In 2017, US intelligence was trying to identify the person behind an anonymous cyberattack. They had only one piece of evidence: a few lines of comments hidden among thousands of lines of code and a readme file. The question was: can a person's writing style give them away?
The same question is identical to the one asked by Jan Svartvik in 1968 in the Timothy Evans case: Did the statement taken at the police station really belong to the defendant, or was it written by someone else? Svartvik's linguistic analysis led to Evans's exoneration 16 years after he was executed, but one question has always remained on the agenda: Does language really carry a fingerprint?
Today, the answer to this question is no longer solely the concern of linguists, but also of artificial intelligence researchers, cybersecurity experts, and legal professionals. Large language models can generate a text; but how do we know who wrote this text? If a suspect writes a fake suicide note, how do we distinguish it from a real one? In an era where identities are blurred in the digital environment, forensic authorship analysis is perhaps in a more critical position than ever before.
In this three-week course, we will address the topic of forensic authorship analysis, the most central subfield of forensic linguistics, within a framework adapted to the context of Turkey. Each session is built upon the triad of theoretical foundation, real case analysis, and practical application.
Dr. Hülya Kocagül received her bachelor's degree from the Department of Linguistics at Dokuz Eylül University. She later earned her master's degree in Linguistic Studies at the University of Essex, UK, and her PhD in Forensic Linguistics from Aston University. She is currently continuing her work as a faculty member in the Department of Linguistics at Istanbul Medeniyet University.
The course will take place online via Zoom on Sundays, March 15, 29, and April 5, between 16:00 and 18:00. It will last approximately 6 hours in total. All sessions are recorded and shared with the participants. The sessions can be accessed indefinitely after the course ends.
Program Outcomes
Participants who complete this program will be able to:
- Evaluate the foundations and limitations of forensic linguistics.
- Grasp the difference between author profiling and author comparison.
- Explain the concept of idiolect and the methods of qualitative stylistic analysis.
- Apply qualitative authorship analysis on real case materials.
- Recognize linguistic intervention markers in statement and declaration texts.
- Evaluate the new challenges and opportunities of authorship analysis in the digital age.
- Discuss the new challenges that AI-generated texts pose for forensic linguistics.
Session 1 (March 15): Foundations of Forensic Authorship Analysis Adli Yazar Analizinin Temelleri
What is forensic linguistics, and where does authorship analysis stand within this field?
- The three pillars of forensic linguistics.
- Idiolect theory and the "linguistic fingerprint" debate.
- The three main tasks of authorship analysis.
- Stylistic markers and intra-linguistic variables.
- Svartvik (1968) and the Evans Case: The birth of forensic linguistics.
Session 2 (March 29): Linguistic Analysis in Statement Texts İfade Metinlerinde Dilsel Analiz
Is a text real or fake? Was it written by the police or the suspect?
- Detecting linguistic intervention in police statements.
- Types of fake texts and their detection.
- Threat language analysis.
Session 3 (April 5): Forensic Authorship Analysis and Artificial Intelligence in the Digital Age Dijital Çağda Adli Yazar Analizi ve Yapay Zekâ
How is authorship analysis transforming in the digital text world?
- Authorship analysis in digital communication.
- Forensic authorship analysis in the age of artificial intelligence.
- Forensic authorship analysis in Turkish.
- Ethics and expertise in forensic linguistics.
- The future of forensic linguistics in Turkey: Opportunities and challenges.
The course fee is 2450 TL per person, including VAT. Payment details will be shared via email with approved participants as soon as possible. The language of instruction will be Turkish.
