Conference on Forensic Linguistics / Language and Law 2004 /4th - 5th July, 2004

Cardiff University will host a 2-day conference on forensic linguistics/language and law on the 4th and 5th of July 2004.

The conference will begin after lunch on 4th July and will finish on the afternoon of the 5th July and will be held at Gregynog Hall, a beautiful stately home in mid-Wales owned by the University of Wales.

Papers will deal with all aspects of forensic linguistics/language and law, including:

• courtroom discourse/police interviews
• courtroom interpreting and translating
• the readability/comprehensibility of legal documents
• the analysis/interpretation of legal texts
• the comprehensibility of the police caution issued to suspects
• interviews with children in the legal system
• the communicative challenges of ‘vulnerable’ witnesses
• the use of linguistic evidence in court
• authorship/speaker identification
• the teaching/testing of forensic linguistics/language and law

Contact:

Dr Janet Cotterill

Centre for Language
and Communication
Cardiff University
PO Box 94 Cardiff
CF10 3XB United Kingdom

Tel: +44 (0)2920 876393
Fax: +44 (0)2920 874242

Email: cotterillj@cf.ac.uk

Programme

July 4th - Day 1

Parallel Sessions

July 5th - Day 2

Parallel Sessions

Abstracts

Michelle Aldridge
Centre for Language and Communication, Cardiff University

Vulnerable witnesses' perception of the criminal justice service in North Wales

In 2002-2003 a study was undertaken to determine the public's perception of the Criminal Justice Service in North Wales. Views were elicited from a wide range of the public including known victims, witnesses and convicted offenders and members of the public who had no experience of the system. This paper reports on part of the finding namely the perceptions of those people who would fall into the category ' vulnerable'. Striking differences were recorded between these people and the rest of the sample. As examples, vulnerable respondents rated all of the agencies within CJS as more likely to treat them unfairly as both victims and accused, they had a greater fear of assault and race/hate crime and in most cases, they were less likely to report crime. The implications of these findings are discussed.

Kelly Benneworth
Loughborough University, UK

A discursive analysis of police interviews with suspected paedophiles: The implications of ‘open’ and ‘closed’ interviewing for admission and denial.

This study examines the discursive interaction between police officers and suspected paedophiles in the investigative interview. Paedophiles talk about their offences in terms of conventional relationships, personal bonds and emotions whilst being discrete about the sexual aspects of their activities. In the investigative interview, police officers must encourage paedophiles to discuss their criminal activities in terms of direct, sexual detail. Given these two distinct approaches to the description of unlawful sexual contact, there is potential for difficulties to arise in the elicitation of confession. Eleven audiotaped interviews conducted at Leicestershire Police Constabulary were transcribed. The offenders were male and aged between 34-54 years and the victims were male (n=5) and female (n=6) and aged between 5-13 years. Discourse analysis explored how the police officers and paedophiles negotiated an account of the offence whilst managing conflicting descriptions of sexual relationships between adults and children. The analysis revealed two distinct styles of police interviewing with implications for confession, ‘open’ interviewing which maximised suspect admission and ‘closed’ interviewing which encouraged denial. In addition to offering a distinctive qualitative understanding of how language clashes shape the progression of the police interview, the findings support the development of a discourse-based training programme to assist police officers in the interviewing of paedophiles. The findings also provide a vocabulary for interviewing officers to reflect on their own practices and communicate their skills to less experienced officers.

Ronald R. Butters
Duke University, Durham, NC, USA

Evidence of the Rehearsal of a Videotaped Confession as Support for a Diminished Capacity Defense in USA Death-Penalty Trials

Although sentenced to death in 1977, Johnny Lee Gates remained on death row until 2003, when a new trial attempted to determine if he was too severely mentally retarded to be legally executable. The later trial could not consider Gates’s guilt or innocence, but a forensic psychologist was prepared to testify that evidence of confession-rehearsal bore vitally on Gates’s putative mental retardation. (Weeks after the crime, police had taken Gates to the scene and obtained a videotaped confession wherein he walked through his purported role while giving incriminating answers to a policeman’s questions.) Linguistically, I determined, four types of pragmatic and discourse evidence supported the psychologist’s testimony:

• Gates was often asked questions that presupposed information that had not yet been established in the interview.
  
• Gates gave many responses that went well beyond what the questions called for.
  
• Questions were always “safe” questions; Gates could easily have been coached to give his answers; obvious informational questions were never asked, however, if the answers would have been contradictory or difficult for one to answer who had not committed the crime.
  
• Some answers were exculpatory, in which case the questioner recycled the questions until he received an incriminating answer.

Carole Chaski
Institute for Linguistic Evidence, Inc., Georgetown, DE, USA

Recent Validation Results for the Syntactic Analysis Method of Author Identification

The Syntactic Analysis Method of Author Identification has been briefly presented in Chaski (1997, 2001). This method differs from other non-lexical approaches to author identification (Stamatatos' et al. 2001) through its reliance on the linguistic principles of headship and markedness in grammar. Stamatatos' work is most comparable; using his parser to break the text into chunks through which he obtains 22 features, his method is able to achieve, using a discriminant function analysis, 81% accurate assignment of texts to ten authors, with the total words per author ranging from 17,000 to 52,000 words. Using a discriminant function analysis of 17 features per text, three texts of the four authors previously reported on in Chaski 2001 were assigned with 100% accuracy. When ten authors were tested, the method obtained 99% accuracy (one author-author text resulted in the DFA not obtaining enough variables to run). When thirteen authors were tested, the syntactic analysis method obtained 91% accurate assignment of the 72 documents to the correct author; the 9% that were not accurately assigned was due to the DFA not having enough variables to run, and all but one of those documents belonged to one author. These results demonstrate that the Syntactic Analysis Method has a high degree of reliability within the parameters of forensic application, a few and often short documents.

Janet Cotterill
Cardiff University

Answers in the Affirmative?:: Yes, No and Maybe in the British Courtroom Context

This presentation draws on a 1-million word sub-corpus of witness examinations and cross-examinations taken from trials held in the late 1990’s in the UK, and applies a combination of corpus linguistic and discourse analytic approaches to study the semantics, pragmatics and discursive characteristics of yes and no in the courtroom setting. These short responses are, on the face of it, semantically transparent and unambiguously affirmative or negative in their orientation. But this most frequent type of witness response is not as pragmatically simple as it might at first seem. This paper reports a study investigating the contextual semantics and pragmatics of this common form of witness response, and the subtle nuances of meaning which are exploited by both barristers and witnesses during examination and cross-examination. It explores the grey area between yes and no and suggests some of the discursive consequences when witnesses do not provide a definitive affirmative or negative response to a yes/no-type question.

Malcolm Coulthard
University of Birmingham

A Guided Tour around Narrative Analysis

Solan (1998) addresses the problem which is specific to linguistic experts, the fact that the judges of fact, whether they be actual judges or jury members are seen, for most purposes, to be their own experts on language. Even so, in such a context, he argues, there is a role for the linguist, which is to explain facts about language and usage, as a result of which the lay judges will then be in the same position as the linguist and able to make their own informed decisions.

In this paper I will focus on a short confession narrative in which an accused purportedly confesses to being present when a murder took place. The accused claims that the narrative is a complete fabrication; the police that it is a verbatim transcript of what he said. I want to explore, using insights from narrative analysis and Gricean maxims, ways of showing a jury why the text cannot be an accurate, let alone a verbatim transcript of what the accused said.

Reference

Solan L 1998 ‘Linguistic experts as semantic tour guides’, Forensic Linguistics 5, ii, 87-106.

Betsy Evans
Centre for Language and Communication, Cardiff University, UK

Imitation and speaker identification: a case study

It has been previously assumed by sociolinguists that speakers can “perform” only gross stereotypical characteristics of other varieties, simply borrowing from their own repertoire of sounds (e.g. Labov 1994). This implies that people cannot perform imitations ‘accurately’. This study explores ‘accuracy’ in a phonetic context via a case study that examines the phonetic match between an imitation and known phonetic characteristics of the imitated dialect. Recordings were made of a sociolinguistic interview during which the respondent read a word list and reading passage in his “usual,” that is, “not-imitating,” speech and in what he perceived to be U.S. “West Virginia” speech. Acoustic analysis of the data revealed notable differences in the first and second formants of the vowels in these two modes and considerable consistency in the imitation and known phonetic features of the target. This paper also examines the reactions of native West Virginians to Noah’s speech in order to determine whether or not the imitation is successful perceptually. Results showed that the majority of respondents perceived the imitation as authentically West Virginian. The acoustic and perceptual success of this respondent’s imitation has important implications for forensic speaker identification and indicates that more research of this kind is necessary for a better understanding of imitation.

Marta Carretero
Department of English Language and Linguistics, Facultad de Filología, Universidad Complutense, Madrid

The role of evidentiality and epistemic modality in a solicitor's submission within divorce proceedings

It is well-known that evidential and epistemic expressions are reliable indicators of the stance taken by language users in discourse. In this paper I will illustrate this statement by analyzing a 5,000-word spoken text from an English corpus. This text is a fragment of a submission by the solicitor who represents the husband in a divorce case, with occasional interruptions by the judge. The use of epistemic and evidential expressions by both speakers reflects their different stances: the solicitor takes a clear position (namely that the main cause of the divorce lies in the wife's behaviour), hence her frequent use of epistemic expressions conveying a high degree of personal commitment, often combined with evidentials indicating a wide array of attitudes towards different sources of evidence. On the other hand, the judge displays a much more cautious attitude, which accounts for the relatively weak epistemic expressions and for the scarce evidential expressions in his turns.

References

Chafe, W. (1986) Evidentiality in English conversation and academic writing. In W. Chafe & J. Nichols (eds.) Evidentiality: the Linguistic Coding of Epistemology. (Advances in Discourse Processes, XX.) Norwood. N.J.: Ablex. 261-272.

Dendale, P. & L. Tasmowski (2001) Introduction: evidentiality and related notions. Journal of Pragmatics 33.3, 339-348.

Downing, A. (2001) "Surely you knew!" Surely as a marker of evidentiality and stance. Functions of Language 8.2, 253-285.

Marín-Arrese, J., Hidalgo, L. & S. Molina (2002) Evidentiality and modality in English and Spanish: connections and interactions. In R. Facchinetti (ed.) English Modality in Perspective: Genre Analysis and Contrastive Studies. Verona: Università di Verona. 135-146.

Nuyts, J. (2000) Epistemic Modality, Language and Conceptualization. A Cognitive-Pragmatic Perspective. Amsterdam / Philadelphia: John Benjamins.

Plungian, V. (2001) The place of evidentiality within the universal grammatical space. Journal of Pragmatics 33.3, 349-357.

Svartvik, J. and Quirk, R. (eds.) (1980) A Corpus of English Conversation. Lund: C.W.K. Gleerup.

Taverniers, M. (1998) Towards a topology of interpersonal meaning. Paper presented at the 10th Euro-International Systemic Functional Workshop, University of Liverpool, July 1998.

Thompson, G. & Y. Ye (1991) Evaluation in the reporting verbs used in academic papers. Applied Linguistics 12.4, 364-382.

White, P.R.R. (1998). An introductory tour through appraisal theory. Paper Presented at the 10th Euro-International Systemic Functional Workshop, University of Liverpool, July 1998.

Maya David
Faculty of Languages and Linguistics, University of Malaya, Malaysia

Language Choice, Law, and Power in a Bilingual Setting

Malay is the National Language of Malaysia. English is an important second language. Many of the older lawyers are proficient in English as English was the medium of instruction when they were in school. The younger lawyers had Malay as the medium of instruction and are consequently more proficient in Malay. As in most other settings in Malaysia, (see Omar, 1982; David, 2001, Morais, 1998) a mixed discourse appears to have become a regular feature in this formal setting too. The aim of this research is to determine the functions of code switching in Malaysian courts of law and was based on the premise that as Malaysians have differing levels of fluency in different languages they would consequently be obliged to code switch to effectively convey meaning. In a study on code switching in Malaysian service encounters David, (1999) found that language convergence rather than language divergence was rampant as negotiation was the goal of both the buyer and seller. The fact that rampant asymmetrical or divergent language choice also occurs in formal settings like the courts indicates that non-accommodation of language choice implies power and control of one party over the other. Language choice and the use of a mixed discourse in the Malaysian courts are triggered to achieve a range of strategic and professional objectives. The reasons for such mixed discourse in the legal setting are discussed. How this analysis empowered the teaching of a relatively new course "Communicative Skills for Professionals" - a course for final year undergraduates in the University of Malaya, is also described. It is imperative that all ESP practitioners conduct action research to empower their teaching.

Lise Fontaine
Centre for Language and Communication, Cardiff University, UK

Your money or your life : Safety and Corporate Crime in the workplace

“The 26 Canadian miners who were killed at the Westray Mine had two choices: to accept their employer's negligent and unsafe workplace practices or to lose their jobs.” [1]

This paper has two main purposes. The first is to introduce the area of Safety and Corporate Crime in the workplace, including a discussion of the amendments made in November 2003 to the criminal code in Canada: Bill C-45. “It (had) already (been) possible for a corporation to be convicted of criminal negligence under the Criminal Code. … Bill C-45 is specifically directed to making convictions for criminal negligence occur more readily for complex organizations, including corporations.” (Edwards and Collins, 2003) The amendments to the code include: new duties, an increased breadth of application, and new “criminal” offences. Specific attention will be given to the most important defense available in such cases, due diligence, and to the role of linguistic evidence.

The second focus of this paper is on a reprisal case where an employee filed an “unlawful reprisal application” [2] against his employer where he claims he was fired for reporting safety violations. Using the judgement in the case and the official transcripts of the hearing, this study analyses the referring expressions used by the various participants with respect to safety, the company’s attention to safety in the workplace and official attitudes as stated in the judgement.

References

Bill C-45 – Amendments to the Criminal Code Affecting the Criminal Liability of Organizations http://canada.justice.gc.ca/en/dept/pub/c45/CCL_english.pdf

Cotterill, J. (2001). “Domestic Discord, Rocky Relationships: Semantic Prosodies in Representations of Marital Violence in the O.J. Simpson Trial.” Discourse & Society, 12(3), 291-312.

Edwards, Cheryl A. and Ryan J. Conlin (2003) “Workplace Safety Amendments to Criminal Code Introduced” http://www.sbhlawyers.com/repo/index.asp?action=edit&report_cat_id=21

Slapper, Gary (2000) Blood in the Bank: Social and Legal Aspects of Death at Work. Aldershot: Ashgate Publishing.

[1] http://www.ccriw.com/campaign.html
[2] Section 50 of the Occupational Health and Safety Act of Ontario states:

(1) No employer or person acting on behalf of an employer shall,

(a) dismiss or threaten to dismiss a worker;

(b) discipline or suspend or threaten to discipline or suspend a worker;

(c) impose any penalty upon a worker; or

(d) intimidate or coerce a worker,

because the worker has acted in compliance with this Act or the regulations or an order made thereunder, has sought the enforcement of this Act or the regulations or has given evidence in a proceeding in respect of the enforcement of this Act or the regulations or in an inquest under the Coroners Act. R.S.O. 1990, c. O.1, s. 50 (1).

Mark Garner, Northumbria University and Edward Johnson, Wolfson College, Cambridge

Operational communication theory as a framework for research into emergency call-handling by the police

Emergency call-handling confronts the researcher with a double challenge. The relatively limited research that has been done in this field has focussed on linguistic issues through discourse analysis, but a more complex and broadly-based approach is necessary if the outcomes of our current research are to be theoretically valid, professionally acceptable, and practically useful by the end users, the police. Whilst it is necessary to analyse the exchanges between caller and call-handler as linguistic interactions, recommendations about the linguistic aspects of the communication must conform to the evolving operational requirements of the police force, including training and development of appropriate IT support, and must meet the needs of the public who make the calls.

We outline a theoretical framework for this sort of institutionally and culturally situated analysis, presenting a series of basic principles for operational communication research that have evolved in the course 25 years of experience of work in maritime, air traffic, and inter-agency communication, employing a range of written, spoken and visual media. These principles go beyond conventional linguistic analysis to embrace an ecological view of the multi-stranded nature of communication within institutional settings. We demonstrate how they are being applied within our on-going research into the development of national call-handling standards.

Tim Grant and Jessica Woodhams
Forensic Section, School of Psychology, University of Leicester

Categorising rapists’ utterances

Whilst rape victim statements frequently refer to dialog between offender and victim, the language offenders use in the course of their offence is an under-researched aspect of offence behaviour. Holstrom and Burgess provided descriptions and some categorisation of rapists’ talk, since then only Dale, Davies and Wei attempt to provide a systematic typology of rapists’ utterances. Application of this typology demonstrated both strengths and weaknesses and led to the development of the current system. This draws more directly on speech act theory to categorise utterances as directives, constatives, commissives, expressives and interrogatives and makes use of subscales to further differentiate utterances. High reliability is reported for both systems, the new system however categorises a greater proportion of utterances in the statements. Current applications include examining differences between statements which are maintained as true from those which are withdrawn as false and examining behavioural consistency for the purpose of case linkage.

Gill Grebler
Cultural and Linguistic Forensics, Santa Monica, CA

False confessions: Fictional tellings with legal consequences

A false confession can be said to have a kind of ‘narrative truth’ although it is not actually true. At first glance this fictional telling displays a convincing verisimilitude. It convinces judge and jury that a suspect is guilty of a crime.

I look at a set of vivid and detailed accounts from proven false confessions, of crimes told by innocent suspects as if they had committed them. These suspects were forced (usually after extended struggle to assert their innocence) to pretend to remember, to create an account - vivid and first-hand - of something they did not do. I look at the linguistic coercion that brought an innocent person to ‘remember’ having committed a crime and at the process of co-telling that was used to frame and shape their ‘post-admission narratives’.

During the co-construction process interrogators seek to fill in the details required of a legally valid confession, and the specifics of the particular crime. They feed information, the suspect draws upon various sources of ‘memory’, and aspects of a suspect's story-telling ability is activated. A workable story is produced although there is no actual memory to base it on.

Mark Griffiths
Centre for Language and Communication, Cardiff University

Forensic earwitness descriptions: predictability, consistency and reliability

Whilst it can be safely assumed that trained linguists may be best placed to provide and solicit descriptions of voices, a knowledge of linguistics and its meta-language rarely extends beyond those who have been specifically trained. It follows then, that in the legal context, only a tiny minority of earwitnesses and police officers will have any such training, yet lay-linguist descriptions are regularly solicited by lay-linguist police officers, with these descriptions forming a vital part of police evidence.

Moreover, a substantial tradition of research has indicated that lay linguists may have a variety of, often predictable, responses to regional and social accents, manner of speech and perceived speaker power, which applied to the forensic context, raises important questions regarding the predictability, consistency and reliability of earwitness descriptions. Drawing on the results of a pilot study of 150 1st year undergraduate students’ responses to recorded voices of suspected criminals, this paper aims to illustrate how lay voice descriptions and attributions of guilt may be qualitatively affected i/. by attitudinal variations towards speech varieties and manner of speech ii/. by variations in questionnaire design and voice line-up iii/. by earwitness desire to give socially acceptable responses.

Phil Hall
Appen Speech Technology and Macquarie University, Sydney, Australia

Interactional constraints on the capacity of suspects to invoke rights

While interrogative practices oriented to the retrieval from suspects of the specific types of information considered constitutive of confession are easily recognisable as characteristic of police questioning, there are other interactional practices whose links to the function of the interview are less readily apparent, but that nonetheless have the capacity to influence the course of the interview with potentially significant consequences for the interviewee/suspect. This paper will examine several examples of failed attempts by suspects to invoke legal rights, and will ask whether interviewers' attempts at "rapport building" practices contributed to the failure of these attempts by establishing a relationship between the interview participants in which suspects were constrained from unequivocal invocation of their rights by adherence to interactional norms.

Kerry Ham
University of Florida, USA

Conversational Maxims in Encounters with Law Enforcement Officers

This paper analyses the use of spoken language by suspects and witnesses being questioned by law enforcement officers. The aim is to discover whether, and to what extent, Grice’s (1975) Cooperative Principle and Brown and Levinson’s (1987) Politeness Principle are in evidence, and how adherence to or breaking these conversational maxims helps or hinders the interrogation process.

Beginning with an analysis of segments of the popular American reality-television show “Cops”, the analysis then draws on experiences gained whilst participating in numerous ‘ride-longs’ with the Alachua County Sheriff’s Office and the Gainesville Police Department, both located in North Florida.

The resultant analysis considers possible explanations as to why some suspects are arrested or taken into custody, whereas others are allowed to go free. The possibility will be considered that successful manipulation Grice’s Cooperative Principle (and the associated Conversational Maxims, in which Politeness is included for the purpose of this study) may have some degree of subconscious influence on police officers’ decisions regarding a suspect’s guilt or innocence. From the media analysis and associated experiences, conclusions are drawn as to the extent to which adherence to or breaking each Conversational Maxim may influence the outcome of an encounter with law enforcement officers.

Silvia Hansen-Schirra and Stella Neumann
Institute of Applied Linguistics & Institute of Law and Informatics, Saarland University, Germany

Comprehensibility in English and German court decisions

In the discussion of how to make the law transparent to the citizens who are subject to it, efforts are made to “translate” legal texts for the lay public. One initiative in this context is the publication of press releases by the German Federal Constitutional Court reporting on each of the Court’s decisions. The present work describes features of the language of law in these decisions in comparison to their realisation in the press releases.

Using corpus linguistic methods such as frequency counts, collocations, part-of-speech tagging and parsing we investigate a corpus of court decisions, press releases as well as a register neutral reference corpus. The analysed features comprise mean sentence length, type-token-ratio, complexity of NPs, nominalisations as well as formulaic expressions. We show how the features are modified or deleted in order to make the decisions comprehensible to a lay public. While the analysis mainly focuses on the German corpus we also draw comparisons to a corpus of British Privy Council decisions in order to find out whether there are any contrastive similarities. This serves our long term goal to develop a system for the computer-assisted translation of legal texts.

Chris Heffer
Nottingham Trent University, UK

Others’ Stories in Other Words: Narrative Strategies in the Judicial Summing-up

In reviewing the evidence in a criminal case during their summing-up to the jury, Crown Court judges in England and Wales find themselves in a particularly complex communicative situation. They inevitably must deal with stories in their summaries of the evidence: stories of the crime seen from different perspectives; competing stories of the investigation and trial proceedings presented by counsel in closing speeches; the narrative schemas held by most jurors. At the same time, they are not meant to be constructing their own subjectively viewed stories, but only assisting jurors in applying the law to the many tales and tellers that appeared in the trial. Drawing on a large corpus of transcripts of summings-up from English jury trials, this paper explores the remarkably different strategies employed by judges in presenting these complex and partial narratives, and considers the extent to which those judges actually adopt narrative techniques themselves.

Alison Johnson
School of Education, University of Birmingham, UK

Producing the voice of the child victim: the role of the interviewer in police interview interaction

Literature on interaction between adults and children in the early acquisition of language notes the importance of questions in scaffolding talk and eliciting conversation. Maternal and caregiver speech expands child utterances (Galloway and Richards, 1994) and children make use of this input in their language growth. In a similar way interviewers perform an important role in the production of the narrative voice of the child witness in police cases involving child abuse victims. This paper looks at child directed speech in police interviews with child witnesses, investigating the interactional relationship between interviewer and interviewee. A corpus of six interviews with children is examined; all cases are located in the U.K. and were carried out by six different interviewers with six children. Analytical perspectives come from an approach that combines the resources of (forensic) discourse analysis and first language acquisition within a legal setting. I suggest (through comparisons made with interviews with adults) that particular questioning sequences are crucial to the growth of a full narrative account of the crime being investigated. And- and so-prefaced questions, in particular, have important eliciting, clarifying and upgrading roles in the interaction. They are crucial in avoiding breakdown and achieving successful creation of narrative in these interview conversations. I conclude that there is a relationship between the interviewer’s interactional behaviour and the child’s production and performance. Producing, or leading forward, the voice of the child demands sensitivity to the cognitive and communicative pressures of the narrative task; the interviewer’s role is scaffolding production in the co-produced narrative.

Cecilia Joseph
Centre for Language and Communication, Cardiff University and University of Malaya, Malaysia

Are you listening to Me? Constructing Meaning and Responding to Children’s Accounts in Forensic Interviews

This paper is part of a Doctoral research that I am developing at Cardiff University. The focus of the research is on the role of listening in police interviews with children. In the UK, it has been the practice since 1992 to use video recordings of forensic interviews with child witnesses as evidence in criminal proceedings dealing with sexual or violent offences. Data from 20 police interviews of children ranging from the ages of 5 to 8 years old are used to explore how, in practical terms, police interviewers listen in these interviews. Drawing upon a holistic comparative analysis, this study examines how police officers respond to and construct meaning from the children’s accounts in these interviews. The findings suggest several avenues of improvement in the interviewing styles currently practised.

Lucinda Kalinski
University of Sheffield, UK

Taking Witness Statements: A Case-Study Analysis of Spoken and Written Discourse

When taking witness statements police officers are engaged in a complex task, acting as an information gatekeeper between the witness and future legal audience. Much interactional currency is involved as the officer asks questions, listens to narrative, comprehends concepts and transfers utterances into a written format. The resulting statement contains features of ‘policespeak,’ blends of register, and legalese to satisfy official crime criteria. A type of co-authorship arises with issues of comprehensibility and authenticity, however unintentional. Witness statements contain elements of the officer’s interpretation of meaning which may be explicitly defined through negotiation sequences, particularly relating to number references (height, distance etc). Through the statement-taking process there is potential for information to be omitted, altered or added through seemingly insignificant pragmatic changes, including hedge-use and vague language. This is comparable with the role undertaken by court interpreters who may portray a witness’s testimony as more or less certain than originally intended. Specifically, witness statement process-to-product information transfer can be categorised:

This research highlights the importance of appropriate police training and awareness-raising about the potential for statements to be subtly misrepresented, and legal implications involved.

Silke Kirschner
Assistant Crime Analyst, CENTREX (Central Police Training and Development Authority), National Crime and Operations Faculty, Serious Crime Analysis Section (SCAS)

Linguistics applied – The work of SCAS and the role of an assistant crime analyst

The presentation will first give a brief overview of the work of the Serious Crime Analysis Section within the National Crime and Operations Faculty before talking specifically about the analytical aspects of the work. Starting with historical aspects on how the section was founded in relation to the Yorkshire Ripper enquiry, I will then explain the structure and function of SCAS. In relation to this, the work of Behavioural Investigative Advisers and Geographic Profilers will be mentioned. In order to understand the function of SCAS as a service for police forces, I will talk about its role in ongoing enquiries on serious crimes. The underlying principle of “behavioural consistency” and its impact on investigative leads and criminal profiling will be discussed. This leads to the more specific work undertaken by the assistant crime analyst and crime analysts and how it relates to the detection of serial offenders. After giving a brief insight into the SCAS database and how linguistic research has been used for coding verbal interaction between the victim and offender, the vital importance of accuracy in data input and how this relates to an ongoing process of interpreting language will be illustrated. Finally, I will outline potential areas of future research and job opportunities in forensic linguistics.

Krzysztof Kredens
Department of English, The University of Birmingham

Forensic authorship attribution and language typology

On May 1st 2004, the European Union became a political structure of as many as 25 official languages. Of the many linguistic aspects of law-enforcement in this multilingual community, forensic authorship attribution deserves a particularly close scrutiny. It is the case that the same scientific issue, viz. the authorship of incriminating spoken or written substance - which one would expect to be resolved with the same results irrespective of the legal system - is treated differently in different European jurisdictions. English judges, for example, view linguistic evidence on authorship with great caution, whereas in Polish courts it is normally recognised as tenable and, on the assumption that the speech patterns of an individual are unique, can be used to convict.

Notwithstanding its legal and procedural determinants, such a state of affairs certainly requires a linguistic investigation. In this paper, I will explore the issue of different languages providing different stylistic material for analysis. Making a case for a contrastive forensic linguistics, I will seek to answer the question to what extent the individuating potential of style markers is dependent on the linguistic code of the language user. I will ask whether authorship evidence can have probative value in the case of language A, but not B, whether stylistic patterns are affected by language type or language family, and about the role of social and pragmatic factors in identifying authorship in different languages. I will illustrate with examples from English, Polish and Spanish.

Since an overwhelming majority of research in forensic authorship attribution has been conducted for the English language, I will also investigate whether stylometric techniques developed for English can be successfully employed for other languages.

Marijke Malsch
Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), Leiden, The Netherlands

The comprehensibility of court decisions

The use of legal jargon impedes an adequate communication between judges, prosecutors and defence lawyers on the one hand, and parties, defendants, witnesses and the public at large on the other. This may reduce the confidence in the legal system. This paper discusses findings of a recently conducted experiment, in which court decisions in Dutch criminal cases were transposed into every day language. In an experiment, lay persons, academics and lawyers assessed the compre­hen­sibility of both the original and the adjusted court decisions. Lay persons and academics found the sentences that had been rewritten more readable and comprehensible than the original ones. The paper discusses the findings of the research as well as its implications for the operation of the criminal justice system.

Mª Luz Vázquez Maroño
Department of English Philology, Universidade da Coruña, Spain

Deceptive Language Under Cross-Examination: Maxine Carr at the Soham Murder Trial

Maxine Carr, the female defendant at the Soham murder trial, openly admitted the accusation of persistent liar. However, she managed to escape a lengthy sentence in a trial with a background of deception prevailing throughout the process. This paper provides an analysis of two aspects. Firstly, what strategies the defendant followed in her struggle to be found not guilty by the jury; Maxine Carr presents a carefully designed lexical choice in an attempt to minimize responsibility for allegedly having perverted the course of justice and assisted an offender. Secondly, it is also described how the controlling and coercive methods of cross-examination are aimed at undermining her credibility. Thus, a non-neutral language is deliberately used by both witness and cross examiners who consciously engage in a question-answer process. Since linguistic features related to the defendant’s words reveal a significant changing attitude in two different temporal dimensions, the findings will show contradictory statements from previous police interviews and from answers given during the trial, which suggest when the defendant was really using deceptive language.

John Olsson
Forensic Linguistics Institute

Designing and building software to measure text style internally

I will examine the extent to which software can reasonably be expected to determine whether a text contains style variation and whether, if such variation is found, ways of determining its significance. The latter question is much harder to provide meaningful answers for, since different levels of significance can be proposed. Also, to what extent should analyst rely on the software to answer questions of internal style variation, and to what extent should we rely on own judgement?

Outline Of Talk:

Own experiences at building software; own limitations; how I became interested in this question; what I mean by 'text style' as opposed to author style; what is meant by 'internally'; what is 'variation'; different kinds of variation; how to classify them for detection by software program; keeping it simple yet effective - (e.g. all forms vs single form - impact of this on 'unique word total'); finding 'abstract' categories (e.g. passives, nominalisations, negation, plural, referencing) as opposed to items (e.g. specific words) and metrics (e.g. lexical density). Making software affordable to students.

Evelyn Perry
CICLaS Research Group, Université de Paris – Dauphine, France

Authorship and the nominal group : the case of the Eloquent Peasant

On the basis of certain similarities in the spheres of lexis and syntax, Hans Goedicke has advanced a theory of common authorship for a number of unsigned Egyptian texts from the Middle Egyptian period (2150 - 1300 BCE). Egyptologists have neither accepted nor attempted to refute this theory : they have simply ignored it. It is the purpose of this paper to explore Goedicke's suggestion through an analysis of two of the longest and best-preserved of the literary texts from ancient Egypt, The Shipwrecked Sailor and The Eloquent Peasant.The courtroom setting and technical legal language of the latter make it particularly relevant. Working within a Systemic Functional Linguistics framework, I will analyse the two works with respect to both the number and function of verbal and non-verbal sentences in each of the texts and the structure of the nominal group. Special attention will be given to post-modification in the nominal group as an indicator of authorship.

References

Goedicke, Hans The Report about the Dispute of a Man with His Ba, Baltimore : The Johns Hopkins University Press, 1970.

Perry, Evelyn. A Critical Study of the Eloquent Peasant. Doctoral thesis, presented at the Johns Hopkins University, 1986. University of Michigan Doctoral Dissertation Series.

Frances Rock
Department of English and Modern Languages, The University of Surrey Roehampton, UK

‘Now these statements always start the same’: one police officer’s work in interviews

Interactions between police officers and witnesses, engaged in producing written police statements through talk, are informing a developing body of literature. This paper will begin by briefly reviewing studies of this creative talk, seeking themes across those jurisdictions and processes which have, so far, attracted attention.

The session will then consider this text production as it is accomplished by one male police officer in combination with different witnesses across a number of encounters. This focus on a single officer makes it possible to explore the extent to which the officer undertakes joint text production consistently, irrespective of such factors as interactant, setting, cotext, context and crime. The paper will close by considering the implications of this examination for forensic linguistic work on such topics as verballing.

M. Teresa Turell
Institut Universitari de Lingüística Aplicada, Universitat Pompeu Fabra, Barcelona, Spain

The disputed authorship of electronic mail: linguistic, stylistic and pragmatic markers in short texts

Many cases of disputed authorship have to do with handwritten letters, usually short, which have to be deciphered for what they say, what they mean – although deciding on the semantics of a text is not always welcome by the Court – and, most importantly, who wrote them. In this paper I present evidence that was used by the defence in a case in which a company had been accused of unfair dismissal by an employee who had tried to extort them. The extortion was expressed in one of the e-mail letters (the disputed texts) written in Spanish, whose authorship he denied later on, and that he supposedly sent the company in a period of time during which he was also writing some faxes (the undisputed texts).

In order to make my case I turn to several working principles which are at the base of sociolinguistic and pragmatic approaches to language. The micro linguistic markers of authorship attribution that were used in the analysis are linguistic, stylistic and pragmatic. Among the linguistic ones: the use of the first person singular pronoun (1PS), the choice of which is useful to discuss the nature of authorship markers, idiosyncratic markers derived from language contact, and finally, language contact, grammatical and orthographic errors and non-standard forms. They all reveal linguistic similarity between the disputed and undisputed texts. The rarity or expectancy of the choices made in the texts was indicated through the use of the Spanish Corpus of the Real Academia Española (http://www.rae.es).

Abstracts by Theme

Authorship Analysis	Speaker Identification/Description
Emergency Calls	Police Interviews, Witness Statements, and Confessions
The Language of the Courtroom	Vulnerable Witnesses/Children in the Legal Process
Language of Judges

Authorship Analysis

Carole Chaski
Institute for Linguistic Evidence, Inc., Georgetown, DE, USA

Recent Validation Results for the Syntactic Analysis Method of Author Identification

The Syntactic Analysis Method of Author Identification has been briefly presented in Chaski (1997, 2001). This method differs from other non-lexical approaches to author identification (Stamatatos' et al. 2001) through its reliance on the linguistic principles of headship and markedness in grammar. Stamatatos' work is most comparable; using his parser to break the text into chunks through which he obtains 22 features, his method is able to achieve, using a discriminant function analysis, 81% accurate assignment of texts to ten authors, with the total words per author ranging from 17,000 to 52,000 words. Using a discriminant function analysis of 17 features per text, three texts of the four authors previously reported on in Chaski 2001 were assigned with 100% accuracy. When ten authors were tested, the method obtained 99% accuracy (one author-author text resulted in the DFA not obtaining enough variables to run). When thirteen authors were tested, the syntactic analysis method obtained 91% accurate assignment of the 72 documents to the correct author; the 9% that were not accurately assigned was due to the DFA not having enough variables to run, and all but one of those documents belonged to one author. These results demonstrate that the Syntactic Analysis Method has a high degree of reliability within the parameters of forensic application, a few and often short documents.

Malcolm Coulthard
University of Birmingham

A Guided Tour around Narrative Analysis

Solan (1998) addresses the problem which is specific to linguistic experts, the fact that the judges of fact, whether they be actual judges or jury members are seen, for most purposes, to be their own experts on language. Even so, in such a context, he argues, there is a role for the linguist, which is to explain facts about language and usage, as a result of which the lay judges will then be in the same position as the linguist and able to make their own informed decisions.

In this paper I will focus on a short confession narrative in which an accused purportedly confesses to being present when a murder took place. The accused claims that the narrative is a complete fabrication; the police that it is a verbatim transcript of what he said. I want to explore, using insights from narrative analysis and Gricean maxims, ways of showing a jury why the text cannot be an accurate, let alone a verbatim transcript of what the accused said.

Reference

Solan L 1998 ‘Linguistic experts as semantic tour guides’, Forensic Linguistics 5, ii, 87-106.

Krzysztof Kredens
Department of English, The University of Birmingham

Forensic authorship attribution and language typology

On May 1st 2004, the European Union became a political structure of as many as 25 official languages. Of the many linguistic aspects of law-enforcement in this multilingual community, forensic authorship attribution deserves a particularly close scrutiny. It is the case that the same scientific issue, viz. the authorship of incriminating spoken or written substance - which one would expect to be resolved with the same results irrespective of the legal system - is treated differently in different European jurisdictions. English judges, for example, view linguistic evidence on authorship with great caution, whereas in Polish courts it is normally recognised as tenable and, on the assumption that the speech patterns of an individual are unique, can be used to convict.

Notwithstanding its legal and procedural determinants, such a state of affairs certainly requires a linguistic investigation. In this paper, I will explore the issue of different languages providing different stylistic material for analysis. Making a case for a contrastive forensic linguistics, I will seek to answer the question to what extent the individuating potential of style markers is dependent on the linguistic code of the language user. I will ask whether authorship evidence can have probative value in the case of language A, but not B, whether stylistic patterns are affected by language type or language family, and about the role of social and pragmatic factors in identifying authorship in different languages. I will illustrate with examples from English, Polish and Spanish.

Since an overwhelming majority of research in forensic authorship attribution has been conducted for the English language, I will also investigate whether stylometric techniques developed for English can be successfully employed for other languages.

John Olsson
Forensic Linguistics Institute

Designing and building software to measure text style internally

I will examine the extent to which software can reasonably be expected to determine whether a text contains style variation and whether, if such variation is found, ways of determining its significance. The latter question is much harder to provide meaningful answers for, since different levels of significance can be proposed. Also, to what extent should analyst rely on the software to answer questions of internal style variation, and to what extent should we rely on own judgement?

Outline Of Talk:

Own experiences at building software; own limitations; how I became interested in this question; what I mean by 'text style' as opposed to author style; what is meant by 'internally'; what is 'variation'; different kinds of variation; how to classify them for detection by software program; keeping it simple yet effective - (e.g. all forms vs single form - impact of this on 'unique word total'); finding 'abstract' categories (e.g. passives, nominalisations, negation, plural, referencing) as opposed to items (e.g. specific words) and metrics (e.g. lexical density). Making software affordable to students.

Evelyn Perry
CICLaS Research Group, Université de Paris – Dauphine, France

Authorship and the nominal group : the case of the Eloquent Peasant

On the basis of certain similarities in the spheres of lexis and syntax, Hans Goedicke has advanced a theory of common authorship for a number of unsigned Egyptian texts from the Middle Egyptian period (2150 - 1300 BCE). Egyptologists have neither accepted nor attempted to refute this theory : they have simply ignored it. It is the purpose of this paper to explore Goedicke's suggestion through an analysis of two of the longest and best-preserved of the literary texts from ancient Egypt, The Shipwrecked Sailor and The Eloquent Peasant.The courtroom setting and technical legal language of the latter make it particularly relevant. Working within a Systemic Functional Linguistics framework, I will analyse the two works with respect to both the number and function of verbal and non-verbal sentences in each of the texts and the structure of the nominal group. Special attention will be given to post-modification in the nominal group as an indicator of authorship.

References

Goedicke, Hans The Report about the Dispute of a Man with His Ba, Baltimore : The Johns Hopkins University Press, 1970.

Perry, Evelyn. A Critical Study of the Eloquent Peasant. Doctoral thesis, presented at the Johns Hopkins University, 1986. University of Michigan Doctoral Dissertation Series.

M. Teresa Turell
Institut Universitari de Lingüística Aplicada, Universitat Pompeu Fabra, Barcelona, Spain

The disputed authorship of electronic mail: linguistic, stylistic and pragmatic markers in short texts

Many cases of disputed authorship have to do with handwritten letters, usually short, which have to be deciphered for what they say, what they mean – although deciding on the semantics of a text is not always welcome by the Court – and, most importantly, who wrote them. In this paper I present evidence that was used by the defence in a case in which a company had been accused of unfair dismissal by an employee who had tried to extort them. The extortion was expressed in one of the e-mail letters (the disputed texts) written in Spanish, whose authorship he denied later on, and that he supposedly sent the company in a period of time during which he was also writing some faxes (the undisputed texts).

In order to make my case I turn to several working principles which are at the base of sociolinguistic and pragmatic approaches to language. The micro linguistic markers of authorship attribution that were used in the analysis are linguistic, stylistic and pragmatic. Among the linguistic ones: the use of the first person singular pronoun (1PS), the choice of which is useful to discuss the nature of authorship markers, idiosyncratic markers derived from language contact, and finally, language contact, grammatical and orthographic errors and non-standard forms. They all reveal linguistic similarity between the disputed and undisputed texts. The rarity or expectancy of the choices made in the texts was indicated through the use of the Spanish Corpus of the Real Academia Española (http://www.rae.es).

Speaker Identification/Description

Betsy Evans
Centre for Language and Communication, Cardiff University, UK

Imitation and speaker identification: a case study

It has been previously assumed by sociolinguists that speakers can “perform” only gross stereotypical characteristics of other varieties, simply borrowing from their own repertoire of sounds (e.g. Labov 1994). This implies that people cannot perform imitations ‘accurately’. This study explores ‘accuracy’ in a phonetic context via a case study that examines the phonetic match between an imitation and known phonetic characteristics of the imitated dialect. Recordings were made of a sociolinguistic interview during which the respondent read a word list and reading passage in his “usual,” that is, “not-imitating,” speech and in what he perceived to be U.S. “West Virginia” speech. Acoustic analysis of the data revealed notable differences in the first and second formants of the vowels in these two modes and considerable consistency in the imitation and known phonetic features of the target. This paper also examines the reactions of native West Virginians to Noah’s speech in order to determine whether or not the imitation is successful perceptually. Results showed that the majority of respondents perceived the imitation as authentically West Virginian. The acoustic and perceptual success of this respondent’s imitation has important implications for forensic speaker identification and indicates that more research of this kind is necessary for a better understanding of imitation.

Tim Grant and Jessica Woodhams
Forensic Section, School of Psychology, University of Leicester

Categorising rapists’ utterances

Whilst rape victim statements frequently refer to dialog between offender and victim, the language offenders use in the course of their offence is an under-researched aspect of offence behaviour. Holstrom and Burgess provided descriptions and some categorisation of rapists’ talk, since then only Dale, Davies and Wei attempt to provide a systematic typology of rapists’ utterances. Application of this typology demonstrated both strengths and weaknesses and led to the development of the current system. This draws more directly on speech act theory to categorise utterances as directives, constatives, commissives, expressives and interrogatives and makes use of subscales to further differentiate utterances. High reliability is reported for both systems, the new system however categorises a greater proportion of utterances in the statements. Current applications include examining differences between statements which are maintained as true from those which are withdrawn as false and examining behavioural consistency for the purpose of case linkage.

Mark Griffiths
Centre for Language and Communication, Cardiff University

Forensic earwitness descriptions: predictability, consistency and reliability

Whilst it can be safely assumed that trained linguists may be best placed to provide and solicit descriptions of voices, a knowledge of linguistics and its meta-language rarely extends beyond those who have been specifically trained. It follows then, that in the legal context, only a tiny minority of earwitnesses and police officers will have any such training, yet lay-linguist descriptions are regularly solicited by lay-linguist police officers, with these descriptions forming a vital part of police evidence.

Moreover, a substantial tradition of research has indicated that lay linguists may have a variety of, often predictable, responses to regional and social accents, manner of speech and perceived speaker power, which applied to the forensic context, raises important questions regarding the predictability, consistency and reliability of earwitness descriptions. Drawing on the results of a pilot study of 150 1st year undergraduate students’ responses to recorded voices of suspected criminals, this paper aims to illustrate how lay voice descriptions and attributions of guilt may be qualitatively affected i/. by attitudinal variations towards speech varieties and manner of speech ii/. by variations in questionnaire design and voice line-up iii/. by earwitness desire to give socially acceptable responses.

Silke Kirschner
Assistant Crime Analyst, CENTREX (Central Police Training and Development Authority), National Crime and Operations Faculty, Serious Crime Analysis Section (SCAS)

Linguistics applied – The work of SCAS and the role of an assistant crime analyst

The presentation will first give a brief overview of the work of the Serious Crime Analysis Section within the National Crime and Operations Faculty before talking specifically about the analytical aspects of the work. Starting with historical aspects on how the section was founded in relation to the Yorkshire Ripper enquiry, I will then explain the structure and function of SCAS. In relation to this, the work of Behavioural Investigative Advisers and Geographic Profilers will be mentioned. In order to understand the function of SCAS as a service for police forces, I will talk about its role in ongoing enquiries on serious crimes. The underlying principle of “behavioural consistency” and its impact on investigative leads and criminal profiling will be discussed. This leads to the more specific work undertaken by the assistant crime analyst and crime analysts and how it relates to the detection of serial offenders. After giving a brief insight into the SCAS database and how linguistic research has been used for coding verbal interaction between the victim and offender, the vital importance of accuracy in data input and how this relates to an ongoing process of interpreting language will be illustrated. Finally, I will outline potential areas of future research and job opportunities in forensic linguistics.

Emergency Calls

Mark Garner, Northumbria University and Edward Johnson, Wolfson College, Cambridge

Operational communication theory as a framework for research into emergency call-handling by the police

Emergency call-handling confronts the researcher with a double challenge. The relatively limited research that has been done in this field has focussed on linguistic issues through discourse analysis, but a more complex and broadly-based approach is necessary if the outcomes of our current research are to be theoretically valid, professionally acceptable, and practically useful by the end users, the police. Whilst it is necessary to analyse the exchanges between caller and call-handler as linguistic interactions, recommendations about the linguistic aspects of the communication must conform to the evolving operational requirements of the police force, including training and development of appropriate IT support, and must meet the needs of the public who make the calls.

We outline a theoretical framework for this sort of institutionally and culturally situated analysis, presenting a series of basic principles for operational communication research that have evolved in the course 25 years of experience of work in maritime, air traffic, and inter-agency communication, employing a range of written, spoken and visual media. These principles go beyond conventional linguistic analysis to embrace an ecological view of the multi-stranded nature of communication within institutional settings. We demonstrate how they are being applied within our on-going research into the development of national call-handling standards.

Police Interviews, Witness Statements and Confessions

Kelly Benneworth
Loughborough University, UK

A discursive analysis of police interviews with suspected paedophiles: The implications of ‘open’ and ‘closed’ interviewing for admission and denial.

This study examines the discursive interaction between police officers and suspected paedophiles in the investigative interview. Paedophiles talk about their offences in terms of conventional relationships, personal bonds and emotions whilst being discrete about the sexual aspects of their activities. In the investigative interview, police officers must encourage paedophiles to discuss their criminal activities in terms of direct, sexual detail. Given these two distinct approaches to the description of unlawful sexual contact, there is potential for difficulties to arise in the elicitation of confession. Eleven audiotaped interviews conducted at Leicestershire Police Constabulary were transcribed. The offenders were male and aged between 34-54 years and the victims were male (n=5) and female (n=6) and aged between 5-13 years. Discourse analysis explored how the police officers and paedophiles negotiated an account of the offence whilst managing conflicting descriptions of sexual relationships between adults and children. The analysis revealed two distinct styles of police interviewing with implications for confession, ‘open’ interviewing which maximised suspect admission and ‘closed’ interviewing which encouraged denial. In addition to offering a distinctive qualitative understanding of how language clashes shape the progression of the police interview, the findings support the development of a discourse-based training programme to assist police officers in the interviewing of paedophiles. The findings also provide a vocabulary for interviewing officers to reflect on their own practices and communicate their skills to less experienced officers.

Gill Grebler
Cultural and Linguistic Forensics, Santa Monica, CA

False confessions: Fictional tellings with legal consequences

A false confession can be said to have a kind of ‘narrative truth’ although it is not actually true. At first glance this fictional telling displays a convincing verisimilitude. It convinces judge and jury that a suspect is guilty of a crime.

I look at a set of vivid and detailed accounts from proven false confessions, of crimes told by innocent suspects as if they had committed them. These suspects were forced (usually after extended struggle to assert their innocence) to pretend to remember, to create an account - vivid and first-hand - of something they did not do. I look at the linguistic coercion that brought an innocent person to ‘remember’ having committed a crime and at the process of co-telling that was used to frame and shape their ‘post-admission narratives’.

During the co-construction process interrogators seek to fill in the details required of a legally valid confession, and the specifics of the particular crime. They feed information, the suspect draws upon various sources of ‘memory’, and aspects of a suspect's story-telling ability is activated. A workable story is produced although there is no actual memory to base it on.

Phil Hall
Appen Speech Technology and Macquarie University, Sydney, Australia

Interactional constraints on the capacity of suspects to invoke rights

While interrogative practices oriented to the retrieval from suspects of the specific types of information considered constitutive of confession are easily recognisable as characteristic of police questioning, there are other interactional practices whose links to the function of the interview are less readily apparent, but that nonetheless have the capacity to influence the course of the interview with potentially significant consequences for the interviewee/suspect. This paper will examine several examples of failed attempts by suspects to invoke legal rights, and will ask whether interviewers' attempts at "rapport building" practices contributed to the failure of these attempts by establishing a relationship between the interview participants in which suspects were constrained from unequivocal invocation of their rights by adherence to interactional norms.

Kerry Ham
University of Florida, USA

Conversational Maxims in Encounters with Law Enforcement Officers

This paper analyses the use of spoken language by suspects and witnesses being questioned by law enforcement officers. The aim is to discover whether, and to what extent, Grice’s (1975) Cooperative Principle and Brown and Levinson’s (1987) Politeness Principle are in evidence, and how adherence to or breaking these conversational maxims helps or hinders the interrogation process.

Beginning with an analysis of segments of the popular American reality-television show “Cops”, the analysis then draws on experiences gained whilst participating in numerous ‘ride-longs’ with the Alachua County Sheriff’s Office and the Gainesville Police Department, both located in North Florida.

The resultant analysis considers possible explanations as to why some suspects are arrested or taken into custody, whereas others are allowed to go free. The possibility will be considered that successful manipulation Grice’s Cooperative Principle (and the associated Conversational Maxims, in which Politeness is included for the purpose of this study) may have some degree of subconscious influence on police officers’ decisions regarding a suspect’s guilt or innocence. From the media analysis and associated experiences, conclusions are drawn as to the extent to which adherence to or breaking each Conversational Maxim may influence the outcome of an encounter with law enforcement officers.

Lucinda Kalinski
University of Sheffield, UK

Taking Witness Statements: A Case-Study Analysis of Spoken and Written Discourse

When taking witness statements police officers are engaged in a complex task, acting as an information gatekeeper between the witness and future legal audience. Much interactional currency is involved as the officer asks questions, listens to narrative, comprehends concepts and transfers utterances into a written format. The resulting statement contains features of ‘policespeak,’ blends of register, and legalese to satisfy official crime criteria. A type of co-authorship arises with issues of comprehensibility and authenticity, however unintentional. Witness statements contain elements of the officer’s interpretation of meaning which may be explicitly defined through negotiation sequences, particularly relating to number references (height, distance etc). Through the statement-taking process there is potential for information to be omitted, altered or added through seemingly insignificant pragmatic changes, including hedge-use and vague language. This is comparable with the role undertaken by court interpreters who may portray a witness’s testimony as more or less certain than originally intended. Specifically, witness statement process-to-product information transfer can be categorised:

Category	Item Spoken	Realised?	Item Written
1	yes	→	yes
2	yes	→	(yes)
3	yes	→	no
4	no	→	yes

This research highlights the importance of appropriate police training and awareness-raising about the potential for statements to be subtly misrepresented, and legal implications involved.

Frances Rock
Department of English and Modern Languages, The University of Surrey Roehampton, UK

‘Now these statements always start the same’: one police officer’s work in interviews

Interactions between police officers and witnesses, engaged in producing written police statements through talk, are informing a developing body of literature. This paper will begin by briefly reviewing studies of this creative talk, seeking themes across those jurisdictions and processes which have, so far, attracted attention.

The session will then consider this text production as it is accomplished by one male police officer in combination with different witnesses across a number of encounters. This focus on a single officer makes it possible to explore the extent to which the officer undertakes joint text production consistently, irrespective of such factors as interactant, setting, cotext, context and crime. The paper will close by considering the implications of this examination for forensic linguistic work on such topics as verballing.

The Language of the Courtroom/Hearings

Janet Cotterill
Cardiff University

Answers in the Affirmative?:: Yes, No and Maybe in the British Courtroom Context

This presentation draws on a 1-million word sub-corpus of witness examinations and cross-examinations taken from trials held in the late 1990’s in the UK, and applies a combination of corpus linguistic and discourse analytic approaches to study the semantics, pragmatics and discursive characteristics of yes and no in the courtroom setting. These short responses are, on the face of it, semantically transparent and unambiguously affirmative or negative in their orientation. But this most frequent type of witness response is not as pragmatically simple as it might at first seem. This paper reports a study investigating the contextual semantics and pragmatics of this common form of witness response, and the subtle nuances of meaning which are exploited by both barristers and witnesses during examination and cross-examination. It explores the grey area between yes and no and suggests some of the discursive consequences when witnesses do not provide a definitive affirmative or negative response to a yes/no-type question.

Maya David
Faculty of Languages and Linguistics, University of Malaya, Malaysia

Language Choice, Law, and Power in a Bilingual Setting

Malay is the National Language of Malaysia. English is an important second language. Many of the older lawyers are proficient in English as English was the medium of instruction when they were in school. The younger lawyers had Malay as the medium of instruction and are consequently more proficient in Malay. As in most other settings in Malaysia, (see Omar, 1982; David, 2001, Morais, 1998) a mixed discourse appears to have become a regular feature in this formal setting too. The aim of this research is to determine the functions of code switching in Malaysian courts of law and was based on the premise that as Malaysians have differing levels of fluency in different languages they would consequently be obliged to code switch to effectively convey meaning. In a study on code switching in Malaysian service encounters David, (1999) found that language convergence rather than language divergence was rampant as negotiation was the goal of both the buyer and seller. The fact that rampant asymmetrical or divergent language choice also occurs in formal settings like the courts indicates that non-accommodation of language choice implies power and control of one party over the other. Language choice and the use of a mixed discourse in the Malaysian courts are triggered to achieve a range of strategic and professional objectives. The reasons for such mixed discourse in the legal setting are discussed. How this analysis empowered the teaching of a relatively new course "Communicative Skills for Professionals" - a course for final year undergraduates in the University of Malaya, is also described. It is imperative that all ESP practitioners conduct action research to empower their teaching.

Mª Luz Vázquez Maroño
Department of English Philology, Universidade da Coruña, Spain

Deceptive Language Under Cross-Examination: Maxine Carr at the Soham Murder Trial

Maxine Carr, the female defendant at the Soham murder trial, openly admitted the accusation of persistent liar. However, she managed to escape a lengthy sentence in a trial with a background of deception prevailing throughout the process. This paper provides an analysis of two aspects. Firstly, what strategies the defendant followed in her struggle to be found not guilty by the jury; Maxine Carr presents a carefully designed lexical choice in an attempt to minimize responsibility for allegedly having perverted the course of justice and assisted an offender. Secondly, it is also described how the controlling and coercive methods of cross-examination are aimed at undermining her credibility. Thus, a non-neutral language is deliberately used by both witness and cross examiners who consciously engage in a question-answer process. Since linguistic features related to the defendant’s words reveal a significant changing attitude in two different temporal dimensions, the findings will show contradictory statements from previous police interviews and from answers given during the trial, which suggest when the defendant was really using deceptive language.

Marta Carretero
Department of English Language and Linguistics, Facultad de Filología, Universidad Complutense, Madrid

The role of evidentiality and epistemic modality in a solicitor's submission within divorce proceedings

It is well-known that evidential and epistemic expressions are reliable indicators of the stance taken by language users in discourse. In this paper I will illustrate this statement by analyzing a 5,000-word spoken text from an English corpus. This text is a fragment of a submission by the solicitor who represents the husband in a divorce case, with occasional interruptions by the judge. The use of epistemic and evidential expressions by both speakers reflects their different stances: the solicitor takes a clear position (namely that the main cause of the divorce lies in the wife's behaviour), hence her frequent use of epistemic expressions conveying a high degree of personal commitment, often combined with evidentials indicating a wide array of attitudes towards different sources of evidence. On the other hand, the judge displays a much more cautious attitude, which accounts for the relatively weak epistemic expressions and for the scarce evidential expressions in his turns.

References

Chafe, W. (1986) Evidentiality in English conversation and academic writing. In W. Chafe & J. Nichols (eds.) Evidentiality: the Linguistic Coding of Epistemology. (Advances in Discourse Processes, XX.) Norwood. N.J.: Ablex. 261-272.
Dendale, P. & L. Tasmowski (2001) Introduction: evidentiality and related notions. Journal of Pragmatics 33.3, 339-348.
Downing, A. (2001) "Surely you knew!" Surely as a marker of evidentiality and stance. Functions of Language 8.2, 253-285.
Marín-Arrese, J., Hidalgo, L. & S. Molina (2002) Evidentiality and modality in English and Spanish: connections and interactions. In R. Facchinetti (ed.) English Modality in Perspective: Genre Analysis and Contrastive Studies. Verona: Università di Verona. 135-146.
Nuyts, J. (2000) Epistemic Modality, Language and Conceptualization. A Cognitive-Pragmatic Perspective. Amsterdam / Philadelphia: John Benjamins.
Plungian, V. (2001) The place of evidentiality within the universal grammatical space. Journal of Pragmatics 33.3, 349-357.
Svartvik, J. and Quirk, R. (eds.) (1980) A Corpus of English Conversation. Lund: C.W.K. Gleerup.
Taverniers, M. (1998) Towards a topology of interpersonal meaning. Paper presented at the 10th Euro-International Systemic Functional Workshop, University of Liverpool, July 1998.
Thompson, G. & Y. Ye (1991) Evaluation in the reporting verbs used in academic papers. Applied Linguistics 12.4, 364-382.
White, P.R.R. (1998). An introductory tour through appraisal theory. Paper Presented at the 10th Euro-International Systemic Functional Workshop, University of Liverpool, July 1998.

Lise Fontaine
Centre for Language and Communication, Cardiff University, UK

Your money or your life : Safety and Corporate Crime in the workplace

“The 26 Canadian miners who were killed at the Westray Mine had two choices: to accept their employer's negligent and unsafe workplace practices or to lose their jobs.” [1]

This paper has two main purposes. The first is to introduce the area of Safety and Corporate Crime in the workplace, including a discussion of the amendments made in November 2003 to the criminal code in Canada: Bill C-45. “It (had) already (been) possible for a corporation to be convicted of criminal negligence under the Criminal Code. … Bill C-45 is specifically directed to making convictions for criminal negligence occur more readily for complex organizations, including corporations.” (Edwards and Collins, 2003) The amendments to the code include: new duties, an increased breadth of application, and new “criminal” offences. Specific attention will be given to the most important defense available in such cases, due diligence, and to the role of linguistic evidence.

The second focus of this paper is on a reprisal case where an employee filed an “unlawful reprisal application” [2] against his employer where he claims he was fired for reporting safety violations. Using the judgement in the case and the official transcripts of the hearing, this study analyses the referring expressions used by the various participants with respect to safety, the company’s attention to safety in the workplace and official attitudes as stated in the judgement.

References
Bill C-45 – Amendments to the Criminal Code Affecting the Criminal Liability of Organizations http://canada.justice.gc.ca/en/dept/pub/c45/CCL_english.pdf
Cotterill, J. (2001). “Domestic Discord, Rocky Relationships: Semantic Prosodies in Representations of Marital Violence in the O.J. Simpson Trial.” Discourse & Society, 12(3), 291-312.
Edwards, Cheryl A. and Ryan J. Conlin (2003) “Workplace Safety Amendments to Criminal Code Introduced” http://www.sbhlawyers.com/repo/index.asp?action=edit&report_cat_id=21
Slapper, Gary (2000) Blood in the Bank: Social and Legal Aspects of Death at Work. Aldershot: Ashgate Publishing.

[1] http://www.ccriw.com/campaign.html
[2] Section 50 of the Occupational Health and Safety Act of Ontario states:

(1) No employer or person acting on behalf of an employer shall,
(a) dismiss or threaten to dismiss a worker;
(b) discipline or suspend or threaten to discipline or suspend a worker;
(c) impose any penalty upon a worker; or
(d) intimidate or coerce a worker,
because the worker has acted in compliance with this Act or the regulations or an order made thereunder, has sought the enforcement of this Act or the regulations or has given evidence in a proceeding in respect of the enforcement of this Act or the regulations or in an inquest under the Coroners Act. R.S.O. 1990, c. O.1, s. 50 (1).

Vulnerable Witnesses/Children in the Legal Process

Michelle Aldridge
Centre for Language and Communication, Cardiff University

Vulnerable witnesses' perception of the criminal justice service in North Wales

In 2002-2003 a study was undertaken to determine the public's perception of the Criminal Justice Service in North Wales. Views were elicited from a wide range of the public including known victims, witnesses and convicted offenders and members of the public who had no experience of the system. This paper reports on part of the finding namely the perceptions of those people who would fall into the category ' vulnerable'. Striking differences were recorded between these people and the rest of the sample. As examples, vulnerable respondents rated all of the agencies within CJS as more likely to treat them unfairly as both victims and accused, they had a greater fear of assault and race/hate crime and in most cases, they were less likely to report crime. The implications of these findings are discussed.

Ronald R. Butters
Duke University, Durham, NC, USA

Evidence of the Rehearsal of a Videotaped Confession as Support for a Diminished Capacity Defense in USA Death-Penalty Trials

Although sentenced to death in 1977, Johnny Lee Gates remained on death row until 2003, when a new trial attempted to determine if he was too severely mentally retarded to be legally executable. The later trial could not consider Gates’s guilt or innocence, but a forensic psychologist was prepared to testify that evidence of confession-rehearsal bore vitally on Gates’s putative mental retardation. (Weeks after the crime, police had taken Gates to the scene and obtained a videotaped confession wherein he walked through his purported role while giving incriminating answers to a policeman’s questions.) Linguistically, I determined, four types of pragmatic and discourse evidence supported the psychologist’s testimony:

• Gates was often asked questions that presupposed information that had not yet been established in the interview.
• Gates gave many responses that went well beyond what the questions called for.
• Questions were always “safe” questions; Gates could easily have been coached to give his answers; obvious informational questions were never asked, however, if the answers would have been contradictory or difficult for one to answer who had not committed the crime.
• Some answers were exculpatory, in which case the questioner recycled the questions until he received an incriminating answer.

Alison Johnson
School of Education, University of Birmingham, UK

Producing the voice of the child victim: the role of the interviewer in police interview interaction

Literature on interaction between adults and children in the early acquisition of language notes the importance of questions in scaffolding talk and eliciting conversation. Maternal and caregiver speech expands child utterances (Galloway and Richards, 1994) and children make use of this input in their language growth. In a similar way interviewers perform an important role in the production of the narrative voice of the child witness in police cases involving child abuse victims. This paper looks at child directed speech in police interviews with child witnesses, investigating the interactional relationship between interviewer and interviewee. A corpus of six interviews with children is examined; all cases are located in the U.K. and were carried out by six different interviewers with six children. Analytical perspectives come from an approach that combines the resources of (forensic) discourse analysis and first language acquisition within a legal setting. I suggest (through comparisons made with interviews with adults) that particular questioning sequences are crucial to the growth of a full narrative account of the crime being investigated. And- and so-prefaced questions, in particular, have important eliciting, clarifying and upgrading roles in the interaction. They are crucial in avoiding breakdown and achieving successful creation of narrative in these interview conversations. I conclude that there is a relationship between the interviewer’s interactional behaviour and the child’s production and performance. Producing, or leading forward, the voice of the child demands sensitivity to the cognitive and communicative pressures of the narrative task; the interviewer’s role is scaffolding production in the co-produced narrative.

Cecilia Joseph
Centre for Language and Communication, Cardiff University and University of Malaya, Malaysia

Are you listening to Me? Constructing Meaning and Responding to Children’s Accounts in Forensic Interviews

This paper is part of a Doctoral research that I am developing at Cardiff University. The focus of the research is on the role of listening in police interviews with children. In the UK, it has been the practice since 1992 to use video recordings of forensic interviews with child witnesses as evidence in criminal proceedings dealing with sexual or violent offences. Data from 20 police interviews of children ranging from the ages of 5 to 8 years old are used to explore how, in practical terms, police interviewers listen in these interviews. Drawing upon a holistic comparative analysis, this study examines how police officers respond to and construct meaning from the children’s accounts in these interviews. The findings suggest several avenues of improvement in the interviewing styles currently practised.

The Language of Judges

Silvia Hansen-Schirra and Stella Neumann
Institute of Applied Linguistics & Institute of Law and Informatics, Saarland University, Germany

Comprehensibility in English and German court decisions

In the discussion of how to make the law transparent to the citizens who are subject to it, efforts are made to “translate” legal texts for the lay public. One initiative in this context is the publication of press releases by the German Federal Constitutional Court reporting on each of the Court’s decisions. The present work describes features of the language of law in these decisions in comparison to their realisation in the press releases.

Using corpus linguistic methods such as frequency counts, collocations, part-of-speech tagging and parsing we investigate a corpus of court decisions, press releases as well as a register neutral reference corpus. The analysed features comprise mean sentence length, type-token-ratio, complexity of NPs, nominalisations as well as formulaic expressions. We show how the features are modified or deleted in order to make the decisions comprehensible to a lay public. While the analysis mainly focuses on the German corpus we also draw comparisons to a corpus of British Privy Council decisions in order to find out whether there are any contrastive similarities. This serves our long term goal to develop a system for the computer-assisted translation of legal texts.

Chris Heffer
Nottingham Trent University, UK

Others’ Stories in Other Words: Narrative Strategies in the Judicial Summing-up

In reviewing the evidence in a criminal case during their summing-up to the jury, Crown Court judges in England and Wales find themselves in a particularly complex communicative situation. They inevitably must deal with stories in their summaries of the evidence: stories of the crime seen from different perspectives; competing stories of the investigation and trial proceedings presented by counsel in closing speeches; the narrative schemas held by most jurors. At the same time, they are not meant to be constructing their own subjectively viewed stories, but only assisting jurors in applying the law to the many tales and tellers that appeared in the trial. Drawing on a large corpus of transcripts of summings-up from English jury trials, this paper explores the remarkably different strategies employed by judges in presenting these complex and partial narratives, and considers the extent to which those judges actually adopt narrative techniques themselves.

Marijke Malsch
Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), Leiden, The Netherlands

The comprehensibility of court decisions

The use of legal jargon impedes an adequate communication between judges, prosecutors and defence lawyers on the one hand, and parties, defendants, witnesses and the public at large on the other. This may reduce the confidence in the legal system. This paper discusses findings of a recently conducted experiment, in which court decisions in Dutch criminal cases were transposed into every day language. In an experiment, lay persons, academics and lawyers assessed the compre­hen­sibility of both the original and the adjusted court decisions. Lay persons and academics found the sentences that had been rewritten more readable and comprehensible than the original ones. The paper discusses the findings of the research as well as its implications for the operation of the criminal justice system.

• General info
• Programme
• Abstracts by name
• Abstracts by theme