Applying Service-Dominant Logic to Translation Service Provision

Translation is commonly regarded as a service both in translation industry and within Translation Studies (TS), but the question of what makes translation a service has not been widely explored. This conceptual paper looks at nonliterary translation as a service, applying a paradigm of Service-Dominant S-D logic (S-D logic) to the field. Practices in translation service provision are analysed using the Facilities-Transformation-Usage framework (FTU framework), designed on the premises of S-D logic, as a tool. The paper shows that translation practices in general comply with this theoretical perspective, making translation, by definition, a service, and opens a window into the aspects that make it a service. Some current practices in the field do not, however, meet the criteria of an ideal service. These practices are discussed briefly in order to pinpoint, from the service theoretical point of view, where the problems lie.


Introduction
During the first decades of the 21 st century, many studies have focused on translation service provision rather than translations as texts. Studies looking into the social aspects of translation have covered issues such as the status and habitus of translators, rates of pay, working conditions, professional organisations, translators' networks, working and quality control procedures, cooperation, and translators' relationships with other agents in the service provision (Chesterman 2009: 16-18). Translators, interpreters and translation project managers are observed and studied in authentic working situations, providing insights into the complexity and diversity of translation as a service provision from sociological, cognitive and ergonomics perspectives (Risku et al. 2017). Service provision competence -the skills and knowledge needed for successful operation in the translation market, for example to 'analyse and critically review language services and policies and suggest improvement strategies' (EMT 2017: 11) -is also regarded as one of the key goals of translator training (Gambier 2009, EMT 2017, Chodkiewicz 2012. Accordingly, most European translator training institutes with EMT status offer courses that prepare students for professional working life (Pym/Torres-Simόn 2017: 9). However, while several theoretical service-related approaches have been applied to analyse translation services, and aspects of service provision have been incorporated in translator training, studies applying theories of services to translation are scarce. As Koskinen (2019: 146) points out, even though translation is commonly labelled as a service, Translation Studies (TS) literature offers little support for understanding what makes translation a service, calling for studies to fill this gap. This paper strives to answer that call by introducing a theoretical framework from the field of service research for the analysis of translation. The questions addressed in this paper are the following: What is a service? What characterises services? To what extent do practices in the translation field have the characteristics of a service? Looking from a theoretical perspective at the various components of a service provides tools not only for a critical evaluation of the current practices in the field, but also for promoting new practices and enhancing the ability of different stakeholders in the service provision process to identify the aspects and components that they can influence.
This paper focuses on one approach to services presented in the services research -that of Service Dominant logic (S-D logic) (e.g. Vargo/Lusch 2004a, 2004b. This approach is well-established in the field of service research: over the past decade the two foundational articles of S-D Logic have been cited over 13,500 times (Wilden et al. 2017: 345). Furthermore, S-D Logic can be regarded as an appropriate philosophical foundation for the development of service science, an emerging discipline concerned with value aspects in service systems (Maglio/Spohrer 2008, Maglio et al. 2009). In addition to introducing the S-D logic approach, I will present a framework that is designed on its premises by Moeller (2008). This Facilities-Transformation-Usage (FTU) framework provides a conceptual apparatus with which practices in the translation market can be analysed and discussed.
First, the paper provides an overview of service-related studies which have been carried out in TS. This is followed by a brief introduction to service research and different approaches to the concept of 'a service', with a more detailed account of S-D Logic and the FTU framework. In section 4, results from earlier studies into translation practice are interpreted from the perspective of the S-D logic and the FTU framework. To conclude, ideas and a justification for further conceptual and empirical studies into translation service provision from a service theoretical perspective are suggested.

Research into translation practice
This section provides an overview of service-related studies into translation practice from different theoretical standpoints. For a more comprehensive account of studies that look at the professional reality of translators in the workplace from different theoretical frameworks, see Risku et al. (2017Risku et al. ( , 2019b, for example.

Service-related approaches to translation practice
The professional reality of translators in the workplace and the translation service provision process has been studied from various theoretical perspectives. In the context of this paper, studies looking into non-literary translators' working lives are of particular interest, since the paper focuses on the translation industry services. The literary translation market is excluded from this discussion due to its different characteristics as a service provision process, for example in terms of management, agents involved, and remuneration. Studies of translation practice from a sociological perspective have been conducted since the turn of the millennium (Kumiko 2014: 135). The application of Bordieu's sociological concept of habitus on translation has shed light, for example, on the level of autonomy of various agents (Vorderobermeier 2014: 12) as well as on the field of translation as a social structure (Abdallah 2014: 111). Latour's actor-network theory and the notion of agency, defined in the context of translation as 'the willingness and ability to act' (Kinnunen/Koskinen 2010: 6), have also been applied to analyse the social practice of translation (Abdallah 2014: 112; for studies adopting actor-network theory, see Risku et al. 2016: 2). In addition, structures of translation networks and translators' and clients' positions in them in terms of centrality and periphery have been investigated by means of Social Network Analysis . Concepts from the sociology of work have also been applied to translation by Kuznik/Verd (2010), for example. The authors investigated the activities carried out by in-house staff of a Polish translation agency, revealing the variety of tasks and disturbances they handle in the translation service provision process (Kuznik/Verd 2010: 41-42).
Recently, sociological perspectives have intertwined with cognitive approaches to translation practices (Risku et al. 2017). Within these socio-cognitive studies, translation processes have been examined in terms of translation networks, actors and environments, with the aim of uncovering the interaction between translators and their environments and the resulting dependencies (Risku et al. 2014). The concept of situated and embedded cognition is central, highlighting the fact that the translation process is carried out within a network of actors in a dynamic setting (see Risku 2010). Socio-cognitive studies have also touched on the ergonomics of translation. Teixeira and O'Brien (2017), for example, focus on the interaction between ten translators employed by a language service provider and the tools they deploy. Ehrensberger-Dow/Massey (2017), in turn, examine the way in which translators work within socio-technical systems in authentic settings: tools, equipment, and computer interfaces as well as other actors in their professional environment and networks.
Management issues have also raised the interest of translation scholars. Risku (2016) and Risku et al. (2019), for example, report on developments in the management of translation projects, based on the findings of a longitudinal study carried out in an Austrian translation agency in 2001, 2007. Alonso (2016 discusses the digital management of professional translation projects, focusing on conflicts in the management of translation projects, especially due to technology. Pym (2015) looks at risk management in translation, whereas Risku et al (2010) and Risku (2013) apply notions from the field of Knowledge Management, exploring the question of how 'knowledge involved and embedded in professional translation forms a key factor in value creation in organizations' (Risku et al. 2010: 84). Translation as project management has been discussed by Dunne/Dunne (2011), for example.
Translation as a profession and as a service provision have also been described and modelled within TS (e.g. Gouadec 2007, Drugan 2013, Risku 2016. The perspective of service provision is apparent in recent contributions discussing quality as a dynamic rather than a static concept (e.g. Drugan 2013, Gouadec 2010). The quality aspect lies at the core of contributions discussing links between functional translation theories and practices in the translation industry (see Calvo 2018 for an overview), the emphasis being on the notion of skopos and the role of the translation brief in the production of quality work (Calvo 2018). Studies looking at translation from the perspective of economics, in turn, account for changes that have taken place in the field in recent decades, seeking coping strategies for translators working in an increasingly automatised field with industrialised services (e.g. Gambier 2014, Kuznik 2014, Biel/Sosoni 2017, Moorkens 2017).

Service theoretical approaches to translation practices
Studies applying theories from service research to translation are, to date, scarce. Aarikka-Stenroos (2010: 3) analyses translation as a professional service, shedding light on the way translation services are produced, marketed and bought. She suggests reasons why marketing, selling and buying a service may be more complex than doing business with goods, applying the characteristics of services -intangibility, perishability, heterogeneity and inseparability -to translation (Aarikka-Stenroos 2010: 7). Aarikka-Stenroos characterises translation as a knowledge-intensive professional service, which is complex, intangible and highly customised, typically carried out by qualified personnel who have a higher, often academic education (Aarikka-Stenroos 2010: 9, 11). Havumetsä (2012) also discusses characteristics of services in general, introducing ideas from services research and reflecting them to translation service provision, with an emphasis on the notion of quality in services and co-operation as a prerequisite for quality. Havumetsä's focus is on clients' views on quality of non-literary translation, their expectations regarding translators and their own role in the process. Kuznik's (2014) conceptual paper also deals with translation as an economic activity, presenting characteristics of services and the service sector in general, and concluding that the field of translation shares all the characteristics of services in general, as identified by Sadler (1997: 98-102, in Kuznik 2014. For example, translation services cannot be produced and stocked beforehand and can be evaluated only after their production, and service provision requires constant contact with the customer. Koskinen (2019) also links the service perspective to studies of translation, discussing the abstract concept of 'a service'. Depending on the viewpoint it can refer to a standardised, controlled process, such as the delivery of electricity, or as a co-operative, customised process between a client and a service provider -and various other types of product delivery. Koskinen's focus is on the concept of service design and contemplating how translation services could be tailored for both clients and end users alike; hence, she does not offer a theoretical apparatus with which translation as a service could be analysed in detail. This paper contributes to the papers above by introducing such a tool.

Service research
We live in a service society. Very generally, performing a 'service' means that people are paid for the activities they perform for clients (Wharton 2016: 330). Nowadays, services cover a wide variety of activities, from house cleaning to insurance and consultation services, transportation to medical services, and preparing food to designing a house. As pointed out by Wharton (2016: 331), service jobs can be found in any business or industry, and they are highly varied in nature: while some are well-paid professional positions, others are characterised by low wages and little respect. A service, then, is a concept with such a huge variety of characteristics that its definition is still a matter of debate in services research (Moeller 2010: 359, Brax 2013 for an overview of service definitions, see e.g. Brax 2013: 10-32). At the time services established their position as a domain of academic research, the focus of analysis was largely on the characteristics of services that distinguish them from physical goods (Brown et al. 1994, in Brax 2013. Four characteristics of services were identified and considered relevant from the perspective of marketing, for example: intangibility, heterogeneity, inseparability and perishability (often referred to as IHIP characteristics) (e.g. Zeithaml et al. 1985: 33-35). In a contemporary world, however, making a distinction between goods and services is found to be challenging or even impossible, and the approach has been criticised as unsatisfactory (e.g. Lovelock/Gummesson 2004: 32, Moeller 2010: 359, Brax 2013. After the turn of the 21st century, a new paradigm called Service-Dominant logic (S-D logic) (Vargo/Lusch 2004a, 2004b emerged in the field of service marketing, positing that the distinction between goods and services is unnecessary because all business exchange can -and should -be viewed as servicebased (Moeller 2008: 198, Brax 2013. S-D logic (Vargo/Lusch 2004a, 2004b) emphasises the production process rather than its output, thus conceptualising services from a completely different perspective than previously. Instead of the division into services and goods, a distinction is made between direct and indirect service provision (Vargo/Lusch 2004a: 8, 2004b. What is traditionally referred to as goods manufacturing (i.e. indirect service provision) is also considered as a type of service in this paradigm, since goods, too, provide a solution to a specific demand; only the process of delivering the solution is different (Vargo/Lusch 2004a: 9). In what follows, the premises of S-D logic will be discussed in more detail. (Vargo/Lusch 2004a, 2004b provides an alternative to the traditional goods-centred approach to economic exchange (Vargo/Akaka 2009). In the S-D approach, a service is a process that can be provided directly or through a good (Vargo/Akaka 2009: 36). Specialised skills and knowledge (Vargo/Lusch 2004a: 2) are always exchanged between customers and companies, and a service is defined as 'the application of specialized competences (knowledge and skills) through deeds, processes, and performances for the benefit of another entity or the entity itself' (Vargo/Lusch 2004a: 2). S-D logic is built on ten foundational premises (Vargo/Lusch 2004a: 6-12, 2008: 213, Vargo and Akaka 2009. The premises underlie the key ideas and concepts in S-D Logic, of which the most central is value (Vargo/ Lusch 2012: 1).

Service-dominant (S-D) Logic
All ten premises of S-D logic either directly address value or are indirectly linked to some aspect of it. Value is defined as the benefit for some of the actors in the process (Vargo/Lusch 2012: 2), and it is always uniquely and phenomenologically determined by the beneficiary; it cannot be 'standardised'. Moreover, the company cannot create value alone but only offer value propositions: it is the beneficiary who either finds value in the service or does not (Vargo/Lusch 2012: 4─5). Hence, the customer is always a co-creator of value. Quoting Vargo/Lusch (2012: 4─5), 'value creation […] is only possible when market and other offerings are integrated, in context -that is, when it contributes to the wellbeing of some actor in the context of his or her life.' Thus, the notions of co-creation of value, value-in-use and value-in-context are at the heart of this paradigm. Customers and companies collaborate in creating value (Lusch/Vargo 2006, Vargo/Lusch 2004a, Lusch et al. 2007, but value-in-use (or value-in-context) can only be determined by the customer (Vargo/Lusch 2004a).
The concepts of customer resources, co-production and resource integration are also cornerstones of S-D logic (Vargo/Lusch 2004a, Vargo 2008, Vargo/Lusch 2012. Customer resources refer to whatever customers bring to the service production process (Moeller 2008: 201). For a haircut, customers bring themselves; for a house renovation, customers bring their house; for tax advice, customers bring their personal data; for a translation service, customers bring (at least) their documents. Co-production represents the co-operation between the company and the customer in the creation of service (Vargo 2008: 211). Resource integration, in turn, is multi-directional: all parties in the service process can integrate multiple resources for their own benefit, or for the benefit of others (Vargo 2008: 211). Resource integration is paramount to value creation: according to Vargo/Lusch (2012: 36), value is created only when the beneficiary of the service (i.e. the customer) integrates the resources of the service provider with other resources in a specific context.
In S-D logic, all business exchanges are considered as either direct or indirect services (Vargo/ Lusch 2004a: 8, 2004b. In the latter, goods act as distribution mechanism of knowledge, whereas in the first, the application of knowledge is directly rendered via services. Differences between direct and indirect service provision are elaborated by Moeller, who argues that the main difference between the two is in the level of customer integration (Moeller 2008: 200─201). Customer integration, according to Moeller (2008: 202), refers to the combination of customer resources and company resources in order to provide a solution to a demand. A direct service is not possible without customer resources; the customer needs to bring in resources in the production process. To elaborate on the level of customer integration in direct and indirect services, Moeller (2008) proposes the Facilities-Transformation-Usage (FTU) framework, which identifies three stages in the service provision process. Each stage can be investigated from the perspective of resources, decision-making and values. This multi-dimensional approach to service provision provides a conceptual apparatus with which practices in the translation market can be analysed and discussed. The FTU framework will be introduced in the following subsection. Moeller (2008Moeller ( , 2010, building on the ideas of Vargo/Lusch (2004a, 2004b, presents a framework in which the level of customer integration can examined at different stages of service provision from three different perspectives. Moeller identifies three stages of service provision: facilities, transformation and usage (Moeller 2008(Moeller , 2010. Each of these stages can be looked at, first, from the resource perspective, which indicates whose resources are integrated in the service provision process at different points: those of companies, of customers, or both. Second, the stages can be analysed for the interdependency of companies and customers in decision-making. Decisionmaking at different stages can be company autonomous, customer autonomous or integrative (Moeller 2008: 204). Finally, the value perspective on each stage makes explicit when customers act as co-creators or the sole creators of value. This FTU framework is depicted in Table 1 below and each stage is explained in more detail in what follows. Table 1. Stages in service provision according to the FTU framework, adapted from Moeller (2008: 198).

Facilities
Facilities make the service accessible (Mayer et al 2003, in Moeller 2008 and are a prerequisite for a service to take place (Edvardsson et al. 2000). As far as resources are concerned, this stage involves company resources only in both types of service provision; customer resources do not integrate in the process at this stage. Facilities comprise all resources that enable service provision, including machines, employees and know-how. The facilities of present-day translation service providers 1 typically include technological solutions for creating translation memories, aligning texts, managing terminology, assuring quality, accessing and searching electronic corpuses, creating subtitles and managing translation projects (Drugan 2013: 81─102, Risku et al. 2019. The key facility for translation service provision is knowledge (Aarikka-Stenroos 2010, Koskinen 2019, Risku et al. 2010: 84), intellectual capital that is comprised of a service provider's collective knowledge, experience and competences (Risku et al. 2010: 84). This intellectual capital can be divided into human capital (referring to individuals' competences, experiences and knowledge), structural capital (referring, for example, to a company's processes, information systems and data-bases) and relational capital (referring to an organisation's relationships with its customers and stakeholders) (Risku et al. 2010: 85). Human capital as a resource for translation service provisions includes, for example, translators' knowledge of the working languages and of translation, field-specific knowledge, and knowledge related to information search skills and translation tools, whereas structural capital is the knowledge stored in translation memories, terminological databases and automatised processes (Risku et al. 2010: 85). Relational capital is linked to the way service providers and customers conceptualise their own roles and the roles of each other in the translation service provision and how they show trust towards each other in the process : 85, also Abdallah/Koskinen 2007.
According to the FTU Framework, the facilities stage is also company-driven from the perspectives of decision-making and values in both types of service provision. Customers do not interfere with decisions regarding facilities, and the stage exhibits only potential value to customers. No matter what the facilities of a translation service provider, as such they are of no value to the customer.

Transformation
In the FTU framework, the transformation stage appears as the most significant difference between production of direct and indirect services. At this stage, resources become a product. From all three perspectives, this stage in direct service provision is characterised by a higher level of customer integration (Moeller 2008). Accordingly, in direct service provision the stage is called customer-induced transformation, whereas in indirect service provision it is labelled company-induced transformation (Moeller 2008: 202─203). In translation service provision, transformation corresponds to the stages of transfer and quality controls (proofreading, revision) (see Gouadec 2007: 25).
With regard to resources, a company providing indirect services uses its own materials and machinery in the production without any input from the customer, whereas in direct service provision the customer is involved at the transformation stage, acting as a co-producer. A solution to a demand cannot be found without some level of customer participation (Moeller 2008: 203). On this criterion, translation is a direct service, since, at the very least, a source text provided by the customer needs to be integrated in the process. The customer can also provide, for example, specification as to the intended use and purpose of translation (i.e. the skopos, Reiss/ Vermeer 1984), other material related to the source text, a translation memory, style guide, or their knowledge on the subject matter, for co-production of value (Gouadec 2007: 23, 65─66). Transformation in direct services concerns the customers themselves (e.g. massage, surgery, haircut), their possessions (e.g. house renovations), or documents (e.g. translation services), etc. In this type of transformation, the effects are produced directly for the customers: company and customer resources are combined to create the core offering (Moeller 2008: 204).
Decision-making at this stage is company-autonomous in indirect service provision, since companies operate on their own resources, independent of customer preferences or resources (Moeller 2008: 204). In contrast, direct services are characterised by joint decisions at this stage: because customer resources are integrated in the process, the company cannot make autonomous decisions but must take the customer into account (Moeller 2008: 204). On the other hand, the customers are also restricted in their decision-making, since they are dependent on the company who provides the facilities for the transformation. Hence, in customer-induced transformation, companies and customers are dependent on each other in service provision, and they need to make integrative decisions (Moeller 2008: 204). In translation, customers may, for example, have preferences regarding terminology/phraseology, the language variant (British/American English), or the level of quality of the translation.
In terms of the perspective of value, the transformation stage in indirect services only exhibits potential value, which may or may not change into perceived value if the offering is accepted by the customer at the usage stage (Moeller 2008: 205). Customers have no role to play in value creation. In general, direct services also exhibit potential value at this stage, with the difference being that this value is jointly created by the service provider and the customer. In other words, customers purchasing direct services act as co-creators of value (Vargo/Lusch 2004a). In translation, for example, the level and nature of customer integration at the transformation stage combined with the service provider's resources create the potential value of the service. This principle is at the very core of the ISO definition of translation service provision, according to which it is 'the result of interaction between client and a translation service provider' (ISO 17100:2015).
Some direct services can also exhibit perceived value for customers, even at the transformation stage. Value-in-transformation can be perceived in offerings in which customers themselves are the resources which are transformed -such as in spa treatment. Such an offering may also have value-in use due to its recreation effects which continue to benefit the customer after the transformation stage (Moeller 2008: 205).

Usage
The usage stage in indirect service provision refers to a marketable product that is being sold. In other words, goods act as distribution mechanisms for a company-induced transformation (Moeller 2008: 203). The usage stage in direct service provision, in turn, takes place when customer resources exit the company's sphere and customers, or their belongings, are no longer integrated in the transformation process (Moeller 2008: 204). The customer, for example, leaves a beauty parlour or a bank, or receives a translation from the service provider.
In both types of services, the company has done its share from the perspectives of resources, decisions and value creation at this stage: company resources or the (more or less) combined resources have resulted in a product which is now to benefit the customer, and it is up to the customer to make decisions on the usage. Customers perceive value by using goods or outcomes of direct services, turning potential value created by the company alone or jointly by the company and the customer into perceived value-in-use (Moeller 2008: 203).
In translation, the usage stage can be very different in nature, since translations are needed for various purposes: to inform about business (e.g. annual reports), to sell products (e.g. websites, advertisements), to grant people rights in a foreign country (e.g. personal documents), to enlarge the audience of a film (subtitling, dubbing), to promote multilingualism and equal status of languages (EU translation, translation practices regulated by law in countries with more than one official language), to distribute cultural ideas (literal translation), to inform people about the usage of a product (e.g. user manuals), and so on (see e.g. Gouadec 2007: 12─13, Drugan 2013.
Based on the above parallels between the principles of S-D logic, FTU framework and the models of translation service provision presented within TS, translation seems to fall within the category of direct services. In the following section, the focus is shifted from theory to translation practices, and the findings of some service-related studies carried out within TS are reinterpreted and discussed from the service theoretical perspective.

A service theoretical view on practices in translation
In the following, practices in the translation market are reflected on the theoretical account of services presented above. The goal of this section is to bring forth some of those practices in the current translation field that do not meet the criteria of an ideal service, and to analyse, from the service theoretical point of view, where the problems lie. The section is organised around the three stages in the FTU framework: facilities, transformation and usage. Since in theory translation service provision seems to meet the criteria of direct services, practices are reflected on these specific criteria.

Facilities in translation service provision
Translation as a professional field is changing. Several major developments seem to have had an influence on the nature of translation as a service: rapid technological advances, increased pressure on productivity and outsourcing (see e.g. Dunne 2012: 144, Biel/Sosoni 2017, Joorkens 2017, Gambier 2014, which all affect the facilities stage in translation service provision. The trend in outsourcing can be traced back to the 1980s and 1990s, when international hardware and software companies concluded that outsourcing translation is a cost-efficient solution (Dunne 2012: 144). As a result of this, companies providing language services emerged. In this scenario, a customer requiring translation subcontracts the project to an LSP, which can subcontract it to another LSP or to a freelance translator (Dunne 2012: 144─145). Resources of different agents in the translation market vary to a great extent: resources of a large LSP comprises not only the human, structural and relational capital ) of the company itself, but also the knowledge and competencies of the other LSPs and freelancers it co-operates with. Selfemployed translators, in turn, rely on their own resources in economic exchanges with direct customers. When working as a subcontractor for a translation agency, they may benefit from the resources of the agency. From the perspective of resources, customers do not contribute to the facilities stage in service provision (Moeller 2008). In general, this applies to translation service provision as well. However, some recent developments, such as the commercial use of machine translation (MT) solutions, seem to challenge this notion. LSPs of various sizes are offering customised machine translation systems to global enterprises for instant translation of corporate communication. For example, one of the largest global language service providers, SDL, offers 'enterprise-scale machine translation solutions' for 'appropriate quality translations for business content that will enable faster decisions and more effective communication', or for a resource to supplement human translation to achieve 'the best-of-both-worlds outcomes'. An enterprise with a customised machine translation solution can indeed contribute to the facilities stage when looking for services from human translators by providing translators with their MT.
In such a scenario, decision-making at the facilities stage is also integrated, not companyautonomous, as suggested by the FTU framework, since the decision to integrate the MT solution into the process is made by the customer, not the service provider. An example of a customer contributing to the facilities stage is the European Union, in which a customised eTranslation system has been trained using over 1 billion sentences in the 24 official EU languages produced by translators of the EU institutions over recent decades (European Commission). This customer facility is also made available to freelance translators providing services to the EU.

Transformation in translation service provision
This stage in direct services is characterised by resource integration. This applies to translation, too, making translation, by definition, a direct service. Translation service provision is not initiated without the customer sending, at least, the documents to be translated. However, analysing this stage of translation service provision with the theoretical tools presented in this paper is not without problems due to ownership issues, which will be presented below. Following this, the transformation stage in translation service provision is discussed from the perspectives of resources, decision-making and value.

On the ownership of resources
According to the theory, the ownership of resources integrated in the service production is clear: there are customer resources and company resources, which integrate for the co-production of (potential) value (Vargo/Lusch 2004a, Vargo 2008, Vargo/Lusch 2012: 6, Moeller 2008. In translation, however, it is not always clear to whom the resources that are integrated in the service provision belong (Moorkens/Lewis 2019).
The question of ownership concerns the use and reuse of translated material in the form of translation memories (TMs) as well as the material used to train MT systems (Moorkens 2017: 472, Moorkens/Lewis 2019). As a resource, a TM used at the transformation stage can belong to a freelance translator, an LSP or a customer (Smith 2008: 23, in Moorkens/Lewis 2019: 472, Sakamoto et al. 2017). This issue was brought up two decades ago by Topping (2000). However, neither consensus nor regulations as to their ownership exist even today (Moorkens/Lewis 2019). According to Sakamoto et al. (2017), there is no standard practice in the industry with regard to who manages a translation memory after a translation project has finished. In other words, it is not clear whose resource it is after it has been created, and how it is used in further the events of translation service provision. Hence, a TM can potentially become a customer resource in a similar way to customised MT systems.
Moorkens/Lewis (2019: 473─474) also discuss the reuse of existing translations in the training of NMT systems, concluding that tracing the intellectual contribution of a single translator in this process is practically impossible. Thus, many of the resources integrated at the transformation stage are ultimately products of joint efforts, allowing for no straightforward division into customer and company resources. This applies to translation memory and project management tools as well, since many translation projects are carried out from start to end in virtual environments, in which the resource 'no longer belongs to a single entity nor constitutes a domain of fixed installations on individual computers' (Gambier 2014: 3).
The question of whose resources are eventually integrated in the translation provision process is also prompted by studies looking into actor networks in translation. Risku (2014: 344─345), for example, describes a process network in a setting involving a translator and a direct client, showing that what may initially seem as a simple, straightforward setting can turn out to be a complex network of actors. These actors are also likely to provide resources that are integrated in the transformation. Moreover, a translator searching for information during the transformation uses a variety of open resources that can be found, for example, by googling. These resources do not belong to the customer or the service provider, although the knowledge needed to search for information can be attributed to the service provider's resources (see Risku 2010: 89─90) for tools to develop translators' knowledge as a resource).

Resources, decision-making and value
Despite the problems in establishing ownership, various types of resources are integrated at this stage for the co-production of the core offering. Decision are made jointly, and both the customers and the company contribute to the creation of the potential value of a translation service.
The transformation stage in service provision seems delicate and complex due to the following factors: first, the heterogenous nature of resources integrated in co-production; second, the decisions as to which and whose resources should be integrated in order to create value; and third, the possibly conflicting views of what will be perceived as value at the usage stage. All these may, at least partly, be attributed to information asymmetry, characterising professional services such as translation (Aarikka-Stenroos 2010, Havumetsä 2012, Koskinen 2019, Pym 2015: 69─70): customers have less information about the nature of languages, cultures and translation, and the lack of information or misconceptions may, in the worst-case scenario, result in conflicting ideas of value. Therefore, trust between parties is key to functional resource integration and coproduction (see e.g. Abdallah /Koskinen 2007, Pym 2015. The customer needs to trust the translator as an expert of interlingual and intercultural communication, and the translator must trust the customer as an expert in their field.
Information asymmetry can become present at the transformation stage in various ways. Some customers may fail to recognise the role of translation in business communication in general, seeing it as an impractical burden (Risku et al. 2019). Therefore, they do not identify themselves as co-producers of the core offering, and do not respond to queries, or send the source texts in the agreed time (Risku et al. 2019). In some cases, a customer's uncertainty as to what resources to provide can be attributed to their ignorance of translation in general or the variety of translation (Havumetsä 2012, Risku et al. 2019. Risku et al. (2016Risku et al. ( , 2017Risku et al. ( , 2019, for example, find differences between the customer's and the translator's expectations regarding, for example, responsibilities for terminology work and layout, and quality management.
All resource integration and decision-making at the transformation stage aims at value creation. Even the best translator cannot succeed in creating value without the customer's contribution, and on the other hand, even the ideal customer contribution does not suffice without translation expertise. Value -whether it is perceived or not at the usage stage -is created at the transformation stage. Therefore, it is imperative for a service provider to know at the transformation stage what is likely to be perceived as value at the usage stage. The very first indication of this is a translation brief. Hence, a translation brief can be considered a key resource in translation service provision. In practice, it is not uncommon that customers fail to provide the translator with explicit and clear briefs with an explanation of the translation's purpose and use (Havumetsä 2012: 131, Calvo 2018: 20, Alonso 2016. All in all, different ideas of whose resources -the company's or the customer's -should be integrated and to what extent seem to be common in the translation business (Havumetsä 2012, Risku et al. 2019. While the service provider may consider a clear translation brief, readiness to provide resource material, and willingness to answer translators' questions as customer resources that clearly contribute to the value creation, the customer may have other ideas (Gouadec 2010: 270). Based on the results of Havumetsä (2012: 131─132), customers are not always aware of their role as a co-producer of the offering and a co-creator of value: for example, only 40% of the 103 respondents in her survey allow the translator to use existing translations or lists of terms concerning the same subject matter, and 41% do not provide the translator with the name of a contact person for questions about the subject matter.
Some customers also seem to overlook translators' knowledge as a resource. Abdallah (2011: 185), for example, reports that translators as actors in production networks have limited agency, which restricts the way their knowledge is integrated in the process. The networks show a lack of trust between parties (Abdallah/Koskinen 2007, 677─678). The integration of translators' knowledge and their decision-making can also be limited by customers' insistence on relying on a translation memory provided by the customer (Ehrensberger-Dow/Massey 2017), regardless of the context (LeBlanc 2013: 9). When decision-making is shifted away from the translators, they may become demotivated, feeling 'forced to always be online and beholden to the tool imposed by the client' (Gambier 2014: 6).
In principle, co-production in translation service provision should result in value-in-use for customers and/or value-in-use for end users. In practice, knowing what constitutes these and which one is preferred by parties is found to be an obstacle in translation service provision. Customers and service providers have been found to disagree on the question of value (usually referred to as 'quality') (e.g. Abdallah 2010: 22). Some customers may even find the value of translation in transformation rather than in use. In the case of value-in-transformation, customers have the translation done purely out of necessity, for example due to national regulations, failing to see any other value in the service. Reluctant customers in Risku et al.'s study (2016) may represent such an attitude: translations' potential value-in-use for example as a means to enable business communication is ignored.

Usage in translation service provision
At the usage stage in translation service provision, the potential value of the outcome of the translation service -a personal document, a user manual, a patient leaflet, a brochure, a subtitled film, a company's annual report, or any other text produced at the previous stage -is judged by the customer and/or the user. In translation, the customers ordering translation services are often different from the users of translations (e.g. Koskinen 2019). Subtitles are needed by different TV channels and companies distributing films, but they are eventually used by TV viewers and cinema-goers. Websites and brochures are read by the wider public, and novels are consumed by book enthusiasts. In many cases, there is no conflict in the way customers and end users perceive value in the service. Sometimes, however, what the end user perceives as value may differ from the perceptions of the customer.
This dual tie is often recognised by the translator, resulting in conflicts in decision-making at the transformation stage: one might wish to create value for the end user (see Suojanen et al. 2015 on user-centred translation), but is compelled to go along with the wishes of the customer, creating value for them (see Koskinen 2019 for potential reasons for different value perceptions). While a translator may find this irritating (e.g. LeBlanc 2013: 9), it complies with the service theoretical perspective presented in the paper: 'Value, in the end, is customer-determined and cocreated benefit' (Vargo/Lusch 2008: 212, my emphasis).

Special characteristics of translation as service provision
In the above, translation is considered as direct service provision. However, translation service provision can also be envisaged as a process during which customer and service provider resources are integrated into a tangible product -a translated text, a physical object. Written translations, then, can be viewed as 'distribution mechanisms for service provision' (Vargo/Lusch 2004a: 8), in which knowledge and skills are transferred indirectly by embedding them in objects (Vargo/Lusch 2004a: 9). Although this view can be compromised on closer inspection -while translations may be tangible objects, the value of a translation service is not created by the paper that translations are printed on or even by the words as such, but by the knowledge that materialises on the paper (see Vargo/Lusch 2004: 9 for the discussion on goods) -there are certain practices in the translation market that seem to support it.
One of these practices is the established way of pricing translation services, namely the adoption of the word as the base unit of work and cost estimates for translation (Dunne 2012: 151). For a direct service, word-based pricing seems unjustified. While the word can -and isused as a proxy of cost, it is not necessarily a proxy for quality, or value (Dunne 2012: 151). A translation service is not constituted of using a certain amount of characters, words, rows or pages, but of integrating intangible resources for creating value in co-operation with the customer. Yet translation service providers are often paid by the specified number of words -tangibles (Dunne 2012: 152) -for a service that is highly intangible in nature. This discrepancy is, according to Dunne (2012: 152), 'one of the most fundamental problems facing the industry today'. This does not only undermine the nature of translation as a knowledge-intensive service, but also serves to strengthen the information asymmetry between customers and translation service providers.
Word-based pricing may contribute to the 'commodification of translation' (Moorkens 2017: 470). Gambier (2014: 6) envisages a similar regression in the conceptualisation of translation to nothing but a simple, mechanical and countable transfer of text segments, attributing this to the increased use of technological tools in the industry. This line of thinking seems to guide some customers in the industry, for whom price is the differentiating feature between service providers instead of their potential to create value (Dunne 2012: 154, Moorkens 2017. These customers buy words, not knowledge-based services. They, as well as the service providers who are willing to lower their prices only to succeed in biddings, do not appreciate one of the basic premises of S-D logic, according to which 'knowledge is the fundamental source of competitive advantage' (Vargo/Lusch 2004a: 9).

Conclusion
The main goal of this paper was to approach translation from a service theoretical perspective. The S-D logic and the FTU framework provided a useful set of concepts for the analysis, shedding light on the question posed at the beginning of this paper: What makes translation a service? While in general the key features of services such as customer integration, co-production and co-creation of value are readily applicable to translation service provision, the level of customer integration with regard to resources and decision-making in translation service provision may differ from that defined in the FTU framework. This points to the need to look at the nature of translation as a service more closely: Is translation a unique type of service, and if not, what types of services in the vast service sector share its characteristics? In addition, the paper looked at translation practices in order to establish what stages and aspects, in terms of a service theory, are most critical in translation service provision. This paper only scratched the surface, and each stage and aspect of service provision should be looked at in more detail, both theoretically and empirically. For example, the question of resourcesits nature, ownership, origin and integration -is of vital importance, since the complexity of resources appears as a (potentially unique) feature characterising translation as a service. As a result of automatisation, the field is becoming more polarised, and there is demand for highly automatised services as well as for customised, more creative ones (Koskinen 2019: 140). A thorough analysis of resources used for the (co-)production of different translation services would contribute to an understanding of the heterogeneous nature of translation as a service business, making differences between services explicit. Studies incorporating the perspective of value in offerings could shed light on questions such as when and how technological resources (e.g. CAT tools) benefit the service provision, and who the beneficiary is (see Koskinen 2019). Pricing practices in the translation market merit more attention, too, since the current situation seems unsustainable from the service theoretical perspective. This paper focused on written non-literary translation. Other subfields within translation, such as literary translation, interpreting and non-professional translation also deserve attention from the service theoretical perspective. Studies focusing on the variety of service providers as well as customers with regard to their views on issues such as resource integration and value creation might be thought-provoking and beneficial for all parties involved in service provision. The role of end users as beneficiaries would be another point of interest, since the extent to which customers of translation service providers seek to create value for the end users of translations remains unclear. Case studies focusing on individual service providers of different sizes are also needed, to enhance understanding of translation as a service. Hopefully, these and other topics will be taken up in future service-theoretically-oriented studies, and more interest in service approaches would consequently also be shown in students' MA and PhD theses.