Judicial Systems in Contact: Access to Justice and the Right to Interpreting/Translating Services among the Kichwa of Ecuador

Session 2B, 7/13, 9-10:15, Mary Gates Hall 231

The right to a court-appointed interpreter has been established in a number of countries (e.g., Australia, Canada, Guatemala, South Africa, U.K., U.S.A.). Yet in many nations, such rights are far less clearcut. In Latin America, countries having substantial indigenous populations that vary in their command of the official language(s) of the State often display a divergence between official policy and de facto policy. Ecuador is a case in point.

Interviews with 93 Ecuadorians (judges, public defenders, private attorneys, prosecutors, interpreters/translators, justices of the peace, police officers, commisioners specializing in domestic violence cases, elected officials, officials working in NGOs and human rights agencies) have revealed that the political leadership of the Kichwa--the largest indigenous group in Ecuador--on the whole prefers not to use the State justice system. However, when the Kichwa do have recourse to it, they send bilingual community leaders to the State courts to act as interpreters. This includes restraining order hearings. When judicial proceedings are carried out within the borders of the Kichwa territories, proceedings are conducted bilingually, that is, defendants and witnesses speak in either Spanish or Kichwa, and the leaders who try these cases interact with them in both languages. This sociolinguistic configuration has important repercussions for cases in which jurisdictional issues involving people of different ethnic background remain unresolved.

The Ecuadorian case is particularly interesting from a forensic linguistic standpoint, in the light of recent (1998) reforms to Ecuador’s constitution giving indigenous and Afro-Ecuadorian groups a large measure of judicial autonomy with respect to administration of justice.

Susan Berk-Seligson (and Bethany Dumas), Shuy Panel Participant Center for Latin American and Iberian Studies Vanderbilt University United States Special Session, 7/13, 1-3:45, Johnson 075

Carl S. Bjerre University of Oregon, School of Law United States

The Force Dynamics of Contract Law

Session 4B, 7/13, 4:15-5:30, Mary Gates Hall 231

This paper uses the linguist Leonard Talmy’s idea of “force dynamics” to show an underlying unity between otherwise disparate aspects of contract law.

For Talmy, physical notions of the exertion of force also apply in non-physical realms, including internal psychological states and interpersonal interactions. E.g., in “I refrained from responding,” one part of the self is conceptualized as exerting force against another, in what Talmy calls “onset causation of rest.” Similarly, in “she indulged his verbosity,” one person is conceptualized as yielding to force by another, in what Talmy calls “extended letting of motion.” Taken together (with permutations between onset and extended forces, causation and letting, and motion and rest), many abstract concepts are unified within this fundamental semantic category of force dynamics.

I show that much of the discourse of contract law reduces to similar exertions of force. The nature of offers and acceptances and rejections, bargaining power, choice and duress, breach, and mitigation of damages, to take just a few examples, are all shown to be explainable in terms of Talmy’s permutations. In fact, the standard rhetoric of freedom of contract is all about onset causation of motion (“your signature obligates you”), extended letting of rest (“the law leaves you free to choose deals”), and the like. It is well established that legal concepts are constructed, but this linguistic framework suggests that many of those constructions are subconscious, fairly primitive, and uniform.

Susan Blackwell University of Birmingham United Kingdom

Defining Denial: The Language of Legislation Concerning the Holocaust

Session 8A, 7/14, 2:30-3:45, Communication 226

Holocaust denial is currently illegal in 11 countries: Austria, Belgium, the Czech Republic, France, Germany, Israel, Lithuania, Poland, Romania, Slovakia and Switzerland. A British citizen, David Irving, recently served a prison sentence in Austria for speeches made there, and some politicians and organisations are now campaigning for such a law to be introduced in England and Wales. An obvious question which presents itself to linguists is how such a law might be worded in order to be enforceable.

This paper reviews and compares the wording of the relevant legislation in the countries concerned and identifies a number of problems. Several of the existing laws refer specifically to the crimes of German National Socialism, while those of Poland group together the crimes of nazism and communism, and the Swiss law refers inclusively to "genocide or other crimes against humanity". The degree to which the victims are defined is equally variable. Some of the legislation criminalizes "minimalising" or "trivialising" the acts concerned as well as merely "denying" them, and in some cases "seeking to justfiy" or "expressing sympathy with" the perpetrators is also punishable.

It emerges that this particular "language crime" is extremely slippery to define. Some examples will be given of how the wording of the legislation is interpreted in practice when cases come to court. The paper concludes that there are linguistic grounds, as well as ethical and political ones, for opposing the introduction of new laws of this kind.

Ronald R. Butters (co-authors Tyler Kendall and Phillip Carter)/Shuy Panel Participant Duke University United States

Discourse Analysis of Instant Messages Used as Incriminating Evidence in “Sexual Predator” Prosecutions

Special Session, 7/13, 1-3:45, Johnson 075

One of Roger Shuy’s greatest contributions to forensic linguistics is his extensive and detailed analysis of surreptitiously recorded evidence used in “language crimes” prosecutions. Our paper builds upon Shuy’s methodology to analyze prosecutorial allegations about Instant Message exchanges between an unsuspecting middle-aged man and an adult who was pretending to be a 13-year-old boy. The case, which led to sexual enticement and solicitation convictions, is one of hundreds of “mark-and-decoy” IM conversations instigated by agents of Perverted Justice, an organization dedicated to the exposure, arrest, and conviction of “sexual predators.” “Dateline,” the NBC news magazine, has paid Perverted Justice hundreds of thousands of dollars to employ online decoys; “Dateline” reporters then conducted surprise “interviews” upon the marks as they arrived at what they were led to believe were the homes of the “youths.” The web site Perverted-Justice.com publishes transcripts of over 150 such IM exchanges.

Anyone who has seen the “Dateline” broadcasts may wonder what sort of defense could be possible in the face of such overwhelming-seeming evidence of sexual crimes against underage youths. It is not our purpose to dispute the decision of the trial judge in the case under analysis here, who found the mark guilty of two felony charges and sentenced him to six years in federal prison. We do seek to demonstrate, however, that the interpretation of the evidence could have been significantly altered by the use of discourse analysis of the IMs and the “Dateline” broadcast—none of which was presented to the court.

Elisabeth Kate Carter University of Essex United Kingdom

Laughing Matters: Investigating Officer and Suspect Laughter in the Police Interview on a Conversation Analytic Level

Session 12B, 7/15, 1-2:15, Mary Gates Hall 251

This conversation analytic research uses police interviews as data in investigating the role of laughter by police officers and suspects in managing aspects of police interviews. Laughter in police interviews has not been the subject of extensive research in this field, unlike that of ordinary conversation. Amongst other uses, laughter in ordinary conversation can be used as a tool to enable talk about troubles and create intimacy. Analysis of laughter used in this institutional setting identifies similarities and differences in its use from ordinary conversation. This research finds that the uses typical of ordinary conversation creep into this institutional talk, however the constraints the parties are operating under mean they get sanctioned for doing precisely that, therefore highlighting what these constraints are.

The two-tiered approach to the analysis finds the basic actions performed by the laughter (such as in response to a ridiculous comment in the prior turn) are uniform across participants. Attention is then given to the way it is used by the suspect to support his/her position, for instance that of innocence, or to deflect contradiction of the officer. On the part of the officer, laughter is used to frame a 'time out' from the constraints he/she is operating under. This suggests that laughter may be used for similar interactional purposes by both the suspect and the officer, but that they can be directed toward differential second order objectives due to the divergent positions, rights and obligations of the two parties.

Luciana Carvalho** Universidade de São Paulo São Paulo, Brazil

Copyright Issues in Bilingual Legal Dictionaries: A Case Study

Session 2C, 7/13, 9-10:30, Mary Gates Hall 251

This paper aims at describing copyright issues raised in a legal suit involving copyright infringement by the authors and publishers of a bilingual legal dictionary in Brazil. This suit was filed by the publisher of another bilingual dictionary in 2004 and, therefore, was still in course at the time of writing. We have been closely following the case for research purposes seeing that this suit addresses a number of relevant legal [and] language issues such as plagiarism, expert opinions, translation and terminology.

We shall begin by elucidating how the Brazilian civil law system operates in this regard focusing on the sources of expert opinion admissible in and/or required by the court. We shall then proceed to examine the questions the plaintiffs and the defendants posed to the court-appointed expert and the opinion issued by the latter. Lastly we shall discuss how the experts appointed by each one of the parties responded to the opinion written by the court-designated expert. And, to conclude, we shall seek to establish points of convergence and divergence involving the three expert opinions at issue.

Luciana Carvalho Universidade de São Paulo São Paulo, Brazil

Translating Binomial Expressions in Common Law Agreements: A Corpus-Based Study

Session 11A, 7/15, 10:30-11:45, Mary Gates Hall 231

This paper aims at presenting the final results of last year’s presentation at the IAFL Barcelona conference.

We have spent the last couple of years studying binomial expressions in common law agreements in light of Corpus Linguistics in an attempt to provide translators with the necessary linguistic elements that will enable them to render a natural translation.

Binomial expressions are formed by two words belonging to the same grammatical category and joined by and or or. Some examples are: terms and conditions, any and all, due and payable, action or proceeding, agreement or obligation.

To study binomials in English legal language and provide elements for translators to arrive at their own translation into Brazilian Portuguese, we compiled and explored a bilingual comparable corpus consisting of authentic agreements and contratos, totaling, approximately, 1 million words—705,744 in English and 289,984 in Brazilian Portuguese—made up of 5 samples of 28 different kinds of contratos and agreements, a total of 140 documents in each language.

Exploring such a corpus depended on principles and tools of Corpus Linguistics. To tag the corpus we used UCREL’s CLAWS 7 (the Constituent Likelihood Automatic Word-tagging System) and to explore the corpus we used Mike Scott’s WordSmith Tools.

We were, therefore, able to retrieve 1,065 binomial candidates of which 819 were confirmed to be binomial or multinomial expressions formed by various grammatical categories. We found 7 binomials formed by determiners, 54 formed by prepositions, 102 formed by adjectives, 490 formed by nouns, 11 formed by pronouns, 15 formed by adverbs, and 140 formed by verbs.

Krisda Chaemsaithong University of Houston, Downtown United States

Self-Face in Early Modern Witchcraft Trials and Contemporary Trials

Session 2A, 7/13, 9-10:15, Communication 326

Drawing data from Early Modern English witchcraft trials and the Salem witchcraft trials in the sixteenth and seventeenth centuries, this paper aims to show the strategies used by examinees in the trials from the perspective of linguistic politeness.

However, politeness as used in this paper, refers to self-politeness—politeness done to save one’s own face, rather than others’ face, I claim that the examinees’ responses were produced primarily for them to save their own face, though in some cases to save the magistrates’ face as well. Such politeness strategies for self can be explained and attributed to the differences in legal paradigms and procedures in which guilt was presupposed during such a period and in which capital punishment of being hanged was reserved for those found to be witches (whereas contemporary Anglo American trials adopt the innocent-until-proven-guilty paradigm). Self-politeness is important in witch trials because it helps the accused to achieve two tasks at the same time: first, they could respond to the arbitrates’ questions as well as possible in ways that are pragmatically co-operative, and, second, functioning as a defense attorney for themselves, they had to produce optimal responses that might lead them to acquittal or partial and more lenient punishment (although it could not be guaranteed whether those optimal responses would work to satisfactory ends). Finally, I compare the examinees’ self-politeness strategies in witch trials with some data from contemporary trials to show that the strategies for self politeness in the Early Modern period were much more diverse due to socio-legal differences.

Carole E. Chaski, Panel Chair/ Panel on the Linguistics of Deception ALIAS Technology LLC United States

An Overview of Research in the Linguistics of Deception Detection

Session 5A, 7/14, 9-11, Communication 120

The interrogation is essential to both civil, criminal and security investigations. Detecting deception within the interrogation and throughout the investigative process enables investigators to allocate resources efficiently, thereby making it a highly valued skill. As an introduction to the following talks, I focus on previous research which has attempted to define, operationalize or in any way measure the skill of detecting deception through linguistic cues, including the work of Sapir, DePaulo, Pennebaker, Hartwig and others.

Carole E. Chaski, Panel Participant/ Panel on Authorship Attribution ALIAS Technology LLC United States

Empirically Testing the Uniqueness of Aggregated Stylemarkers

Session 7A, 7/14, 1-5:15, Communication 120

One of the central claims of forensic stylistics is that each person has a unique set of stylemarkers (McMenamin 2002). McMenamin (2002) provides no empirical evidence for this claim, but he does list the kinds of stylemarkers which have been used within forensic stylistics. Following Foster (2000), Koppel and Schler (2003) tested 99 stylemarkers on large writing samples of 11 authors, obtaining only 67% classification accuracy when they used these stylemarkers with two very powerful classification algorithms. Schler, Koppel, Argamon and Pennebaker (2006) have recently made available a corpus of almost 20,000 bloggers. Some of the bloggers who have produced approximately 150,000 words. From this corpus, I have extracted the writing samples of bloggers who have produced less than 3,000 words, so that the amount of data is forensically feasible. This study reports the frequencies of aggregated punctuation stylemarkers. Even though forensic stylistics includes other features besides punctuation, in every forensic stylistic analysis I have ever read, punctuation plays a key role. This study provides empirical results as well as a discussion of presence versus rate in stylemarkers aggregation.

Carole E. Chaski ALIAS Technology LLC United States

Multilingual Forensic Author Identification through N-Gram Analysis

Session 12B, 7/15, 1-2:15, Mary Gates Hall 251

In many criminal and civil disputes, the issue of authorship identification is relevant. Current approaches can be forensically difficult because the quality of the required text is too large or the approach requires complicated and time-consuming natural language processing or the approach is domain-dependent. N-gram analysis, in which sequences of various n lengths, has proven its utility in part-of-speech tagging, parsing, information retrieval and authorship identification (Keselj et al. 2000, Peng et al. 2001, Koppel et al. 2006). Whether this technique can also be used in forensic author identification is explored in this paper. Further, we discuss n-gram analysis in terms of doman-and language-dependence. We hypothesize that different language groups requires different n-gram lengths for the best accuracy. We compare the results of using Keselj et al.’s (2000) Relative Distance Metric versus Frantzeskou et al.’s (2007) Overlap Metric. Finally, we report a series of experiments using the two metrics with different n-gram lengths under varying conditions of textual quality for English, Russian and Arabic.

Sgt. Wes Clark / Panel on the Linguistics of Deception Connecticut State Police United States

The Practical Application of Investigative Statement Analysis

Session 5A, 7/14, 9-11, Communication 120

This talk focuses on the practical application of investigative statement analysis within the context of criminal investigations. I will show how various principles and practices of statement analysis techniques have proven to be a valuable tool for law enforcement in specific cases by helping investigators identify missing information and assist with distinguishing between truthful and deceptive communication.

Effie Papazikou Cochran (and Alice H. Deakins) City University of New York United States

What Legal Writers Should Know: A Syntactic Analysis of a Legal Brief

Session 6B, 7/14, 10:30-11:45, Mary Gates Hall 231

Among the many types of written expression, the legal register is fraught with denseness and dryness, which posing a challenge to writers of legal texts. How much of this is syntactic? And how does syntax interact with content?

In this paper, the editing process on the first paragraph of three drafts of a U.S. legal brief will be explored and then connected to the content of the paragraph. After an introduction of the analytic system, the syntactic changes over the three drafts will be examined, revealing the superiority and accessibility of the final version. With its fewer sentences, reduced number of different subjects, and simpler predicates, the final version of the brief unpacks the density of the previous drafts by dropping or shifting information. It also uses the full spectrum of sectors in the English sentence by adding information in “extra” insert positions. In terms of content, the syntax of the final version interacts with the substance of the brief by presenting a drama which uses only three different subjects: the court, the jury, and the defendant. Each subject/agent has a role that as assessed via judgment words, making them, respectively, impartial, decisive, and weak. The plaintiff becomes a wise observer.

Through examination of the changes in syntax over the three versions of the brief, a movement from obfuscation to clarity and simplicity becomes apparent. The syntactic structures and the content of the brief work in synergy to create a simpler yet powerful and sophisticated final version.

G. Burns Cooper University of Alaska Fairbanks United State

Stability of Certain Stylistic Features Over Time

Session 3A, 7/13, 10:30-11:45, Mary Gates Hall 231

A persistent problem in authorship analysis is that authors’ styles may change over time; comparisons of documents written within the same relatively short time frame are thought to be most reliable. Complicating this problem, we do not know in any detail which features are most likely to change, how much they may change, or exactly why they change. It would be useful to have much more information about how consistent the process of change is between authors, and about which aspects change most (or least) over time. The present study contributes to building the necessary base of knowledge to address these problems. It examines a series of letters written by the same authors over the course of many years. It focuses on the parameter of semantic density, because psycholinguistic and medical evidence has suggested that it is relatively stable over time (Snowden, et al. 1996). The present study expands the data set used in Cooper 2007 (in press) to make cross-year comparisons more useful. The document type, Christmas letters sent to family and friends, ensures that topics and audiences are similar, and using actual letters as opposed to documents written expressly for research purposes eliminates the question of whether such documents are truly comparable to real-life ones. The comparability of semantic density to other relatively simple stylistic measures will also be discussed briefly.

Silvia Dahmen, Panel Participant, Authorship Attribution Cologne University Germany

Frank Albrecht, Silvia Dahmen, Hannes Kniffka

On the History and the Current State of Forensic Linguistic Expert Testimony in Authorship Analysis Cases in Germany

Session 7A, 7/14, 1-5:15, Communication 120

The three papers by Frank Albrecht, Silvia Dahmen and Hannes Kniffka address the history and development of forensic linguistic expert testimony in and for German (criminal) courts.

The overall hypothesis is that we want to give a detailed account of how linguistic or “linguistic” expert testimony was given in Germany some 40 years ago, of some of the (undisputedly) “bad” examples of forensic linguistic expert testimony that can be stated, and of the improvements, corrections, and achievements that have been made to date. In short, we want to present data-driven “good” and “bad” examples of forensic linguistic expert testimony. (Needless to say, the “good” ones are from our own, the others are from other experts’ testimony.)

It is also part of the overall hypothesis that it needs thorough elaboration and discussion of what we did wrong, what mistakes were made some 40 years ago, what mistakes we don’t make any more (and which ones we might still make today). We feel that forensic linguists act half of the time as if it were self-evident what mistakes we don’t make any more and which ones we do. We act as if we know precisely which results, which “solutions”, which methods were/are “real good” or “real bad”, or something in between. In fact, we don’t have much detailed hard empirical evidence even on “real bad” solutions, let alone just “bad” answers, arguments and examples produced by forensic linguistic experts. We three are convinced that it is useful and not at all trivial to gain a more precise picture of what has been done wrong or more or less wrong in forensic linguistic expert testimony in the past and of where we stand now. As you will see, it is not trivial, , , also when dealing with data from other than , , , your own native language, e.g. from an exotic language like German.

Silvia Dahmen Cologne University Germany

An Example from the “Dark Side” of Forensic Linguistic Expert Testimony in a German Court: A Case Study

This paper symptomatically illustrates data used as evidence and arguments by a “semi-linguist” who on his own initiative gave forensic linguistic expert testimony in the 1980s, but ever since the 1990s has been considered a charlatan by practically all forensic linguists and investigating authorities in Germany. He is still doing (and advertising) the services of a “linguistic detective” and seems to be able to live on that. The point of this paper is not to give a full-fledged description of a charlatan and his potentially dangerous work. Rather, some of the details of items and arguments that were used in a symptomatically selected expert opinion of his will be illustrated, trying to figure out why courts and some linguists in those days were swayed by his arguments. How could that happen?

Alice H. Deakins (and Effie Paptzikou Cochran) City University of New York United States

What Legal Writers Should Know: A Syntactic Analysis of a Legal Brief

Session 6B, 7/14, 10:30-11:45, Mary Gates Hall 231

Among the many types of written expression, the legal register is fraught with denseness and dryness, which posing a challenge to writers of legal texts. How much of this is syntactic? And how does syntax interact with content?

In this paper, the editing process on the first paragraph of three drafts of a U.S. legal brief will be explored and then connected to the content of the paragraph. After an introduction of the analytic system, the syntactic changes over the three drafts will be examined, revealing the superiority and accessibility of the final version. With its fewer sentences, reduced number of different subjects, and simpler predicates, the final version of the brief unpacks the density of the previous drafts by dropping or shifting information. It also uses the full spectrum of sectors in the English sentence by adding information in “extra” insert positions. In terms of content, the syntax of the final version interacts with the substance of the brief by presenting a drama which uses only three different subjects: the court, the jury, and the defendant. Each subject/agent has a role that as assessed via judgment words, making them, respectively, impartial, decisive, and weak. The plaintiff becomes a wise observer.

Through examination of the changes in syntax over the three versions of the brief, a movement from obfuscation to clarity and simplicity becomes apparent. The syntactic structures and the content of the brief work in synergy to create a simpler yet powerful and sophisticated final version.

Sandra Ferrari Disner Linguistic Consultant Los Angeles, United States

Speaker Identification and the Reassigned Spectrogram

Session 10B, 7/15, 9-10:15, Mary Gates Hall 251

A relatively new method for imaging the time-frequency content of speech, variously referred to as the Hilbert Transform, or more colloquially, the reassigned spectrogram, is applied to the problem of speaker identification. Reassigned spectrograms show the instantaneous frequencies of signal components as well as the occurrence of impulses with increased precision compared to conventional spectrograms. A small pilot study has shown that students of linguistics and engineering who received one hour of familiarization with the technique were able to match pairs of reassigned spectrograms correctly 50 out of 52 times, yielding a reliability measure that matches the most optimistic estimates of conventional spectrography. Importantly, the reassigned spectrogram requires only a brief (50 ms, or just about four pulsations of the vocal chords) sample of speech, which makes it useful in examining utterances consisting of only a few words. At the very least, reassigned spectrograms may be presented in conjunction with conventional imaging techniques in forensic cases, possibly yielding more reliable results than either technique on its own.

Bethany Dumas (and Susan Berk-Seligson), Shuy Panel Participant University of Tennessee, Knoxville United States Special Session, 7/13, 1-3:45, Johnson 075

Diana Eades University of New England Armidale, Australia

“Carrying on silly” or “antics”?: Recontextualisation and Dialectal Subtleties

Session 4C, 7/13, 4:15-5:30, Mary Gates Hall 251

An ongoing concern of forensic linguistics in Australia relates to dialectal difference between general Australian English and the varieties of Aboriginal English spoken by most Aboriginal people in their dealings with the law. Previous research and expert evidence has examined issues involved in fabricated confessions, as well as pragmatic features affecting intercultural communication in the legal process. In this paper, it is lexical se, mantics which is central to a case study on dialectal difference in the criminal justice process. The paper will examine subtle transformations which occurred when police officers summarised in general Australian English the story of a suspect speaking a closely related dialect of Aboriginal English. Seemingly unaware of key lexical semantic differences between the two dialects, the police officers produced a Fact Sheet which removed a central element of the suspect’s story. Theoretically, the paper will be situated in current research in linguistic anthropology and sociolinguistics on the recontextualisation of stories in the legal process (e.g. Trinch, Cotterill, Rock, Matoesian). While cross-examination makes much of the “ideology of inconsistency” in the way in which an individual might re-tell their story, other legal contexts seem to find the transformation of individuals’ stories unproblematic. The paper concludes by considering the possibility of training police in basic discourse analysis.

William G. Eggington Brigham Young University United States

The Linguistic Component of Hate Crimes

Session 12A, 7/15, 1-2:15, Mary Gates Hall 231

In recent times, hate crime legislation has been enacted in many Western nations. Hate crimes are defined as “any of various crimes … when motivated by hostility to the victim as a member of a group (as one based on color, creed, gender, or sexual orientation).” “Hostility to the victim” is often evidenced by the offender’s use of language or symbols. In this sense, defining the nature and degree of that hostility becomes a matter of linguistic inquiry. The potential centrality claim of linguistic analysis applied to hate crimes is further enhanced when we consider the American Psychological Association’s notion that “hate crimes are message crimes. They are different from other crimes in that the offender is sending a message to members of a certain group that they are unwelcome.” The crime is in the message and often the message is conveyed through language. This paper examines the following compounded questions. Is it a hate crime when an offender uses (and is exposed to) hate crime qualifying phrases not only when referring to the victim, or the victim’s group, but also when referring to friends, acquaintances, and the general public; and when that marked language encompasses a wide range of register variation including popular media culture; and when that language is accepted as normal usage in the offender’s and the victim’s speech communities? These questions become more serious when asked in a context that, if answered in the affirmative, will result in the offender facing the death penalty.

Sabine Ehrhardt Bundeskriminalamt KT54-Author Identification Germany