26 April 2018: Gea de Jong-Lendle (Marburg) – “Forensic phonetics: Language identification from a foreign accent in German – Where does the kidnapper come from?”


In the middle of the day the son of a rich industrial family in Germany was kidnapped. A ransom was demanded for the victim to be released and exact details provided of how the money transfer should take place. Surprisingly, the victim was released the next morning, left in the woods, without the money transfer having taken place. As it concerned a 50-year old mentally-disabled man, a psychological expert in the case stated that based on his limited mental abilities any judgment or identification concerning his kidnapper would not be admissible in court. No one seemed to have witnessed the event: it was only when the victim did not return from his usual walk, that his care-takers raised the alarm. However, high-quality recordings existed of three phone calls made by the kidnapper(s) contacting the victim´s family and a hotel reception. The caller speaks German fairly fluently, but clearly with a foreign accent.


The case led to a number of interesting but also highly complicated research questions. Can a speaker´s native language be derived from his foreign (L1) accent in another language (L2)? How does a speaker`s existing phoneme set influence the pronunciation of new sounds? Can such articulations be predicted? Related to this: how well do speakers learn new phonemes? Are particular phonemes more difficult than others? Are certain speakers better than others in acquiring new articulations? The production of certain phonemes seems unusually variable. Does this indicate, that the phoneme in question does not belong to his native set of phonemes, or could it be related to his manner of speaking; the caller is partially reading the message and could be confused by the orthography? Or is the variation just lexical, i.e. the speaker learnt particular expressions from other non-native speakers/colleagues in German. From an articulatory and perceptual point of view: how much variation in the production of a particular sound is “allowed”?


Phonetic methods used in this case are demonstrated and the associated findings presented. Finally, questions are discussed, that are still left unanswered.

Perhaps … you can help?


Short Bio:

Since 2010 I am a senior scientist/lecturer in phonetics at the University of Marburg in Germany. My main research interests focus on the area of forensic phonetics and I have undertaken forensic investigations since 1994, for both prosecution and defense. I have provided voice analysis services for a number of UK institutes/forces like the Forensic Science Service, Metropolitan Police and Liverpool Police and assisted the Dutch police. Currently, I am leading a forensic phonetics team in Marburg conducting forensic analyses in Germany. Other academic posts include Senior Research Associate at the University of Cambridge and Lecturer at City University in London. I hold a PhD in linguistics from the University of Florida and an M.Phil. in Computer Speech and Language Processing from the University of Cambridge.