingstoke: Palgrave Macmillan, 2011.

Just, M.A., and P.A. Carpenter. “A Theory of Reading: From Eye Fixations to Comprehension.” Psychological Review 87.4 (1980): 329-54.

Kuhl, Patricia. “The Linguistic Genius of Babies.” TED Talks. Lecture, October 2010.
Lakoff, Robin Tolmach. Language and Woman’s Place: Text and Commentaries. Ed. Mary Bucholtz. New York: Oxford University Press, 2004.

Levý, Jiří. “Drama Translation.” The Art of Translation. Eds. Patrick Corness, and Zuzana Jettmarova. Amsterdam and Philadelphia: John Benjamins, 2011. 129-66.

Mackay, Donald G., et al. “Relations between Emotion, Memory, and Attention: Evidence from Taboo Stroop, Lexical Decision, and Immediate Memory Tasks.” Memory and Cognition 3 (2004): 474.

Marinetti, Cristina. “Translation and Theatre: From Performance to Performativity.” Target 25.3 (2013): 307-20.

Marslen-Wilson, William. “Activation, Competition and Frequency in Lexical Access.” Cognitive Models of Speech Processing: Psycholinguistic and Computational Perspectives. Ed. Gerry T. M. Altmann. Cambridge, Mass.: MIT Press, 1990. 148-72.

Marslen-Wilson, William D. “Functional Parallelism in Spoken Word-Recognition.” Cognition 25.1-2 (1987): 71-102.

Mence, David. Convincing Ground. Unpublished script, workshopped by the Melbourne Theatre Company in 2009. Revised edition 2013.

Meyer, Michael. “On Translating Plays.” Twentieth Century Studies 11 (1974): 44-51.
Morgan, Edwin. “Language at Play. In Conversation with Joseph Farrell.” Stages of Translation. Ed. David Johnston. Bath: Absolute Press, 1996. 45-55.

Nencioni, Giovanni. “Parlato-Parlato, Parlato-Scritto, Parlato-Recitato.” Strumenti critici 29 (1976): 1-56.

Nikolarea, Ekaterini. “Performability Versus Readability: A Historical Overview of a Theoretical Polarization in Theater Translation.” Translation Journal 6.4 (2002) http://translationjournal.net/journal/22theater.htm (accessed May 2016)

Pavis, Patrice. Theatre at the Crossroads of Culture. London and New York: Routledge, 1992.
Pinker, Steven. The Stuff of Thought: Language as a Window into Human Nature. New York: Viking, 2007.
—_. “The Stuff of Thought: Language as a Window into Human Nature.” Lecture for the Royal Society for the encouragement of Arts, Manufactures and Commerce (RSA), London, 2010.

Rayner, Keith, and Susan A. Duffy. “Lexical Complexity and Fixation Times in Reading: Effects of Word Frequency, Verb Complexity, and Lexical Ambiguity.” Memory and Cognition 14.3 (1986): 191-201.

Robinson, Douglas. Translation and Empire: Postcolonial Theories Explained. Manchester: St Jerome, 1997.

Segre, Cesare. Teatro e romanzo. Turin: Einaudi, 1984.
Serón Ordóñez, Imaculada. “Theatre Translation Studies: An Overview of a Burgeoning Field (Part I: Up to the Early 2000s).” Status Quaestionis. Rivista di studi letterari, linguistici e interdisciplinari 5 (2013): 90-129.

Serpieri, Alessandro. “Semantics and Syntax in Translating Shakespeare.” Theatre Translation in Performance. Ed. Silvia Bigliazzi, Peter Kofler and Paola Ambrosi. New York: Routledge, 2013. 50-60.

Short, Mick. “Discourse Analysis and the Analysis of Drama.” Language, Discourse, and Literature : An Introductory Reader in Discourse Stylistics. Ed. Ronald Carter and Paul Simpson. London and Boston: Unwin Hyman, 1989. 139-70.
—_. “From Dramatic Text to Dramatic Performance.” Exploring the Language of Drama. From Text to Context. Ed. Jonathan Culpeper, Mick Short and Peter Verdonk. London and New York: Routledge, 1998.

Taylor, Rebe. “Savages or Saviours? - The Australian Sealers and Aboriginal Tasmanian Survival.” Journal of Australian Studies 24.66 (2000): 73-84.

Trim, J.L.M., illustrated by Peter Kneebone. English Pronunciation Illustrated. Cambridge: Cambridge University Press, 1965.

Tymoczko, Maria. “Translation and Political Engagement: Activism, Social Change and the Role of Translation in Geopolitical Shifts.” The Translator: Studies in Intercultural Communication 6.1 (2000): 23-47.

Venuti, Lawrence. The Translator’s Invisibility: A History of Translation. London and New York: Routledge, 1995.

Vitevitch, M. S., and P. A. Luce. “When Words Compete: Levels of Processing in Perception of Spoken Words.” Psychological Science 9.4 (1998): 325-29.

Zwitserlood, Pienie. “The Locus of the Effects of Sentential-Semantic Context in SpokenWord Processing.” Cognition 32.1 (1989): 25-64.