Professionals’ views on the concepts of translation: the challenge of categorisation

ABSTRACT

In this article, we analyse and discuss translation professionals’ views on the concepts of their trade: the concept of translation itself but also adjacent notions such as subtitling, interpreting, localisation and transcreation. Through focus groups with translators and translation project managers working for a translation agency, we explore how practitioners understand the concepts of translation, how they categorise and why they categorise as they do. Data were analysed by means of prototype theory, and results show that the focus-group participants understand prototypical translation rather narrowly as source-text oriented, written, interlingual transfer, but also that they have a broad and nuanced understanding of translation as a category. Beyond the prototype, they use ‘translation’ as a cover term for a wide variety of tasks that involve different degrees of intervention, transformation and creativity. Interestingly, the analyses show that the practitioners in the study categorise quite differently than translation scholars do. They draw on a fluctuating set of criteria for categorisation, some of which are highly pragmatic (e.g. pricing), and their definitions are much more floating and context-sensitive. In fact, they operate with different concepts depending on whom they interact with: clients, colleagues – or scholars.

KEYWORDS:

• Translation practice
• translation concepts
• prototype theory
• categorisation
• definition

1. Introduction

This article builds on and develops a previous study on translation professionals’ views on the concepts of their trade: the concept of translation itself but also adjacent notions such as interpreting, subtitling, localisation and transcreation (Dam and Korning Zethsen 2019). The central purpose of the study was to explore how practitioners understand and use these concepts in practice. The present article has similar aims but presents new data and digs deeper into a topic that proved to be a challenge to the practitioners in our previous research: the issue of categorisation. Before we unfold the specific aims of the present study, a more detailed description of our previous research and its findings helps setting the scene.

1.1. Background: the STP-study

Our 2019-study was based on focus groups with fifteen translation professionals – eight translators and seven translation project managers – employed in the same company: Sandberg Translation Partners (STP), a large UK-based translation agency that specialises in the Nordic languages. The focus-group participants were given 12 keywords to, firstly, describe and characterise and, secondly, categorise as either ‘Translation’ or ‘Not Translation’. The keywords were: Translation; Localisation; Transcreation; Subtitling/Dubbing; Interpreting; Revision; Proofreading; Editing; Glossary creation; Content writing; Intralingual translation; Intersemiotic translation.

Analyses of the focus-group sessions soon showed that the data did not lend themselves to categorisation in the classical, Aristotelian sense. Overlapping definitions and permeable conceptual boundaries were salient throughout the discussions, and there was abundant evidence of doubt, contradiction and disagreement. To make sense of the data, we therefore resorted to the concept of family resemblances (Wittgenstein 1953/1958), a way of categorising that allows for conceptual overlaps and fuzzy categories. For the representation of results, we drew on prototype theory based on Rosch’s (1973) ideas of categories as idealised cognitive representations and graded category membership. According to Rosch, categories are mentally represented by means of schemata of their most characteristic members – the prototypes – whereas other members of a category are more peripheral in nature (see also Halverson 1999, 2000, 2002 for analyses of translation as a prototype concept). By means of prototype theory, we were able to draw a model of translation as a category reflecting the conceptualisations of STP-professionals (see Figure 1).

Figure 1. A prototype model of translation: STP-professionals’ views (reproduced from Dam and Korning Zethsen 2019, 214 by permission of Taylor & Francis Group).

As shown in Figure 1 (with only a selection of the focus-group keywords represented), STP- professionals agreed that the most characteristic and central member of the category of translation – prototypical translation – was a rather source-text oriented, written translation between two clearly delineated languages. Also interpreting, subtitling and dubbing were seen as central members of the category, albeit not as the prototypes, as reflected by their position in the inner circle but not at the very centre. With increasing distance to the prototype, we find localisation in the first and transcreation in the second outer circle, with adaptation (a keyword proposed not by us but by the focus-group participants themselves) in between. Not surprisingly, the scholarly concepts of intralingual and intersemiotic translation were considered with some scepticism by the practitioners: intralingual was seen as a borderline case, the most peripheral member of the translation family at best; intersemiotic was dismissed as a non-member, an outlier.

The STP-study was evidently a snapshot; tied to a particular context, a given time and place: a UK-based translation agency in 2017. The study we shall report on in this article is a partial replication but set in a different context: a translation agency with similar characteristics but located in Spain and with data collected five years later, in 2022. Like translation studies, the translation industry is border-crossing and international by nature and presumably has a shared frame of reference across geographies, but it is also characterised by ‘rapid and radical changes’ (Gambier 2016), suggesting that the time factor could be a vector of change. A comparison of the two studies – the STP-study and the current research – may therefore allow us to identify some of the major movements that have taken place in the translation field over the past five years. The article does include an overall comparison (see section 3.8), but comparing is not its main aim, nor is it its most important contribution.

1.2. Aim and rationale

This paper delves into a main insight derived from the STP-study but not explored in detail in that study. As indicated, the difficulty of categorisation was a salient result. The STP-professionals virtuously joggled the many concepts of translation in their every-day dealings with clients and colleagues, yet in their conversations with us they were struggling with definitions, delimitations and criteria for inclusion and exclusion. In this partial replication of our previous research, the overall aim remains unaltered: to examine how translation professionals understand the concepts of their trade. More specifically, however, we ask how they categorise and why they categorise as they do.

The larger purpose of our inquiry is to bring in the perspectives of practitioners on a topic that could potentially benefit from a dialogue between research and practice but on which practitioners’ voices are rarely heard. Translation scholars have a deeply rooted ‘definitional impulse’ (Tymoczko 2007, 57), a ‘metatheoretical awareness’ (Halverson 2010, 384) that translation practitioners do not necessarily share (or publicise), and conceptual discussions of translation have traditionally been the domain of academics. This is perfectly natural and legitimate. However, the negative depiction of practitioners that sometimes accompanies this debate is, we suggest, counterproductive and calls for a closer examination of practitioners’ views. Let us explain.

A common stance among contemporary translation scholars is that the concept of translation is a broad one that encompasses all sorts of mediation activities (e.g. Dam, Nisbeth Brøgger, and Korning Zethsen 2019; Halverson and Muñoz Martín 2020). Language- and mode-transcending phenomena like audio-description, respeaking and speech-to-text interpreting have become part and parcel of modern translation studies, and today the discipline actively embraces the Jakobsonian concepts beyond interlingual translation – intralingual and intersemiotic translation. The expansion of the territory has been accompanied by a growing recognition of the complexity of the object of study, and paradigms have shifted too. Early objectivist approaches to translation have been replaced by more relativist views (Halverson 2010), and previous reductionist understandings of translation as ‘equivalence-based interlingual transfer’ (van Doorslaer 2021, 3) have given way to approaches that privilege dynamics, variation and transformation.

It has been suggested that while many translation scholars nowadays subscribe to a richer and more nuanced understanding of translation, the world beyond translation studies continues to consider translation ‘as a mechanical process, a word-by-word substitution, a problem of dictionaries’ (Gambier 2016, 887). Such simplistic views are presumed to be dominant among the general public, in other academic disciplines and – importantly in this context – in the translation industry, hence in translation practice (Gambier 2016; van Doorslaer 2019, 2021; Gambier and van Doorslaer 2016; Zwischenberger 2019; see, however; Marais 2022 for an alternative analysis). van Doorslaer (2021) cautiously tags the view of translation as simple linguistic transfer as ‘fairly dominant in the world of translation practice’ (van Doorslaer (2021, 3)), whereas Zwischenberger (2019) goes as far as to talk about ‘a gap’ (Zwischenberger (2019, 263)) between translation studies and translation practice resulting from incompatible views of what translation is. She describes continued scholarly efforts to distance the discipline from ‘simplistic, equivalence-oriented and prescriptive notions of translation, as well as from the myth of translators and interpreters as unproblematic conveyers of an author’s or speaker’s intended messages’ (Zwischenberger (2019, 264)). Yet translation practice, she suggests, still propagates this myth.

In the same vein, the translation industry has been criticised by scholars for introducing concepts that devalue the notion of translation (e.g. Gambier and Munday 2014). Localisation and transcreation are prime examples as they are promoted by industry (and sometimes by scholars too) as ‘more than translation’ (e.g. Schäler 2010; Rike 2013; Jiménez Crespo 2019; Katan 2021; for a discussion of definitions of these and other translational concepts, see; Dam and Korning Zethsen 2019). Some scholars have warned against the promotion of alternative labels that ‘manipulate a very restrictive concept of translation, keeping aspects like cultural adaptation for themselves’ (Pym 2004, 51, on the industry discourse on localisation), thus devaluing the concept of translation. As Schäffner (2012) points out in her discussion of transediting, when such labels are used instead of or even in opposition to the term translation, ‘there is the danger that translation continues to be understood in a narrower sense of a purely word-for-word transfer process’ (Schäffner (212, 881)).

Empirical studies on translation practitioners’ understandings and uses of the concept(s) of translation are few and far between. Previous research on the topic has mainly analysed industry meta-discourse: marketing material drawn from the websites of translation/localisation/transcreation agencies (e.g. Munday in Gambier and Munday 2014, 20; Pedersen 2014) or from the websites of influential professional associations (FIT, AIIC and similar) who are assumed to speak on behalf of translation professions around the globe (e.g. Zwischenberger 2019). By contrast, the research reported in this article draws on data elicited in face-to-face interactions with the actual players of the field, translators and translation project managers. Such data allow us to examine not how the services of translation companies are marketed with a view to increasing sales, nor how professional associations depict translation, but how translation professionals themselves understand and use the concepts they work with every day. As such, the study has the potential of adding nuances to our understandings of the practitioners who inhabit our field of study and who all too often are depicted as proponents of translation as simple word-for-word transfer.

To reiterate, the aim of the study is to explore how practitioners understand the concepts of translation, how they categorise and why they categorise as they do. After a presentation of the methodology in section 2, we address the two first research questions in the results section 3. A discussion of how practitioners’ definitions and categorisations compare with those of academia, in section 4, serves to elucidate the third research question. The concluding remarks in section 5 address the presumed gap between research and practice.

2. Methodology

Similar to the STP-study, data for the current research were mainly collected by means of focus groups, a participant-centred method whose reliance on group interaction among peers allows the researcher to collect rich, qualitative data on group views, understandings and norms – in this case, on the shared attitudes and beliefs (but also differences of opinion) of translation professionals.

2.1. Participants

To mirror the methodology of the STP-study and ensure group homogeneity, two focus groups were conducted: one group with four translators and a second group consisting of three translation project managers. All participants worked for the same translation agency: Comunica.dk – Translations S.L (‘Comunica’ for short). Comunica is based in Spain but, similar to STP, specialises in the Nordic languages, and the owner and managing director is Danish. Like STP, Comunica’s portfolio of clients includes both other agencies and end-clients. Comunica is a small company (a ‘boutique translation agency’, as they call themselves on their website) with twelve staff employees, of whom five are translators and four are project managers. In addition to in-house translators, Comunica draws on a large network of freelance translators. Some of the freelancers are closely linked to the company and have a workspace of their own on the premises, whereas some of the in-house translators frequently work from home, indicating somewhat blurred boundaries between employment categories, not uncommon in small agencies.

The participants were recruited through the managing director, whom we asked to select as many as possible for each group. The only criterion we set for participation was a minimum of one year of experience with professional translation. As Comunica is a small company, the choice was not large. All four project managers were invited and agreed to participate. To match the size of the project-manager group, four translators among those (few) who regularly worked on the premises were recruited for the other focus group. Among the project managers, one fell ill, reducing this group to three participants. All four translators showed up. Group sizes were thus relatively small, but discussions were nevertheless intense and lively.

All four participants in the translator group were female; two were Swedish and two were Danish. They all had an MA in translation, and their experience with professional translation ranged from 2 to 12 years, with an average of 6.4 years. One was a staff translator and three were freelancers who worked mainly or exclusively for Comunica.

The three participants in the project-manager group were also female; two were Spanish and one was French. Their educational backgrounds were more mixed: one had an MA in translation, one had a BA in foreign languages applied to international trade and one had an MA in translation and localisation of new technologies. Their experience with professional translation ranged from 5 to 20 years, with an average of 13.3 years. As indicated, all project managers were staff employed.

2.2. Procedure

The two focus groups were conducted on two consecutive days in March 2022 and took place on the premises of Comunica in Fuengirola, Spain. Apart from the focus-group participants, two researchers (the authors) were present: a moderator who guided the discussions and an observer who took notes and occasionally asked complementary questions.

The moderator opened the discussions with a brief explanation of the project and the ground rules of focus groups, following Morgan (1996, 48–54) (e.g. that she was not there to interview them but to propose topics for them to discuss among themselves, that she was there to learn from them and not the other way around, that all experiences, stories and viewpoints were equally valid, that there were no right or wrong answers, etc.). After that, for ice-breaking purposes, the participants introduced themselves and talked briefly about their professional backgrounds and the tasks and projects they were currently engaged in.

The discussion topics in the interview guide were generally structured according to the funnel model (Morgan 1996, 41–2), going from very open questions that invite participants to discuss freely, to more specific – and sometimes challenging – questions. The first discussion topic was the concept of translation itself, which was addressed in some detail with questions such as: how would you describe/characterise/define translation, what are its core features, how to distinguish it from other types of text production or communication and – not least – which of these adjacent types are NOT translation, and why? The participants were then given some examples of intralingual translation (e.g. expert-to-layman rewriting, UK-to-US conversion) and were asked if these could be considered translation or not. The reason for zooming in on intralingual translation – identified as a borderline translational phenomenon in our previous study – was that this same research had shown that marginal concepts and examples tend to spark particularly intense and highly revealing discussions, not only of the borderline categories themselves but also of the very prototype, laying bare what is considered to be its core traits. After an open, participant-lead discussion of the question of intralingual translation, the moderator adopted a more directive style and presented a series of arguments in favour of considering it a type of translation (in particular, that intra- and interlingual translators draw on the same set of tools and skills, cf. section 4), thus prompting participants to reconsider their initial (dismissive) stances and not least to explain and justify them, which served to further illuminate participants’ categorisation criteria.

The next step was to introduce the prototype model of translation derived from the STP-study (Figure 1) to the groups. A paper copy of the model was placed on the meeting table and, after an explanation of the general tenets of the model, the participants were asked to reflect on and discuss if they agreed with the way the STP-participants had mapped the translation field, if they would organise the concepts differently, discard some, include new ones, and why.

This exercise served as a warm-up to the core component of the focus groups, which was to have the Comunica employees discuss the various concepts of translation and produce their own map of the field. The concepts selected for discussion were: Translation; Localisation; Transcreation; Subtitling; Interpreting; Intralingual translation; Intersemiotic translation; Copywriting; Revision; Proofreading. The selected concepts mirrored those used in the STP-study but were adapted to the new context. Services offered by STP but not by Comunica were for example not included (‘Glossary creation’ and ‘Editing’), and the STP-keyword ‘Subtitling/Dubbing’ was reduced to ‘Subtitling’ as Comunica offers subtitling but not dubbing. Also, what STP sells as ‘Content writing’ was changed to ‘Copywriting’ to reflect Comunica’s term of choice. Each concept was captured on a separate keyword poster, and the 10 posters were placed on the meeting table so that participants could move them around as their discussion developed. They were instructed to place the most central or typical translation concepts in the middle and less central or typical ones further away, on a continuum along the lines of the STP-model (which had been removed for this exercise); they were also invited to discard concepts if they felt some did not belong in a map of translation. To encourage participants to be ‘discursively explicit’ (Halkier 2009, 10 – our translation), they were asked to talk out loud while constructing the model and to motivate their choices. When a group had produced their model – after much negotiation and moving keyword posters around – we invited a participant of the group’s own choice to explain it to us.

A final discussion topic was the usefulness of the many terms and concepts of the translation field, a topic that invited groups to reflect on the advantages and disadvantages of distinguishing between, say, localisation and translation and to discuss whether we could more usefully have just one cover term or concept.

The duration of each focus group was two hours.

2.3. Analyses

The analyses relied on three complementary data sources: (1) audio-recordings of the focus-group sessions, with full transcriptions of key passages (relating to the translation concepts discussed, definitions and categorisations) and content summaries of the rest; (2) observer’s notes; and (3) photos of the models generated by the focus groups. Prototype theory served as the analytical framework, and thematic analysis was used to identify and group all statements concerning the various translation concepts discussed. The analyses were conducted manually and in several iterations by the authors.

3. Results

In this section, we present the main findings, showing how the participants in the study describe the concepts of translation, the criteria by which they categorise and how they map their professional field. Figures 2 and 3 show the models of translation produced by the translators and project managers, respectively. The source of the data extracts that accompany the analyses below is indicated by means of abbreviations: PM for project managers and T for translators. No further differentiation is made to safeguard the anonymity of the participants.

Figure 2. The translators’ model of translation.

Figure 3. The project managers’ model of translation.

3.1. Translation

Both the translators and the project managers agree that translation is basically a transfer that takes place between two languages. It is ‘putting one language into another language’ (T), and as one of the project managers says: ‘For me translation is really, is really to … you have a source text and then you do a target text’ (PM). They also speak about ‘loyalty to the source’ (PM) as particular for translation. They later add that translation is about ‘transferring a message from one language to another’ (T), and that ‘everything that means conveying a message from one source to a target for me is translation’ (PM). In more general terms they speak about translation as ‘communication’ (PM) and ‘helping people to understand each other’ (PM). During the focus groups it becomes clear that the word ‘translation’ is used both as an umbrella term and as a prototype.

Interestingly, they introduce a very pragmatic parameter for categorisation, namely payment: ‘If I get paid, like, per word that’s a translation’ (T), and it is indicated that if creativity is required in a task, then it is not prototypical translation. As one of the translators says about a concept (inaudible which concept, but probably intersemiotic translation): ‘that is not even translation for me – that’s too creative’ (T). However, they acknowledge without hesitation that localisation, transcreation and subtitling are all translational activities. The parameter of creativity seems to be used as synonymous with degree of change in relation to the source text. Technical translation is thus mentioned as the most prototypical type of translation because it is ‘very literal’ (T), and transcreation is seen as creative (see section 3.5), i.e. they operate with a continuum ranging from literal to creative.

There is general agreement, then, that prototypical translation is a loyal written transfer of a message that takes place between two languages, i.e. between a source text and a target text. The task is not too creative, meaning that it is not too far removed from the source text, but is usually complex as it requires deep knowledge about language, culture and subject. In the beginning of the focus groups, the participants are somewhat categorical in their attempt to delimit prototypical translation from other activities, but once they are deeper into discussions they agree that in fact there is no such thing as ‘just’ translation (see section 3.3 on subtitling). From a pragmatic business point of view, they recognise a translation task by being paid by word!

When establishing the prototype model, both translators and project managers place interlingual, written, source-oriented ‘translation proper’ (Jakobson 1959/2012, 114) right at the centre (see Figures 2 and 3).

3.2. Interpreting

Interpreting is not much discussed, but everybody agrees that it is prototypical translation almost on a par with written, interlingual translation, because two languages are involved.

When establishing the prototype model, both translators and project managers place interpreting at the centre because they see it as almost identical with written translation (see Figures 2 and 3). They compare it to translation and say: ‘This is like written and this is oral, but the thing is the same’ (PM). One of the translators then comments that you could also have intralingual interpreting and even though this does not change the prototypical status of interlingual interpreting, they are clearly annoyed by the fact that they cannot necessarily categorise on the basis of the two-languages requirement as they seem to consider if intralingual interpreting does after all belong in the centre or in the more peripheral position of intralingual translation (see section 3.6).

3.3. Subtitling

Although they do not have much experience, if any, with the task, the translators immediately agree that subtitling is translation because of the two-languages criterion. The project managers see subtitling as an emerging market because of increased streaming and the use of videos for marketing. They indicate that subtitling may be more than translation. One of the project managers starts by saying ‘I would consider it [subtitling] kind of translation but you know it is like, you need to add something else’ (PM). They seem to mean that subtitling is not ‘just’ translation, but then they give examples pointing out that this is also the case with other kinds of translation such as for instance medical translation, where you need specialisation, and conclude that there is in fact no such thing as ‘just’ translation. Subtitling clearly remains in the prototypical centre of translation.

When establishing the prototype model, the project managers place subtitling at the prototypical centre (albeit a bit further out than translation, interpreting and localisation), but the translators, who also started by placing it at the centre, end up putting it in the first outer circle together with transcreation because of the creativity involved (see Figures 2 and 3).

3.4. Localisation

Both the translators and the project managers agree that localisation is if not prototypical translation, then very nearly so. According to one of the translators ‘localisation and translation often go hand in hand’ (T), and the project managers expect that all good translators will automatically localise when necessary. Localisation is not seen as an independent task, but rather as part of translation, especially in connection with technical translation. As one project manager says: ‘localisation is everywhere’ (PM).

When establishing the prototype model, the project managers place localisation right at the centre, whereas the translators place it on the border between the centre and the first outer circle (see Figures 2 and 3).

3.5. Transcreation

Both translators and project managers consider transcreation to be translation, although not prototypical translation. It is described as rewriting and interestingly (when considering the fact that they maintain its translational status) they liken it to copywriting and intralingual translation. They agree that creativity is required and therefore ‘not every single linguist can do that’ (PM). In addition, an example is given of technical translators who may not enjoy transcreation due to the creativity. Because of the creativity required, transcreation is considered to be translation with an additional layer and is thus more expensive than ‘a simple translation’ (T). It is paid by the hour, not by word, which is seen as another reason for its more peripheral status.

When establishing the prototype model, both the project managers and the translators place transcreation in the first outer circle due to the extra layer of creativity (see Figures 2 and 3).

3.6. Intralingual translation

When asked about the concept of intralingual translation, the translators instinctively say that it is not translation, but adaptation (or perhaps even revision). They only know the term intralingual from university and seem to call the activity adaptation in practice. However, as the project managers point out, adaptation is not paid by word but by the hour, and this business perspective seems to be an argument for it not being translation (though in the case of transcreation the same fact did not prompt them to consider it non-translation).

The project managers say that the translators are never asked to do intralingual tasks, but their initial reaction is different from that of the translators as some of the project managers acknowledge the translational status of intralingual translation:

I also agree with [xx], you have a point. This is kind of a translation of course. (PM)

Yeahh, I would say it could be translation. (PM)

However, one of the project managers disagrees:

For me it is still not translation per se [it is adaptation]. (PM)

During the focus-group sessions, both the translators and the project managers repeatedly return to the concept of intralingual translation. What clearly bothers them is the fact that when the activity is further exemplified and the similarities with interlingual translation are pointed out by the moderator, they increasingly see the similarities and logically they would consider it translation, but especially the translators do not feel that it is. The following excerpt from the translators’ discussion reflects their doubts and categorisation difficulties:

If we see translation as the transfer of a message from one language to another language, be it a dialect or expert language, then it would fall into the category of translation. (T)

I have never thought about it like that. (T)

It is a good, good question. (T)

‘It makes logical sense, but I am also like, it is not translation’ (T). In the discussion that follows, this translator recognises that it is definitely her ‘gut-feeling’ that makes her say so.

They realise that a crucial point would be how language is defined, but even though they are open towards the idea that language could be more than a national language, they disagree on what would in fact constitute a language:

Even like, between really like different dialects, I could still consider it translation. (T)

For me it [expert-lay] just does not say translation even though it makes perfect sense the way you lay it out. (T)

At some point one of the translators says:

You might not have to be a trained translator [to do intralingual translation]. (T)

And perhaps this explains their ‘gut-feeling’ that it is not translation.

When establishing the prototype model (see Figures 2 and 3), both the project managers and the translators place intralingual translation in the second outer circle, i.e. they end up considering it a translational activity, albeit a peripheral one. The translators place intralingual translation together with revision as it is also within the same language. Their gut-feeling may still be that it is not translation, but as one of the translators explains:

No, it’s just that intralingual translation is just growing on me. (T)

The translators then start discussing whether two languages are involved in the different tasks, and this becomes the main criterion for categorisation, and they seem completely to forget other criteria like transfer of a message and the existence of a source text. They agree that interpreting belongs in the centre because two languages are involved. And then one comments that you could also have intralingual interpreting and, as mentioned above, it clearly annoys them that they cannot categorise solely on the basis of the two-national-languages requirement after all (which would have been convenient since they disagree on a definition of language cf. above).

The project managers place intralingual translation alongside intersemiotic translation, reflecting the fact that they consider both concepts to be the most alien ones among the concepts which they have chosen to include in the translation model. Even if they were more accepting than the translators of the translational status of intralingual translation from the beginning, they still feel that inclusion in the model may be a result of the focus-group discussions and insights:

Maybe, if you had asked us to do it before you talked about intra we would have put it outside, maybe, maybe. (PM)

3.7. Intersemiotic translation, revision, proofreading, copywriting

Both translators and project managers agree that pure proofreading and copywriting do not belong in the prototype model. Proofreading is seen as a natural part of a translation task and of other tasks such as copywriting, but not as a translational task in its own right. Copywriting they see as ‘in the same vein’ as translation (T). The project managers also place revision outside the model, whereas the translators see it as a translational task, but place it in the most peripheral position together with intralingual translation as mentioned above. Intersemiotic translation was hardly discussed during the focus groups, and the translators place it outside the model whereas the project managers think that it belongs in a peripheral position together with intralingual translation, presumably acknowledging the theoretical affinity between the two concepts.

3.8. Synthesis of results and comparison with previous research

The models of translation generated by the two focus groups are reproduced in Figures 2 and 3. As we can see, the two models are very similar. If anything, the project managers tend to be less conservative than the translators: they place more concepts in the central circle of their model, more right at the prototypical centre and are willing to include intersemiotic translation in their map of translation; to the translators, intersemiotic translation is an outlier. On the other hand, the project managers see revision, included in the second outer circle in the translators’ model, as non-translation. The translators’ more accepting view of revision may be explained by the fact that it constitutes a large part of their everyday routines as translators. In other words, the different professional roles of the two groups may play a role for their conceptualisations, and so may their different demographic profiles. As we saw in section 2.1, the project managers have more work experience and more varied educational backgrounds than the translators.

A comparison of the results of the present research, the Comunica-study, and those of our previous research, the STP-study (Dam and Korning Zethsen 2019, cf. section 1.2 and Figure 1), calls for caution because of the different contexts and slightly different methods of the two studies; note also that Figure 1 is our interpretation of the data from the focus groups at STP, whereas Figures 2 and 3 have been generated by the translation professionals at Comunica. However, some salient similarities and differences deserve to be pointed out. First and foremost, there is full agreement between the practitioners at STP and Comunica about what constitutes prototypical translation, namely a rather source-text oriented, written, interlingual translation. There is also consensus that interpreting and subtitling are highly central members of the translation category (though the Comunica-translators are less sure about subtitling than the other groups). Interestingly, both localisation and transcreation are viewed as more central, hence prototypical, in the Comunica-study than in the five years older STP-study, suggesting a gradual movement towards the prototypical centre of newcomer concepts and tasks. The different results for intralingual translation in the two studies (seen as borderline by STP-employees but included in the second outer circle by Comunica-professionals) could well be a result of the different methods used to elicit opinions on this topic: a neutral, non-directive moderator style in the case of STP and a more directive and challenging style in the case of Comunica.

4. Discussion

An important insight from the results is the fact that the translators and project managers clearly categorise in a different way than the scientific community does. Researchers need and are trained in stipulative definitions (see Robinson 2011) with a set of criteria for inclusion in a category, but this may not be how people categorise in practice. In other words, the research community may need a definition which might not be in complete agreement with common usage, but which may serve as the basis of a scientific taxonomy. Hill-Madsen and Korning Zethsen (2016) express the need for a stipulative definition within translation studies, i.e. of the translational phenomena which constitute the field, and suggest a broad, encompassing definition (a slightly adjusted version of Zethsen’s 2007 model), but nevertheless a definition which will indeed specify criteria for membership, not to make it meaningless. The definition is inspired by Toury’s very influential, broad and highly pragmatic definition of translation (Toury 1985, 1995), by Chesterman’s subsequent discussions of this definition, especially of the central concept of ‘similarity’ (Chesterman 1996, 1997, 1998) and by the work of first Wittgenstein on family resemblances (Wittgenstein 1953/1958) and then Tymoczko (1998, 2005) on translation as a cluster concept.

According to this definition, a translation is a text which conforms to the following conditions:

• A source text (verbal or non-verbal) exists or has existed at some point in time.

• The target text has been derived from the source text (resulting in a new product in another language, genre, medium or semiotic system).

• The resulting relationship is one of relevant similarity, which may take many forms depending on the skopos.

(Hill-Madsen and Korning Zethsen 2016, 705 - our emphasis)

As can be seen from our analyses of the Comunica focus groups, the participants do set up criteria for something to be translation which resemble those within a more scholarly definition, namely the existence of a source text, a derivational relationship between the source text and the target text resulting in a product of relevant similarity (the ‘transfer of a message’) in another language. However, they find it harder to accept that apart from ‘another language’ the target text could be another ‘genre, medium or semiotic system’ or that language itself could have a broader meaning than what they refer to as ‘national’ language. As far as ‘relevant similarity’ is concerned, they require a high degree of actual similarity for something to be at the centre of prototypical translation. At the same time, they introduce some extra criteria such as form of payment (by word or by hour) and the degree of literalness and creativity in a task. It is somewhat surprising that they do not categorise more systematically on the basis of the de facto tools of translation required for the translational tasks, cf. e.g. Zethsen (2024) which compares an interlingual and an intralingual translation and documents that the same microstrategies (i.e. both tools and skills) are used by the translators. The focus-group participants draw on additional criteria and they fluctuate between the importance of each criterion for categorisation. For instance, intralingual and interlingual translation are not compared by their similarities and differences in actual fact, but on the basis of the restrictive national-language-criterion and ‘gut-feeling’. However, it is important to note that the criteria applied may make perfect pragmatic sense! To the focus-group participants, the definition of a prototypical translation is as follows:

It has a source text in one national language, the message of which is transferred loyally into another national language, it is non-creative and is paid by the word.

This prototypical definition does not mean that the focus-group participants have an inflexible view on translation, but rather that this is the mental representation on the basis of which they compare and gauge translational tasks. Both translators and project managers are willing to accept that tasks which fall outside this prototypical description may still be a kind of translation in accordance with the above definition, but the further away from the prototype the more the acceptance is based on reflection and not on what they call ‘gut-feeling’. In addition, everybody agrees on the most prototypical kinds of translation, but are much more varied in their opinions when it comes to the more peripheral members of the category. This tendency was already noted by Leech (1981), according to whom people generally agree about what constitutes a prototypical member of a category, whereas disagreement and uncertainty is common when it comes to establishing peripheral members (ibid.: 84).

Our findings can be explained by the different needs of scholars and practitioners respectively. Halliday and Christian (1999) stress the divergence between scientific taxonomies and so-called folk ones. They point out that the scientific approach differs from the folk one not only in the delicacy of categorisation, but also in terms of classificatory criteria:

The move from folk taxonomies towards scientific ones involves both an increase in steps of delicacy and a change in the criteria used for classification. […The change is one] from overt criteria accessible to the naked eye to covert criteria available only through the application of scientific techniques. (Halliday and Christian 1999, 85–86 as cited in Hill-Madsen and Korning Zethsen 2016, 697)

These so-called covert criteria explain the change in the perception of the participants during the focus groups and the task of agreeing on a prototype model. They have ‘gut-feelings’ and a fluctuating set of criteria ‘accessible to the naked eye’ which determine their initial categorisations, and then when they are exposed to research which for instance explains that systematic comparison reveals that intralingual and interlingual translation to a large degree are based on the same set of translational tools (‘covert criteria’), they adjust their perceptions.

However, it is interesting to note that all focus-group participants feel that the various terms, albeit not based on stipulative definitions but much more floating, are very useful indeed! They constantly negotiate the meaning of the various concepts in their interactions with clients and thus there does not seem to be a need for stricter definition, i.e. a stipulative definition. As a matter of fact, such a definition, even a very broad one, might well be much too static, or too broad, to be pragmatically meaningful. Whatever his motives, Robinson (2011) may be right when he says that:

Such resistance of our common instinct to the first requisite of science [to ask what is meant by a certain concept] never dies. The scientist is still saying ‘it depends what you mean by … ’ and the layman is still resenting that answer. (Robinson 2011, 69)

But there may be good reasons for the practitioner (the scholarly layman) to resent the answer! Or put differently, to resist a stipulative definition. During the focus groups it became evident that both the translators and the project managers operate with three different discourses depending on the target group:

• A scholarly discourse which they know of but do not necessarily use (for instance they say of intralingual translation ‘it is a university term’ (PM))

• An internal discourse where they have an in-house understanding of the various concepts. In practice this means that they often use some of the ‘university terms’, but with the meaning which has evolved in-house (or perhaps in interaction with other agencies)

• A discourse used with clients, especially end-clients who do not know any of the translational concepts and who may ask for a translation but really mean something more specific: ‘ah, transcreation’ (PM), as the project manager may think when listening to a client’s description of the task. The interaction with end-clients will thus be translated into in-house terminology.

Again, these three discourses illustrate the flexibility of pragmatic, negotiated definitions. Scholars may be tempted to teach their definitions, concepts and categorisations to the translation industry (and of course there may be much pedagogical gain in such teaching), but we find that the main point in describing these differences between the research community and practice is to create awareness so that we do not compare apples and pears when doing research. Having said that, it was evident that the participants reflected and reached new and revised insights when exposed to scholarly definitions. Prototype theory is extremely useful when explaining concepts in practice, but for the delimitation of a research field we will need stipulative definitions. These can be very broad and inclusive, but still stipulative not to make a research area meaningless.

5. Concluding remarks

The discussion in section 4 centred on what the practitioners in this study conceive of as prototypical translation. Its focus on features like ‘non-creative transfer’ and ‘equivalence’ (‘loyalty’, as these practitioners say) can easily, but erroneously, be taken to reflect the simplistic view of translation that the profession is sometimes criticised for propagating. By way of conclusion, we therefore wish to remind our readers that these features belong to the reign of the prototype. To the practitioners in this study, translation is much more than the prototype. It is an umbrella term, one that embraces a wide variety of tasks: interpreting, subtitling, localisation, transcreation and more, all with different degrees of creativity and transformation and with different constraints and conditions of production. As they point out, there is no such thing as ‘just’ translation. All translation tasks require deep knowledge about language, culture and subject, and even for seemingly straightforward tasks like written translation of technical texts, an additional layer of knowledge or skills is needed (e.g. specialisation for medical translation).

As described in section 1, translation scholars sometimes warn against the proliferation of labels that replace or compete with the notion of translation. To the practitioners in this research, there is no competition. Names like ‘localisation’ or ‘transcreation’ are not considered ‘alternative labels for translation’ (cf. the title of van Doorslaer 2021) but seen as additional names and concepts that serve to complement and in fact enrich the notion of translation, emphasising its multifaceted nature. The many names and concepts enable the professionals at Comunica to navigate the many different tasks they are confronted with every day and, as no small feat, allow them to make very practical business decisions e.g. about pricing. Perhaps most importantly, the variety of labels and concepts testifies to the complexity of their trade.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Helle V. Dam

Helle V. Dam, PhD, is Professor in Interpreting and Translation Studies at Aarhus University, where she directs the master program in conference interpreting and is a founding member of the research program Communication in International Business and the Professions. She is also Adjunct Professor at Ilisimatusarfik, University of Greenland. Her research covers a broad spectrum of topics in both interpreting and translation, with the sociology of translation as a particular area of interest. The translation profession and translators as a social and professional group have been salient themes in her research production. Another topic of interest is the conceptual mapping of translation and translation studies.

Karen Korning Zethsen

Karen Korning Zethsen is Professor of Translation Studies at Aarhus University where she co-heads the research programme Communication in International Business and the Professions. Her primary research interests include translation studies, in particular intralingual translation, as well as health communication, especially expert-lay. She has published extensively in journals such as Target, the Translator, TTR, Meta, Across, Jostrans, Text & Talk, Communication & Medicine, Qualitative Health Communication and the Journal of Pragmatics.