The Role of Linguistic Impression Management: The Case of Hungarian Courtroom Discourses

This paper aims to present various linguistic strategies of impression management used by Hungarian judges, defendants and witnesses, taking into consideration the characteristics of Hungarian culture, language and legal system. In courtroom proceedings the way the participants construct an impression depends greatly on their institutional roles. Hungarian law belongs to the family of continental systems and the judge has an active role in the questioning. In contrary to the expansively investigated Anglo-American cross-examination where witnesses are questioned by the opposing parties, in Hungarian courtroom proceedings the judge is the first to ask questions from the witnesses and the defendants in an unbiased way. The judges are able to motivate the witnesses to cooperate with them by defending their positive impression construction. In addition, the witnesses can be placed in a more advantageous position as a result of an impression management strategy applied by the judges. The research is based on a corpus collected by the author. It consists of ten Hungarian trials recorded using a voice recorder. Based on the corpus, five types of linguistic impression management strategies were investigated: raising or conveying interest, emotional identification with another’s assertion or attitude, highlighting socially approved values, self-superiority and protection, as well as repair work. This research aims to demonstrate the key roles of linguistic impression management strategies in Hungarian courtroom questioning where the judge questions the participants unbiasedly.


Introduction
This paper aims to point out the important role of linguistic impression management strategies used by judges, defendants and witnesses in a continental type of courtroom proceedings, taking into consideration the characteristics of Hungarian culture, language, and legal system. This article focuses equally on defensive strategies (usually associated with classical politeness theories) and assertive strategies (Tedeschi and Melburg 1984). The theoretical framework of the research presented is linguistic impression management theory based on social psychological considerations. One of its main assumptions is that people attempt to act in the way they want to be perceived by others in social interactions (Goffman 1959;Schlenker 1980;Leary and Kowalski 1990;Nemesi 2000). Though impression management is a universal phenomenon, the strategies used to perform linguistic impression management can be considered language-and culture-specific, as well as contextdependent (Nemesi 2000(Nemesi , 2004(Nemesi , 2011Ting-Toomey and Oetzel 2001;Németh T. 2004). Nemesi (2000Nemesi ( , 2011) elaborated a non-taxative dynamic list, mainly based on everyday conversations in Hungarian. This article investigates the occurrence of these linguistic impression management sub-strategies and tools in Hungarian courtroom discourses.
The paper is guided by two hypotheses: (1) compared to witnesses, defendants generally use more linguistic impression management strategies, because their goal is to avoid penalties or to receive less severe ones 1 and (2) in the main part of the questioning where the questions appertain to the crime, judges attempt to minimise the threat by defending the positive impression construction of the person questioned in order to ensure cooperation.
It is necessary to define briefly the concept of cooperation. In this article I consider cooperation as a matter of degrees, a continuum. It is really important at a trial that an interrogated person should not cooperate only in a formal way, but her/his contribution should also be a sufficient contribution to the conversation (Liao and Sun 2017). In Gu's (1994: 181) terminology, there are two kinds of cooperation: a pragmatic and a rhetorical cooperation. In this sense rhetorical cooperation is the crucial one at a trial, because this means that the speaker and the hearer exchange information with each other by adopting each other's goals, while pragmatic cooperation involves only the participants' willingness to listen to each other and their interpreting of each others' message, and usually it is goal-insensitive and situationinsensitive. A witness testimony is one of the most significant evidence; therefore it is vital that a witness will be really cooperative and give all the necessary information (Bócz 2008).

3
The Role of Linguistic Impression Management: The Case of… The structure of the present paper is as follows. Section 2 outlines the theoretical background of the presented research, which involves the demonstration of the current researches in impression management at the courtroom, the presentation of the concept of impression management, the theoretical frameworks, namely, the Two-Component model and the linguistic impression management theory, the relevant characters of courtroom context, and the linguistic impression management strategies rooted in social psychological tactics are also given. 2 Section 3 introduces the Hungarian legal corpus collected by the author. Section 4 presents the analysis of the trials, and the last section summarizes the results.

Theoretical Background
The State of the Art Strategic management of behaviour with the aim of influencing how others perceive us naturally appears in any kind of social interaction (Metts and Grohskopf 2003: 357;Terkourafi 2008: 47;Harris 2011: 103). Consequently, several research traditions can be found in the literature concerned with this phenomenon, such as self-presentation (Goffman 1959;Jones and Pitman 1982), impression management (Goffman 1959;Schlenker 1980;Tedeschi and Melburg 1984;Metts and Grohskopf 2003;Nemesi 2011), facework (Goffman 1955(Goffman /1967, and politeness with two main approaches in pragmatics, namely, classical politeness theories (Lakoff 1973;Leech 1983;Brown and Levinson 1978) and discursive 3 (im)politeness theories (Spencer-Oatey 2005;Culpeper 2005;Terkourafi 2008;Arundale 2010;Kádár 2011;Mills 2011). 4 The followings highlight the most relevant international findings related to courtroom interaction. 5 Hobbs (2003) investigates impression management strategies in lawyers' opening statements and closing arguments, and she points out the importance of the lawyers' attempt to construct a shared identity with jurors to persuade them to affiliate with the lawyer's point of view. Kurzon (2001) analysed American and English appellate judges' verbal behaviour in their written opinions in the framework of Brown and Levinson's politeness theory (1978). He found an important cultural difference in how the judges and their colleagues behave on the same bench and in lower courts: he found far more politeness strategies in the English texts. Sanderson's (1995) analysis demonstrates judges resort to a specific form of deferential address such as Mr. with the person's last name, Sir and likewise overt politeness signals such as please and thank you which are able to mitigate the witnesses' face threats. It is equally important to show to the witness that interviewing them is a cooperative act and the response is not coerced, although they have no real possibility of not answering the question. In regard to this, Sanderson emphasizes the role of negative politeness as per Brown and Levinson's (1978) concept. Harris (2011) also examines lexemes such as please and thank you in the trial of Dr. Shipman. She argues that the consistent usage of these expressions by the court helps them to maintain civility, by indicating that face-aggravation is not deliberately personal or insulting. These lexemes used by judges have a highly mitigative effect, and Harris does not consider they use it sarcastically or as 'mock' politeness. However, it is possible that these forms are the part of the latter behaviours. She supposes defendants may use these lexemes and apologies as 'mock' politeness in some cases, but the analysed transcripts do not involve any indication of prosodic features, therefore this is only a tentative statement (Harris 2011: 91-92). Lakoff (1989), Penman (1990), Archer (2011) and Harris (2011) argue for the role of self-directed strategies besides other-directed facework. These strategies allow the face to be aggravated or depreciated, threatened, protected and also mitigated or enhanced. They could operate simultaneously when a speaker has multiple goals and could play an important role in a courtroom context. Lakoff (1989) states the first step a researcher should make when starting to analyse a type of a discourse is to understand its function. With regards to the courtroom context, face enhancement and face threat are components, not the primary goal. For example, when a lawyer questions their own client-who is the defendant-and they enhance the defendant's face, the lawyer's primary goal is to prove the innocence of the defendant (Lakoff 1989;Archer 2011).
The above mentioned results are based on Anglo-American cross-examination where witnesses are questioned by the opposing parties. In Hungarian courtroom proceedings the judge is the first to ask questions from the witnesses and the defendants in an unbiased way; therefore the judge has an active role in the questioning. For this reason, this article focuses on the strategies used by judges and the paper also investigates the impression management strategies of the defendants and witnesses.

Linguistic Impression Management and Its Social Psychological Basis
According to Verschueren's (1999: 55-69) approach, language use involves continuously making linguistic choices in a conscious or unconscious manner. These choices are made at any level of linguistic form (phonetic/phonological, morphological, syntactic, lexical, semantic) and choices are also made between linguistic 1 3 The Role of Linguistic Impression Management: The Case of… strategies. 6 The phenomenon of impression management has an important role in the process of choosing between the possible alternatives, because the way someone uses language can influence what others think about them (Kiss 1995: 145;Nemesi 2000: 425). People have the ability (so-called ethos) to display a personal character that will serve to enhance their credibility (Rosenberg 2005: 2;Hobbs 2008: 243;Ramírez 2013). The speaker's credibility can be equated with their believability which has a great importance in the courtroom (Archer 2018). A defendant's main motivation to be believed is to avoid penalties or to receive less severe ones. In the case of witnesses, perjury is punishable by imprisonment, hence they also have a good reason to be credible. As for the judges, they have the possibility to shape the laymen's impression which comes from their hierarchical position, and their own impression is also exceptionally important: they represent the court, therefore they should adhere to strict moral rules in order to maintain society's trust in the court. Consequently, in a courtroom, interaction plays an essential role in how people manage the impressions others form of them.
Impression management (or presentation of the self) as a general concept appeared in Goffman's work (1959Goffman's work ( /1967, who introduced it with the following words: "when an individual appears in the presence of others, there will usually be some reason for him to mobilize his activity so that it will convey an impression to others which it is in his interest to convey." On the basis of Goffman's work, Schlenker further develops the concept of impression management, and defines it as a "conscious or unconscious attempt to control images that are projected in real or imagined social interactions" (Schlenker 1980: 6). In his work he differentiates impression management from self-presentation, which is a narrower concept: the latter covers only the impressions which are projected from the self, while impression management involves the goal-directed control of the outer image of not only the actor, but of other persons, associations, entities, and ideologies (Leary 1995;Nemesi 2011Nemesi , 2004. However, later works use these concepts interchangeably, because it is rare for a projected image to ignore all kinds of self-content (Nemesi 2011).
There is a slightly more significant difference between classical politeness strategies originate in Goffman's face concept and impression management. The latter involves not only the face-saving acts from the hearer's perspective, it also highlights the importance of the interplay of private self-processes and interpersonal factors. It includes individual face constructions or self-identifications with all kinds of processes, means and results involved in showing oneself to be a particular type of person, thereby specifying one's identity (Nemesi 2004(Nemesi : 13, 2011. For this purpose, impression management behaviours can be divided into two main categories: defensive and assertive tactics (Tedeschi and Melburg 1984;Ellis et al. 2002;Archer 2018). Defensive impression management tactics are associated with the notion of corrective facework and remedial work, both of which are able to protect or repair one's image by limiting the loss of positive self-image when an individual feels it is under threat (Archer 2018). Assertive impression management tactics are promotional behaviours which are adopted to acquire and promote favourable impressions (Tedeschi and Norman 1985;Ellis et al. 2002). These tactics can mutually overlap; for instance, admissions of responsibility have a defensive nature, but can be categorised as assertive tactics too, because they can contribute to one's positive self-presentation by reflecting positive moral traits (Schlenker and Weigold 1992;Archer 2018).
The concept of impression management describes a general human ambition, but it does not determine what would be an individual's desired image of him/herself to present in a concrete situation. Numerous social psychological research studies have focused on providing an answer to this question (Tedeschi and Reiss 1981;Jones and Pittman 1982;Leary and Kowalski 1990). In this paper, I use the Two-Component model (Leary and Kowalski 1990;Leary 1995) for my analysis, which considers role constraints that have a great importance in the highly institutionalised courtroom conversation.

The Two-Component Model and Its Elements
Leary and Kowalski divide impression management into two main components: impression motivation and impression construction. 7 Impression motivation is determined by the goal-relevance of the impressions, the value of the desired outcomes and the discrepancy between one's desired versus one's current social image. It is worthy of note that another person can never know for sure what someone else's degree of impression motivation is. In spite of this fact, we can investigate goalrelevance which is a determinative factor.
The second main component of impression management is impression construction, which involves (1) decision making about how the impression should be created and (2) five decisive factors. The first two are intrapersonal variables. These are self-concept and the desired and undesired identity images which represent the private self's considerations. People prefer to suggest impressions which harmonise with their self-concept. The other three factors are interpersonal variables: firstly, role constraints such as the judges', the defendants' and the witnesses' roles,

3
The Role of Linguistic Impression Management: The Case of… secondly, target values, which mean that people tailor their public images to the perceived values and preferences of their communicational partners, and thirdly, the current or potential social image, since the impressions people try to create are affected both by how they think they are currently regarded by others and by how they think others may perceive them in the future.

Courtroom Context
In the courtroom context role constraints are decisive elements of impression construction. Judges require politeness, sincerity and cooperation from witnesses. The first is provided by the hierarchical situation, which also depends on the social status, the institutional status and the individual's role in the concrete conversation (Fairclough 1989;Hámori 2006;Nemesi and Szombathelyi 2009). The requirement for sincerity and cooperation can be derived from Hungarian criminal procedural law, civil procedural law and criminal substantive law. 8 As for the defendants, the required characteristic is also politeness, but there is an interesting difference regarding sincerity. If a defendant lies, there are no criminal consequences. 9 On the contrary, any witness who gives false testimony before the authority concerning an essential circumstance of a case, or suppresses evidence is guilty of perjury. Furthermore, defendants are not obliged to make a testimony during a courtroom trial. If a defendant refuses to make a testimony, the presiding judge reads out or presents the testimony which they had already made in the pretrial investigation. In general, defendants refuse to testify, and if this is the case, no further questions may be asked from the defendant in respect of the criminal offence s/he refused to testify. In other words, defendants cannot be forced to cooperate or tell the truth, cooperation and truthfulness are not required from them (Bócz 2008;Belovics 2016).
Let us turn to the institutional role and the constraints it places on the judges. In the Hungarian legal system the judge's role is to interrogate the defendants and question the witnesses in an unbiased way. The judges should adhere to stricter moral rules than generally accepted ethical norms. Act CLXII of 2011 on the Legal Status and Remuneration of Judges involves constitutional requirements. The following points present only the most relevant obligations, such as impartiality, fairness and independence: (1) Judges shall act impartially in all cases and shall not be influenced in any way.
(2) During proceedings, judges shall conduct themselves in a fair and impartial manner with their clients. (3) Judges shall conduct themselves every time in an impeccable manner worthy of their office, and refrain from any manifestations which would undermine the trust in judicial proceedings or the authority of the court.
Furthermore, ethical rules can be found in the Code of Judicial Conduct. 10 The most cardinal and topic-relevant characteristics of judges presented in the ethical code are as follows. A judge should be independent, impartial (from parties and politics), polite, respectful, dignified, patient, diligent, prepared, humble, cooperative, rule-governed and non-discriminative.

Linguistic Impression Management Strategies
On the basis of Leary and Kowalski's (1990) social psychological categorisation and classical linguistic politeness theories, Nemesi gives a non-taxative dynamic list of linguistic impression management sub-strategies which people usually use in everyday conversations to present themselves linguistically in the way in which they would like to be seen (Nemesi 2000(Nemesi , 2011Németh T. 2004: 404). The content of these sub-strategies is mainly based on conversations from Hungarian everyday situations. The five main categories are: (1) Raising or conveying interest: The speaker tries to speak about interesting topics and/or resort to interesting linguistic expressions such as hyperbole or expressive linguistic elements.
(2) Emotional identification with another's assertion or attitude: includes an attempt to reach a consensus between conversational participants. The expectations related to making a consensus are culture-specific and refer to the hearer confirming, affirming and accepting the speaker's utterance, or agreeing with the speaker.
(3) Highlighting socially approved values: People have to adapt themselves to written and unwritten expectations, and also have to be observant to the norms of a group in order to stand out. This strategy involves different communication attempts, which could be both universal and specific to a concrete speech community. This strategy means, among other things, the presentation of intelligence, education, diversity, humour, intellect, playfulness, politeness and modesty. (4) Self-superiority and protection: This involves enhancing the speaker's own face while aggravating the partner's face. It means querying the other's opinion or disparaging the partner. There is an interesting observation about expressions such as "not really" which is typically a euphemism with regard to the speaker, but sounds ironical when used for another person (e.g. 'I am not really good at cooking' vs. 'She is not really good at cooking'). (5) Repair work 11 : This strategy is based on Goffman's remedial work (1971) and Suszczynska's repair work (2003). The speaker breaks a social norm with his/ her behaviour, they are made responsible for it, and then try to compensate for it. In this strategy two intentions can be recognised: (1) the speaker tries to repair 1 3 The Role of Linguistic Impression Management: The Case of… the face damage they have suffered from a face-threatening act, or (2) tries to redress the harm that they have caused to the conversational partner. Nemesi (2011: 122) differentiates two sub-categories in this strategy: (a) Admission of responsibility: apologising and making excuses, self-accusation, lack of intent, justification of the partner, expression of disorder, offer of reparation, promise of future restrain from similar actions. (b) Denial of responsibility: recognition of the fact without recognition of responsibility, disclaimers to avoid responsibility, providing justification, offering plausible reasons. Nemesi (2011: 123) emphasises that this list is only one categorisation from the possible alternatives; furthermore, it is not a closed, taxonomic and culture-independent classification. As mentioned in Sect. 2.1, some researchers (e.g. Tedeschi and Melburg 1984;Ellis et al. 2002;Archer 2018) make a dual distinction and differentiate assertive and defensive strategies. However, this categorisation is also not a strict one. As an illustration, admissions of responsibility are generally associated with a defensive strategy, but can be included in the group of assertive tactics too, because it can contribute to one's positive self-presentation by reflecting positive moral traits (Schlenker and Weigold 1992;Archer 2018). In addition, this observation highlights the multiplicity of these strategies, and also the way in which participants interpret these strategies.
The next section describes the corpus of this research, then Sect. 5 turns to the analysis of interrogation extracts.

The Legal Corpus
The analysed corpus consists of the recordings and written notes of six Hungarian criminal trials and four civil trials from 2017, all held in Hungary. The criminal cases involve offences including battery, money laundering, budget fraud, and the civil cases involve placement under custodianship, claim arising from contract, as well as payment of a pecuniary debt and usage fee. Regarding to my research permit, I was allowed to observe the proceedings personally and record these trials with a voice recorder (Olympus WS-831), as well as to write notes on non-verbal communication in the courtroom. The excerpts from these trials have been anonymised and transcribed using the Conversation Analysis method (Jefferson 1984). 12 Because of the Secrecy Obligation, below the ten trials are described generally. There were eight defendants, fourteen witnesses, two presiding judges and three single judges involved in the ten trials altogether.
(1) Interrogations of eight defendants: six persons refused to testify, but all of them were cooperative in answering questions about their personal circumstances (e.g. marital status, public duties, financial circumstances etc.). The refusal has no consequences in terms of their right to question or make objection or motion. The other two defendants did not refuse to testify; therefore they answered the substantial questions too which connected to the criminal offence.
(2) Testimonies of fourteen witnesses: If there is no obstacle to the testimony, the presiding judge (criminal case) or the single professional judge (civil case) shall question the witness. All of these fourteen witnesses had to answer the questions put to them. (3) Two presiding judges in criminal cases and three single professional judges in civil cases: The judge plays an active role in the proceedings, s/he shall question the witnesses and in criminal proceedings the defendants too.
The next section turns to the analysis and presents the most frequently applied linguistic impression management sub-strategies and the linguistic tools used in courtroom discourses, based on the collected data.

Impression Management Strategies Applied by the Defendants and the Witnesses
At first the analysis regarding Hypothesis (1) will be introduced: Defendants generally use more types of linguistic impression management strategies than witnesses, because their goal is to avoid penalties or to receive less severe ones.
Impression management strategies appear in two frequently occurring situations involving the witnesses. Witnesses usually use a defensive strategy when they feel their own responsibility in the case. In the first example the judge is questioning the security guard who works at the pub where the crime occurred.
'Well, maybe his side, actually not exactly, I was more interested in asking him whether he was feeling okay.' For this yes/no question the security guard did not want to say explicitly that he did not ask the victim for this piece of information. Instead, he uses a euphemistic selfdefensive expression and he says: "actually not exactly." This is an answer denying responsibility with providing justification: he did not ask this question specifically, but he points out it was not necessary because he was inquiring about the victim's well-being. The second type of linguistic impression management strategy used by witnesses is highlighting socially approved values in situations in which the witnesses themselves do not feel cooperative or sincere enough.
'Why did you claim then, during the evidentiary procedures, that the two men attacked each other simultaneously?' Witness: Nem tudom, nem tudom. Próbálok én segíteni.
'I don't know, I don't know. I am really trying to help.' In this case the witness explicitly indicates that she is telling the truth and trying to help to the best of her knowledge. In the courtroom, cooperativity is vitally important, so the witnesses need to show willingness to cooperate.
Contrary to these findings, the defendants use many more linguistic impression management strategies, as for example, in the following excerpt: ( In Hungary, the maximum penalty for the crime in the example above is five years, therefore it is clear that the defendant is using hyperbole, which could incline the hearer to identify with him emotionally (Nemesi 2004(Nemesi , 2009). The defendant applies the strategy of emotional identification with other's assertion or attitude: he would accept even 100 years as a sanction if he really had committed the crime he is accused of. This would be "fair," therefore he would not appeal. But then he denies responsibility with providing justification why it is obvious that he could not have committed the crime. He actually lacks the necessary knowledge of the law and the economy. Nevertheless, he has "common sense"; therefore he would not have made such a bad deal if he had known how much the item was actually worth. Beside this, he shows respect with the overt politeness signal "Your Honour" and apologises for his incompetency in law and economics. In example (4) the judge calls for silence several times, because the defendant consistently interrupts the proceeding. The defendant uses a hyperbole ("I will not say a word anymore") to show his conformity to the rules. It is a hyperbole, because defendants have the right to ask questions, make objections and motions during the trial; therefore it is not required that they keep quiet till the end of the trial. 13 The defendant was advised to these rights by the judge several times, but evidently defendants are not allowed to disturb the court procedure. In short, the defendant tries to highlight socially approved values with showing respect to the rules and using overt politeness signals ("Your Honour" and "thank you for your patience"). In fact, his respect is only formality because he cannot keep quiet and he interrupts the proceeding over and over.
'I will not say a word anymore, Your Honour, thank you for your patience.' Defendants often express respect toward judges with overt politeness signals such as thank you or Your Honour, etc. This explicitness in the expression of politeness is not so salient in the case of witnesses. Defendants also frequently express regret about what they did, as in the fifth example: (5) Defendant: Én nem győzöm hangsúlyozni, nagyon sajnálom, hogy Géza bácsi meghalt, de én nem tartom magam ezért felelősnek.
'I can't emphasise enough how sorry I am that uncle Géza died, but I believe I am not responsible for that.' The defendant uses the defensive strategy of denying responsibility by recognising the fact without accepting responsibility: he acknowledges the death of uncle Géza, but does not take responsibility. He expresses his solidarity ("I can't emphasise enough how sorry I am") and uses the form "uncle Géza," a form which expresses emotional identification. Defendants frequently use the denying responsibility strategy by explaining their acts or making reference to an external cause. For example, in excerpt (6) the defendant says he only wrote in the inventory what his boss told him to write.
'I only wrote in the inventory what I was told to write.' As a last example of Hypothesis (1), the defendant agrees with the statement of the prosecutor, and therefore uses the strategy of emotional identification with another's statement.

3
The Role of Linguistic Impression Management: The Case of… In the first utterance of excerpt (7), the defendant not only agrees with the statement of the prosecutor, the utterance could be interpreted as complimenting the prosecutor's good work, so the defendant enhances the prosecutor's face. In his second utterance, referring to the prosecutor's statement, he wonders why the court thinks he has got ten million forints, if the prosecutor stated that everything was a fiction (although what the prosecutor is referring to here are the accounts).

Impression Management Strategies Applied by the Judges
Now let us move on to Hypothesis (2): Judges attempt to minimise the threat by defending the positive impression construction of the person questioned in order to ensure cooperation.
In the corpus, I did not find any impression management strategies used by judges in order to facilitate cooperation during the testimonies of the defendants. The reason is that defendants have right to silence and not to incriminate themselves, and these rights mean a general restriction in the enforce of cooperation (Belovics 2016). These are the rights of all defendants, and as excerpt (8)  We can see that the judge did not apply any strategies here to facilitate the cooperation of the defendant, but the defendant uses several linguistic impression management strategies: making excuses and apologising ("I apologise", "I am sorry", "I regret everything", "I am going to cry"), showing socially approved values such as repentance and regret, and also admitting responsibility. Furthermore, he emphasises his lack of intent ("I still have no idea how this whole thing happened", "We just wanted to go out") and also indicates that his lack of intent was due to the good relationship between them ("he was my best friend") which they have preserved ("We stayed good mates", he addresses the victim as "bro"). He justifies his friend (the victim was good to him, therefore he does not deserve the harm he caused-"he even brought me a present"). We can only say that the judges use overt politeness signals during the questioning of the defendants and the witnesses, too. The most frequently applied linguistic elements used by judges to express politeness are the following: (1) They use the form tetszik + inf. 'like + inf.' as an attitude deixis which expresses the social relation between the conversational partners (Tátrai 2010: 218;Csontos and Dér 2018: 30). They use it independently of their own and the interrogated partners' gender and age. For instance: Hol tetszik most lakni? 'Where do you like to live now?' or Ilyen nincs, hogy arról állítunk ki számlát, ami nincs. Tetszik érteni? 'It is not allowed to write invoices for things which not exist. Do you like to understand?' In Hungarian this construction can express a high level of formal politeness. This linguistic construction presents the judges' politeness and respectfulness independently of the person being questioned, and also shows their non-discriminative attitude.
(2) The formal you in Hungarian has two forms: ön and maga (and the verb conjugates in the third person) to express the institutionally polite relation between the participants. These elements can be omitted in Hungarian; however, these formal pronouns are an organic part of judges' questions. For example: Ezt a tíz számlát akkor Ön valóban kiállította ennek az ismerősnek? 'Did you actually write these ten invoices for your acquaintance?' 14 Mit tudott maga? 'What were you aware of?' Milyen problémája lett ebből magának? 'What problems did it cause to you?' Hol tetszik, melyik büntetés-végrehatási intézetben van most? 'Where do you like to be, where are you now, in which prison?' In the corpus judges consequently used the more formal ön variant with witnesses, while both of them appeared when they talked to the defendants-i.e. with the same judge and the same defendant. The ön form is more formal and distanced, while maga is still formal, but more familiar (Tátrai 2010). One possible explanation is that the more distanced ön defends the negative face of the witness in the sense of Brown and Levinson (1978) and therefore it might be able to ensure cooperation. Similarly, Sanderson (1995) emphasises that the interviewee should feel their response is not coerced, even though there is no real possibility to avoid the answer. On the contrary, defendants have no obligation to answer questions, they can say at any time during the interrogation that they will refuse to answer. The trial aims to decide whether the defendant is guilty or not, and it will happen even without the contribution of the defendant. Therefore the more familiar maga is also more neutral than the ön in this sense in the courtroom context.
However, I also found two linguistic impression management strategies used by the judges during the questioning of the witnesses. The witness is young and a bit shy. The judge uses the expression "this much we are sure of, aren't we?" meaning himself and the witness, using the we inclusive person deixis which expresses a sense of unity and belonging to the hearer (Wodak et al. 2009: 45;Tátrai 2010: 216;Csontos and Dér 2018: 30;Varga 2019). The judge expresses mutually gathered information in an understanding way and uses emotional identification with the assertion or attitude of a partner strategy. Using this strategy the judge expresses her cooperation and mitigates the threat of the obligation to answer.
The last example shows the second type of strategy. The judge tries to enhance or defend the witnesses' positive face in the interest of cooperation. She makes a repair: denying responsibility with an explanation. We can see in (10) the judge saying to the witness, who is a paramedic, that it is not a problem if he cannot answer the question, because it is not part of his job to know.
'But is it only large quantities of alcohol that has an effect on blood coagulation or not?' Witness: Hát egész kicsi (1.0) szerintem annyira nem, de ezt mint csak mentős mondom.
'Well, it's very small (1.0), so I think not really, but I'm only speaking as a paramedic.
'Do you mean to say that alcohol has less of an effect on blood coagulation or the exact opposite?' Witness: Hát nem tudom pontosan.
'You don't know exactly. Okay, that is not a problem, it is not your job to know, I was just asking.'

Conclusions
This paper has investigated linguistic impression management strategies in the context of the characteristics of the Hungarian legal system, culture and language from two perspectives: (1) strategies used by defendants and witnesses to start off their impression motivation in the courtroom, and (2) strategies used by judges to minimise the threat by maintaining and defending the positive impression construction of the person questioned in order to heighten the cooperation. The analysis contributes to emphasise the importance of impression motivation in applying an impression management strategy. Defendants use substantially more types of linguistic impression management strategies than witnesses, because their goal is to avoid penalties or to receive less severe ones. Witnesses use impression management strategies (1) to protect themselves when they feel their own responsibility in the case lies mostly in providing a justification and offering a reason or (2) to highlight socially approved values such as sincerity and cooperation. As for defendants, (1) they use all of the sub-categories of repair work such as denying responsibility and admitting responsibility, too, and (2) they highlight socially approved values such as formal politeness and respect with overt politeness signals and solidarity, although in some cases (e.g. excerpt 3) showing incompetence could be a part of arguing why they have not committed the crime, and (3) to emotionally identify with another's assertion or attitude could be a defensive strategy (example 5) or an attempt to enhance another's face in order to benefit them (example 7). As far as the second hypothesis is concerned, the judges use linguistic impression management strategies to defend and enhance the face of the witness in order to ensure cooperation (1) through emotional identification with another's assertion or attitude and (2) by denying their responsibility in the case to defend and/or enhance the witness' face. In contrary to this, in the corpus judges do not use linguistic impression management strategies in order to ensure cooperation with defendants, but this phenomenon greatly depends on the rules of criminal law, because the defendants' right to silence and not to incriminate themselves makes obligations to the judge regarding the facilitation of cooperation during the testimony of the defendants. Although the presented research is limited to ten trials, the results highlight some significant aspects of linguistic impression management executed by the participants in a continental legal system. A considerable part of the previous international literature on impression management at the courtroom investigated cross-examination, where judges have different role in the proceedings. Consequently, one of the main aims of the presented article was to put focus on the strategies and tools that judges apply during the testimony in order to increase the cooperative willingness of witnesses, and give motivations for further investigations in the topic, because witness testimony is one of the most important evidence.