Interpreting the discourse of reporting : The case of screening interviews with asylum seekers and police interviews in Finland

This paper examines the ways in which the discourse of reporting, i.e. the implicit and explicit presence, production, and usage of written texts in public administration reifies monolingual and monolithic language ideologies in interpretermediated police interviews and screening interviews with asylum seekers. The goal is to provide new insights into the analysis of complex networks of power relations that determine whether human rights can actually be exercised through public service or community interpreting. The paper derives from ethnographic data emanating from participant observation as an interpreter for migrants, asylum seekers, and international offenders using French and English in the Helsinki metropolitan area in Finland. In addition to being an interpreter, the author is also a researcher informed by critical discourse studies and sociolinguistic theory. The main argument of the article is that many problems related to public service interpreting that are thought to stem from cultural differences or the interpreter’s general lack of competence can be interpreted as resulting from language ideologies, reified in the practices in which they appear. One of the most important of such practices is the discourse of reporting, effectively blurring the distinction between written and oral language and denaturalizing “spontaneous” speech of interpreter-mediated communicative encounters. The paper suggests that a critical reflection on the nature and function of language and multilingualism and the consequences of language use is necessary in order to allow the interpreter to occupy subject positions from which power relations can be negotiated and linguistic equality delivered. Such a reflection should be part of both interpreter and service provider training.


Introduction
The study of language ideologies, i.e. cultural conceptions of the nature, purpose, and function of language (Gal & Woolard, 1995, p. 130;Woolard & Schieffelin, 1994) has been an important field of inquiry in sociolinguistics, linguistic anthropology, and critical discourse studies for over thirty decades.To some extent, language ideologies have also been studied in the context of public service interpreting, i.e. community interpreting, and legal and courtroom interpreting in particular.Thus, many studies have argued that monolithic, monolingual language ideologies, in which generalizations about the nature and function of language are based on monolingual and institutionalized contexts, in fact dominate legal services.These ideologies have also been identified as a major source of linguistic injustice in interpretermediated encounters (see e.g.Angermeyer, 2008Angermeyer, , 2014;;Berk-Seligson, 2008;Haviland, 2003;Maryns, 2006).However, in interpreting studies, larger social phenomena and institutional constraints that link the interpreter's performance to language ideologies have not attracted much attention.
This paper contends that many problems and phenomena related to public service interpreting that are thought to emanate from cultural differences or the interpreter's general lack of competence (see e.g.Hale, 2004, p. 238;Pöllabauer, 2006;Rudvin, 2006) or lack of accuracy due to the interpreter's omission of words and discourse markers or other pragmatic information (Berk-Seligson, 1999;Hale, 2004, p. 239) can be interpreted as resulting from language ideologies.Furthermore, the paper argues that monolingual and monolithic language ideologies need to be explained in connection with the practices in which they appear and become reified.While ideologies can be defined as sets of beliefs or ideas having an object, typically a contested concept such as language, the practices reifying them can be conceived as discourses, i.e. systematic ways of using language in a particular way, directing the formation of meanings, creating a prototypical set of oral, written, and multimodal genres and texts, and enacting, reifying, and enforcing ideologies within a field of activity or an institution (Määttä, 2014;Määttä & Pietikäinen, 2014).One of the most important practices within the public service is transcribing and reporting in the written form, based on the particularities of written language and specific written genres.This discourse of reporting characterizes all encounters between public service providers and service users in much of the world today and therefore also in public service interpreting.
Both discourse and ideology are essentially contested, polysemic concepts.While the way in which these notions are used in this paper may seem unorthodox from the viewpoint of translation and interpreting studies, certain links to descriptive translation studies (e.g.Toury, 2012) can be identified.Thus, grosso modo, ideologies correspond to value systems lying beyond observable norms that govern translational and interpreting activities.However, norms do not pertain to the same perspective or level of analysis with discourse and ideology.Norm can be used as a tool to describe and analyse the ways in which and the reasons why people perform tasks such as translating or interpreting in a particular way.Discourse and ideology, on the other hand, are useful tools when the analysis focuses on the effects and consequences of language use, on the one hand, and on the ways in which meanings are created and concepts and objects defined and reified, on the other hand.Therefore, critical discourse studies and sociolinguistics, which form the conceptual and theoretical framework of this paper, differ from descriptive translation studies both in terms of the questions asked and the theoretical and methodological tools used.
This paper suggests that the specific features of the discourse of reporting and its relation to monolithic, monolingual language ideologies may provide new insights into the analysis of complex networks of power relations that determine whether human rights can actually be exercised through public service interpreting.The data consists of participant observation by the author as a public service interpreter and translator between Finnish, on the one hand, and French, English, and Spanish, on the other, in the Helsinki metropolitan area in Finland.This observation extends from 2004 to the present day and has produced a large set of ethnographic data comprised of hundreds of encounters in different settings: legal, health care, mental health, education, and social work.In this paper, the focus will be on interviews conducted by law enforcement agents in which the interpreter and the migrant or the asylum seeker communicate in French or English.
The paper therefore stems from participant observation in a double role: the observer is both the interpreter and the researcher.As a result, the data does not include recordings, observations as an outsider, or surveys.However, this does not mean that the observation consists of mere intuition and impressions.Rather, it is grounded in the theoretical and methodological traditions of sociolinguistics and critical discourse studies and a long-standing scholarly inquiry into language policy, language ideologies, and language variation.From the above-mentioned double position, the following research questions concerning the interplay between writing and speaking in community interpreting emerge: What is the role played by written texts in interpretermediated encounters between service providers and migrants, asylum seekers, and international offenders?What kind of links can be identified between the discourse of reporting and monolingual language ideologies?What are the consequences of the discourse of reporting in community interpreting from the viewpoint of the service provider, the migrant, and the interpreter?These questions have guided my systematic practice of taking field notes during and after each encounter in which I have been an interpreter since January 2013.
While my data collection methods may be regarded as biased and unscientific when compared to those traditionally used in interpreting studies, they are perfectly valid in other disciplines such as linguistic anthropology (see e.g.Haviland, 2015).In addition, there are advantages related to this data collection method.In fact, other methods would not have allowed me to answer the research questions presented above.Furthermore, had I not been an interpreter in these encounters, my data would not even exist.Indeed, I have observed situations that are hermetically closed for most researchers, such as screening interviews with asylum seekers, asylum interviews and police interviews with minors, and interviews with torture survivors and alleged perpetrators of sexual aggression.Hence, my approach has allowed me to observe the consequences and material effects of language use in these situations.For example, I have been able to survey the itinerary of several persons from initial police interviews and screening interviews to the moment in which a decision was taken by a court of law or the immigration service.This observation has focused essentially on the role, function, and interpretation of the written texts used, evoked, produced, and summarized during the process.
In order to protect the subjects and in accordance with the ethical code of Finnish community interpreters published by the Finnish Association of Translators and Interpreters (SKTL, 2013), details such as the exact date, place, age, origin, literacy level, or language resources of the subject are not provided when discussing the examples.Indeed, since French and English are worldwide lingua francas, migrants using them in Finland have diverse origins.However, certain backgrounds are more common than others.Thus, in this data, some 99 percent of the migrants using French and 90 percent of the migrants using English come from Central and Western African countries, the vast majority of them from the Democratic Republic of Congo and Nigeria, respectively.In the vast majority of the cases, English or French is not the first language of the migrant.
In the following section, I will present the context of public service interpreting and recent demographic changes in Finland in order to explain the reasons why problems related to the discourse of reporting may be particularly salient.Subsequently, I will analyse the characteristics of the discourse of reporting and the ways in which structural and institutional constraints embodied in the discourse of reporting may in fact explain some of the phenomena that previous literature has analysed in terms of alterations the interpreter makes to the original message.Focus will be on processes in which translating and interpreting occurs not only between two languages but also between spoken and written modalities of language use.To conclude, I will discuss the ethical dimension of the interpreter's agency and examine possible solutions.

Particularities of public service interpreting and recent demographic changes in Finland
Public service interpreting in Finland is based on several laws.Thus, asylum seekers are entitled to use an interpreter in all matters related to the asylum process.In addition, interpreting is always provided in cases in which an alien is refused entry into the country or is deported, and when procedures are initiated by the authorities.Thus, when a person has insufficient knowledge of the language of the administration, interpreting is also provided in hearings conducted by law enforcement agents and in courtroom trials.In addition, since the law stipulates that the clients of healthcare and social services are entitled to be informed about matters concerning them in a language that they can understand, these services also use interpreters when needed (see e.g.City of Helsinki, 2014a).
While there are no comprehensive nation-wide statistics, data provided by interpreting agencies and city administrations suggests that Somali, Russian, and Arabic are by far the most important languages in public service interpreting in Finland.Thus, one of the largest community interpreting agencies in the Helsinki metropolitan area reports that in 2013, of over 60,000 interpreted hours, the most important languages were Somali (17%), Russian (17%), and Arabic (14%).Sorani, Albanian, Vietnamese, Estonian, Turkish, Farsi, Mandarin, Dari, and Thai all represented between 3% and 5% of the total number of hours.English represented 4%, French 3%, and Spanish 1% (Hänninen, 2014).Nevertheless, in order to get a picture of the total number of hours interpreted, one has to take into consideration that in the Helsinki metropolitan area alone there are two other major agencies and several smaller ones.In addition, many service providers, courtrooms in particular, recruit interpreters directly.
Rather than specializing in one or two fields, public service interpreters in Finland tend to work in a variety of domains: physical and mental healthcare, social work and the integration of immigrants, schools and day-care centres, courtrooms, interviews conducted by law enforcement agents, and the asylum process.Interpreters usually work as freelancers for a private or public interpreting agency, most of which are specialized in public service interpreting and translation of written documents related to public service interpreting.
The linguistic and cultural itineraries of Finnish public service interpreters are quite varied.While most interpreters are native speakers of the language spoken by their migrant clients and second-language speakers of the language of the administration, there is a considerable variation contingent upon the language combination (cf.Hale et al., 2011, p. 49).Thus, most interpreters of languages that one can study widely in Finnish universities such as French or English are of Finnish descent, typically with university degrees in philology, language studies, or translation and interpreting studies.The body of languages requiring interpreting that are not widely studied at universities is more varied.For example, among interpreters with Spanish as one of their working languages, there are both native speakers of Spanish and second-language speakers or speakers who grew up learning both languages.Russian, the most widely spoken foreign language in Finland, is a special case: while Russian can be studied in most universities, the majority of public service interpreters are bilinguals who grew up learning both Finnish and Russian either in Finland or in Russia, or immigrants who learned Finnish after they came to Finland.But some of the most important languages in public service interpreting today, such as Somali, Persian, Thai, or Vietnamese, cannot be studied at university with the exception of a few courses.Most interpreters of these languages are migrants or they have grown up in Finland learning both Finnish and the language of their migrant parents.While these interpreters have no extensive training in interpreting studies or general language studies, most of them have attended short ad hoc training courses offered by interpreting agencies.Today, public service interpreting may also be studied at certain professional institutes.However, no official accreditation system exists and no professional qualification is required to work as a community interpreter.
The diverse backgrounds of interpreters have a major impact on the field: interpreters represent a wide array of linguistic cultures and linguistic and cultural itineraries and repertoires.Public service interpreters do not form a real community.Thus, whilst the quality of public service interpreting is a constant concern in public discussion, it is difficult to tackle these problems because the Finnish Association of Translators and Interpreters has managed to attract only very few public service interpreters.Indeed, the interpreter section of this association is traditionally dominated by conference and business liaison interpreters and although most public service interpreters who studied translation and interpreting at university are members of the association, very few other public service interpreters are.The degree of professionalization reflects both the lack of prestige of public service interpreting in general as well as ethnic divisions within the body of interpreters.
However, this ethnic divide also reflects the fact that while public service interpreting is a relatively stabilized field of activity in Finland, the practice is fairly recent, as until the early 1990s Finland was not a migrant destination.Since then, important changes have come about in the population structure: in today's Finland up to 90% of population growth is due to the arrival of new residents whose first language is not one of the national languages (i.e., Finnish, Swedish, or one of the three Sámi languages).In 20 years, there has been a fourfold increase in the number of resident aliens and a six-fold increase in the number of residents born abroad.At the same time, approximately 70,000 people have been naturalized.In addition, over 18,000 people have been granted refugee status (Official Statistics of Finland, 2014).
Official statistics related to the languages spoken by people residing in Finland are particularly illustrative in this respect: while there were nearly 25,000 people whose first language was not one of the national languages in 1990, there were almost 290,000 persons in that category at the end of 2013, i.e., a 12-fold increase.This change is most visible and audible in the Helsinki metropolitan area, where the number of people speaking a foreign language as their first languages has doubled since the year 2000: this population now represents about 13% of the total population of the city and up to 25% in certain districts (City of Helsinki, 2014b [statistical databases available only in Finnish]; Official Statistics of Finland, 2014).However, the appearance of new languages is not the only consequence of immigration: there are also new forms of multilingualism, new kinds of linguistic itineraries, and unforeseen combinations of linguistic resources and repertoires, including new kinds of dynamics between written and spoken resources.For the general public and service providers alike, this diversity remains mostly invisible due to the monolingual language ideologies that are prevalent both in the public administration and society as a whole, a monolithic view equating one language with one culture, and failing to acknowledge the fact that bilingualism and multilingualism are never 'stable.'These invisible dimensions of multilingualism are particularly important when considering the links between interpreting and the discourse of reporting.

The discourse of reporting
Written language is such an important part of Western culture that is easy to forget how much it has shaped the very conception of language that we have.Many scholars (e.g.Linell, 2004) argue that the notion of language developed by linguistics is biased because it is based on written language.Writing has also made possible the techniques of reporting and registration which according to Foucault (1977, pp. 190-191) emerged during the Age of Enlightenment towards the late 18 th century, allowing for the individual to become an object of knowledge and generating techniques of disciplinary writing.These techniques made it possible to produce objects such as descriptions, biographical narratives, files, and records, ultimately permitting the production of knowledge and leading to a circular relation between knowledge and power (ibid., pp. 219, 224-227).As a result, writing is one of the most important tools through which knowledge is accumulated and state power is exercised in contemporary society (Bourdieu, 1998, p. 45;de Certeau, 1984, p. 134).
Reporting, recording, and registration techniques based on writing are powerfully foregrounded in public service interpreting.While interpretermediated communication between service providers and service users (migrants, asylum seekers, and border-crossing offenders) is by definition spoken, interpreters are often required to interpret written texts that are read aloud.In addition, oral sight translations of written texts are not uncommon, although some ethical codes such as that used in Flanders prohibit this practice.(In Finland this is not the case.)However, interpreter-mediated encounters are also characterized by an indirect presence of written language: written documents such as laws, regulations, and reports of previous encounters are referred to during the encounter or they are present implicitly because the encounter and the entire activity are based on them.In addition, the goal of the encounter is almost invariably the production of a written text such as a report, a referral note, or a decision.Finally, even 'spontaneous' or 'naturally occurring' speech in interpreter-mediated encounters is often infiltrated by features typical of written rather than oral communication.This is visible not only in the terminology but also in the sentence structure and the ways in which genres and text types typical of written communication colonize spoken interaction.Examples of this colonization include the usage of nouns to express 'doing' and 'happening' instead of using the prototypical category, i.e. a verb.For example, nominalizations of this kind are a typical feature of the (written) standard questions that comprise the screening interview of asylum seekers in Finland, so that instead of asking "What would you do or think if you were not allowed to enter the country?" the question is "What would your reaction be to refusal of entry?"The rights and duties of the suspect or the witness, read or explained at the beginning of an interview and deriving directly from the Criminal Investigation Act, are also examples of this infiltration.More examples will be given in sections 4 and 5.
Consequently, the distinction between oral and written texts is in fact blurred in interpreter-mediated encounters (see also Blommaert, 2013).As a result, quite sophisticated textual and discursive resources are required in order to communicate effectively and exercise one's rights in such intertextual chains (cf.Blommaert, 2001;Gumperz, 1982a;Hymes, 1996, p. 213).In fact, familiarity with the discourse of reporting is necessary for a successful contextualization without which there is no understanding nor full participation (see Gumperz, 1982aGumperz, , 1982b, pp. 178-179), pp. 178-179).While people who have lived their whole lives in a country in which the discourse of reporting prevails in public administration may have difficulties with the textual world generated by this discourse, newcomers are confronted with particularly serious problems because the importance of written texts is typically accentuated in a migration context, and in asylum cases in particular.Thus, encounters between public service providers and migrants or asylum seekers are characterized by pretextual gaps: service users fail to produce what is expected of them (Maryns, 2006, p. 7;Maryns & Blommaert, 2002).However, the public administration rarely considers that migrants may not be familiar with reporting and related practices such as interviewsand if they are, they may not be familiar with the specific characteristics of these practices as performed in a local context (Jacquemet, 2009, p. 528).
The role of writing and specific discourses establishing the power of an institution has been widely studied in sociolinguistics and linguistic anthropology (e.g.Gumperz, 1982aGumperz, , 1982b)).The power of writing is based on the set of ideas or ideology according to which writing guarantees neutrality and transforms random and ambiguous spoken language into fixed and permanent written records.Therefore, a key moment in this process is the inscription by which spoken text is extracted from its original context and written down.In other words, spoken language is entextualized and becomes a written text (Bucholtz, 2009;Park & Bucholtz, 2009).In legal contexts such as the asylum procedure, such a written record functions as an essential instrument for the expression of truth (Maryns, 2006, p. 316).Once the written record exists, it becomes the central piece of intertextual networks linking it with new discourses and genres and engendering new texts.
However, whereas institutional and situational constraints, including the presence of written texts, are recurrent themes in interpreters' conversations among colleagues, such constraints have not been widely analysed in interpreting studies.Rather, focus has been on the deconstruction of the conduit metaphor according to which the interpreter is merely a neutral translation machine through which messages pass between two linguistic codes.Thus, many scholars have concluded that interpreters are in fact active participants in the communicative events mediated by them (see e.g.Angelelli, 2004, pp. 16-21;Berk-Seligson, 2002, p. 96;Eades, 2010, p. 71;Kinnunen, 2010, pp. 127-128;Wadensjö, 1998).Nevertheless, as Angelelli (2008, p. 151) notes, interpreters may not be fully aware of their agency and the consequences of the act of exercising it.
Focus on the interpreter's agency may also contribute to the perseverance of the research tradition of concentrating on alterations made by the interpreter in the translation process.Paradoxically, such stances may echo the views of interpreters who complain about the insufficient language and interpreting skills of their colleagues, service providers who are not aware of the adverse conditions in which interpreters work, and sociolinguists who, while working within the paradigm in which traditional notions of language are deconstructed, appear to see the interpreter through the lens of the conduit metaphor (e.g.Jacquemet, 2009).It is therefore instructive to approach the interpreter's agency from the perspective of the structural and institutional constraints that may affect the performance of interpreting skills.Among the practices through which such constraints are enacted, the discourse of reporting is perhaps the most important.

The discourse of reporting in interpreter-mediated police interviews and screening interviews
While the discourse of reporting is present in all domains of public service interpreting, it is most clearly visible in the legal domain, including the asylum procedurein intertextual chains that start with interviews conducted by law enforcement agents and end with the courtroom trial in criminal cases and with the asylum interview and eventually a hearing in the appellate court in asylum cases.In Finland, interpreters work at all stages of the process: during the interviews, in meetings between the migrant and his or her lawyer, and in the trial or the final asylum interview.
In pre-trial investigations, interviews with suspects and witnesses may be conducted by police, border guards, or customs agents, depending on the nature of the suspected offence.These interviews may last from one hour to several days.In addition, in Finland, the police are in charge of the first interview with the asylum seeker, i.e. the screening interview, the goal of which is to establish the person's identity and verify his or her itinerary as well as the reason for the asylum claim.The actual asylum interview focusing on the grounds for the application is conducted approximately six months later by the immigration services.Screening interviews conducted by the police typically take two hours whereas the asylum interview by the immigration service usually lasts a couple of hours but may also take several days.In an accelerated procedure, the interview conducted by the immigration service may follow the screening interview immediately.
The interpreter does not receive any background material for these interviewsin fact, in many instances no background material is available.Thus, unless the interpreter has been present in previous interviews with the same person, he or she typically knows only that the encounter is an interview related to an offence or an asylum claim, knows the language, and sometimes also the name and country of origin of the person being interviewed.Therefore, while the service provider and other participants such as the legal counsel have access to written background material, helping them in the process of contextualization, the interpreter relies on previous experiences and ad hoc knowledge of the situation.
Written texts are strongly present in all these interviews.To commence, the interviewer explains the rights and duties of the person being interviewed.Sometimes these are read aloud from a written text, sometimes the officer gives the text to the interpreter so that he or she can translate it on sight.Some interviewers require a faithful word-to-word translation whereas others ask the interpreter to summarize the text to the migrant client.In fact, many interpreters keep a copy of the translation they have made previously so that they can just read it aloud.Such practices are greatly appreciated by interviewers and by legal representatives if they are present, for they reduce the time needed for the interview significantly and give the impression that the translation is particularly accurate.However, since the rights and obligations contain complicated legal terms and their syntactic structure is quite complex, an accurate, i.e. word-to-word or literal translation, often means that the message is not understood by the interviewee.Literal translations of legal terms may also be problematic when the interviewer makes references to different laws on which the interview itself is based.
Usually only one interviewer is present: the same person conducts the interview and takes notes.The interviews are not recorded.The interviewee has the right to request a witness (another agent) and a legal counsel to be present.Interviewees almost never request a witness.A legal counsel is present in many screening interviews with asylum seekers but only in fewer than half of the interviews related to suspected offences.If the interviewee is a minor, both a legal guardian and a legal counsel are present.
During the course of the interview, written documents such as copies of airline tickets, mobile phone tracking records, photographs, text messages, or copies of passports may be shown so that the interviewee can comment on them.Reference to other written documents such as laws is made as well.Finally, the transcript is printed so that the interpreter can translate it orally.The importance of this transcript is crucial when the matter is forwarded to the authorities in charge of the later stages of the process, the chief investigator and the prosecutor in the case of suspected offences and immigration service officials in asylum cases.Transcripts corresponding to different interviews therefore become key documents in the pre-trial investigation file or the asylum file.They are summarized in the application for a summons prepared by the prosecutor or in the asylum file sent to the immigration service.In addition, these transcripts are quoted on several occasions during the trial or the actual asylum interview and summarized in the decision prepared by the court or the immigration service.Transcripts of previous interviews are also used as a reference in interviews that consist of several parts.In fact, the production of the written record, the transcript, is the main objective of the interview.
As previous studies on police interviews for example have shown, the written transcript can be very different from what was actually said (see e.g.Bucholtz, 2009;Gumperz, 1982b, p. 166).Firstly, while the interviewee may speak in a way that is quite different from the norms of standard written language, such features are standardized when filtered first through the interpreter and then through the person writing the transcript (see also Maryns, 2006).In addition, features typical of spontaneous spoken language such as false starts, pauses, truncated sentences, or hesitations are inevitably simplified and normalized either by the interpreter or the officer writing the transcript.Long, incoherent stories are summarized and rendered coherent.Erroneous transcriptions of names, dates, and other information are not uncommon.While there are studies about the ways in which such alterations happen and the kind of alterations that may occur in interpreter-mediated interviews (e.g.Gallez & Maryns, 2014;Jacquemet, 2009;Pöllabauer, 2004), it is also useful to reflect on the reasons why these alterations take place.This will be the main theme of the next section.

The discourse of reporting as a normalizing device explaining alterations in interpreter-mediated police and screening interviews
The language ideology of neutrality and uniformity is certainly the overarching principle behind alterations in interpreter-mediated interviews: language use in legal settings is based on the premise that the passage from oral to written text is a neutral one and that fixity achieved by its recording in written form is in fact a guarantee of neutrality.Besides, the unconscious framing that is part of the contextualization process (Gumperz, 1982a, pp. 170-171) plays an important role.Thus, summarizing is typical in screening interviews if the asylum seeker has already claimed asylum in another EU country or comes from a country considered to be relatively peaceful or when the grounds for the asylum claim are thought to be weak for other reasons.In interviews with suspects of an offence, however, the amount of evidence gathered to build the case or unconscious ideas about the veracity of the suspicion may guide the transcription process.Sometimes, summarizing is due to external factors: the officer is in a hurry because he wants to watch an important hockey match, the lawyer wants to speed up the interview because she wants to go to her summer cottage for the weekend.
While simplification, normalization of non-standard speech, summarizing, and omissions are often related to the interpreter's action, they are not necessarily due to the interpreter's lack of competence (cf.Hale, 2004;Jacquemet, 2009;Pöllabauer, 2004).In fact, many omissions and simplifications are due to interpreter fatigue, for interpreters are rarely offered enough pauses.It would be particularly important for the interpreter to have a break before the oral translation of the transcript.In addition, summarizing is one of the most fundamental strategies interpreters use in order to maintain concentration.Since the interviewer not only interviews, but also writes the transcript, the interpreter is oriented towards an output that is easy to transcribe.Furthermore, the interpreter knows that he or she has to translate the transcript orally after the interview is over and may unconsciously seek words, sentence structures, and a logical cohesion that will be easier to translate in a way that is understood by the interviewee.Indeed, the interpreter automatically makes assumptions about the migrant client's language skills and discursive skills, i.e. his or her familiarity with the linguistic features typical of the discourse of reporting.Therefore, what is conceived as a matter of interpreter competence is often related to the discourse of reporting embedded in the institutional structure of the interview.
At the end of the interview, the interviewee and (in face-to-face encounters) the interpreter acknowledge the accuracy of the transcript by signing it.This leads us to the question as to why the interviewee or the interpreter do not notice or do not want to point out the alterations that the transcript may contain.Several explanations are possible: The interviewee may be tired, may not remember everything accurately, or may not care about the accuracy of the transcript.He or she may also have invented a story that is not true and the story may change several times during the investigation.Thus, in courtroom trials, alleged offenders often blame incompetent interpreters for the inaccuracies in the transcripts of the interviews when they tell a different version of the story to the court.Such claims are common also in asylum interviews conducted by the immigration services and in appellate hearings of asylum cases in the administrative court.However, my data also suggests that many interviewees are not at all familiar with the characteristics and importance of the discourse of reporting; the interview and the transcript of the interview as a genre may be alien to them.For example, they may not understand that during the oral translation of the transcript, the question-answer phase is over and that they are only supposed to listen in order to point out any inaccuracies of the transcript.(This occurred in a police interview with a French-speaking suspect interpreted over the phone in September 2014; the police officer questioned my interpreting skills because I could not convince the migrant (whose French was rather rudimentary) that the question-answer phase of the interview was over).
When the interviewer relates the inadequacies in the interviewee's story to the interpreter's incompetence, it increases the pressure the interpreter feels to make the interviewee's story more coherent, cohesive, and logical in translation.For example, offenders may concentrate on the accuracy of features that are important to them, but irrelevant from the viewpoint of the authorities, and ignore the importance of features that have an essential impact on the outcome of their case.This has been the case in more than a half the police interviews I have interpreted in French and English and interviewers have sometimes questioned my interpreting skills in these situations.Such problems are particularly salient in interviews conducted via telephone and with persons whose literacy skills or resources in the language they use are very modest.Interviewees may not realize the importance of the interview and its transcript in the intertextual chain in which the transcript becomes a vital link, where it becomes a key document dictating the way in which their case will be handled subsequently.Further, the interviewer may refuse to take into account the interviewee's observations concerning the accuracy of the transcript, as I will illustrate later in this section.
The interpreters' code of ethics and general ethical consciousness dictate that inaccuracies that are clearly due the interpreter's own error must be corrected.In many cases, however, the issue is not clear and the interpreter's agency is often restricted.For example, it may be quite difficult to know which one of the several possible interpretations of a polysemic word is the most accurate in a given situation (cf.Jacquemet, 2009, p. 538).In addition, 'errors' related to a choice made by the interpreter in translating a polysemic word may be explained by the constraints of the situation: the interpreter has to make rapid decisions to provide an accurate translation even when the lack of cohesion and coherence in the migrant's story constitutes a major hindrance for a successful contextualization guaranteeing full understanding.The speaker may also be incoherent and illogical for reasons unknown to the other participants and perhaps also to him or herself.For example, interviewees may not perceive the features that appear to others' eyes as inconsistency, inaccuracy, or an illogical course of events, for their story makes sense in their own minds (cf.Blommaert, 2001Blommaert, , 2006).Nevertheless, it can be quite difficult to convince law enforcement agents that features that do not exist in the language of the administration may also cause semantic ambivalence.In an interpretation between Finnish and French, for example, such features include grammatical gender, because Finnish has no gender.Thus, for example in a police interview that I interpreted by phone in October 2014 the officer asked the interviewee several questions related to a person who had a unisex name.When I asked the officer whether the person was male or female so that I could use adequate pronouns in French, he replied that I was not supposed to know this.
The interpreter may also feel that his or her remarks are regarded as arrogant or inappropriate, or sense that the interviewer considers comments and clarifications to be a demonstration of interpreter incompetence.One strategy of hidden resistance interpreters may adopt in order to raise the migrant's awareness about the insufficiencies or inaccuracies of the record is to modify their tone or look at the interviewee to indicate that something is wrong while orally translating a transcript in which there is a blatant error.However, when interpreting remotely, interpreting inaccuracies in the transcript have to be indicated by questions, thereby constituting a direct intervention, as exemplified by the notes I took in April 2013 after a telephone-interpreted police interview with an English-speaking asylum seeker who wanted to add details to his asylum file.As we can see, the intervention was unsuccessful because the officer refused to correct the transcript: ( Sometimes interpreters may also take a risk and claim that the error was made by the interpreter or even by the interviewer, saying that the transcript does not correspond to what they said.This strategy is illustrated by the following notes, taken after a screening interview with a French-speaking minor in January 2015.In this case, the intervention was successful: (2) Asylum seeker: "On a pleuré quand on a compris que c'était notre soeur." [We cried when we realized that she was our sister.]Transcript : Itkimme, kun tajusimme, että nainen oli meidän tätimme.[We cried when we realized that the woman was our aunt.]Interpreter: "En tulkannut "täti" vaan "sisko."[I did not use the word "aunt" in my interpretation, I used the word "sister."]Indeed, the fact that the interpreter has to sign the transcript increases the interpreter's agency.At the same time, the signing implies increased responsibility.In telephone-interpreted interviews, on the other hand, this agency and responsibility are missing because the interpreter does not sign anything.
The transcript may also contain alterations such as omissions, inconsistencies, or addendums of modifiers such as 'naturally', 'of course', or 'evidently' which create a form of cohesion that the spoken story did not have and may give an unfavourable image of the interviewee's case.It is not always possible to render such subtle nuances in the oral translation, although general ethical responsibility would require it, for languages have different means for conveying them.For example, in Finnish nuances can be created not only by adding or omitting actual words but also by altering the word order.Metalinguistic comments about such alterations are not possible either: the interpreter is not recognized as a language specialist with expertise extending beyond the meaning of words.
Furthermore, the interpreter is supposed to translate faithfully without interfering with the content of the message, as stated in the Code of Conduct of Community Interpreters in Finland (SKTL, 2013).Therefore, the interpreter is not responsible for the inaccuracy of the transcript unless it is clearly the result of an inaccurate translation of which the interpreter is conscious and certain.In fact, the person responsible for these problems is the migrant's legal counsel.If the interpreter corrects the transcript in such a situation, the lawyer may tell them that they are not responsible for correcting the errors.For example, in a screening interview with a French-speaking asylum seeker in February 2014, the lawyer made this remark although my intervention actually saved her client from accelerated procedure followed by expulsion.
Thus, the decision to correct an error in a transcript may result in a conflict between the requirements of professional ethics, general ethical responsibility towards a fellow human being, and the interpreter's own sense of professionalism and professional future in the sense of less work if the service providers feel that the interpreter intervenes too much and report this to the interpreting agency.Therefore, the consequences of the interpreter's decisions do not only affect the migrant but also the interpreter in the form of increased ethical stress, general work stress, and potential vicarious trauma.
However, interpreters are not immune to contributing to the framing and stereotyping that are embedded in the contextualization process and human interaction in general: otherwise it would be impossible for them to process the continuous, chaotic flow of information in interpreted interaction.As studies on literary translation of sociolinguistic variation have shown (Määttä, 2004), it is impossible to find exact equivalences between non-normalized varieties and idiolects in two different languages, for the ideological constellations and cultural connotations of such varieties can never be the same in two different cultures and contexts.Therefore, it is inevitable that phenomena such as "styles" or "authentic features" (Gallez & Maryns, 2014) are altered in the process of interpretation and transcription.Similarly, it is impossible to interpret incoherent, spontaneous oral speech accuratelyeven in simultaneous interpreting with headphones and microphones this would be quite difficult.Therefore, codes of conduct such as the one used in Finland (SKTL, 2013), requiring an accurate translation of colloquialisms, proverbs, and "cultural messages," describe an ideal state that cannot always be achieved.
In brief, the transcript, which is supposedly a neutral account of the suspect's, witness's, or the asylum seeker's actual speech, is in fact a polyphonic product composed of the voices of the speakers, including the interpreter's, as well as voices originating from adjacent written texts such as laws, regulations, and pieces of evidence referred to during the interview.All these are filtered through framing and stereotyping processes that are embedded in human communication abilities on the one hand, and the normalizing, summarizing, and selection processes that are embedded in the techniques through which oral language is transformed into a written text on the other hand.
The interpreter is a crucial agent in the process in which oral texts and written texts intersect.The question then arises regarding the extent to which the interpreter is able to create a context and is responsible for creating a context that enables the understanding not only of the words of the message but also its function and consequences.In other words, to what extent can interpreters also interpret the discourse of reporting and bridge pretextual gaps within the limits of their agency?One solution would be a code of ethics that forbids orally translating written documents.In the short run, this solution would certainly make the interpreter's task easier.At the same time, it would mean that interpreters have less work, for (in Finland) the oral translation of written documents is often a substantial part of each assignment.Translators, on the other hand, would have more work.However, from the service providers' point of view, such a solution would be disastrous, for it would make judicial processes significantly more complicated, prolonged, and costly, at least in a country like Finland where incoming migrants almost never speak the language of the host country.In addition, current regulations require police officers and immigration officials to use interpreters even in cases in which the service provider and the migrant or the asylum seeker share a common code, the English language for instance.
An additional problem related to increased costs caused by the written translation of all transcripts could be drop in interpreter income.In fact, increased translation costs could easily result in attempts to decrease interpreting costs and therefore also interpreters' earnings.
From the migrant's point of view, written translations would not necessarily entail an improvement either.Firstly, processes such as the asylum procedure and the pre-trial investigation would become lengthier and therefore more stressful.Secondly, while written translations tend to be more accurate than oral translations, increased accuracy of legal terminology, syntax, and other typical features of legal texts does not automatically imply that the migrant or the asylum seeker understands the text better because they may not be familiar with such features.This problem is particularly salient in situations in which the migrants use a language that is not their language of primary socialization, which is the case for most migrants using French and English in Finland.
Paradoxically, prohibiting oral translation or 'interpretation' of written documents would, on the one hand, emphasize the delegation of agency from persons to the written document.Sustainable solutions maximizing the agency of the interpreter, on the other hand, depend on the action of the service provider.Thus, the interpreter would be in a better position to provide overall accurate translations that reflect not only words but also the function and the consequences of the message, i.e. the discourse, if the service provider produced sufficiently simple written documents and used them in ways that would allow the interpreter to concentrate on translation between two languages rather than between two modes of communication, written and oral.Indeed, if we want interpreters to occupy subject positions from which linguistic justice can be enacted (cf.SKTL, 2013) and power relations negotiated, a critical reflection on the nature and function of language and multilingualism as well as the consequences of different ways of using language is in order.Such a reflection should be part of both interpreter and service provider training.

Conclusions
In this article, I have argued that the discourse of reporting, i.e. the explicit and implicit presence of written texts, can be a useful tool for explaining many of the problems present in interviews conducted by law enforcement agents in criminal and asylum settings.In the analysis, I have tried to combine the points of view of the interpreter, the migrant or the asylum seeker, and the service provider.However, my main focus has been on the interpreter.
This article is also an attempt to pinpoint possible interfaces between sociolinguistics and interpreting studies.In fact, while the language of law in interviews conducted by law enforcement agents in police and asylum contexts has been analysed in both disciplines, the approaches are rarely the same.Interpreting studies typically focus on issues such as the interpreter's agency and the accuracy of the interpretation whereas sociolinguistics concentrates on larger social constraints that affect language use, on the one hand, and the viewpoint of the migrant, on the other.The intersection of these two (and other) disciplines could constitute a particularly enriching source of knowledge.
In summary, the article claims that oral, written, and multimodal genres typical of the discourse of reporting, such as interviews, transcripts, and reports, rooted in practices such as recording and transcribing, are powerful instruments in the reification of language ideologies that emphasize the illusion of neutral, uniform language as a guarantee of equal treatment.Reified through the discourse of reporting, such ideologies seriously limit the interpreter's agency.Such ideologies also completely obliterate the fact that language is actually the instrument through which power relations are negotiated.An attempt to deconstruct these ideologies would constitute an attack against the very foundations of fields of activities such as law enforcement, interpretingand interpreting studies.Perhaps this article constitutes such an attack.But it is a necessary attack, for it is only through confrontation that a reflection on these ideologies can be initiated.A thorough, critical reflection focusing on language, which represents the very essence of the interpreting practice, is necessary if sustainable changes are sought.
1) Interpreter (orally translating the transcript): They told me that if I don't do what they tell me to do, they will threaten my sister in ---.Is this correct?Asylum seeker: Just one point: it was written "my sister would be threatened."I said they would kill my sister in ---, I would never see her again.Interpreter: Yks juttu, siinä sanotaan, että siskoani uhattaisiin.Minä sanoin, että he tappaisivat siskoni, enkä enää näkisi häntä.[One point: it was written "my sister would be threatened."I said they would kill my sister in ---and that I would never see her again.]Police officer: Niinhän tässä sanotaan.[That's what it says here.]Interpreter: That's what it says on the paper.