Contribution to the Analysis of Witness Statements in the Croatian Language

Abstract

Forensic linguistics might be defined as a branch of applied linguistics interested primarily in the interaction among language, crime and the law. The term itself was coined by Svartvik in 1968. Initially, the growth and development of forensic linguistics was relatively slow. However, in the past 15 years there has been a significant growth in the frequency with which the courts in many countries have called on the expertise of linguists. Being a relatively new discipline, it has not yet been strictly defined which sub- disciplines belong to its ambit. Forensic discourse analysis is concerned with institutional discourse in a variety of legal settings and its intersection with lay and social meaning, including selection and use of grammar and vocabulary in certain legal genres. This paper deals with witness statements in litigation proceeding before municipal courts in Croatia and analyses their syntactical and lexical features in order to determine which functional style(s) they belong to. In the Croatian language, witness statements are characterized by a functional style hybridity, since they exhibit characteristics of the administrative and the conversational functional style. In some countries, such and similar analyses were carried out on larger corpora of witness statements in order to draw a style profile that enables observing inconsistencies that might be used as a clue in veracity analysis of witness statements and as an aid to investigation. In order to obtain a comprehensive image of the witness statement legal genre in the Croatian language, which could, among other things, be used as an indication in establishing veracity, a significantly larger corpora, as well as witness statements from various types of legal proceedings and various types of courts, should be collected and analysed. [PUBLICATION ABSTRACT]

Headnote

Forensic linguistics might be defined as a branch of applied linguistics interested primarily in the interaction among language, crime and the law. The term itself was coined by Svartvik in 1968. Initially, the growth and development of forensic linguistics was relatively slow. However, in the past 15 years there has been a significant growth in the frequency with which the courts in many countries have called on the expertise of linguists. Being a relatively new discipline, it has not yet been strictly defined which sub- disciplines belong to its ambit. Forensic discourse analysis is concerned with institutional discourse in a variety of legal settings and its intersection with lay and social meaning, including selection and use of grammar and vocabulary in certain legal genres. This paper deals with witness statements in litigation proceeding before municipal courts in Croatia and analyses their syntactical and lexical features in order to determine which functional style(s) they belong to. In the Croatian language, witness statements are characterized by a functional style hybridity, since they exhibit characteristics of the administrative and the conversational functional style. In some countries, such and similar analyses were carried out on larger corpora of witness statements in order to draw a style profile that enables observing inconsistencies that might be used as a clue in veracity analysis of witness statements and as an aid to investigation. In order to obtain a comprehensive image of the witness statement legal genre in the Croatian language, which could, among other things, be used as an indication in establishing veracity, a significantly larger corpora, as well as witness statements from various types of legal proceedings and various types of courts, should be collected and analysed.

Doprinos proucavanju iskaza svjedoka u hrvatskome jeziku

Forenzicka se lingvistika moze definirati kao grana primijenjene lingvistike koja se ponajprije bavi me|usobnim prozimanjem jezika, zlocina i prava. Termin potjece iz 1968. godine (Svartvik). U pocetku se forenzicka lingvistika razvija razmjerno sporo, no u zadnjih je petnaest godina zabiljezen znacajan porast broja slucajeva u kojima sudovi niza zemalja konzultiraju lingviste i pozivaju ih da svjedoce u svojstvu sudskih vjestaka. Buduci da se radi o relativno novoj disciplini, nije jos strogo odre|eno koje joj poddiscipline pripadaju. Forenzicka se analiza diskursa bavi institucionalnim diskursom u pravnom okruzenju te njegovim isprepletanjem s laickim i drustvenim znacenjem, ukljucujuci odabir i uporabu gramatickih struktura te leksika u pojedinim pravnim tekstnim vrstama. Ovaj se rad bavi sintakti~kim i leksickim zna~ajkama iskaza svjedoka u parnicnim postupcima na opcinskim sudovima u Hrvatskoj, kako bi se utvrdilo kojemu funkcionalnom stilu, odnosno funkcionalnim stilovima oni pripadaju. U hrvatskome jeziku iskaz svjedoka karakterizira hibridnost jer sadrzava karakteristike i administrativno-poslovnoga i razgovornoga stila. U nekim su zemljama ovakve i slicne analize provedene na velikim korpusima iskaza svjedoka kako bi se utvrdili stilski profili koji omogucuju uocavanje nepodudarnosti, sto se mòe rabiti kao pokazatelj u analizi vjerodostojnosti iskaza i pomc u istraznom postupku. Kako bi se dobila potpuna slika pravne tekstne vrste iskaz svjedoka u hrvatskome jeziku, koja bi izme|u ostaloga mogla posluziti kao pokazatelj prilikom utvrdivanja vjerodostojnosti, potrebno je prikupiti i analizirati znatno veci korpus te iskaze svjedoka iz razlicitih vrsta sudskih postupaka i s razlicitih vrsta sudova.

Key words: witness statements, legal genres, administrative style, conversational style, forensic linguistics, Croatian.

Kljucne rijeci: iskazi svjedoka, pravne tekstne vrste, administrativni stil, razgovorni stil, forenzicka lingvistika, hrvatski jezik

1. Introduction

In the past few decades the study of language has significantly increased at the centres of higher learning throughout the world. One of the consequences is that linguistics became increasingly concerned with the application of its theories and knowledge to everyday language. One area that has greatly benefited from this approach is forensic linguistics that can be roughly defined as interaction between language and the law. The term forensic linguistics is nowadays widely accepted1 and refers to the branch of applied linguistics interested primarily in the interaction among language, crime and the law. The term itself was coined by Svartvik (1968)2. He analysed the statements of Timothy Evans who was hanged for murder of his wife and baby and posthumously pardoned. In his analysis he demonstrated that incriminating parts of four statements made to police officers by Timothy Evans had a grammatical style measurably different from that of uncontested parts of the statements. Initially, the growth of forensic linguistics was relatively slow. However, in the past 15 years (Coulthard and Johnson, 2007: 5) there has been a significant growth in the frequency with which the courts in a series of countries have called on the expertise of linguists. Consequently, there is now a developing methodology and a growing number of linguists who act as expert witnesses, primarily in English speaking countries. Moreover, there are two professional associations, the International Association of Forensic Linguists (IAFL)3, and the International Association for Forensic Phonetics and Acoustics (IAFPA)4.

Being a relatively new discipline, it has not yet been strictly defined which sub-disciplines belong to its ambit. According to Coulthard and Johnson (2007: 1-5) the central concerns of forensic linguistics are falsification, plagiarism, trademarks5 and hoax, as well as the language of legal documents6, the language of police and law enforcement7, interviews with children and vulnerable witnesses in the legal system8, courtroom interaction9, linguistic evidence and expert witness testimony in courtrooms10, authorship attribution11, as well as forensic phonetics12 and speaker identification13. According to Olsson (2004: 3-5) forensic linguistics, among other sub-disciplines, encompasses authorship identification, legal interpreting and translation, transcribing verbal statements, the language of discourse and courtroom, language rights, statement analysis, forensic phonetics and textual status. Shuy (2004: 4) excludes the areas such as document analysis, handwriting analysis and type-token analysis from the scope of forensic linguistics. McMenamin (2002: 86) argues that the classification of areas in forensic linguistics usually follows existing classification in the structure and function of language and that it evolves as the field develops. However, sometimes even narrower specification is necessary. Furthermore, in some forensic studies a whole range of tools from overlapping linguistic disciplines is used. According to McMenamin (2002: 86) the solution to the classification problem is to allow the research and casework of forensic linguists to define the field and thereby develop its taxonomy.

Forensic linguistics is best developed in English-speaking countries, such as Great Britain, the USA, Canada, Australia and the like, whereas it is in still in its infancy in Croatia, where out of all its sub-disciplines predominantly forensic phonetics research is conducted. The only prominent Croatian phonetician seriously dealing with forensic phonetics is Varo{anec-[kari}, who has published several articles on methods and practices in forensic voice identification14. Moreover, as a phonetician she introduces linguistic comparisons along with phonetic analyses into forensic research on real criminal cases15.

1.1. Legal genres

Since the publication of An Introduction to Discourse Analysis (Coulthard, 1997) descriptive linguistics has undergone many changes, mostly as a result of the creation of massive corpora16 and the rapid development of the World Wide Web. Coulthard and Johnson (2007: 7) explain that the term discourse analysis is nowadays preceded by a whole range of adjectives, such as anthropological, cognitive, educational, feminist and forensic discourse analysis. Clearly, discourse analysis has branched offinto a number of sub-domains, one of them being forensic discourse analysis. It is concerned with institutional discourse in a variety of legal settings and its intersection with lay and social meaning, including selection and use of grammar and vocabulary in certain legal genres. Coulthard's (1994) groundbreaking study of the alleged statement of Derek Bentley is considered to be the birth of forensic discourse analysis. Olsson (2004: 15) claims that Coulthard's study can be regarded as a combination of different linguistic fields, such as speech act theory, corpus linguistics, register and psycholinguistics.

Coulthard and Johnson (2007: 55) define genres17 as conventional, repeated and distinctive features of texts that arise from their communicative purposes. Brinker (2005: 144) defines text types as conventional forms of complex language actions that can be described as typical combinations of contextual (situational), communicative-functional and structural (grammatical and thematic) features. Coulthard and Johnson (2007: 55) stress that genres are domain specific, having lexical and grammatical features associated with them, as well as specific stages usually named and sequenced.

The majority of us are familiar with the major fictional genres and their style and interpretation. However, when it comes to professional genres, naming and defining becomes much more difficult. Professional genres are to some extent closed to lay interpretation. Legal genres can broadly be defined as genres used in legal settings. According to Coulthard and Johnson (2007: 61) style is one of the key defining aspects of legal genres, which are to some extent characterized by functional hybridity, that is, a constant awareness of and interaction with the legal statutes and laws they relate to.

1.2. Functional styles of the Croatian language

The Croatian standard language is multifunctional, which is one of the crucial features of a standard language. The most orderly description of the functional styles of the Croatian language derives from Silic and the articles he published in Kolo journal and subsequently in his book Funkcionalni stilovi (2006). He divides the Croatian standard into five functional styles that differ mutually to a greater or lesser extent regarding their relationships to the norm and the levels of individual freedom. These are: literary-artistic, conversational, journalistic, administrative and scientific functional styles18. However, all authors analysing functional styles stress the hybridity of functional styles, i.e. mutual overlapping of functional styles and strong influences they exhibit on one another. Silic (2006: 74) describes the strong influence of the administrative style on the conversational style, while Francic, Hudecek and Mihaljevic (2005: 253) stress the strong influence of the administrative style on all other functional styles. However, regardless of certain overlapping and mutual influence, the five functional styles of Croatian exhibit specific characteristics judging from which they should be relatively easily distinguished.

The main characteristics of the administrative and conversational style will be briefly outlined here, since they are the most relevant to our paper and analysis.

The administrative style is, according to Sili} (2006: 65-74), topic-related, simple, clear, precise, complete, uniform, explicit, economic, analytical, concrete, objective, short, determined, specific, terminological, non-emotional and rich in clichés.

Silic (2006), Francic, Hudecek and Mihaljevic (2005), as well as Tosovc (2002) have described the main features of the conversational style. The conversational style is primarily characterized by informal, unaffected, spontaneous, mostly oral and dialogical way of expression, carried out in direct or indirect contact in which the participants simultaneously or successively send and receive messages.

2. Hypothesis and methodology

The aim of this paper is to determine characteristics of witness statements as a genre, thus contributing to the analysis of legal genres. The corpus is made up of witness statements written in the Croatian language. Based on the characteristics of the genre witness statements, it will be determined which functional style, that is, functional styles they belong to. The starting hypothesis is that, as a result of the interaction-situational context, that is, the place and way of testifying, witness statements will exhibit elements of the administrative style, as well as of the conversational style. Coulthard and Johnson (2007: 62) explain that the study of context is vital to the understanding and interpretation of legal texts and forensic linguistic analysis.

The corpus consists of 34 court transcripts including 50 witness statements from litigation proceedings. The witness statements are authentic and have been obtained for the purpose of scientific research from the Municipal Court in Labin, Croatia. All personal data of the participants involved in the litigation proceedings in the above mentioned 34 court transcripts have been deleted from the witness statements. The analysis has been focused on the syntactical and lexical levels19, being one of the key criteria in determining the functional style(s) of a particular genre.

3. Analysis and discussion

Witness statements from Croatian municipal courts are written records of witness' sworn evidence given in court. They are part of the court transcripts in litigation proceedings. Taking into consideration the way witness statements in Croatian municipal courts are produced, the term itself might be somewhat misleading. Witnesses, laypersons who find themselves in courtroom settings give evidence; the judge dictates their words to court clerks who write their words down. Thus, there are three participants included in the production of witness statements, which is reflected in the syntactical and lexical level of the witness statements analysed. Regardless of the fact that witness statements in Croatian municipal courts do not reflect the exact words of the sworn evidence given, they are 1st person narratives.

U predmetnoj kuci smo zivjeli ja, moja braca te roditelji. Otac je umro 1941. godine, a mi smo ostali sa majkom. Pamtim da je u istoj kuci sa nama zivjela i baka, majka oca, te da je ona umrla kada sam ja imao cca. 5-6 godina. Pamtim da smo tamo drzali stoku. I takoder smo dugo godina obra|ivali zemlju. Pamtim da je kucu sa okuznicom i nek. kupio moj pok. otac. Ja sam imao cca. 18 godina. Tuzenike sam i osobno poznavao buduci su prije nego je otac kupio kucu dolazili na imanje jednom tjedno. Kako smo obra|ivali zemlju za njih, te drzali stoku to smo ostvarivali pravo na 1/2 dobivene vrijednosti20.

3.1. Characteristics of the administrative style

The characteristics of the administrative style, including its legal sub-style, prevail. The most prominent characteristics of the administrative style found in witness statements are "unintentional" pleonasms and semi-copulative and/ or copulative verbs with complements.

3.1.1. Pleonasms

A pleonasm can be defined as an instance of redundancy, that is, the use of more words than required to express an idea. Sili} (2006: 68) distinguishes between "intentional" and "unintentional" pleonasms. "Intentional" pleo nasms are an integral part of the administrative style, whereas "unintentional" pleonasms are to be considered as a sign of one of the most severe diseases of the administrative style and are in the majority of cases the result of average or insufficient language competencies21. Apart from several examples of "intentional" pleonasms our corpus contains numerous "unintentional" pleonasms. The vast majority of them are related to expressing time relations.

... a to je bilo prije otprilike desetak godina.

On je to zahtijevao tijekom prosle zime jer ...

... kada sam ja imao cca. petnaestak godina.

Other "unintentional" pleonasms include double use of conjunctions and/ or adverbs with the same or very similar meaning, as well as incorrect usage of words with Latinate origin.

Pamtim da smo tamo drzali stoku. I tako|er smo dugo godina obra|ivali zemlju.

Mi smo ostvarivali pravo na 1/2 dobivene vrijednosti, no me|utim brat njezinog oca je koristio jedan dio parcele i ...

Meni je bilo poznato da su oni izme|u sebe bili najoptimalnije podijeljeni ...

Apart from being an important feature of the administrative style, "unintentional" pleonasms and their use might in some instances be attributed to the desire of the speaker (in our case the judge dictating the statement) to underline the importance of certain parts of the statement. "Unintentional" pleonasms are found in over three-quarters of the witness statements, moreover witness statements in which they have been found contain two or more examples of "unintentional" pleonasms.

3.1.2. Semi-copulative or copulative verbs with complements

Functional styles can broadly be divided into nominal and verbal functional styles. According to Sili} (2006: 65-66) the former is characterized by "objects", meaning that description and nouns prevail, whereas the latter is characterized by "actions", meaning that narration and verbs dominate.

In accordance with the nominality of the administrative style, semi-copulative and copulative verbs requiring a complement occur frequently in the analysed witness statements.

Niti sam kasnije dobio to saznanje.

Isti nikada nisu bili u posjedu ni jedne od sporednih cestica.

Isti sjedi na stolici i u mogucnosti je dati iskaz, ali nije u mogucnosti ...

Each witness statement contains three or more examples of semi-copulative and/or copulative verbs with complements. The most prominent example of a copulative verb with complement, found at least once in every witness statement analysed, is the use of the expression biti poznato (be familiar with) instead of znati (know).

... i nije mi poznato i ne sjecam se tko je jos tamo bio.

Nista mi nije poznato vezano za ....

... po pricanju mi je poznato da su bili bogatasi iz Labina.

3.1.3. Tenses

The tenses used in witness statements are present and past, as well as rare instances of pluperfect. The statements analysed refer to the events from the past, such as who was born where, who lived where, who inherited what from whom, who used which property for what purpose and for how long, who renovated or built what and the like. Consequently, there are no examples of future tense in our corpus.

... ja vec dugi niz godina stanujem u Umagu.

Ne sjecam se kada ...

... od tada zivim u Presici.

Nikada nam nitko nije osporavao koristenje ...

... ja sam otisao u partizane ...

U toj kuci stanovali su roditelji ...

Ja sam 1941. godine bio pozvao ...

... ali da je u jedno vrijeme kuca bila izgorjela.

3.2. Characteristics of the legal style

Apart from the above described characteristics of the administrative style, the witness statements exhibit certain characteristics that can be attributed to one of the sub-styles of the administrative style, that is, to the legal style (Kovacevic and Badurina, 2001: 480).

3.2.1. Legal vocabulary

Lexical choice is a direct consequence of communicative activity and purpose, meaning they have a direct impact on word choice and in our corpus result in the distinctive legal register, that is, legal vocabulary. The most frequent core legal vocabulary in our corpus includes general legal terminology such as dati iskaz (give sworn evidence), ostavinska rasprava (probate proceedings) and the like, as well as for the largest part terms closely related to litigation proceedings: predmetna nekretnina (property in question), imovinskopravno stanje (property-rights situation), imovinskopravni odnosi (property-rights relations), osporavati posjed (contest property), and optant (optant).

Nisam o~ekivao da }e me pozvati da dam iskaz ...

Kada je moj otac umro kod ovog suda je iza njega provedena ostavinska

rasprava i donijeto je rje{enje ...

Rodio sam se u predmetnoj ku}i ...

...na drugi na~in nisam uspio rije{iti imovinskopravno stanje ....

... niti nam je osporavao posjed.

.. nisu trebali biti progla{eni za optante jer nisu optirali.

In addition, the repetition of lexical items, rather than the use of pronoun substitutes, produces frequency information that underlines the key semantic fields present in witness statements in litigation proceedings. When the lexical items are not repeated, the pronoun isti (the aforesaid) is used.

Rodio sam se u predmetnoj ku}i ... i od kada pamtim ista je u na{em posjedu.

Nikada nam nitko nije osporavao kori{tenje nekretnina pa smo se na istima

uvijek pona{ali kao vlasnici.

3.2.2. Long, complex, multi-clause sentences

Long, complex, multi-clause sentences are one of syntactic features that are consistently used in legal genres (Coulthard and Johnson, 2007: 39), which is the case in our corpus as well. Depending primarily on its length, each witness statement contains a minimum of two long, complex, multi-clause sentences.

Nakon {to je isti odlu~io ove nekretnine darovati meni iz razloga da uredimo ku}u i sa~uvamo imanje, te nakon {to je mij sin po~eo ure|ivati ku}u utvrdili smo da nek. nisu u zemlji{ne knjige upisane kao na{e vlasni{tvo.

Sto se ti~e Milana on je napustio zemlju 1944. godine kada je oti{ao u vojsku, i vi{e se nije vra}ao, ali isti nije po mom sje}anju bio optant, ve} je zivio u Italiji, ali je prije toga bio u njema~kom logoru kao i brat Bruno.

Osim toga ja nikada nisam znala da u te nekretnine koje sam ja dobila ulazi i nekretnina koja je bila od pok. Ivana jer da sam to znala ja to sigurno nikad ne bi u~inila.

3.2.3. Conjunctions

Coordinating conjunctions prevail, whereas the subordinating ones are represented to a lesser extent.

Rodio sam se u predmetnoj ku}i i od kada pamtim ista je u na{em posjedu. Ja sam sa majkom ìvio u predmetnoj ku}i, a nakon toga 1973. godine sam odselio u Labin.

... ja sam oti{ao u partizane, dok je zavr{io Drugi svjetski rat.

Frequently used are compound subordinate conjunctions, primarily expressing time, cause or providing additional explanation. The frequency of conjunctions in a particular witness statement is closely related to the length of the statement, that is, to the number of long, complex, multi-clause sentences.

Tek kada je moj sin zatràio z.k. izvadak za nekretninu a radi dizanja kredita, utvrdio je da nekretnine nisu u vlasni{tvu tùitelja.

... nisam tràio nikakve z.k. izvatke iz razloga {to je moja majka bila vlasnik predmetnih nekretnina ... Budu}i da na drugi na~in nisam uspio rije{iti imovinsko pravno stanje podigao sam tùbu ...

... Ivan i Anton bili suvlasnici te parcele s time {to je na djelu parcele koju je koristio Ivan ... I donijeto je rje{enje o naslje|ivanju s time da je on meni pustio ...

3.3. Characteristics of the conversational style

The above described characteristics of the administrative style and its sub-style, the legal style, are the result of the legal settings and the fact that judges dictate to court clerks what has to be written down. Nevertheless, since witnesses are in the vast majority of cases laymen the witness statements analysed exhibit several features of the conversational style. The most prominent ones are the use of dialect words, frequent use of idioms and numerous mistakes of various types.

3.3.1. Dialect words

When used, dialect words are usually written in inverted commas. Although found in 70% of the witness statements, they are very frequent in those statements where found. One possible explanation might be that some witnesses use dialect words so often, thus making it very difficult for judges, who otherwise strive to avoid dialect words when dictating to court clerks, to completely omit them from witness statements.

... znam da je jedan »kus« parcele bio od barba Ivana

... »ku}a vela izvana, a iznutra jo{ »stora«

... zemlja na{eg noneta

3.3.2. Idioms

Idioms are used in even less statements than dialect words. However, when used in a certain statement, they occur frequently. Interestingly, witness statements abundant in dialect words also contain many idioms. The probable cause might be a witness, although in administrative settings, using predominantly conversational style, thus strongly influencing a judge and his or her choice of words when dictating a witness statement.

... na licu mjesta oni su znali ...

... ta je povr{ina ostala do dan danas...

... i to od temelja pa do krova.

3.3.3. Mistakes

Numerous mistakes found in our corpus are probably the result of the circumstances in which court clerks have to write down what judges dictate to them; on the one hand they have to be quick, on the other hand, judges might change their minds and rephrase certain sentences several times. The mistakes can be divided into three larger groups: typing errors, syntactical mistakes and orthographic mistakes.

... te nakon {to je mij sin po~eo...

... podgiao sam tùbu ...

... i smatrala se kao vlasnicom.

... mi se nismo podjelili ja samo znam da ...

3.4. Discussion

Table 1 provides the most prominent characteristics of the witness statements exemplified and described in the above paragraphs. It is divided into three main parts; the first part containing characteristics of the administrative style, the second part with characteristics of the legal style, that is a sub-style of the administrative style, and the third part comprising characteristics of the conversational style. The first column names the characteristic, the second column refers to the frequency of a particular characteristic in the entire corpus i.e. in how many witness statements the characteristic is found. Since the listed characteristics are of different nature and cannot be statistically measured and described in the same way, the third column includes remarks on each characteristic. Apart from the number of witness statements in which a certain characteristic of either the administrative or the conversational style is found, it is also important how often these characteristics appear in particular witness statements.

In brief, the presented characteristics of the analysed witness statements can be attributed to the participants in the process of giving and producing witness statements in Croatian courts, their interrelationships and the courtroom settings itself. The dominance of the administrative style stems partly from the superior role of the judge in the process of statement producing and party from the circumstances witnesses find themselves in. It would appear that witnesses, who otherwise rarely use the administrative style, in particular the legal style, make an effort to adjust to the circumstances and use the style appropriate for the courtroom setting. Nonetheless, some witnesses are either unaware of or not affected by the setting or not knowledgeable about the appropriate style, which results in several statements rich in characteristics of the conversational style.

The above described and presented features make the texts we have analysed instantly identifiable as an example of the genre of a written witness statement in litigation proceedings in Croatian courts.

Similar and different analyses of various legal genres have been conducted in other languages, mostly in English-speaking countries. Olsson (2004: 121-139) outlines the most important characteristics of well-formed witness narratives and reports in Great Britain and in the United States23. They include the categories of time, place, sequence, superfluity and tense. Events take place in a given time period without gaps, key places are introduced as they appear, events are described in sequence, superfluous details are absent and simple past tense is used. Olsson suggests (123-124) that the above mentioned characteristics could be used to judge the speaker's commitment and appraise veracity in language. If the speaker's commitment is high, his or her relationship with veracity is said to be high. However, if the commitment is low, this might signal inconsistency and cast doubt on veracity. Such and similar analyses have rarely been accepted in courts and their pivotal function and merit is to facilitate the investigation by providing useful clues.

There are cases which revolve around disputes about the accuracy of the written record of an interaction between the police and the accused. Although converting the spoken to the written language is not an unproblematic task, police officers usually do not have explicit guidelines about the procedures they should use and what could or should be omitted. Unlike the previously mentioned veracity analysis which adopts the top-down approach, the case of Robert Burton described by Coulthard and Johnson (2007: 132-137) demanded the bottom-up approach. Robert Burton, who was caught trying to steal trailers loaded with whiskey, claimed he had been forced to commit robbery by undercover police officers. In order to prove it he claimed that the records of telephone calls with undercover police officers were too accurate and accordingly could not have been produced from memory. Although the police denied the existence of the tape-recordings of the telephone calls, the linguistic analysis revealed the occurrence of two sets of features in the written records: the appearance of spoken discourse items and the use of a "step" word24, thus confirming that either the police officers had an amazing ability to recall conversations verbatim or that there had been tape recordings. After having tested the ability of laymen to recall conversation and showing that it was impossible to remember verbatim what was said, the linguistic evidence was accepted unchallenged. However, the appeal failed on other grounds.

The experience and results of described top-down or bottom-up approaches in analysing legal genres are not to be directly applied to the legal genres in the Croatian language. Firstly, language differences should be taken into account and even more importantly, the context, that is, the Croatian legal system and setting differs from the British and American. It is considered that the above described and similar analyses of the legal genres in the Croatian language contribute to better understanding of how language functions and of how written discourse is structured and organized. Furthermore, they could be successfully applied in police work and in courts, provided they are drawn up and conducted and the findings are evaluated in cooperation with experts and practitioners from police and legal system.

4. Conclusion

Legal genres, their styles and modes of interaction, and the social practices, roles and participant relationships that they produce constitute complex interrelationships between text and context. Legal genres are the result of the communicative practices that they employ and the functions and purposes that they serve in legal and world context. Witness statements in litigation proceedings in Croatia, compiled in accordance with valid laws and regulations, represent written records of sworn evidence witnesses give to judges who dictate them to court clerks. They are subsequently used by judges when making a ruling on a particular case

The analysis has proven that witness statements from Croatian municipal courts are strongly influenced by the interaction-situational context and its givenness, which to a large extent determines its syntactical and lexical features. Accordingly, they exhibit characteristics of the administrative and of the conversational functional style of the Croatian language, whereas the characteristics of the administrative style are dominant. The characteristics of the administrative style include "unintentional" pleonasms, semi-copulative and/or copulative verbs with complements, present and past tense, frequently-occurring legal vocabulary, long, complex, multi-clause sentences, as well as frequent use of coordinating and subordinating conjunctions. The characteristics of the conversational style are frequent use of dialect words and idioms, as well as numerous language mistakes. It can be concluded that witness statements are to some extent, as many other genres, characterized by the functional style hybridity.

In some countries large corpora of witness statements have been collected and surveyed in order to draw a style profile and establish a model to which all statements should conform to a greater or lesser extent. This enables observing and recording inconsistencies and might be used as a clue in veracity analysis of witness statements. Such analyses have so far been primarily regarded as an aid to investigation and courts have accepted them in rare instances and only if there was plenty of additional supportive evidence.

In order to obtain a comprehensive image of the witness statement in the Croatian language, which could, among other things, be used as an indication in establishing veracity, a significantly larger corpora, as well as witness statements from various types of legal proceedings and various types of courts, should be collected and analysed.

Footnote

1 Some linguists have expressed certain concerns about the term itself (Shuy, 2006: 3-4) claiming that when one does forensic linguistics, one simply does linguistics, a type of applied linguistics, urging practitioners never to neglect the foundations of linguistics they use in their work as forensic linguists. Moreover, McMenamin stresses (2002: 84) that a forensic linguist must first of all be a good linguist, not intending to minimize the importance of acquiring certain knowledge about the judicial system itself, nonetheless signaling the significance of the connection forensic linguistics has to scientific theories and methods of general and applied linguistics. However, they have also accepted the term as a convenient way for lawyers and linguists alike to refer to the area they are concerned with.

2 The term forensic English was used in 1949 by Philbrick in the title of his book on legal English Language and the Law: the Semantics of Forensic English, but the phrase was never taken up.

3 See http://www.iafl.org.

4 See http://www.iafpa.net.

5 On how linguistics analysis works in trademark cases see Shuy (2002).

6 See Tiersma (1999).

7 See Shuy (2005), as well as Tiersma and Solan (2005).

8 See Walker (1999), as well as Smith and Tilney (2007).

9 See Cotterill (2003).

10 See Shuy (2006).

11 See McMenamin (1993) and McMenamin (2002).

12 See Nolan (1983), Baldwin and French (1990), as well as Rose (2002).

13 See Nolan (1983), as well as Baldwin and French (1990).

14 See Varosanec-Skaric and Bicanic (2007), Varosanec-Skaric, Stankovic and Safaric (2008) and Varosanec-Skaric (2008).

15 See Varosanec-Skaric and Kisicek (2010).

16 See Stede (2007).

17 Although Coulthard and Johnson (2007: 55) state that the terms genre and text type are synonyms, the term genre prevails in English speaking countries, whereas the term text type is predominantly used in German speaking countries. Croatian linguists (Ivaneti}, 2003 and Glovacki-Bernardi, 2004) have adopted the term text type, that is, tekstna vrsta.

18 This division has been accepted in Croatian linguistics, thus Francic-Hudecek-Mihaljevic (2005: 230) stress that, regardless of its drawbacks, this division is all-encompassing and operational.

19 Syntactical and lexical levels are part of the linguistic microstructure, which according to Adamzik (2004: 59) along with situational context, function and topic belongs to the traditional elements of a text analysis.

20 Regardless of various types of mistakes, the examples from witness statements have not been altered and are authentic.

21 See Silic (2006: 65-70).

22 The total of the tense use does not add up to 100% due to rare instances of predicates in pluperfect.

23 It should be stressed that witness narratives and reports are not taken and produced in the same way in Great Britain and in the United States. The differences in legal system and practice should be taken into account when conducting such and similar analyses in real police or courtroom cases.

24 A step word is a word which a speaker learns to produce automatically to disguise the fact that he or she is having difficulties with the articulation of other words.

References

References

Adamzik, K. (2004) Textlinguistik. Eine Einführende Darstellung, Tübingen: Max Niemeyer Verlag.

Baldwin, J.R. and French, P. (1990) Forensic Phonetics, London: Pinter Publishers.

Brinker, K. (2005) Linguistische Textanalyse: Eine Einführung in Grundbegriffe und Methoden, Berlin: Erich Schmidt Verlag.

Cotterill, J. (2003) Language and Power in Court: A Linguistic Analysis of the O.J. Simpson Trial, Houndmills: Palgrave Macmillan.

Coulthard, M. (1977) An Introduction to Discourse Analysis, London: Longman.

Coulthard, M. (1994) On the use of Corpora in the analysis of forensic texts, u Forensic Linguistics I (1), 125-141.

Coulthard, M. and A. Johnson (2007) An Introduction to Forensic Linguistics, Language in Evidence, New York: Routledge.

Francic, A., Hudecek L. and Mihaljevic, M. (2005) Normativnost i visefunkcionalnost u hrvatskome standardnom jeziku, Zagreb: Hrvatska sveu~ili{na naklada.

Glovacki-Bernardi, Z. (2004) O tekstu, Zagreb: [kolska knjiga.

Ivanetic, N. (2003). Uporabni tekstovi, Zagreb: Zavod za lingvistiku Filozofskog fakulteta Sveu~ili{ta u Zagrebu.

Kovacevic, M. and Badurina L. (2001) »Jezi~ni paradoksi administrativnoga stila« in Drugi hrvatski slavisti~ki kongres, Zbornik radova I., pp. 479-487.

McMenamin, G. (1993) Forensic Stylistics, Amsterdam: Elsevier.

McMenamin, G. (2002) Forensic Linguistics, Advances in Forensic Stylistics, London: CRC Press.

Nolan, F. (1983) The phonetic bases for speaker recognition, Cambridge: Cambridge University Press.

Olsson, J. (2004) Forensic linguistics, An Introduction to Language, Crime and the Law, New York: Continuum.

Philbrock, F.A. (1949) Language and the Law: the Semantics of Forensic English, New York: Macmillan.

Rose, P. (2002) Forensic Speaker Identification, London: Taylor & Francis.

Shuy, R.W. (2002) Linguistic Battles in Trademark Disputes, Houndmills: Palgrave Macmillan.

Shuy, R.W. (2005) Creating Language Crimes, How Law Enforcement Uses and Misuses Language, New York: Oxford University Press.

Shuy, R. W. (2006) Linguistics in the Courtroom, A Practical Guide, New York: Oxford University Press.

Sili}, J. (2006) Funkcionalni stilovi hrvatskoga jezika, Zagreb: Disput.

Solan, L.M. and P.M. Tiersma (2005) Speaking of Crime, The Language of Criminal Justice, Chicago: University of Chicago Press.

Smith, K. and S. Tilney (2007) Vulnerable Adult and Child Witnesses, New York: Oxford University Press.

Stede, M. (2007) Korpusgestützte Textanalyse Grundzüge der Ebenen-orientierten Textlinguistik, Tübingen: Gunter Narr Verlag.

Svartvik, J. (1968) The Evans Statement: A case for forensic linguistics, Gothenburg: University of Gothenburg Press.

Tiersma, P. (1999) Legal language, Chicago: University of Chicago Press.

Tosovic, B. (2002) Funkcionalni stilovi, Graz: Institut für Slawistik.

Varosanec-Skaric, G. and Bicanic, J. (2007) "A comparison of indices of difference and similarity, based on LTASS and tested on voices in real forensic case in controlled conditions" in Proceedings of 16th International Congress of Phonetic Sciences, pp. 2085-2088.

Varosanec-Skaric, G. (2008) »Verifikacija govornika u forenzicnoj fonetici« in Govor - casopis za fonetiku XXV, 1, pp. 31-44.

Varosanec-Skaric, G., Stankovic, D. and Safaric, I. (2008) »Prepoznavanje poznatih glasova studenata fonetike u normalnom govoru, normalnom filtriranom, prikrivenom i prikrivenom filtriranom govoru« in Govor - casopis za fonetiku XXV, 2, pp. 143-170.

Varosanec-Skaric, G. and Kisicek, G. (2010) "Linguistic and forensic phonetics identification of a male speaker (the real case)" in Abstracts for the 19th Annual Conference of the International Association for Forensic Phonetics and Acoustics, pp. 34.

Walker, A.G. (1999) Handbook on Questioning Children. 2nd edition. Chicago: American Bar Association Center on Children Law.

Online sources

The International Association of Forensic Linguists http://www.iafl.org (Accessed on 14/06/2011)

The International Association for Forensic Phonetics and Acoustics http://www.iafpa.net (Accessed on 14/06/2011)