Investigating Real Work Situations in Translation Agencies. Work Content and Its Components

The aim of this article 1 is to analyze work content and its components in translation agencies. In the conceptual part of the article, we refer to concepts taken from the sociology of work and translation studies. In the analytical part, we use data produced by Stelmach in her study carried out in a small translation agency in Poland using the technique of self-observation (Stelmach 2000). The aim of Stelmach's study was to record and analyze all the activities that form part of the production process of a translation service. Because the observation was continual in time, it provided a complete list of all the activities carried out by the permanent staff occupying two internal jobs. Stelmach's approach was quantitative and not focused on specifi c, translation-related work organization. In this article, we reinterpret these activities as the content of the work in translation-related internal positions, and compare it with Gouadec's model of translation service provision process (Gouadec 2002, 2005a, 2005b, 2007). The data analyzed show the importance of outsourcing in the everyday activity of translation agencies and (partly as a consequence of this outsourcing) the magnitude and importance of the management activities carried out by the staff.


Introduction
Work occupies a large part of our lives.Sociologists specialized in studying the living and working conditions of persons in highly developed countries since the 1980s agree that we devote far more time to work than before (Prieto 2004), that we work more intensely (Durand 2004), and that our identifi cation with the content of our work still occupies a central place in our lives (Sanchis 2004).
In the early 21st century, translation is a job performed by many people and mainly provided through translation agencies.In this article, we focus on the work content in one of the many agencies in Poland providing translation services.We study the work content in the translation agency by focusing on the activities carried out by the in-house staff and by observing real work situations.Our focus on the translation agency rather than on the translator gives greater visibility to translation as work and employment.Our aim is to identify the components of the work content and the role played by translation activites in translation agencies.In other words, we will try to answer the question: What do people in translation agencies do and what place does translation occupy in their activities?
This question is answered through a re-analysis of empirical material from a study carried out by Stelmach (2000) in a small translation agency in Poland from 1999 to 2000.The data we have used in the article were not originally intended to form the basis of an analysis of translation activity, but were merely examples of how to measure the real activity (how long activities last, how often they occur and how they are distributed over the day) of the persons occupying the posts in the agency.We used this quantitative information on time use and reinterpreted it in order to identify the components of the work content and the organization of translation activity in the general context of the activity carried out in the agency studied.To carry out this interpretation, we used concepts taken from the sociology of work and translation studies.The former already has a long tradition of studying real work situations, whereas in the latter it is still quite rare.One could say that this article prepares the basis for the sociology of work to be applied to the study of translation-related jobs, just as Monzó Nebot (2006) does for the sociology of professions applied to the study of translation as a profession.
The results of the analysis serve two main purposes.First, for professionals in translation services they provide information on the characteristics and limits of job profi les in companies and institutions.Second, they may help to adapt the training of professionals to real situations and thus increase their employability.

Conceptual bases: work content and organizational context
In simple terms, work content is absolutely everything that people do in their workplace and during working time.All the activities are structured as productive work processes and occur in a particular organizational context -in our case, a translation agency.The following two sub-sections are devoted to defi ning these concepts.

The process of productive work
In the fi eld of the sociology of work, using Situated Activity Theory, the general process of productive work is defi ned as a set of interactions of different types (factual, cognitive, etc.) between social subjects (individuals or groups) and instruments which is carried out in a given space-time situation, in contexts ranging from the narrowest to the widest and involving different spheres (economics, production organization, etc.), with a view to achieving the material or symbolic modifi cation of a product (Verd et al. 2007: 150).
The process of work is thus a human activity that is carried out in a context of personal interactions, in a given place and time (situated activity), in connection with other processes and with an organizational structure, and with the fundamental aim of transforming a material.However, the process is not a simple sum of these elements, but rather a constant and dynamic interaction.
In the fi eld of translation studies, the closest and at the same time most complete approach that meets the description given in the foregoing paragraphs is, in our view, the model of translation service provision proposed by Gouadec (2002Gouadec ( , 2005aGouadec ( , 2005bGouadec ( , 2007)).This model focuses primarily on the last element of the defi nition of the productive work process given above: the material or symbolic modifi cation of a product.
According to Gouadec (2005a), the process of providing translation services is a set of operations performed between the reception of an assignment and the delivery of the translation.All these operations build a production process.The process consists in transforming raw material (the original text) into processed material (the target text).Gouadec develops a procedure composed of highly detailed tasks and operations that must be followed by the people fulfi lling the functions of the translator, the reviser and the work provider (client) to achieve a high quality translation service (Gouadec 2007).He shows how these three functions interact and make a joint contribution to the implementation of the assignment.He also draws up a list of all the documents necessary for the correct management of a translation agency.
In fact, the roles of the translator, the reviser and the work provider should be seen as abstract missions because each one includes an endless number of real actors inside and outside a translation agency.On the translator's side, the following players can be involved (Gouadec 2007: 55-56): -The project manager, responsible for managing the translation on the translator's side of the operation.
-The translation company's marketing service.
-The translation company's accounting or fi nance department (responsible for invoicing the client).
-The terminologist employed by the translation agency or paid by the translator.
-The documentation researcher and manager.
-The reviser hired by the translator or by the translation agency.
-The data or CAP [computer assisted publication] operator hired by the translator or by the translation agency.
-The IT engineer or technician working for the translator.
-Any information provider questioned by the translator, plus anyone else directly involved in the process on the translator's side.
The activities carried out by the translator, or any of the players that represent him or her, and are involved in providing a translation service, are organized according to the work cycle shown in Figure 1.All the operations listed in this fi gure are activities directly related to the carrying out of translations.Gouadec's model does not take into account the other activities (accounting, cleaning, etc.) that belong to the general management of a translation agency, but are not included in the function of the translator.(Gouadec 2007: 15) It can be seen that Gouadec's model considers the activity of the translator far more widely and not only as the activity of performing a translation itself.Among the "other" activities, Gouadec includes all those that precede and prepare the translation (obtaining the translation, receiving and checking the material for translation, etc.) and the activities that follow it (closing the fi le, fi ling the materials and resources consulted, etc.).
In the middle of the whole process is the translation itself (the translation phase), which has three stages: pre-transfer, transfer and post-transfer.Gouadec defi nes the transfer phase as follows (Gouadec 2007: 23): The transfer task is, of course, central to the translating process and it is what all 'translators' carry out.Transfer normally means transferring contents and meaning into a different culture, a different code (linguistic or other), a different communicative set-up, for an audience or users who are different, though homologous, making all necessary adaptations to that effect and purpose.Transfer normally includes self-checks and controls by the translator himself.
The translator is required not only to carry out many activities other than translation, but also to be available if one of his/her contact persons (the reviser or work provider) fails at some point in the process.In this case, the translator must undertake an operation that a priori does not correspond to him/her in order to ensure the continuity of the process (Gouadec 2007: 56).
In the fi eld of the sociology of work, Terssac (1995) studies disturbances during work in continuous production process industries (cement factories, printing presses, chemical and plastics industries).He defi nes disturbances as incidents which lead the components of the work process to abandon its limits of functionality (Terssac 1995: 86-87) and distinguishes the following categories of disturbances: Incidents or interru (1) ptions produced by the physical context of production Accidents or accidental dysfunctions of machines and installations (2) Errors in the control devices of the work process (3) Input or output fl uctuations of the work process and during the workfl ow proper (4) Interferences between tasks (5) Errors in the general organization and documentation of the work process (6) Terssac states (1995: 83-84) that all the formal and explicit regulations, but especially informal and implicit regulations of the teams involved in the work organization, are conceived to ensure the continuity of the production.Séguinot (2008) mentions that there has been a change in the paradigm of research in translation studies: from texts as the object of study, through communicative and cognitive processes, to research focused on the translator.However, the translator is canonically treated as an individual.This is the case in studies of self-employed and in-house translators that fail to take into account the organizational characteristics of their work context.These studies highlight the tasks carried out by translators as representatives of a profession, but they separate them from the context provided by their workplace.

The study of the work content in its context
The rules and imperfections of the organizational context and all the daily situations in which the translator works, are revealed in studies that analyze the work of translators and incorporate the characteristics of the organizations as variables of study (e.g.Lebtahi/Ibert 2000, Hébert-Malloch 2004, Abdallah/Koskinen 2007, Dam/Zethsen 2008), and in those that directly focus on the organization and management of translation services (Gouadec 2002, 2005a, 2005b, 2007, Davico 2005, Risku 2006) and on the cognitive aspects of the work in the organizational, translationrelated context (Risku 2000(Risku , 2002)).
However, empirical research conducted within companies and institutions sometimes reveals surprising realities.Hébert-Malloch (2004) recorded the work of a high-level in-house translator in a translation institution in Canada with a video camera and found that the translator devoted only 41% of all his working time to translating (in the sense of typing a translation on the computer).The rest of the time (59%) is classifi ed in the study as "pauses", or rather "no-translation" (Hébert-Malloch 2004: 974).
Despite the great interest of such studies and their potential to produce new fi ndings, few empirical studies with fi eldwork in labour contexts have been carried out in our discipline.There may be two reasons for this: methodological limitations and lack of a research tradition.
As for the methodological limitations, in translation studies the workplace of translators has tended to be de-contextualized, largely because of the use of the survey method, which is increasingly widespread in studies on translation tasks.Due to their methodological characteristics, surveys typically concern persons outside their work context (Kuznik et al. in press).Another reason for not heeding the organizational context of work might be the obvious diffi culty of gaining access to organizations (Coller/Garvía 2004).Moreover, it is easier to conduct surveys among translators outside their work context than to carry out fi eld work through observation in agencies, organizations, and institutions.
As for the lack of research practice in the past, translation studies have traditionally approached problems related to carrying out translation assignments in the workplace as problems external to the discipline.However, the study of the work situation and the organizational context is gradually gaining ground and defi ning its place within the discipline.
Certain representatives of functional theories (Reiss 1971, Vermeer 1978, Holz-Mänttäri 1984, Reiss/Vermeer 1991, Nord 1988, 1997) approach translation as a communicative and intercultural interaction through texts which are produced and translated.The point of view of Mayoral (2001) is quite similar as he stresses the need to approach translation as the result of a type of organization of the translator's work.Mayoral (2001: 46) defi nes translation as: (1) the cognitive process developed by an individual or the process developed by a machine in the case of machine translation; (2) the social processes of communication; (3) the result or the product of these processes (texts).Each of these sets of processes fi nds its expression in reality: the cognitive processes consist of decision-making techniques and the social-communicative processes consist of the processes of action of translators, understood as "the performance by the translator of the ordered tasks that allow him or her to carry out a given assignment" (Mayoral 2001: 49).These processes of action are carried out in the professional work and business environment.Mayoral approached the business contextualization of the provision of translation services as something being external to translation and proposed an open list of several "conditioning parameters" (Mayoral 2001: 50): the situations of translation, the necessary factors, the original text, the target text, the translation and the specifi c translation requirements.However, he did not study these factors in depth because he considered that there were too many of them, and that they therefore had a very low potential for theorization.
We think that the high number of "external" parameters involved in the translation process is precisely due to the contextualization of this process in its situational coordinates, and that these parameters can be studied theoretically using the general social and economic disciplines that study professions and work.As Mayoral (2001) rightly argues, translation studies are a technological discipline and as such feed on the major theoretical developments of related disciplines.
If we want to place the work of in-house translators in context, we must be aware of all the components of the work situation within the company.As we know from longstanding research in the fi elds of sociology of labour and sociology of organizations, work situation is a very complex reality.Durand (2004) gives us a thorough vision of work situation, which we represent below in Figure 2 (Durand, 2004: 272).(Durand 2004: 272) One of the central concepts of Durand's sociological theory ( 2004) is the concept of workload.He states that, in order to properly address the workload faced by a worker, two dimensions should be put in relation: the individual dimension and the dimension of the work situation.Therefore, if we want to measure the workload, we must take into account the components of the person (the physiological and social characteristics and the subjective perception of the tasks: the sensation of monotony and tiredness) and all the components of the work situation.According to Durand (2004), the work situation is built from elements such as the job or function carried out, the productive techniques, the work organization, the forms of pay and the professional relationships.
The complex nature of the work situation is basically due to two reasons: (1) it is composed of a high number of variables in the sense of the "conditioning parameters" mentioned by Mayoral (2001: 50); (2) it is closely linked to the -rather problematic -delimitation and defi nition of the job, which is the principal element of the work content.The notion of job means no more than "a person working".The term has completely lost its traditional organizational meaning and has been replaced by a far more diluted and changing reality resulting from constant structural changes of markets and companies.For this reason, researchers use such terms as "project" and "fi eld of work" (Sennett 1998), "work unit" (Terssac 1995), and "collective worker" (Castillo et al. 2000).
While productive techniques provide jobs with the repertory of all the activities necessary to produce goods and services and the order of their execution (work process), there is one component of the work situation that seems to be very helpful to defi ne and delimitate the job or function: the work organization, with its individual and internal regulations of teams.The work organization allows us to identify which activities are part of the particular in-house job (work content), how these tasks are articulated in the internal organization of the company and what type of links exist with actors and organizations outside it (tasks outsourced).

Data and methodology
With the aim of studying the internal job content of translation agencies, we re-analyze data from a research project carried out by Stelmach (2000) within the disciplinary sphere of work organization and business management.

The use of secondary data
Stelmach's original aim (2000) was to analyze and assess working time organization.In the conceptual part, she analyzes the notion of time as a legal labour and organizational category.In the empirical part, she uses a quantitative approach to analyse the data gathered in a translation agency.
We use Stelmach's results as secondary data.Cea D'Ancona (1996) defi nes secondary data as those that have been gathered previously by other researchers in a context different to that of their current exploitation.The current analysis may be related to the original purpose for which the data were gathered or to a completely different purpose.In our case, though we also deal with translation services, the objective of our analysis is different from that of the original researcher.Our use of the secondary data to provide an original contribution consisted in: (1) making a selection of data relevant to our research question from all the data gathered by Stelmach (2000); (2) evaluating the results obtained by Stelmach and some methodological aspects of her study for the sphere of translation studies; and (3) interpreting the data and the quantitative results with an approach aimed at determining the translation-related tasks that are carried out in the translation agency and those that are outsourced.
The use of secondary data has a long tradition in sociology (González Rodríguez 1994).Moreover, the increasingly widespread practice of making data sets available to other researchers and the ongoing development of databases and fi les that store all kinds of quantitative and qualitative data are popularizing their use.The main advantage of using this type of information lies in the low production costs.González Rodríguez (1994) also mentions the possibility of stimulating new ideas (serendipity), the possibility of replication and reanalysis aimed at strengthening the fi ndings, and the non-reactivity of the subjects investigated.
One of the main drawbacks of this practice is the age of the data (Cea D'Ancona 1996: 228) and the fact that it can be used acritically and without clear objectives.In our case, we have a good knowledge of the agency studied by Stelmach, so the data are not de-contextualized.
The use of secondary data in translation studies is very rare and does not have as much tradition as in sociology.We feel that this method is underestimated in our discipline, and that it could be a good solution for obtaining evidence from sites with highly restricted access such as business environments.
In view of the lack of tradition of using secondary data in translation studies, in order to give a further insight into our analysis we reproduce part of the original data.As Stelmach (2000) studied a single a translation agency in a case study, we offer a brief description of the company studied below.

Contextualization of Stelmach's study: general information on the translation agency
The empirical work of the study was carried out from 1999 to 2000 in a translation agency specialized in translating non-fi ction, professional texts (technical, fi nancial and medical) from several foreign languages to the mother tongue (Polish) and from the mother tongue into foreign languages (both Indo-European and non-Indo-European).The agency also offered sworn translations and interpretation services during the period of data collection.
The translation agency was located in Silesia, southern Poland, a highly industrialized region devoted to mining, metalworking and chemicals, in the middle of an agglomeration of similar cities with industrial and commercial activity and well-developed contacts with Germany.Its clients were companies in the region of Silesia and in other Polish cities, foreign companies, and private Polish and foreign individuals (Stelmach 2000: 53).
The agency was founded in December 1996 and started operating as a one-person fi rm.At that time, the main income was from translations from German into Polish.Later, the agency began to work with external translators and thus widened the offer of languages.The external translators worked on service contracts.
Due to the increasing volume of services, two workers2 were recruited to manage the offi ce.The founder therefore no longer had to deal with the administrative business and could devote himself exclusively to the -mainly sworn -translations from German into Polish.Both of the workers had vocational training in administration and accountancy, but not in translation; they were permanent employees of the agency and worked on open-ended full-time contracts with a continuous workday from Monday to Friday from 8 AM to 4 PM.They sometimes carried out additional tasks outside the 8-hour day and these activities were recognized as overtime.At the time when the study was carried out, both of these workers were 24 years old and had been working in the agency for just over two years.These two workers are referred to as worker A and worker B in Stelmach's study and in this article.

Results of Stelmach's study
The methodology chosen by Stelmach was self-observation by the workers in the workplace.The data were collected as follows: Worker A and worker B chose fi ve continuous working days from 8 AM to 4 PM, including any overtime.They drew up a self-observation template indicating the time in the left column (in the format hour: minutes) and then fi lled in the corresponding boxes at the time of self-observation in the right column with the activity they had carried out (Table 1, Table 2 and Table 3).The time intervals (in the left column) were 15 minutes.As an example of the kind of information offered by Stelmach, we reproduce three tables with collected data3 : Table 1 covers the fi rst day of worker A; Table 2 the fi fth day of worker A; and Table 3 the fi fth day of worker B. In the case of worker A, our selection of tables was based on offering the records of the fi rst and the last days and above all on the content of special interest for us that those days represented.As we will see below, they are the only days on which worker A did translations personally, though this activity was merely reported as "writing a text" in the fi nal results (Table 4).We chose to reproduce the fi fth day of worker B for four reasons: To be able to compare it with the fi fth day of worker A with regard to the (1) communication (or lack of record of communication) between them.
To demonstrate the difference between the records of worker B and worker A, in (2) the sense that worker B always took down far more information in general, and in particular specifi c information on the translation activity (language combinations of translations, types of text, etc.).
To exemplify the variety of tasks and the type of disturbances that arise in this type (3) of work.
To obtain the most complete example possible of a production cycle of the (4) translation service, because during the fi fth day a translation assignment was received in the morning and the translation done by an external translator was received at the end of the working day.This did not occur on any previous day of either worker.Visit to the house of the sworn translator to pick up the German translation 15:45 cont.
Table 3. Results of the fi fth day of self-observation of worker B, according to Stelmach (2000: 64) Data of the type shown in Table 1, Table 2 and Table 3 were systematized and reduced by the researcher into wider concepts such as "Making breakfast, eating" or "Visit to the house of a translator, time in the house, return".The main activities were identifi ed and labelled, all the records of each activity were summed up, and their incidence during the working day was calculated.For this calculation, each working day was considered separately, so the 8 hours of work represented 100 percent.
Table 4 and Table 5 show the labels of the main activities carried out by workers A and B, respectively, during the fi ve days of self-observation.On the fi rst day, worker A dedicated 31.25% of her working day to "Writing a text" (activity no. 8, Table 4) and 25% to "Printing and copying documents" (activity no.13, Table 4), the two main activities on that day.On the fi fth day, "Writing a text" (18.75%) and surprisingly "Making breakfast, eating" (12.5%: activity no. 1, Table 4) were the main activities of worker A (31.25%).Stelmach ( 2000) thus made a reading of the temporal distribution of the activities in terms of the time used, and identifi ed the activities of greatest and least incidence.She observed a great variability in the results for the fi ve days of self-observation.She also incorporated a brief assessment of the overtime dedicated to the agency (in addition to the 8-hour working day).

Activities of worker
Table 5. Distribution of the activities of worker B during her working day in fi ve days of self-observation, according to Stelmach (2000: 66) Regarding methodology, Stelmach reported that self-observation was a good technique, because the workers did not feel disturbed by the presence of someone from outside the agency (Stelmach 2000: 67).It therefore helped to ensure the ecological validity of the study, though one drawback was detected: the workers were forced to interrupt the activities that were underway every 15 minutes to fi ll in the self-observation form, which added many more interruptions to the ones that already arose in the working situation.

Value and limitations of Stelmach's study data
In relation to our own objectives of research, the greatest value of Stelmach's study (2000) lies in her identifi cation of all the activities carried out continuously during the fi ve days of self-observation by workers A and B and in her attempt to classify them into broader concepts.Between the elementary data recorded in the self-observation templates (Table 1, Table 2 and Table 3) and the labels with names of the main activities that appear in the fi nal results (Table 4 and Table 5), she systematized, reduced, grouped, and abstracted the concepts.The data gathered by Stelmach are of great use for identifying the set of tasks carried out by the workers in the translation agency.Offi ce work and management involve a large variety of tasks.The persons occupying jobs of this type work in different spaces, with different persons, using a wide range of tools and machines to carry out highly varied data and text processing activities.
Stelmach's second contribution, on a methodological level, was to determine the occurrences of each task.She measured the incidence of the activities and then converted it into percentages of distribution of these activities in the working day.However, she only recorded the presence (against the absence) of the tasks in equal intervals of 15 minutes, but not their duration.The percentage value given to each activity therefore informs us of the frequency of the activity and its incidence in the day in comparison with other activities, but it does not identify its exact duration.
Stelmach also showed a simple and objective way of recording working activities, choosing the time axis as her main principle of organization.With the correct application of the working day photography technique, through continuous observation (self-observation), one can cover all the activities carried out, paying more attention to the tasks recognized to be important and the tasks that are not visible due to their routine, apparently unimportant and uncontroversial nature.Despite the value of Stelmach's study, her results imposed some limitations on the present study.First, the fact that the names of the main activities were not always recorded under the same name by worker A and B makes it impossible to draw up a single table for the two workers.Furthermore, the names do not follow a pre-established order, based on a single criterion (alphabetic, chronological, of greater importance, or of greater incidence during the day) and the order was not the same for the two workers, so the overall interpretation of the data was diffi cult.If similar studies are carried out, they might also include an analysis of the daily distribution of each activity, setting the level of activity analyzed at 100% (an extra column on the right of Table 4 and  Table 5).
Second, the data provide fairly poor information on the interaction of persons within the agency.This is not surprising, because the objectives of the study were to analyze the use of working time.One therefore obtains only a partial view of the agency's work organization and, especially, of the collective and distributed nature of the work that emerges from the data gathered.It seems that the main channel of communication was with the manager (activity no.16, Table 4; and no. 7, Table 5) and only in activity no.7 is explicit reference made to communication with the other worker.

Analysis of the work content in the translation agency
Our analysis of the secondary data allows us to identify the main characteristics of work content and the specifi c place occupied by translation activity within the translation agency studied by Stelmach (2000).

Variety of interactions
If the data gathered by Stelmach (2000) are analysed according to the elements of the the process of productive work (Verd et al. 2007: 150) as defi ned above, they show great variety of interactions carried out in the translation agency.
The spaces involved include internal spaces (offi ce, toilet) and external spaces (bank, post offi ce, house of subcontracting translators, stationery shop).The persons with whom the workers interacted were the manager and his wife (for the annual tax declaration), the clients, external translators, the computer technician, the accountant, and the workers themselves.Sometimes the general term "all" was used (Stelmach 2000: 58).The tools and machines referred to were the computer, the external hard drive (with the CDs), the printer, the telephone and material for handwriting.
During the fi ve working days, the workers handled a great variety of documents: the daily plan, the invoices, the VAT forms, the personal income tax forms, the annual tax declaration (of the manager and his wife), bank transfer forms, the accounting documents, the assignment form, the demand for payment, the client database, the service contracts with subcontracted translators, the advertising leafl et of the translation agency, the e-mail and the certifi cate of the sworn translator.
The activities carried out with the texts were also highly varied.We fi nd references to reading (revising, checking data, payments, reading invoices), writing, drawing up, copying (moving from one document to another), photocopying, printing, fi lling in forms (by hand and on the computer), updating the database, fi ling (copying on the external hard drive), faxing, revising and correcting, lending and returning the qualifi cation of the sworn translator, and explaining aspects of the text (technical drawings; activity carried out at 12.45, Table 3).Some of these activities are specifi c to translation management: doing a translation, transcribing on the computer, printing, receiving and sending by fax, checking the completion and legibility of the fax, checking the translation, revising it and correcting it.
On two occasions, we detected activities recorded as waiting: in the fi rst case, worker B waited at the house of a subcontracted translator while he checked his translation; in the second case, worker B waited in the offi ce to receive an urgent translation by e-mail during overtime (Stelmach 2000: 61).
Finally, among this great variety of activities we can distinguish between social and individual activities of a physiological nature (coffee, tea, breakfast, lunch, toilet) and social activities related to work (all the other activities).

Mixture of many translation assignments in different stages of production
Though the data on translation assignments gathered by Stelmach (2000) present the methodological limitations mentioned above, they inform us of the work content of translation services.
The vast majority of the activities carried out by workers A and B are management activities.They represent the different stages of Gouadec's model of translation service provision, including the preliminary step of "acquiring the translation", "pre-translation" and "pre-transfer" and the subsequent step of "post-transfer" or "post-translation" (Gouadec 2007; Figure 1).During the 8-hour working day of workers A and B, all these elements belong to different assignments and are totally intermingled.We have the impression that the activity of these two persons is similar to that of an airport logistics park, which deals with many fl ights at the same time and therefore requires a high degree of concentration and self-control.
The self-observation of worker B provided fairly rich information on translation assignments managed in the agency.Our interpretation of the data gathered by worker B in the tables shows that in the fi ve days of self-observation she dealt with an English interpretation, a translation from English, and three translations into English; a translation from German, two sworn translations from German and three translations into German; a translation from Russian, a translation into Russian and a sworn translation into Russian; a translation from Czech; and a sworn translation from Ukrainian.
These data help to confi rm the general characteristics of the agency described by Stelmach, i.e. its services in terms of language combinations and interpretation, and the frequency of sworn translations from German.A strong presence of Slavic languages (Polish, Russian, Czech, Ukrainian, Croatian, etc.) is also noted.Additional data that were gathered include the type of client (company, private individual; regional, domestic or international), the number of pages of text, and the type of service (urgent or normal).
As stated above, the fi fth day of self-observation of worker B was the only one that provided the possibility of following almost an entire translation cycle (Table 3).It began at 10:45 AM, when worker B received a client who brought a document for a sworn translation from German.At 11:45 AM she left the offi ce and delivered the text to the sworn translator.Between 3:30 and 4 PM she picked up the completed sworn translation.It was a simple job, but during the rest of the time the different elements of the assignments are fairly mixed up.They were carried out at the same time and were all concentrated in the two administrative workers (A and B).

Disturbances in the process of productive work
The ordered and harmonious progress of translation services is often disturbed.Only two disturbances affected the work of both employees: unpaid invoices and a faulty printer.However, the fi rst, third and fi fth day (Table 3) of worker B seem to have been fairly hectic.In chronological order, the problems that she faced and solved on the fi rst day of self-observation were as follows: Repeated faxing of several illegible pages of a translation (1) Checking and seeking the explanation for a client who called to ask about the (2) invoiced price of a translation: they do not agree with the emergency surcharge Desperately seeking a translator who is not at home or at work (3) Introducing changes to the translated text requested by the client, who visited (4) the agency in person requesting an immediate response; printing it and handing it over to the client On the third day, worker B faced and solved the following problems: Call from a company asking them to send a copy of a translation in Russian done a (1) long time ago because it had been lost Call asking for the original qualifi cation of the sworn translator who did a (2) translation; worker B went to the house of the sworn translator, borrowed the document, delivered it to the requested place and returned it to the translator On the fi fth day, worker B faced and solved the following problems: P (1) rolonged search for a translator (09:15-09:45, Table 3) Request for additional explanations on some aspects of a translation, related to the (2) technical drawings (12:45-13:15, Table 3) We know from Stelmach's data that both workers included the tasks of daily planning and organization in the morning but we do not know the tasks included in this planning and the extent to which it was effective.In fact, as shown by the list of incidents and problems described above, all the management work involved dealing with unforeseeable and unexpected situations.

The place and function of the translation activity
The translation activity, situated in Gouadec's model of translation service provision (2002, 2005a, 2005b, 2007; Figure 1), corresponds to the central stage of the cycle, i.e. the transfer.In the two administrative and accounting jobs subjected to self-observation, translation has a dual presence: on the one hand, it is the translation done by one of the workers: on the other hand, it is the translation managed.
During the fi rst and the fi fth day of self-observation, worker A did two translations herself (one per day), without giving us any details of the language combination, type of text or genre.The fi rst translation was 200 pages long (as reported by Stelmach, at 12 AM, Table 1).Worker A took 2 hours 15 minutes to perform this translation task.The second translation was 50 pages long (it started at 1:15 PM, Table 2) and the same worker took 1 hour 15 minutes to do it.The two translation sessions were immediately followed by activities of printing the translation.What is curious is that Stelmach does not call them "translations" in the fi nal results.She calls them "Writing a text" (activity no. 8, Table 4), to distinguish translation from copying (activity no. 7, Table 4).
On the days on which translation itself was recorded, this activity occupied the greatest proportion of the time (31.25% on the fi rst day and 18.75% on the fi fth day), followed by printing.Translation was probably the activity that required the greatest concentration in the long term (more than 1 hour in both cases) because of its highly cognitive and creative nature, which explain its incidence in the day and the lack of interruptions.
Translation activity is also implicitly present in the main activity of the manager of the agency, whose employment position was unfortunately not studied because he did not grant his/her permission (Stelmach 2000: 69).Peneff (1998: 15) states that the great diffi culty of carrying out observation in the service sector is that one must study not only those who provide the services (worker A and B in our case), but also those who manage it.The fact that members of staff, and indirectly the manager or director, are observed necessarily bothers the latter because it affects the hierarchy established in the company.
The translation done is thus marked by a curious triple silence: Wor (1) ker A does not give details of its characteristics.
There is a lack of explicit recognition of its presence in the mind of worker A and in (2) the fi nal results of the study (this was not the objective of Stelmach's work).
It is mainly concentrated in the manager/translator, whose employment position (3) was not observed.
Therefore, we can conclude that the translation activity was performed sporadically by the agency's permanent administrative staff (worker A, who had no specialized training) and continuously -at least the sworn translations from and into German -by the manager of the agency (this second conclusion is our assumption only).
Except for the activity of translation mentioned above "Writing a text" (activity no. 8, Table 4), all the rest of the activities listed in Stelmach's results in Table 4 and Table 5, are considered to be management activities.However, using Stelmach's data, it is not possible to clearly distinguish between activities directly related to carrying out translation assignments and the operations that belong to the general management of a translation agency.

Work content in the organizational context: companies, situations and work processes
The analysis presented in this article shows the importance of doing empirical research inside companies.In this specifi c case, the data show the organization and timing of work in a translation agency, indicating the tasks undertaken by the agency itself and those that are outsourced.The importance of placing persons in a specifi c work situation, as indicated by Durand's model, is thus demonstrated.
Observing real work situations structured on the basis of their main, continuous axis -time -allowed us to place Gouadec's model in its work context.We were therefore able to detect the richness of the activity carried out, its variety, its complexity and its simultaneity: the most intense moments (in which a great variety of activities is concentrated), the least intense moments, and the forced interruption of disturbances in the service provision process.
It must be noted that Gouadec's model is based on the tasks carried out by large companies working with complex multimedia and multilingual products.The agency we studied carries out fairly simple translations and its specifi c products are in many cases sworn translations.However, Gouadec's model fi ts precisely with the data analyzed because it places the whole set of tasks surrounding the basic translation work at a general level and in its organizational context.

Components of management jobs
Our analysis of the secondary data allows us to identify the main features of work content of management jobs in a translation agency.The presence of translation management seems to be deeply rooted in the activities of the permanent staff.This fi nal observation corroborates the opinion of Davico (2005: 46), who, after analyzing many cases of Italian translation agencies (e.g.Studio Essepi of Milan, Arancho s.r.l. of Rimini and Logos S.P.A. of Modena), reached the conclusion that there is a common feature in the organizations analyzed: none of them subcontracts the management because it is at the heart of each translation agency.
The content of management jobs is very heterogeneous and complex, involving the permanent coexistence of the following components: (1) a wide range of interactions with persons, machines, tools and texts; (2) a mixture of different elements of translation assignments at different stages of execution, which require constant monitoring; (3) the detection and immediate solution of alterations that arise in the process of carrying out translation assignments; and (4) combining management tasks with translation tasks.This combination of tasks in the same job requires multi-skilled staff.Mastering such a variety of tasks and disturbances, whilst bearing in mind the centrality of the most important tasks, is the greatest challenge for the management and the individual worker.

The place and characteristics of translation activity
In the data gathered by Stelmach (2000), we have observed the almost residual presence of translation itself in the in-house jobs of the translation agency analyzed.It is somewhat surprising that the central part of the translation process (the translation phase, in Figure 2 based on Gouadec 2007: 15) is of lesser importance in the data set collected.The tasks that Gouadec places in the centre of the translation cycle (transfer) are largely outsourced, whereas those that surround this central phase are kept within the agency.Translation is the main activity within the production process that is subcontracted.Hence the paradox: When the real situations of work in the agency are observed, one notices the management of the translation, but not the translation activity itself.
If members of the staff do translate, this activity is preceded and followed by translation management activities.However, when the staff are actually doing a translation, then it turns into an activity that is largely uninterrupted, probably because it requires a high level of concentration.

Outsourcing of translation as an additional component of management jobs
From the empirical material analyzed, it can be concluded that the outsourcing strategies of translation agencies lead to the need for new management tasks.Outsourcing provides the agencies with greater fl exibility and adaptability to the demands of clients and allows them to better meet the deadlines set by the clients.The constant pressure of time (the need to do the translations in the shortest time possible) and the desire to have a fl exible organization that can meet all the translation demands without requiring a large number of contracted employees, have led companies to outsource much of the work.However, in this type of organization, translation service is the result of a collective effort rather than of a single translator's work.This specifi c work organization leads to the creation of a whole set of new management tasks aimed at dealing with unforeseen circumstances and contingencies arising from the distributed nature of work proper.
To access international literature as diverse as the study of sociology, start here.
ProQuest Sociological Abstracts offers a world of relevant, comprehensive, and timely bibliographic coverage.Over 890,000 easily searchable abstracts enhance discovery of full-text articles in thousands of key journals from 35 countries, along with books, conference papers, and dissertations, as well as citations to reviews of books and other media.This continuously growing collection is updated monthly, and offers backfiles to 1952-plus scholar profiles, browsable indexes, and a searchable thesaurus through the ProQuest Illumina ™ interface.

The ProQuest Sociological Abstracts Discovery Prize.
Tell us how ProQuest Sociological Abstracts has advanced teaching and learning at your institution, and you may win the ProQuest Sociological Abstracts Discovery Prize.Visit: info.csa.com/sociologicaldiscovery

Table 1 .
Results of the fi rst day of self-observation of worker A, according toStelmach 2000: 55

Table 2 .
Results of the fi fth day of self-observation of worker A, according toStelmach 2000: 59

Table 4 .
Distribution of the activities of worker A during her working day in fi ve days of self-observation, according toStelmach (2000: 65)