Forensic linguistics

Forensic linguistics is the application of linguistic knowledge, methods and insights to the forensic context of law, crime investigation, trial, and judicial procedure. It is a branch of applied linguistics. There are principally three areas of application for linguists working in forensic contexts - understanding language of the written law, understanding language use in forensic and judicial processes and the provision of linguistic evidence. The discipline of forensic linguistics is not homogenous; it involves a range of experts and researchers in different areas of the field.

Brief History

The phrase forensic linguistics was only made known in 1968 when Jan Svartvik, a linguistic professor recorded its first mention in a famous analysis of statements by Timothy John Evans.

United Kingdom

During the early days of forensic linguistics in the United Kingdom, many criminal cases questioned the authenticity of police statements. In the past, police statements contained phrases such as 'I then observed', etc. During the interrogation, a police officer will take down notes and type the suspect's statement not in the words of the suspect but in a form that the police custom dictates. Statements given by witnesses are very seldom in a coherent or orderly fashion, with speculation and backtracking done out loud. The delivery is often too fast-paced, causing important details to be left out.

United States

Early work of forensic linguistics in the USA concerned the rights of individuals with regard to the interrogation process.

An early application of forensic linguistics in the United States was related to the status of trademarks as words or phrases in the language. One of the bigger cases involved fastfood giant McDonald's claiming that it had originated the process of attaching unprotected words to the 'Mc' prefix and was unhappy with Quality Inns International's intention of opening a chain of economy hotels to be called 'McSleep.'

Australia

In the 1980s Australian linguists discussed the application of linguistics and sociolinguitics to legal issues.They discovered that phrases such as ' the same language ' is opened to interpretation. Aboriginal people have their own understanding and use of 'English', something that is not always appreciated by speakers of the dominant version of English, i.e., 'white English'. The Aboriginal people also bring their own culturally-based, interactional styles to the interview.

Areas of study

The range of topics within forensic linguistics is diverse, but research occurs in the following areas:

The language of legal texts

The study of the language of legal texts encompasses a wide range of forensic texts, that is, text types and forms of analysis. Any text or item of spoken language can potentially be a forensic text, as long as it is somehow implicated in a legal or criminal context. This includes analysing the linguistics of documents as diverse as Acts of Parliament (or other law-making body), private wills, court judgements and summonses and the statutes of other bodies, such as States and government departments. One important area is that of the transformative effect of Norman French and Ecclesiastic Latin on the development of the English common law, and the evolution of the legal specifics associated with it. It can also refer to the ongoing attempts at making legal language more comprehensible to laypeople.

The language of legal processes

Amongst other things, this area examines language as it is used in cross-examination, evidence presentation, judge's direction, police cautions, police testimonies in court, summing up to a jury, interview techniques, the questioning process in court and in other areas such as police interviews.

Forensic text types

Emergency call

In an emergency call, the recipient or emergency operator's ability to extract primarily linguistic information in threatening situations and to come up with the required response in a timely manner is highly important. Intonational emphasis, voice pitch and the extent to which there is cooperation between the caller and the recipient at any one time are also very important in analyzing an emergency call. Full cooperation include frank and timely responses.

Urgency plays a big role in emergency calls so hesitations, signs of evasiveness and incomplete or overly short answers indicate that the caller might be making a false or hoax call. A genuine call has distinctive interlocking and slight overlap of turns. The recipient trusts the caller to provide accurate information and the caller trusts the recipient to ask only pertinent questions. If the caller uses a rising pitch at the end of every turn, it might represent a lack of commitment; whereas the recipient's use of a rising pitch indicates doubt or desire for clarification. The call ideally moves from nil knowledge on the part of the recipient to maximum amount of knowledge in a minimum possible period of time. This makes the emergency call unlike any other kind of service encounter.

Ransom demands or other threat communication

Threat is a counterpart of a promise and is an important feature in a ransom demand. Ransom demands are also examined to identify between a genuine and a false threat. An example of a ransom note analysis is in the case of the Lindbergh kidnapping, where the ransom note stated: "We warn you for making anyding public or for notify the polise the child is in gute care". From the sentence, the kidnapper makes the claim that the child is in good hands but to make such a claim, the note would have to be written before the perpetrator enters the premises. Therefore, the claim is false since the kidnapper had not even encountered the child when he wrote the note.

Suicide letters

A suicide note is typically brief, concise and highly propositional with a degree of evasiveness. A credible suicide letter must be making a definite unequivocal proposition in a situational context. The proposition of genuine suicide is thematic, directed to the addressee (or addressees) and relevant to the relationship between them. Suicide notes generally have sentences alluding to the act of killing oneself, or the method of suicide that was undertaken. The contents of a suicide note could be intended to make the addressee suffer or feel guilt. Genuine suicide letter are short with less than 300 words in length. Extraneous or irrelevant material are often excluded from the text.

Death row statements

Death row statements either admit the crime (leave the witness with an impression of honesty, forthrightness, etc) or deny the crime (leave the witness with an impression of innocence). Death row statements also denounce witnesses as dishonest, critique law enforcement as corrupt in an attempt to portray innocence and seek an element of revenge in their last moments. Death row statements are within the heavily institutionalized setting of death row prisons, i.e., only someone who is facing imminent death are capable of producing such texts.

Use of linguistic evidence in legal proceedings

These areas of research have varying degrees of acceptability or reliability within the field. Linguists have provided evidence in:

• Trademark and other intellectual property disputes
• Disputes of meaning and use
• Author identification (determining who wrote an anonymous texts by making comparisons to known writing samples of a suspect) (such as threat letters, mobile phone texts, emails)
• Forensic stylistics (identifying cases of plagiarism)
• Voice identification, also known as forensic phonetics, used to determine, through acoustic qualities, if the voice on a tape recorder is that of the defendant)
• Discourse analysis (the analysis of the structure of written or spoken utterance to determine who is introducing topics or whether a suspect is agreeing to engage in criminal conspiracy)
• Language analysis (forensic dialectology) tracing the linguistic history of asylum seekers
• Reconstruction of mobile phone text conversations
• Forensic phonetics

Specialist databases of language (called corpora) are now frequently being used by forensic linguists. These include corpora of suicide notes, mobile phone texts, police statements, police interview records and witness statements.

Author identification

The identification of whether a given individual said or wrote something relies on analysis of their idiolect, or particular patterns of language use (vocabulary, collocations, pronunciation, spelling, grammar, etc). The idiolect is a theoretical construct based on the idea that there is linguistic variation at the group level and hence there may also be linguistic variation at the individual level. William Labov has stated that nobody has found a "homogenous data" in idiolects, and there are many reasons why it is difficult to provide such evidence.

Firstly, language is not an inherited property, but one which is socially acquired. Because acquisition is continuous and life-long, an individual's use of language is always susceptible to variation from a variety of sources, including other speakers, the media and macro-social changes. Education can have a profoundly homogenizing effect on language use. Research into authorship identification is ongoing. The term authorship attribution is now felt to be too deterministic.

The small size of the documents (ransom notes, threatening letters, etc) in most criminal cases in a forensic setting is usually much too short to make a reliable identification. However, the information provided may be adequate to eliminate a suspect as an author or narrow down an author from a small group of suspects.

Authorship measures that analysts use include word length average, average number of syllabus per word, article frequency, type-token ratio, punctuation (both in terms of overall density and syntactic boundaries) and the measurements of hapax legomena (unique words in a text). Statistical approaches include factor analysis, Bayesian statistics, Poisson distribution, multivariate analysis, and discriminant function analysis of function words.

The Cusum (Cumulative Sum) method for text analysis has also been developed.. Cusum analysis works even on short texts and relies on the assumption that each speaker has a unique set of habits, thus rendering no significant difference between their speech and writing. Speakers tend to utilize two to three letter words in a sentence and their utterances tend to include vowel-initial words.

In order to carry out the Cusum test on habits of utilizing two to three letter words and vowel-initial words in a sentential clause, the occurrences of each type of word in the text must be identified and the distribution plotted in each sentence. The Cusum distribution for these two habits will be compared with the average sentence length of the text. The two sets of values should track each other. Any altered section of the text would show a distinct discrepancy between the values of the two reference points. The tampered section will exhibit a different pattern from the rest of the text.

Forensic stylistics

This discipline subjects written or spoken materials (or both), to scientific analysis for determination and measurement of content, meaning, speaker identification, or determination of authorship, in identifying plagiarism.

One of the earliest cases where forensic stylistics was used to detect plagiarism was the case of Helen Keller's short story. The blind American author was accused of plagiarism in 1892 with regard to her published short story, The Frost King. Upon investigation, The Frost King was found to have been plagarised from Margaret Canby's book Frost Fairies which had been read to her some time ago. Keller was discovered to have made only minute changes to common words and phrases and used less common words to put the same point across, suggesting mere alterations to original ideas.

Keller used 'vast wealth' instead of 'treasure' (approximately 230 times less common in the language) 'bethought' instead of 'concluded' (approximately 450 times less common), 'blade them' instead of 'told them' (approximately 30 times less common). Keller used the phrase 'ever since that time' whilst Canby chose 'from that time' (the latter 50 times more common than the former). Keller also used ' I cannot imagine' whereas Canby used ' I do not know'. 'Know' is approximately ten times more common than 'imagine'.

Keller relied on a lexis that is less common when compared to Canby's. The Flesch and Flesch-Kincaid readability test showed that Canby's text showing more originality compared to Keller's. Canby's text obtained a higher grade on the reading ease scale compared to Kelelr's. The distinctions between Keller and Canby's text are at the lexical and phrasal level.

Other examples of plagiarism include the cases between Richard Condon, author of The Manchurian Candidate and English novelist Robert Graves; and between Martin Luther King Jr and Archibald Carey.

Judging by the text in The Manchurian Candidate, Condon's work is seen to be rich in cliches such as "in his superstitious heart of hearts." While Helen Keller took pride in using rare phrases and avoids common source words, Condon was fond of expanding existing words into phrases and existing phrases into more extensive ones. Condon was found to have borrowed from a wide range of Graves' work.

In the plagiarism case of Martin Luther King Jr, almost half of his doctoral dissertation was discovered to have been copied from another theology student. King simply changed the actual names of the mountains and used much more alliteration and assonance.

Carey's and Graves' texts (source texts) were noticeably shorter, pithier and relatively simpler in structure while Condon's and King's texts relied on 'purple' devices, extending the existing text and flourish their language significantly.

Discourse analysis

Discourse analysis deals with analyzing written, spoken, signed language use or any significant semiotic event. Close analysis of a covert recording can produce useful deductions. The use of 'I' instead of 'We' in a recording highlights non-complicity in a conspiracy. The utterance 'yeah' and 'uh-huh' as responses indicate that the suspect understands the suggestion, while feedback markers such as 'yeah' and 'uh-huh' do not denote the suspect's agreement to the suggestion. Discourse analysts are not always allowed to testify but during preparation for a case they are often useful to lawyers.

Linguistic dialectology

This refers to the study of dialects in a methodological manner based on anthropological information. A systematic study of dialects is crucial as dialects are no longer as distinct as they once were due to the onslaught of mass media and population mobility. Political and social issues have also caused languages to straddle across geographical borders resulting in certain language varieties spoken in multiple countries, leading to complications when determining an individual's origin by means of his/her language or dialect.

Dialectology was used during the investigations into the Yorkshire Ripper tape hoax.

Forensic phonetics

The forensic phonetician is concerned with the production of accurate transcriptions of what was being said. Transcriptions can reveal information about a speaker's social and regional background. Forensic phonetics deals with whether the speaker in two or more separate tape recordings is the same. Transcriptions are useful as it allows victims and witnesses to indicate whether the voice of a suspect is that of the criminal.

A man accused of manufacturing the drug Ecstasy was mis-heard by the police transcriber as 'hallucinogenic' The police transcriber heard "but if it's as you say it's hallucinogenic, it's in the Sigma catalogue." However, the actual utterance was "but if it's as you say it's German, it's in the Sigma catalogue."

Another disputed utterance was between a police officer and a suspect. One of the topics of conversation was a third man known as 'Ernie'. The poor signal of the recording made 'Ernie' sound like 'Ronnie'. The surveillance tape presented acoustic problems- an intrusive electronic-sounding cackle, the sound of the car engine, the playing of the car radio, the movement of the target vehicle. The intrusive noise coincided with the first syllable of the disputed name.

Examples

Evidence from forensic linguistics has more power to eliminate someone as a suspect than to prove him or her guilty. Linguistic expertise has been employed in criminal cases to defend an individual suspected of a crime. Linguistic expertise has also been rendered useful even at the investigatory stage of government bodies. Forensic linguists have given expert evidence in a wide variety of cases, including abuse of process, where police statements were found to be too similar to have been independently produced by police officers; the authorship of hate mail; the authorship of letters to an Internet child pornography service; the contemporaneity of an arsonist's diary; the comparison between a set of mobile phone texts and a suspect's police interview, and the reconstruction of a mobile phone text conversation. Some well-known examples include an appeal against the conviction of Derek Bentley and the identification of Theodore Kaczynski as the so-called "Unabomber".

The criminal laboratories Bundeskriminalamt (in Germany) and the Nederlands Forensisch Instituut (in the Netherlands) both have forensic linguists working for them.

Derek Bentley

Forensic linguistics contributed to the overturning of Derek Bentley's conviction for murder in 1998 although there were other non-linguistic issues.

Nineteen-year-old Bentley was hanged in 1953 for his part in the murder of PC Sidney Miles. The fatal shot had been fired by Bentley's sixteen-year-old friend, Christopher Craig, when Bentley was already in police custody. Bentley, who had a mental age of eleven and was functionally illiterate, was convicted partly on the basis of his statement to police, allegedly transcribed verbatim from a spoken monologue.

Linguist Malcolm Coulthard examined the text when the case was reopened, and found a number of features which indicated police co-authorship, and which suggested that at least part of the statement resulted from questions and answers, as Bentley claimed, and was not, as police claimed, a "verbatim record of dictated monologue". One such feature was the use of the word "then", which Coulthard and his colleague David Woolls found to be the eighth most frequently-occurring word in Bentley's text, as compared with the 58th most frequent word in spoken English, and the 83rd most frequent word in English in general (according to the 1.5-million-word Bank of English corpus they were using).

Feeling that the use of that word could be expected to be higher than average in witness statements (which generally report a sequence of events and show concern for accuracy about time), two corpora were compiled, one of witness statements and one of police statements. The word "then" occurred once every 930 words in the former but once every 78 words in the latter. Compared with the Bank of English corpus where it occurred once every 500 words, it occurred once every 53 words in Bentley's text. Coulthard's analysis of the statement was a major factor in Bentley's posthumous pardon. Coulthard's approach is not only a forensic discourse analysis but a combination of insights from different linguistic fields including speech act theory, corpus linguistics, register and even psycholinguistics.

The focus then turned to the use of the word "then". The frequent post-positioning of temporal (time-related) "then" after the grammatical subject ("I then" rather than "then I"), which occurred 7 times in the 582-word text, was also noted. The Bank of English spoken corpus showed "then I" to occur ten times more frequently than "I then", the latter occurring only once every 165,000 words. That structure did not occur at all in the corpus of witness statements but occurred once every 119 words in the corpus of police statements.

These features, combined with many others, contributed to a successful argument that the Bentley "confession" was, in part, the written work of police officers, and not simply a word-for-word transcript of Bentley's spoken statement as the police alleged.

The "Unabomber"

In the case of Theodore Kaczynski, who was eventually convicted of being the "Unabomber", family members recognized his writing style from the published 35,000-word Industrial Society and Its Future (commonly called the "Unabomber Manifesto"), and notified the authorities. FBI agents searching Kaczynski's hut found hundreds of documents written by Kaczynski, but not published anywhere. An analysis produced by FBI Supervisory Special Agent James R. Fitzgerald identified numerous lexical items and phrases common to the two documents. Some were more distinctive than others, but the prosecution (assisted by Vassar Professor of English Donald Foster) successfully argued that even the more common words and phrases being used by Kaczynski became distinctive when used in combination with each other.

During the appeal against the conviction of the Bridgewater Four, the forensic linguist examined the written confession of Patrick Molloy, one of the defendants — a confession which he had retracted immediately — and a written record of an interview which the police claimed took place immediately before the confession was dictated. Molloy denied that the interview had ever taken place, and the analysis indicated that the answers in the interview were not consistent with the questions being asked. The linguist came to the conclusion that the interview had been fabricated by police. Later the conviction against the Bridgewater Four was quashed before the linguist in the case Malcolm Coulthard could produce his evidence.

In an Australian case reported by Eagleson, a "farewell letter" had apparently been written by a woman prior to her disappearance. The letter was compared with a sample of her previous writing and that of her husband. Eagleson came to the conclusion that the letter had been written by the husband of the missing woman, who subsequently confessed to having written it and to having killed his wife. The features analysed included sentence breaks, marked themes, and deletion of prepositions.

Additional forensic linguistics concepts

Linguistic fingerprinting

A linguistic fingerprint is a collection of markers which stamps a speaker/writer as unique. Every individual utilizes languages differently and this difference can be observed as a fingerprint. It is formed as a result of merged language style. A person's linguistic fingerprint can be reconstructed from the individual's daily interactions and relate it to a variety of self-reported personality characteristics, situational variables and physiological markers (eg. blood pressure, cortisol, testosterone).. In the process of an investigation, the emphasis should be on the relative rather than absolute difference between the authors and how we can classify their texts.

Variation

Intra-author variation are the ways in which one author's texts differ from each other. Inter-author variation are the ways in which different authors vary from each other. Two texts by one author does not necessarily translate to less variation than two texts by different authors.

• Genre: When texts are being measured in different genres, considerable variation is observed even though they are by the same author.
• Text Type: Personal letters contain more inter-relationship bonding strategies than an academic article or a term paper.
• Fiction vs. Non-Fiction: Some fiction writers are journalists.Due to the different demands of each medium , the media can be completely different from one another and this results in intra-author variation.
• Private vs Public: A politician writing a political speech is a public text which will differ greatly from a private text to a friend or family member.
• Time lapse as a cause of variation: The greater the time lapse between two works, the greater the likely variation. Language changes more than we realize in a relatively short space of time influencing our susceptibility to language changes around us.
• Disguise as a sort of variation: A writer can publish anonymously hence disguising output to prevent recognition.

Evidence in court

In the United States, the Frye test was used in evidence of a scientific or technical nature. This standard was proven to be flawed in recent years. It is now common for all federal and some state courts to demand evidence which pass the 'Daubert test'. Not all forensic fields can pass the Daubert test. The forensic disciplines include ballistics, forensic phonetics, forensic anthropology etc. The polygraph (lie-detector) analysis is unadmissible under the Fyre test but is acceptable under Daubert.

Forensic Transcription

The two main types of transcriptions are written documents and video and audio records. Accurate ,reliable text transcription is important because the text is the data, it is the available evidence.If a transcription is wrong, the evidence is altered.If there is failure to transcribe the full text, evidence is once again altered unwittingly. There must be emphasis on the text being the evidence. A transcription of an audio file should never be assumed to be completely accurate. Each type of transcription contains its own problems. A handwritten document might contain unusual spellings which result in ambiguous meanings,illegible handwriting and illustrations that are difficult to comprehend etc. A scanned document is tricky too, it must not alter anything from the original. Audio and video documents can include repetitions, hesitation, nonsensical talk, jargons which are hand to understand, speakers mumbling incoherently and inaudibly. Non-linguistic sounds such as crying, laughing may also be included in the audio and video text which cannot be transcribed easily.

See also

• Forensic science
• The International Association of Forensic Linguists (IAFL)
• Stylometry (authorship analysis)
• Writeprint

Further reading

• Baldwin, J. R. and P. French (1990). Forensic phonetics. London: Pinter Publishers.
• Eagleson, R. (1994). 'Forensic analysis of personal written texts: a case study', J.Gibbons (ed.), Language and the Law, London: Longman, 362-73
• Ellis,S. (1994). 'Case report: The Yorkshire Ripper enquiry, Part 1', Forensic Linguistics 1, ii, 197-206
• Fairclough, N. (1989) Language and Power, London: Longman.
• Gibbons, J. (2003). Forensic Linguistics: an introduction to language in the Justice System. Blackwell.
• Gibbons, J., V Prakasam, K V Tirumalesh, and H Nagarajan (Eds) (2004). Language in the Law. New Delhi: Orient Longman.
• Gibbons, J. and M. Teresa Turell (eds) (2008). Dimensions of Forensic Linguistics. Amsterdam: John Benjamins.
• Grant, T. (2008). "Quantifying evidence in forensic authorship analysis", Journal of Speech, Language and the Law 14(1).
• Grant, T. and Baker, K. (2001). 'Reliable, valid markers of authorship', Forensic Linguistics VIII(1): 66-79.
• Hollien, H. (2002). "Forensic Voice Identification". New York: Harcourt.
• Hoover, D. L. (2001). "Statistical stylistics and authorship attribution: an empirical investigation", Literary and Linguistic Comuputing, XIV (4), 421-44
• Koenig, B.J. (1986) 'Spectrographic voice identification: a forensic survey', letter to the editor of J. Acoustic Soc, Am., 79, 6, 2088-90.
• Maley, Y. (1994). 'The language of the law', in J. Gibbons (ed.), Language and the Law, London:Longman,246-69
• McGehee, F. (1937).' The reliability of the identification of the human voice', Journal of General Psychology, 17, 249-71
• McMenamin, G. (1993). Forensic Stylistics. Amsterdam: Elsevier.
• Miller, C. (1984). "Genre as social action." Quarterly Journal of Speech.
• Nolan, F. and Grabe, E. (1996) 'Preparing a voice lineup', Forensic Linguistics, 3 i, 74-94
• Pennycook, A. (1996) 'Borrowing others words: text, ownership, memory and plagiarism', TESOL Quarterly, 30, 201-30.
• Shuy, Roger W (2001). 'Discourse Analysis in the Legal Context.' In The Handbook of Discourse Analysis. Eds. Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton. Oxford: Blackwell Publishing. pp. 437–452.

External links

• ALIAS Technology LLC
• Forensic Linguistics at All About Forensic Science.com
• Cardiff University's MA (or Postgraduate Diploma/Certificate) in Forensic Linguistics
• Unabomber Case and Trial
• International Journal of Speech Language and the Law
• Forensic Linguistics Casebook - an independent wiki collecting cases where forensic linguistics has been used.
• Aston University's Centre for Forensic Linguistics
• International Summer School in Forensic Linguistic Analysis
• Hofstra University's Forensic Linguistics Project
• International Association of Forensic Linguists
• International Association for Forensic Phonetics and Acoustics
• Tiersma, Peter,: Homepage of eminent forensic linguist from Loyola Law School
• Institute for Linguistic Evidence, a nonprofit research organization.
• Language in the Judicial Process, edited by Bethany K. Dumas.
• Reading between the lines - BBC News article on the Centre for Forensic Linguistics
• Foulkes, Dr Paul, a forensic linguist at the University of York- Radio interview.
• Shuy, Roger W.: Homepage of eminent scholar in the field of language and law
• Forensic Linguistics Institute
• Butters, Ronald: Homepage of Professor, scholar, & journal editor in forensic linguistics
• Webcast of a lecture on forensic linguistics given by Roger W. Shuy at the University of Montana School of Law
• Webster’s forensic linguistic home page - links to a wide variety of forensic linguistic [resources
• Forensic Linguistics: Linguist as detective & expert witness - Public lecture by Prof. Malcolm Coulthard