Detecting Deception through RST: A Case Study of the Casey Anthony Trial
Bibliographic record
Abstract
Many researchers have used linguistic analyses to determine if features, such as syntactic patterns or word choice, vary based on the truth or untruth of an utterance. For example, Newman et al. (2003) examined lying in written communication, finding that deceptive utterances used more total words but fewer personal pronouns. However, relatively few studies have focused on speech or writing style, which can be used to aid in authorship attribution and plagiarism identification (Cristani et al., 2012), and would thus seem to prove valuable for detecting deception. Recently, efforts have been made to remedy this by extending the application of linguistic feature analysis. For example, Rubin and Lukoianova (2014) applied Mann and Thompson’s (1987) Rhetorical Structure Theory (RST) to elicited written narratives that participants self-identified as either truthful or deceitful. Their findings suggest that RST relations, illustrative of functional relationships between ‘spans’ of text, vary based on the truthfulness of the narratives. However, this study, like previous studies, relies on researcher-prompted untruths rather than naturally occurring ones. As such, participants have little motivation to make the deception believable, unlike in real-world situations. The present study thus combines linguistic analysis with an examination of naturally occurring deception in the high-stakes setting of the State of Florida versus Casey Marie Anthony , in order to determine if findings like those of Rubin and Lukoianova (2014) are generalizable to deceptive statements in real-world settings. From publically available legal case documents, a corpus of 724 words (65 text segments) was selected and RST relations were coded. While some of Rubin and Lukoianova’s (2014) findings were minimally supported, no strong correlation between relations and the truth value of an utterance were found, suggesting the need for additional research in this area.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".