In the media, the phrase “forensic linguistics” is usually narrowly applied to document interpretation and author identification. This is probably because of the high-profile criminal cases that have involved writing, including Son of Sam, Zodiac killer, JonBenet Ramsey, and Unabomber. Occasionally, someone on TV is testifying about the dialect or pronunciation of another individual in the case. More often, it’s an investigator at a whiteboard comparing various writings. But TV linguists often overstate the importance of similarities and never mention usage patterns, which are critical to author identification.
Experts in forensic linguistics can be a great asset in the kinds of authorship cases you don’t see on TV—for example, in cases involving online harassment, anonymous or libelous complaints, online police sting operations, custody battles, conflicting police reports or witness statements, as well as forged emails and other documents. These kinds of cases hinge on successful identification of authorship for the disputed or anonymous documents in question.
But the scope of a forensic linguist’s expertise extends far beyond authorship of documents. The voluminous (669 pp.) Routledge Handbook of Forensic Linguistics[1] covers dozens of areas in which legal proceedings and litigation require technical language analysis. The papers are concerned with legal language, the linguist’s role in judicial proceedings, police interviews in the judicial process, forensic stylistics, trademark linguistics, and text messaging forensics, to name just a few.
In his classic text, “Forensic Linguistics”, John Olsson notes that “A forensic linguist is sometimes a general practitioner and sometimes a specialist in any of a number of sub-areas…[2] My background is in general and historical English linguistics—including phonetics, phonology, dialectology, lexicography, syntax, semantics, text analysis, and audio transcription. I also have experience in examination of transcripts of police sting operations and content analysis of different versions of the same event.
Every forensic linguist applies a highly-refined sensitivity to the components, construction, and meaning of a text. These skills, coupled with a background in linguistic evolution, make linguistics experts witness a powerful legal asset in five distinct case areas.
1) Determining Authorship of Anonymous, Disputed, or Questioned Documents
Forensic linguists analyze writing samples and compare features, noting their presence or absence. Through this comparison of known and unknown writings, they can offer an expert opinion on the likelihood of particular suspected writers, single-author/forged texts, and other legal/linguistic hypotheses. The method for accomplishing this has been verified in thousands of academic studies and real-world situations and used in legal cases since the 18th century. The analyst examines texts in fine detail, identifying and evaluating similarities and differences in orthography, formatting, lexicon, syntax, idiom, and other aspects of language, as well as the patterns in which they occur. The method is probabilistic: the more similarities, the more likely it is that the documents in question are the product of the same author.
Everybody has a writing style. Some writing styles are more distinct than others. Linguists are trained to make an authorship call based on feature occurrence and pattern similarities, especially in relatively spontaneously-generated texts, such as emails. If the writer is a foreign language speaker, their style will be even more identifiable.
Case Example
In one case I reviewed, an ex-husband’s new wife was writing emails over his signature. I was asked to identify the elements of her style and help resolve the disputed authorship.
2) Interpretation of Contracts, Wills, Laws, and Binding Documents
What was the intention of a text? What, if anything, does the text actually say? Is it ambiguous? If so, what are the possible interpretations of the text, given the semantics, syntax, lexicon, and intent?
Forensic linguists can analyze specific words, phrases, clauses, sentences, and other units to offer expert judgments on a text’s clarity, meaning, comprehensibility, and ambiguity. These judgments can be particularly useful in legal disputes involving contracts, wills, laws, and binding documents.
Prenuptial agreements, for example, and other contracts designed to be air-tight, are exactly where interpretation problems are likely to occur, as people try to execute complex ideas and cover all contingencies.
Case Example
In one case, I identified the obstacles that a second-language speaker would encounter in trying to understand the Miranda warning.
3) Assessing the Likelihood of Plagiarism
In the Internet age, plagiarism—defined as deliberate, dishonest appropriation of another’s words and/or ideas—has been redefined to include any similarity of text, regardless of its (un)originality. Most academics, it seems, have implicitly adopted the new meaning.
In my experience, accusations arise when the traditional definitions are applied. Similarity of text, which is what plagiarism tools find, is not a basis for a charge of plagiarism. It is not unique but prosaic information that is most likely copied, thus undermining the traditional charge of plagiarism. Much more is in the public domain than ever before. Outright theft of originality is rare.
I’ve defended many academic writers, from college students to law professors, against groundless charges of plagiarism. These charges are often based on the similarity of items in the public domain, as identified by Turnitin—an internet-based plagiarism detection service. Over-reliance on machine data has damaged many academic reputations. In addition, many charges of plagiarism can be attributed to misunderstandings of the institution’s quotation or paraphrase rules.
In literary plagiarism allegations, one writer accuses another of stealing ideas. But these similarities usually turn out to be widespread, occurring in typical books of this genre. Courts have ruled that elements of theme and setting are not protectable.
Case Example
In numerous cases in which I have been involved, students are accused of dishonesty because their rendering of basic, mundane factual information, which can be accurately expressed in only a few ways, is too similar to another source.
4) Copyright and Trademark Infringement
Expert forensic linguists can offer informed judgment on the genericity, specificity, and protectability of individual words, phrases, and brand names. I assess linguistic similarities to evaluate infringement claims. The question here is: to what extent is the item already part of the language?
If it is already a lexical unit, it can still be trademarked if it’s “suggestive”, as with brands like “Dawn” and “Joy”. A lexical unit can also be trademarked if it is descriptive. Such is the case with brands like “SimpleGreen” or “EverReady”. But you cannot market a name that already stands for a whole category of items—for example, “Pants”.
A forensic linguistics expert can also prove helpful for cases in which one company’s mark may be too similar to another’s, causing confusion to consumers or clients.
Case Example
In one case, I showed, through dictionary research, that the term “junk(yard) art” was firmly established in the language and thus not protectable.
5) Transcription Evaluation
A linguist’s trained ear can hear a lot in an audio recording—what was said, how it was said, what it says about the speaker, and how it seems to have been perceived. I studied phonetics and later transcribed English and, for my doctoral thesis, Hawaiian dialect tapes and, still later, interviews with executives. I have applied this experience to texts of various lengths. Sometimes the meaning of an entire statement can hinge on the interpretation of a single word or phrase.
Case Example
I reviewed a video interview of a young girl in connection with molestation charges. Although the interviewer was blatantly trying to elicit incriminating replies, I determined that the transcription was in many places inaccurate, and most of the child’s responses showed that she did not understand the question.
[1] London: Routledge, 2010. [2] Forensic Linguistics, London: Contnuum, 1994.
