Hancock et al.'s Study of the Language of Psychopaths

The aim of the study was to test how crime narratives differ between psychopaths and non-psychotic murderers. The psychopathic speech was analysed for indications of an instrumental of predatory worldview, unique socio-emotional needs and a poverty of affect. 

The sample consisted of 14 psychopathic and 38 non-psychopathic male murdered (the control group) who were imprisoned in Canadian correctional facilities. All of the 52 participants had admitted their crime and volunteered for the study. The two groups did not differ on age, and their overall mean age at the time of their current homicide was 28.9 years (range of 14-50 years). A similar mean amount of time had elapsed since their crime across the two groups (psychopaths had a mean length of time since their crime of 11.87 years, non-psychopaths 9.82 years).

Types of Crime (Number of Convinctions):

First-Degree Murder (8) 

Second-Degree Murder (32)

Manslaughter (10) 

Not specified in study (2) 

Total (52)

A self-report method was used which involved face-to-face individual interviews. The Psychopathy Checklist-Revised (PCL-R) was used to assess the participants against 20 different criteria. This clinical diagnostic tool measured affective traits and impulsive/antisocial traits to give individual scores out of 40 (30 or above is the cut-off for diagnosing psychopathy). The assessments were carried out by trained prison psychologists and researchers experienced in coding the test. An inter-rater reliability check was conducted on the PCL-R scores by having a trained graduate student re-code 10 randomly selected case files. At the start of each interview, participants were verbally briefed on the aims and procedure of the study. The offenders were asked to describe their offences in as much detail as possible while being audio-taped. The interviews were carried out individually and lasted approximately 25 minutes each. Participants were prompted by interviews using a standardised procedure known as the 'Step-Wise Interview', which facilitates open-ended questioning. The interviewers were two psychology graduates and one research assistant, who were all 'blind' to the psychopathy scores of the participants. Each interview was then transcribed verbatim and analysed using two text-analysis tools. The first was known as Wmatrix, a programme that was used to analyse parts of speech and semantics contained in the 'corpora' or whole body of the two groups of transcripts. The second tool was the Dictionary of Affect in Language (DAL) software programme, which specifically analysed emotional properties of language such as positivity, intensity and imagery. 

The interviews of psychopaths and control group produced a total 127,376 words. The results showed no significant difference in the average number of words produced by psychopaths and the control group. Instrumental language analysis showed that psychopaths were more likely to describe their homicide using subordinating conjunctions (1.82% of words in the psychopath corpora compared to 1.54% of words in the control corpora). These include words that show cause and effect, such as 'because', 'as' and 'so that'. As the researchers had expected, the crime narratives of the psychopaths emphasised more basic needs, including food, sex and shelter, while on the other hand, the controls focused more on higher-level social needs, such as meaningful relationships and spirituality. Overall, psychopaths referred to basic psychological needs far more as often in comparison to the control group when describing their homicide. Finally, emotional expression in language was analysed and compared. Psychopathic offenders were found to have used more past-tense of verbs and fewer present tense forms of verbs than controls. Psychopaths' language also involved more concrete nouns. This data suggests that psychopaths viewed their crime as more in the past and more psychologically distant than did non-psychopaths. In terms of the content, psychopaths generally used less positive or emotionally intense language. There were more instances of callousness and lack of empathy in their narratives, "I just turned around and stabbed him", language was significantly less fluent (33% more disfluencies) "We got uh, we got high, and had a few beers...we went uh for a swim". 

Hancock et al. conclude that psychopaths describe powerful emotional events in a rational but more primitive way compared to others. Their analysis suggests that compared to non-psychopaths, psychopaths: 

1. tend to view their crimes as the logical outcome of a plan 

2. are more likely to focus on their own basic physiological needs

3. overall, are less emotional and less positive in their speech 

4. are more emotionally detached from their crimes 

The study used the self report method of interviews to compare psychopaths' and non-psychopaths' crime narratives. Interviews have the advantage over questionnaires that they gather large amounts of details and in-depth data about indviduals. However, owing to how time-intensive they are compared to questionnaires, they tend to gather data across fewer participants. 

Large amounts of qualitative data in this study were produced from individual interview recordings. This captured the richness of language used to describe each homicide, allowing the researchers to study the differences between psychopaths and non-psychopaths in great depth. However, in its raw form, the audio-taped narratives were too detailed and varied to allow direct comparison. Each interview was subsequently transcribed and analysed, producing quantitative data that could be more easily assessed. This allowed researchers an overall view of differences in language between the two groups.

Standards of privacy and confidentially were maintained as participants remained unidentified within the study. Researchers took care to gain fully informed consent from participants, using a volunteer sampling method. This took into consideration the circumstances of participants' imprisonment, and how it might affect their abilities to freely consent to contribute to the study. Participants were also given a full verbal brief at the outset of each interview. 

The study has good ecological validity because participants were interviewed about their own real-life crimes, during which they provided detailed, in-depth accounts. Furthermore, the interviewing technique was designed specifically to avoid leading participants towards particular responses to allow them to speak freely. However, in some cases, participants were interviewed more than a decade after their crimes. Social desirability bias might have reduced the validity of the responses as participants might have wanted to appear remorseful. 

Highly valid measures were used to determine psychopathy (PCL-R). The linguistic analysis tool (Wmatrix and DAL) had been tested for validity and were being used extensively across other research, providing concurrent validity. 

Overall, this study was very reliable. The interview procedure was capable of replication, as was the linguistic analysis. Additionally, there were random checks by another researcher, which confirmed a high level of inter-rater reliability. 

Hancock et al.'s sample were unrepresentative, as all participants were Canadian prisoners, which makes the study ethnocentric. Furthermore, all the participants were male, making it difficult to accurately generalise to a female population. The participants were recruited as volunteers, and this may also have introduced bias to the research as the sample is unlikely to represent murderers in the prison who chose not to participate. However, a strength of the sample is that a relatively large number of prisoners were interviewed, considering the time-consuming nature of this research method. This produced huge amounts of data for statistical analysis, which improves the representativeness of the research. 

Understanding the language of psychopaths, in turn, reveals personality and behavioural traits that set them apart from the general population. It has the potential to help identify psychopaths and deal with them effectively. For example, as Hancock's study shows, psychopaths reveal an emotional distance from their crimes and motivations linked to basic needs. These key differences can be considered by those working in the criminal justice system, for example in developing appropriate ways to rehabilitating psychopathic criminals.