Psycholinguistics and Threat Prediction: Analyzing

Psycholinguistics and Threat Prediction: Analyzing the Words That Hurt

The purpose of the study is to conduct research into actual language and see what constitutes a presage of violence and what does not. The applications are valuable and obvious in the field of psycholinguistics. Within the study, several approaches including game theory, threat simulation theory, options theory, lexical features, phonological features and finally syntactic features in order to answer the following questions.

What is threatening language?

Are there specific levels of escalation within threatening language?

How can these levels of escalation be determined through the language used?

How does escalation of language lead to an actual violent act?

Within the scope of the research, there is especially a huge amount of work to be carried out in the area of game and options theory where both epistemic and frequentistic methodology has been applied to the prediction of human behavior (Jaeger, 2008, pp. 407-408). While previously game theory has provided more insightful than quantitative results for psycholinguistics, impressive results from the gambling industry where human behavior has been accurately and profitably predicted based upon the work of Harrah's CEO and former Harvard associate Dr. Gary Loveman. In Dr. Loveman's work, the Harrah's hotel and casino chain scientifically gauges and accurately predicts a specific variable in human behavior: the willingness to spend money (Metters, et. al., 2008, pp. 162-163).

While the Metters article centers more on the specific statistical aspects of Loveman's approach, a practical example can be shown of the effect that the research has upon the functioning of the Harrah's Casino chain on a customer service level. While not a law enforcement example, the revolutionary implications can easily be carried over into the threat arena. For instance, Loveman quotes an example of a casino customer that most of the industry would not consider a typical prospect. However, by using game theory to predict the customer's behavior, they are able specifically tailor the staff's service to match this predicted behavior and keep her coming back many more times to game and to enjoy the hotel's amenities. In this effort, they use an extensive data base that tracks customer purchases and spending in the casino, as well as gaming habits and preferances. By mining this data and using statistical analysis to process it, they have being able to accurately anticipate the customer's behavior. Harrah's can offer attractive options packages based upon that will impress the patron and attract their business (Loveman, 2003, p. 1). If one can use game theory to predict one type of human behavior such as spending and behavior in games of chance, one should be able to predict violent behavior of potential and active perpetrators. Dr. Loveman and his company is not just a scientific, but also a financial success.

It is the opinion of this author that it would be profitable to apply this approach of analysis to provide more accuracy to the field of linguistics and to not just profile, but accurately predict the behavior of a potentially threatening communication, such as an outburst or a phone call from the threatening person. Certainly, if it works in the gaming industry, it has a wide variety of applications in other areas as well by using the same quantitative methodologies. Basically, the above approach would provide a more accurate set of data points to threat simulation theory, options theory, lexical features, phonological features and finally syntactic features so that we can accurately answer the research questions we have raised above, especially when the threat can be positively identified.

Certainly, an incarceration, simple arrest, questioning, or data gathering on individuals such as wire taps would produce a plethora of data that could be used the statistical analysis of potential, real, or existing threats. Some individuals are under surveillance or incarceration for extended periods of time (such as gang leaders, Mafiosi, etc.) and would provide a huge quantity of analyzable data that could be fed to the National Center for the Analysis of Violent Crime (NCVAC) at Quantico, Virginia. Such data storage and analysis facilities feed data to law enforcement agencies such as the FBI. The Bureau is a major researcher into and user of psycholinguistic analysis and products. While much of law enforcement's criminal investigative analysis was designed for and works best in investigations of serial criminal acts (rapes, homicides, or arsons), a more systematic and scientific methodology of data analysis would allow more specific profiling in the case of individual crime (Smith & Shuy, 2002, p.p. 16-17).

In this milieu, threat simulation theory, options theory, lexical features, phonological features and syntactic features all play a part in the when sociolinguistic features such as geographic origins, ethnicity or race, age, sex, occupation, education level and religious orientation are analyzed. Individual psycholinguistics provides narrower categories where our above theoretical categories can be applied in areas such as threat assessment, authorship identification, false allegations and threats, workplace violence, statement analysis and homicides disguised as suicides. All of this is culled from the analysis of the offender's actual language which will answer our above research questions (ibid, pp. 17-20).

Lexical features can also be analyzed to predict threatening behavior in speech. For instance, it has been found that threatening speech is all researchers agree that a generic slowing can be found in threatening speech (Larsen, Mercer, & Balota, 2006, p. 3). Phonological analysis will help us determine where the threatening person is from, especially in the analysis of regional dialect and accent (Smith & Shuy, 2002p. 17).

Before we go further, some more practical examples for theoretical examples will illustrate practical applications of threat analysis of speech and written communications. For instance, lexical Features, phonological features and syntactic features all figure very largely in police bomb threat analysis. In an article by Robert F. Tunkel, he cites an example of a bomb threat to a high school pep rally: