3.1Boundaries of the unit of analysis

According to the situated, 4EA paradigm, cognitive processes are embedded in and dependent on their contexts and cultures. Their significance can only be explained and understood as part of that situation and culture – they mean what they mean due to their role in the situated activity and cultural practice (see Robbins and Aydede 2009Robbins, Philip, and Murat Aydede 2009 “A Short Primer on Situated Cognition.” In The Cambridge Handbook of Situated Cognition, edited by Philip Robbins and Murat Aydede, 3–10. New York: Cambridge University Press.). Suchman (2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press., 23) understands being human “as an unfolding, shifting biography of culturally specific experience and relations, inflected for each of us in uniquely particular ways.” However, within the situated, 4EA cluster, the boundaries of the unit of analysis can be defined differently. Studies investigating cognitive processes from an individual perspective look at cognition from the perspective of individual actors and study how they utilise, produce, and interact with elements in their material and social environments. In the case of translation, they would, for example, look at how translators use certain tools, how translators are affected by their social setting, or how the resources available or social interactions in the situation affect the translation process.

From a DC perspective, in contrast, cognition is taken to be distributed across actors, objects, and the environment (i.e., [potentially multiple] agents and the material world). The cognitive process is not carried out solely by the individuals but by the cognitive system of actors, artefacts, and other objects in their present environment. This stance should not be misinterpreted as a disengagement with individual actors: Hollan, Hutchins, and Kirsh (2000Hollan, James, Edwin Hutchins, and David Kirsh 2000 “Distributed Cognition: Towards a New Foundation for Human-Computer Interaction Research.” ACM Transactions on Computer-Human Interaction 7 (2): 174–196. , 179) maintain that the “ethnography of distributed cognitive systems retains an interest in individual minds, but adds to that a focus on the material and social means of the construction of action and meaning.” However, the “system of brains, bodies, and shared environments for action in interaction” (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 376) or the “brain-body-world system” (389) on a social, organisational, or group level is seen as the unit that remembers and acts. All its components contribute to the dynamics of the heterogeneous cognitive system, and these dynamics include short- and long-time developments at a particular time and through time – from momentary reactions and interactions to social, organisational, or group history. As Groleau (2002Groleau, Carole 2002 “Structuration, Situated Action and Distributed Cognition: Rethinking the Computerization of Organizations.” Systèmes d’Information et Management 7 (2). https://​aisel​.aisnet​.org​/sim​/vol7​/iss2​/2, 18) rather metaphorically and semiotically puts it, Hutchins dissolves “the boundaries of the human body to conceptualize cognition as a series of interactions among media located inside and outside the individual’s skin” so that a larger socio-technical unit like an airplane cockpit or a cargo ship can be studied as a cognitive system. According to DC, a complete understanding of internal mental processes would not suffice to explain how remembering works in real-life situations of action (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 378). From the DC perspective, remembering builds on the whole system as actors and artefacts both resort to the traces of previous activities and create traces for future use for all the actors and artefacts in the system. The DC perspective offers a way to expand the unit and timeframe of analysis in our empirical research and thus enables us, for example, to understand the system-level, socio-cognitive dynamics of remembering. The “boundary around the person” becomes “permeable” (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 388).

The DC approach also stresses that the “distribution of cognitive labor can give rise to supraindividual cognitive effects,” in other words, “cognitive properties that are distinct from the cognitive properties of the individuals who compose the group” (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 377). In research on translation, it would focus, for example, on how translation agencies (with their project managers, in-house and freelance translators, revisers, clients, language technology tools, and other technical and communication resources) translate texts; how texts emerge as products of the publisher, its networks, guidelines, and processes; and how technological infrastructures and human contributors in digital translation platforms produce translations. These examples of complex networks as cognitive units are not specific to translation or the language industry but rather just examples of tight webs of interrelationships that are “typical of real-world cognitive ecologies” (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 390).

To summarise, the choice of objects for empirical study differs depending on the perspective taken. Studies of T&I processes from an individual, person-oriented cognitive perspective concentrate on the influence of the local, immediate context of action on the cognitive processes (in the respective individuals), whereas studies looking at T&I processes from a DC perspective choose a social network, a workplace (e.g., an agency, a language department, an office, a hospital unit, a court), or some other larger unit as their empirical setting and analyse the joint dynamics of, and interactions within, this overall system. We do not, however, picture or claim this to be an ideal or idyllic world: just like all cooperation is prone to conflict, the coupled or joint activities in such larger units are not without frictions, asymmetries, and inequalities between humans and between humans and things (see Ehrensberger-Dow and O’Brien 2015Ehrensberger-Dow, Maureen, and Sharon O’Brien 2015 “Ergonomics of the Translation Workplace: Potential for Cognitive Friction.” Translation Spaces 4 (1): 98–118. ). Neither do we envision a dystopia in which individuals have no choice but to subordinate themselves totally to the overall system (though we are aware that social systems like this are a real threat, also in the language industry; see Fırat 2021Fırat, Gökhan 2021 “Uberization of Translation: Impacts on Working Conditions.” The Journal of Internationalization and Localization 8 (1): 48–75. ). On the contrary, DC should not be understood as obscuring or ignoring the individual actors and their agency but as a means of approaching the cognitive aspects of interaction between human actors, between human actors and material artefacts, and between the past and the present, on the systemic and collective levels (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge.).

3.2Structuration of activity

The situated, 4EA approaches share a critical view of predetermined plans and goals and common rules and agreements that are supposed to be valid independent of the context. As Groleau (2002Groleau, Carole 2002 “Structuration, Situated Action and Distributed Cognition: Rethinking the Computerization of Organizations.” Systèmes d’Information et Management 7 (2). https://​aisel​.aisnet​.org​/sim​/vol7​/iss2​/2, 17) notes, “the intelligibility of action stems from the link between action and context, and not from shared conventions.” Plans and schemata do not steer cognition directly (‘online’, peri-actionally) but provide a possibility for ‘offline’ cognition – to plan an action in advance (pre-actionally) and to explain an action afterwards (post-actionally). Goals and plans thus only help to form and understand the action beforehand and afterwards but do not determine and structure the action itself (Suchman 2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press., 13).

For Suchman (2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press., 13), putting plans (schemes, scripts, etc.) in the centre of our explanation of cognition is a deeply cultural phenomenon:

A central feature of planning […] is that it is among the many everyday practices that we, as participants in Euro-American cultural traditions at least, call out as a foundation of the rationality of our actions. The planned character of our actions is not, in this sense, inherent but is demonstrably achieved.

For observers/others watching from a distance, the activity might seem planned and appear to follow an a priori charted course, but the cognitive process itself is situated and contingent, depending on the cues and resources in the specific context.

Apart from the common situated orientation within the 4EA approaches, we can, again, distinguish different perspectives on the forces that structure intelligent activity. From the individual, person-centred perspective, the respective cogniser navigates in their present environment, depending on the cues and resources it provides. From the distributed perspective, pre-actional plans and post-actional explanations can be positioned on the system level. The socio-cognitive system of actors, objects, and their environment as a whole seems from a distance to act in a goal-oriented way, as the system follows its situated and contingent dynamics, adapts to changes, and is transformed by new elements. The heterogeneous elements in this system all contribute to its activity and dynamics. In activities that involve several actors, “features of a shared task world can contribute to the establishment and maintenance of common ground” (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 385). Through repeated interactions and conversations, team members, for example, contribute to an understanding of the situation “superseding any prior individual understanding” (Groleau 2002Groleau, Carole 2002 “Structuration, Situated Action and Distributed Cognition: Rethinking the Computerization of Organizations.” Systèmes d’Information et Management 7 (2). https://​aisel​.aisnet​.org​/sim​/vol7​/iss2​/2, 26). Depending on the overlap of the individual and the systemic understanding, this can be a highly rewarding and energising process or a constraining and frictional one. On the systemic level, the continuous joint flow of interaction and knowledge construction allows for a certain flexibility and resilience of the team as a cognitive system and helps it to deal with the absence of team members or introduction of new team members (Groleau 2002Groleau, Carole 2002 “Structuration, Situated Action and Distributed Cognition: Rethinking the Computerization of Organizations.” Systèmes d’Information et Management 7 (2). https://​aisel​.aisnet​.org​/sim​/vol7​/iss2​/2, 26).

3.3Role of artefacts

The 4EA approaches generally emphasise the significance of artefacts for cognition. Artefacts are not necessarily or primarily seen as amplifiers of any existing individual cognitive capacity (such as memory or visual perception) but rather change the nature of the cognitive process (Hutchins 1999Hutchins, Edwin 1999 “Cognitive Artifacts.” In The MIT Encyclopedia of the Cognitive Sciences, edited by Robert A. Wilson and Frank C. Kiel, 126–127. Cambridge: MIT Press.). From an individual perspective, they reconfigure the cognitive space and task for the human being who is performing the action (Suchman 2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press.). From a systemic perspective, they transform the whole cognitive system (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 378). Menary and Gillett (2022Menary, Richard, and Alexander Gillett 2022 “The Tools of Enculturation.” Topics in Cognitive Science 14: 363–387. , 366) further emphasise that “one cannot consider the cognitive tool, the physical symbols, in separation from the practices, which govern their manipulation. Cognitive tools are to be thought of conjointly with cognitive practices.” Introducing a specific artefact into a cognitive system potentially results in a different set of components and relationships, and thus in a different system dynamic altogether.

Accordingly, despite the common emphasis of the significance of artefacts in the 4EA cluster, there can be differences in the way their role is described. From an individual perspective, they can be described as external to the cognitive system (i.e., the individual in their situation of action). Artefacts like technologies form part of the context of action and cognition, so that “technology is one of many elements of the context having an influence on its unfolding” (Groleau 2002Groleau, Carole 2002 “Structuration, Situated Action and Distributed Cognition: Rethinking the Computerization of Organizations.” Systèmes d’Information et Management 7 (2). https://​aisel​.aisnet​.org​/sim​/vol7​/iss2​/2, 17). They are resources to which the cognitive unit of the individual actor can resort. They provide and are used as “scaffolds” (Clark 1997Clark, Andy 1997 Being There: Putting Brain, Body, and World Together Again. Cambridge: MIT Press., 45) – reliable environmental properties that can be exploited and leaned on to structure activity.

We can also differentiate from an individual perspective between subjects and objects. By dividing artefacts into objects and cognitive processes into subjects, Suchman (2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press., 269) explicitly seeks to retain the asymmetry or “dissymmetry” between persons and artefacts. She argues that the scholarly stress on granting artefacts agency “arose from well-founded concerns to recover for the social sciences and humanities aspects of the lived world” (ibid.) and to restore the role of technology. However, her project stems from the area of technoscience and engineering where the point of departure is reversed, with technology in the centre and human and social aspects at the margins: “It is the privileged machine in this context that creates its marginalized human others” (Suchman 2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press., 270). This is why she stresses the differences and boundary between humans and non-humans and thus puts the human individual in the centre of the cognitive process.

From a distributed perspective, however, artefacts form part of the cognitive unit itself. Since this unit is composed of actors and the material world, artefacts are, strictly speaking, not external to the cognitive system but integral components thereof. In fact, Garbis (2002Garbis, Christer 2002 The Cognitive Use of Artifacts in Cooperative Process Management. Rescue Management and Underground Line Control. PhD diss., Linköping University., 50) explains that DC “attempts to move beyond the initial concept of internalization and dissolve the boundary between internal and external, between the individual and the social.”

3.4Role of representation(s)

Since the 4EA cluster incorporates different positions on the concept of representations, not all of them can be categorised as non-representational views on cognition such as that advocated, for example, by Chemero (2009)Chemero, Anthony 2009 Radical Embodied Cognitive Science. Cambridge: MIT Press. . Situated, 4EA approaches typically assume “local and action-oriented rather than objective and action-independent” representations (Clark 1997Clark, Andy 1997 Being There: Putting Brain, Body, and World Together Again. Cambridge: MIT Press., 150). Contrary to its predecessors in cognitive science, situated cognition does not describe representations as units that actually represent ‘reality’ (e.g., in a propositional manner, in order to be able to process them by mental rules/algorithms) but rather as patterns generated by one’s own actions (without any representational quality or necessity). They are always constructed partly anew in the process of their activation, depending on the situation, and are thus always at least partly unpredictable.

In the situated, 4EA cluster, representations have been described as residing inside and outside the brain. Gestures, facial expressions, speech, tools, visuals, space, body orientation, and other material and immaterial (ephemeral), persistent, contingent, and provisional structures can all be representations on which the cognitive process unfolds (Hutchins 2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 384). We use them as representations that facilitate memory, coordination, and communication.

The DC perspective assumes assemblages of human and non-human elements that function together as a system. According to Hutchins (2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 389), representations in artefacts, bodies, and brains play an important role here. In his well-known study of an airline cockpit, Hutchins (1995b)Hutchins, Edwin 1995b “How a Cockpit Remembers Its Speeds.” Cognitive Science 19 (3): 265–288. observed how the planned speed of the airplane is calculated before landing. Whereas a person-centred perspective would focus on how the pilots remembered what speed to aim for using the instruments in the cockpit, Hutchins depicts the cockpit as a whole as a distributed cognitive system which functions through the transfer and transformation of representations. As Dahlbäck, Rambusch, and Susi (2012Dahlbäck, Nils, Jana Rambusch, and Tarja Susi 2012 “Distribuerad kognition [Distributed cognition].” In Kognitionsvetenskap [Cognitive science], edited by Jens Allwood and Mikael Jensen, 487–496. Lund: Studentlitteratur., 492; our translation) explain, “in this cognitive system, consisting of the pilots and the artefacts in the cockpit, there are a number of representations and media through which they are transferred and transformed.” They further specify that “some of the representations are observable (e.g., instruments, manuals, utterances from the pilots) whereas others are not (e.g., the pilots’ memory)” (ibid.; our translation). Despite the obvious differences in terms of activities between a cockpit and a translation agency, we could still assume that an analysis of the content of workplace documents, to-do lists on desktops, information in project management systems, and how human actors interact with such resources can provide insights into how the larger unit of a team or workplace operates.

4.1Empirical applications of situated approaches in CTIS

Following the conceptual work done by Risku (2010)Risku, Hanna 2010 “A Cognitive Scientific View on Technical Communication and Translation: Do Embodiment and Situatedness Really Make a Difference?” Target 22 (1): 94–111. and Muñoz (2010)Muñoz Martín, Ricardo 2010 “On Paradigms and Cognitive Translatology.” In Translation and Cognition, edited by Gregory Shreve and Erik Angelone, 169–187. Amsterdam: John Benjamins. , there has been an increase in empirical work that uses situated approaches as its theoretical point of departure. In addition to the research on cognitive ergonomic aspects of translation mentioned in Section 4, situated studies addressing social interaction and technology use in translation workplaces have also emerged in recent years, a selection of which are briefly presented in this section.

As can be concluded from several publications (e.g., Risku 2002Risku, Hanna 2002 “Situatedness in Translation Studies.” Cognitive Systems Research 3 (3): 523–533. , 2010Risku, Hanna 2010 “A Cognitive Scientific View on Technical Communication and Translation: Do Embodiment and Situatedness Really Make a Difference?” Target 22 (1): 94–111. , 2014Risku, Hanna 2014 “Translation Process Research as Interaction Research: From Mental to Socio-Cognitive Processes.” In Minding Translation/Con la traducción en mente, edited by Ricardo Muñoz Martín, special issue of MonTI 1: 331–353. ; Hirvonen and Tiittula 2018Hirvonen, Maija, and Liisa Tiittula 2018 “How are Translations Created? Using Multimodal Conversation Analysis to Study a Team Translation Process.” Linguistica Antverpiensa, New Series: Themes in Translation Studies 17: 157–173.; Korhonen and Hirvonen 2021Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. ; Pleijel 2021Pleijel, Richard 2021 “Translation Teams as Cognitive Systems: Archival Material, Cognitive Artifacts, and Group-Level Cognitive Processes.” Cognitive Linguistic Studies 8 (2): 307–327. ; Sannholm 2021Sannholm, Raphael 2021 Translation, Teamwork, and Technology: The Use of Social and Material Scaffolds in the Translation Process. PhD diss. Stockholm University.), 4EA approaches entail a specific interest in interactions of different sorts. The notion of interaction is made relevant from several different perspectives. It not only concerns interaction between social actors and with various artefacts (in the translation process), it is also used to depict the very nature of cognition as perceived from a 4EA perspective. One case where cognition in the translation process is approached not (solely) as the internal manipulation of mental representations but as the result of interaction between agents and environments is presented by Risku (2014)Risku, Hanna 2014 “Translation Process Research as Interaction Research: From Mental to Socio-Cognitive Processes.” In Minding Translation/Con la traducción en mente, edited by Ricardo Muñoz Martín, special issue of MonTI 1: 331–353. , who analyses a translator’s interaction with the material environment at her workplace. A key observation in this study is that the translator manages different tasks and difficulties by interactively “re-configuring the cognitive space” (Risku 2014Risku, Hanna 2014 “Translation Process Research as Interaction Research: From Mental to Socio-Cognitive Processes.” In Minding Translation/Con la traducción en mente, edited by Ricardo Muñoz Martín, special issue of MonTI 1: 331–353. , 348) and interacting “with self-produced stimuli” (347). Such stimuli include the emerging target text as well as tentative formulations which are voiced aloud. As the translator in question is not interacting with any social others, such actions may be seen as modifications of the immediate environment, conceivably supporting further actions.

Martín de León and Fernández Santana (2021)Martín de León, Celia, and Alba Fernández Santana 2021 “Embodied Cognition in the Booth: Referential and Pragmatic Gestures in Simultaneous Interpreting.” Cognitive Linguistic Studies 8 (2): 277–306. , who apply the embodied cognition perspective in an analysis of the work of simultaneous interpreters, also focus on interactions that can conceivably be seen as modifications of the immediate environment. Observing the different ways in which interpreters use bodily gestures during the interpreting process (and with reference to the research tradition of studying gestures from the perspective of embodied understanding), they suggest that interpreters use gestures to facilitate comprehension and the construction of meaning, which can be related to Risku’s (2014Risku, Hanna 2014 “Translation Process Research as Interaction Research: From Mental to Socio-Cognitive Processes.” In Minding Translation/Con la traducción en mente, edited by Ricardo Muñoz Martín, special issue of MonTI 1: 331–353. , 347) conception of “self-produced stimuli.” Importantly, the interpreters in Martín de León and Fernández Santana’s study do not interact directly with other social actors, which means that the gestures are not used in this case in social interaction as the setup was artificial: the speaker appeared on a video recording, and there was no audience (besides the researchers). They conclude that simultaneous interpreting can be seen as a “multimodal, embodied cognitive activity […] that can extend beyond the interpreter’s organism” (Martín de León and Fernández Santana 2021Martín de León, Celia, and Alba Fernández Santana 2021 “Embodied Cognition in the Booth: Referential and Pragmatic Gestures in Simultaneous Interpreting.” Cognitive Linguistic Studies 8 (2): 277–306. , 302).

In Hirvonen and Tiittula (2018)Hirvonen, Maija, and Liisa Tiittula 2018 “How are Translations Created? Using Multimodal Conversation Analysis to Study a Team Translation Process.” Linguistica Antverpiensa, New Series: Themes in Translation Studies 17: 157–173. and Korhonen and Hirvonen (2021)Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. , the interaction studied is primarily of a social kind. In their studies of audio-description (Hirvonen and Tiittula 2018Hirvonen, Maija, and Liisa Tiittula 2018 “How are Translations Created? Using Multimodal Conversation Analysis to Study a Team Translation Process.” Linguistica Antverpiensa, New Series: Themes in Translation Studies 17: 157–173.) and the comparison of audio-description and commercial specialised translation (Korhonen and Hirvonen 2021Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. ), the authors look mainly at how verbal contributions and responses can be seen as constitutive of collaborative processes in which translation problems are jointly addressed by the different actors. The focus in these studies is on the co-construction of translations between different social actors working together.

Sannholm (2021)Sannholm, Raphael 2021 Translation, Teamwork, and Technology: The Use of Social and Material Scaffolds in the Translation Process. PhD diss. Stockholm University., in turn, focuses on the interactive co-construction of solutions to translation problems as it unfolds in digital communication between translators working for a language service provider and on how such interaction intersects with the use of material resources. He observes that translators, when seeking assistance from colleagues, perform actions to narrow and clarify the frames of problematic source text items. This includes specifying the material resources that have been consulted prior to reaching out for help.

Pleijel (2021)Pleijel, Richard 2021 “Translation Teams as Cognitive Systems: Archival Material, Cognitive Artifacts, and Group-Level Cognitive Processes.” Cognitive Linguistic Studies 8 (2): 307–327. discusses the concept of interaction in relation to collaboration in the translation process. Based on a retrospective reflection on collaboration between different actors involved in the three-decade process of retranslating the Bible into Swedish, he contends that different versions of the emerging target texts and adjacent comments can be understood as cognitive artefacts which communicate the residue of previous decisions and instances of reasoning within the teams. The notion that interim versions of target text items serve to mediate representations within a cognitive system of different actors resonates with the points made by Korhonen and Hirvonen (2021)Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. (see also Sannholm 2021Sannholm, Raphael 2021 Translation, Teamwork, and Technology: The Use of Social and Material Scaffolds in the Translation Process. PhD diss. Stockholm University., 103). Pleijel (2021)Pleijel, Richard 2021 “Translation Teams as Cognitive Systems: Archival Material, Cognitive Artifacts, and Group-Level Cognitive Processes.” Cognitive Linguistic Studies 8 (2): 307–327. also suggests that the translation teams in his study can be viewed as cognitive systems, as the target texts can be seen to result from their joint actions rather than from the work of any one particular individual. In other words, the role of interaction is traced retrospectively by focusing on the ways in which the work of translation team members draws on previous work in the form of target text versions and comments.

Although there are some studies that allude to distributed cognition in their titles and elsewhere (e.g., Dragsted 2008Dragsted, Barbara 2008 “Computer-Aided Translation as a Distributed Cognitive Task.” In Cognition Distributed: How Cognitive Technology Extends Our Minds, edited by Itiel E. Dror and Stevan Harnad, 238–256. Amsterdam: John Benjamins. ; Nurminen 2020Nurminen, Mary 2020 “Raw Machine Translation Use by Patent Professionals: A Case of Distributed Cognition.” Translation, Cognition & Behavior 3 (1): 100–121. ), CTIS publications which explicitly apply DC as a theoretical perspective are rarer than those using individual-centred cognitive perspectives (see Korhonen and Hirvonen 2021Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. , 254–255). References to DC can also be detected in the use of certain terminologies and conclusions drawn from analyses of empirical data (as is the case in Pleijel [2021]Pleijel, Richard 2021 “Translation Teams as Cognitive Systems: Archival Material, Cognitive Artifacts, and Group-Level Cognitive Processes.” Cognitive Linguistic Studies 8 (2): 307–327. ).

The study presented by Korhonen and Hirvonen (2021)Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. draws on the work of Hutchins (1995aHutchins, Edwin 1995a Cognition in the Wild. Cambridge: MIT Press., 1995bHutchins, Edwin 1995b “How a Cockpit Remembers Its Speeds.” Cognitive Science 19 (3): 265–288. ) and approaches interaction between different actors in the translation process from the DC perspective. The authors emphasise the social perspective and label their theoretical lens as socially distributed cognition. Similarly to Pleijel (2021)Pleijel, Richard 2021 “Translation Teams as Cognitive Systems: Archival Material, Cognitive Artifacts, and Group-Level Cognitive Processes.” Cognitive Linguistic Studies 8 (2): 307–327. , they posit that the different actors in the translation process can be seen as a single cognitive system which functions by means of communication (e.g., in the form of draft translations). Going further, they consider how creativity emerges within this overarching cognitive system.

Sannholm (2021)Sannholm, Raphael 2021 Translation, Teamwork, and Technology: The Use of Social and Material Scaffolds in the Translation Process. PhD diss. Stockholm University. also uses the notion of the cognitive system to approach interaction in the translation process but makes a distinction between individual and distributed cognitive perspectives based on the delimitation of the object of study. Focusing on the interactions of individual translators with their environment (e.g., for problem solving), the boundaries of the object of study are taken to encompass the individual plus the respective social and material resources (cf. Korhonen and Hirvonen 2021Korhonen, Annamari, and Maija Hirvonen 2021 “Joint Creative Process in Translation: Socially Distributed Cognition in Two Production Contexts.” Cognitive Linguistic Studies 8 (2): 251–276. , 254–255). However, for certain processes like remembering, Sannholm (2021)Sannholm, Raphael 2021 Translation, Teamwork, and Technology: The Use of Social and Material Scaffolds in the Translation Process. PhD diss. Stockholm University. suggests that the boundaries of the cognitive system may be productively drawn around larger constellations such as translation teams.

5.1Empirical extract

The empirical setting for the conversation in question is the Swedish translation office of a global language service provider, where a team of thirty translators and other staff worked at the time of data collection in 2017. The translators were divided into three teams, each of which focused on a specific set of clients based on the subject matter of the translation jobs (medical technology, general technology, and information technology). The translators featured in this extract form part of the general technology team. Elsa is fairly new to the translation profession (one and a half years) and this workplace (ten months), while Hugo has worked as a translator for sixteen years, fifteen of which have been with this particular language service provider. At the time of data collection, remote work was still rare, and the translators were generally co-present in the office premises. Nevertheless, in addition to interacting face-to-face in the workplace, they also communicated via email and IM.

The main activity in the IM conversation displayed in Table 1 could be constructed as an attempt to agree on the interpretation of a technical source text term. The conversation starts as Elsa reaches out to Hugo for assistance on what appears to be the interpretation of a specific source text item (line 4).

5.2Analysis

We will now use this extract to discuss the analytical implications of using the individual and the distributed perspectives when analysing interaction in the translation process. In doing so, we will structure the discussion along the aspects considered in the previous sections of the article: (1) boundaries of the unit of analysis, (2) structuration of activity, (3) role of artefacts, and (4) role of representations.

5.2.1Boundaries of the unit of analysis

Approaching interactions of the kind displayed in Table 1 from the individual perspective means taking the individual translator as the point of departure. In this scenario, the analysis would focus on Elsa’s interactions with different resources and the way she interactively and discursively modifies the conditions for obtaining relevant input from her colleague. Such modifications can be discerned not only in the way the problematic source text item is contextualised in the conversation, which includes naming the client (line 2), but also in the statement that material digital resources have been consulted and deemed unhelpful (line 6). In other words, it can be inferred that some form of interaction with material resources (specifically the translation memory [TM] and glossary) preceded the action of reaching out to a colleague. Looking at how the social interaction unfolds, it can also be assumed that the meaning ascribed to these statements by Hugo is dependent on the situated and cultural embedding of the activity and a “mutual intelligibility” (Suchman 2007Suchman, Lucy 2007 Human-Machine Reconfigurations: Plans and Situated Actions. 2nd ed. Cambridge: Cambridge University Press., 63) which results from situated interpretation, shared experiences, and the recurring nature of the activity. As a simple illustration, it is not far-fetched to assume that a statement like ‘there is nothing in the memory or the glossary’ (line 6) only makes sense in the given situation and, more specifically, in relation to the preceding remarks, which have served to identify and contextualise the difficulty with the source text.

From the individual perspective, the analysis thus focuses on Elsa’s interactions with Hugo and her implied and actual use of different artefacts. In this scenario, it is still the individual translator’s cognitive processes that take centre stage but – and contrary to traditional conceptions of cognition – these processes unfold interactively as she dynamically navigates through the task at hand.

When approaching the interaction from the distributed perspective, the boundaries of the unit of analysis could be extended to both actors and the material resources involved. It is important to note, however, that these boundaries should not be seen as a given. Indeed, Hutchins (2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 376) states that “rather than assuming a boundary for the unit of analysis a priori, distributed cognition […] attempts to put boundaries on its unit of analysis in ways that do not leave important things unexplained or unexplainable.” Thus, if the research interest concerns, for example, how remembering works in a translation workplace, the material resources surfacing in the interaction in the empirical extract are not reduced to external aids which may support individual memory. They are instead seen as co-constitutive of the cognitive system that does the remembering, which in DC terms happens through communications within the system, both by means of face-to-face and direct digital interaction between people (e.g., via IM, as in the extract) as well as through the use of material resources such as digital systems and documents.

5.2.2Structuration of activity

With regard to the structuration of activity as seen from the individual perspective, the individual actor navigates the present environment using the resources it provides. In our empirical extract, we can see that Elsa has not only sought support from situation-specific material resources, but also pursues different paths towards a possible target text solution by responding to the various resources and responses she receives as the interaction unfolds. For example, she formulates assumptions about the meaning of the source text item (line 15) in response to the introduction of information from the website (line 14) and reconsiders these assumptions (lines 23–24) based on the assumption formulated by Hugo (line 22). Thus, rather than being indicative of a preconceived ‘plan’ for reaching a solution to the problem posed by the source text item, the interaction suggests a dynamic interaction with situation-specific resources as assumptions about meaning emerge and are reformulated.

For the translation team, of which Elsa and Hugo form part, a conceivable goal is to deliver translated texts of a certain quality that (at the very least) do not attract negative attention from reviewers and clients. Taking the distributed perspective, both the disclosure of unhelpful resources (line 6) and Hugo’s insistence that textual and TM resources are to be used (lines 9, 13, 14, and 19) could be taken as indications of how the task at hand – to be performed in this instance by Elsa – unfolds at the system level. Some elements in the interaction are also indicative of previous activities of a similar kind. For example, terms such as “minnet” ‘the memory’ (line 6), “glossaryt” ‘the glossary’ (line 6), and “konkordanssöka” ‘do a concordance search’ (line 19) are used and responded to in a way that suggests previous usage and common ground within the team. Again, as Hutchins (2006Hutchins, Edwin 2006 “The Distributed Cognition Perspective on Human Interaction.” In Roots of Human Sociality: Culture, Cognition and Interaction, edited by Nick J. Enfield and Stephen C. Levinson, 375–398. Abingdon: Routledge., 385) points out, “features of a shared task world can contribute to the establishment and maintenance of common ground.”

5.2.3The role of artefacts

Even a short extract of interaction in the translation process like this allows us to formulate assumptions about the role of artefacts in the translation process, as we can observe how the use of resources like TM systems, glossaries, the source text, and other textual resources surface throughout the conversation, both directly and in the participants’ references to them. Elsa’s efforts to construct an interpretation of the source text item proceed in part through interactions with textual content in digital resources introduced by Hugo, some of which are rejected as unhelpful (lines 6 and 12).

In a similar manner, the distributed perspective would pay attention to translators’ interactions with each other in relation to TMs and other technological resources. However, when considering the team as the object of study, potential research foci would also concern, for example, how information discrepancies in digital systems are managed and how information is maintained, manipulated, and disseminated across resources in the organisation. Such activities imply considerations of different time scales as translators act not only to establish support in relation to immediate needs but also to ensure the long-term accuracy and reliability of information in the organisation. While such actions are not present in the empirical extract above, the fact that the participants orient towards the TM system in different ways is suggestive of the importance ascribed to this resource (Elsa’s evaluation of the TM and glossary [line 6]; Hugo’s suggestion for how to overcome the failure to retrieve TM matches using the concordance search feature [line 19]).

5.2.4The role of representations

When examining this interaction from the individual and distributed perspectives, the conceptualisation of representations differs. From the individual perspective, situation-bound and contingent representations could be assumed to be at work in (and result from) the discursive actions. Indeed, Elsa’s gradual construction of interpretations of the source text item clearly emerges as the interaction proceeds (lines 15, 18, 23, and 24) in response to Hugo’s actions (lines 13, 14, 17, and 22). However, actions in the present also draw on actions in the past, and representations are likely to comprise elements of previous experiences, which in the situation in question encompass numerous aspects related to the client, their text types, guidelines, particular restrictions, etc. For example, the assumption about source text meaning formulated by Elsa (line 15) conceivably arises partly from previous experience and knowledge (of the client, language, translation, etc.) and partly from the action-generated input mediated discursively (lines 13 and 14).

Turning to the distributed perspective, Hutchins and Palen (1997Hutchins, Edwin, and Leysia Palen 1997 “Constructing Meaning from Space, Gesture, and Speech.” In Discourse, Tools, and Reasoning: Essays on Situated Cognition, edited by Lauren B. Resnick, Roger Säljö, Clotilde Pontecorvo, and Barbara Burge, 23–40. Berlin: Springer. , 24; emphasis in original) hold that “communicative behaviors are the representations by which a socially distributed cognitive system does its work.” From this perspective, the analysis looks at communications within the system. In this empirical extract, the communication includes information generated through online searches. Although the current empirical data does not reveal how the online content subsequently introduced into the IM conversation was retrieved, it conceivably resulted from the input of search terms into a search engine. From the distributed perspective, the analysis would ideally examine how communicative trajectories unfold – for example, how negotiations between social actors lead to the storage of content in digital resources, which may serve as more or less permanent representations which function as meaning-making resources in later situations for the same or other social actors.