In criminal proceedings, wiretaps and their transcription are often used as circumstantial evidence. This procedure presents numerous difficulties, as the wiretaps often have poor quality audio and the transcriber is not only required to transcribe the content of the conversation but also to interpret and deduce what is said. The research activity carried out in this paper is focused on the analysis of the problems inherent in the transcription of a spontaneous conversation between individuals. This activity is divided into two phases: a first phase dedicated to the creation of experimental materials and a second in which such materials are listened to and transcribed. The first section of the paper includes an introduction to the legal framework of wiretaps and transcriptions along with a review of scientific studies that have highlighted the presence of perceptual illusions when listening to degraded audio. The creation of the experimental material allows us to compare and identify such perceptual illusions, distinguishing between transcriptions of clearly intelligible speech, where the words are easily recognizable and the risk of error is minimal, and transcriptions of speech that are difficult to understand, where the transcriber must fill the interpretative gaps, with a consequent increase in the risk of perceptual biases and auditory hallucinations. This study proposes an innovative approach to improve the reliability of transcriptions: having the audio segments relevant to the process transcribed by a large number of transcribers and assessing the objectivity of the result on the basis of the percentage of agreement between the transcriptions obtained.
Nel processo penale si utilizzano spesso come prove indiziarie le intercettazioni ambientali e la loro trascrizione. Tale procedura presenta numerose difficoltà, in quanto le registrazioni ambientali spesso presentano un audio di scarsa qualità e il trascrittore si trova non solo a trascrivere il contenuto della conversazione ma anche a interpretare e dedurre ciò che viene detto. L’attività di ricerca svolta in questo elaborato è focalizzata sull’analisi delle problematiche inerenti alla trascrizione di una conversazione spontanea tra individui. Tale attività è articolata in due fasi: una prima fase dedicata alla creazione dei materiali sperimentali e una seconda in cui tali materiali sono stati ascoltati e trascritti. La prima sezione dell’elaborato prevede un'introduzione sull'inquadramento giuridico delle intercettazioni e trascrizioni insieme a una rassegna di studi scientifici che hanno evidenziato la presenza di illusioni percettive legate all'ascolto di audio degradato. La realizzazione del materiale sperimentale consente di confrontare e identificare tali illusioni percettive, distinguendo tra trascrizioni di parlato chiaramente intellegibile, dove le parole sono facilmente riconoscibili e il rischio di errore è minimo, e trascrizioni di parlato difficilmente comprensibile, dove il trascrittore deve colmare le lacune interpretative, con un conseguente aumento del rischio di bias percettivi e allucinazioni uditive. Questo studio propone un approccio innovativo per migliorare l’affidabilità delle trascrizioni: far trascrivere i segmenti audio rilevanti per il processo a un ampio numero di trascrittori e valutare l’oggettività del risultato sulla base della percentuale di concordanza tra le trascrizioni ottenute.
La trascrizione di intercettazioni ambientali nel contesto forense
ARCUDI, FRANCESCA
2024/2025
Abstract
In criminal proceedings, wiretaps and their transcription are often used as circumstantial evidence. This procedure presents numerous difficulties, as the wiretaps often have poor quality audio and the transcriber is not only required to transcribe the content of the conversation but also to interpret and deduce what is said. The research activity carried out in this paper is focused on the analysis of the problems inherent in the transcription of a spontaneous conversation between individuals. This activity is divided into two phases: a first phase dedicated to the creation of experimental materials and a second in which such materials are listened to and transcribed. The first section of the paper includes an introduction to the legal framework of wiretaps and transcriptions along with a review of scientific studies that have highlighted the presence of perceptual illusions when listening to degraded audio. The creation of the experimental material allows us to compare and identify such perceptual illusions, distinguishing between transcriptions of clearly intelligible speech, where the words are easily recognizable and the risk of error is minimal, and transcriptions of speech that are difficult to understand, where the transcriber must fill the interpretative gaps, with a consequent increase in the risk of perceptual biases and auditory hallucinations. This study proposes an innovative approach to improve the reliability of transcriptions: having the audio segments relevant to the process transcribed by a large number of transcribers and assessing the objectivity of the result on the basis of the percentage of agreement between the transcriptions obtained.File | Dimensione | Formato | |
---|---|---|---|
Arcudi_Francesca.pdf
accesso riservato
Dimensione
2.78 MB
Formato
Adobe PDF
|
2.78 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/84970