Forensic Linguistics Overview: Voice ID and Authorship Profiling

An idiolect is an individual's personal style of speaking, characterized by unique patterns in pronunciation, grammar, vocabulary, tone, and word choice. Even when two people speak the same language, their idiolects make them identifiable. This personal variation in language use allows us to recognize individuals by how they speak, even without seeing them.

Idiolect is defined as an individual variety of language emerging from the crossing of collective varieties. It is not merely personal language but results from the intersection of multiple social varieties: regional varieties (dialects based on geographic location), social stratification varieties (based on class and economic status), and chronological varieties (based on historical periods). Each person's language reflects their specific combination of regional origin, social background, and historical period. The speaker explains that idiolect is adapted from sociolinguistics and represents how individual language emerges from collective linguistic varieties.

Sociolect is language variation by age, sex, and profession (e.g., how doctors, engineers, or different generations speak). Like dialects, there are no better or worse sociolects—only different ones. Idiolect is individual variation by circumstances and moments (speaking differently in public vs. with friends, in markets vs. parties). Dialect is the highest level (macro), encompassing sociolect and idiolect. When asked to define dialect, 'any variation of language' is correct because dialect encompasses all other variations.

An idiolect, also known as diafasic variation, is the individual or personal variation in language use. This variation is influenced by multiple factors including sex (men typically have deeper voices while women have higher-pitched voices), age (language evolves as people grow older, such as changing from saying 'mamá' at age five to more mature expressions), and context of communication.

An idiolect is the unique language variety of an individual, encompassing their specific vocabulary, grammar, pronunciation, and speech habits, which differ from dialects that are shared by communities. While the Stanford Encyclopedia of Philosophy defines idiolect as language properties exhaustively specified by an individual's intrinsic properties, this definition excludes social and environmental factors that actually shape language use. Factors such as profession, living environment, family background, workplace jargon, and multilingualism all contribute to making each person's language use distinct. For example, a doctor who watches sci-fi movies would have different vocabulary than a lawyer who reads historical fiction, and even color perception varies between individuals based on personal experience and environment.

Acoustic phonetics studies the physical properties of sound waves in speech, focusing on three fundamental parameters: time (horizontal axis), amplitude (maximum deviation from rest position, vertical axis), and frequency (number of cycles per second). Sound waves require two preconditions for production: a source of energy (such as vocal cord vibration) and a medium (like air) for transmission. Sound waves are categorized into periodic waves (regular, repeating patterns that produce musical sounds with line spectra showing harmonics) and aperiodic waves (irregular, non-repeating patterns that produce noise-like sounds with flat spectra). A spectrum represents sound in the frequency domain, showing constituent frequencies and their amplitudes, while a spectrogram combines all three parameters (time, frequency, and amplitude) for comprehensive analysis. The source-filter theory of speech production, introduced by Müller in 1848, explains that the larynx acts as the sound source generating phonated airstream, while the oropharyngeal chamber functions as an acoustic filter that modifies the sound through resonance and filtering processes.

This comprehensive section establishes the foundational concepts of acoustic phonetics. It begins with vowel classification using three formants: F1 relates to tongue height (higher tongue = lower F1), F2 relates to tongue frontness/backness (more front = higher F2), and F3 relates to lip rounding and rhoticity. The relationship between tongue position and formants is not perfectly parallel - tongue height correlates more strongly with F1 than tongue frontness with F2. The section then explores the harmonic series in vibrating strings, demonstrating how a single vibrating body produces multiple frequencies simultaneously: the fundamental, first overtone (octave), and second overtone (octave plus fifth). The distinction between harmonic and overtone counting is clarified: overtones are always one less than harmonics. Resonance principles are introduced through soda bottle experiments showing that objects have natural frequencies at which they amplify signals dramatically. The source-filter model is introduced, explaining that the vocal folds generate a complex sound (source) with fundamental frequency and many overtones, while the vocal tract acts as a filter that selectively amplifies certain frequencies while attenuating others.

Acoustic phonetics studies the physical properties of speech sounds, particularly how the vocal tract configuration produces different vowel qualities through formants (F1, F2, F3), which are resonant frequencies amplified by the vocal tract. The first formant (F1) relates to tongue height (higher F1 = lower tongue position), while the second formant (F2) relates to tongue advancement (higher F2 = more anterior tongue position). These acoustic properties can be visualized using spectrograms, which display time versus frequency versus intensity, allowing researchers to measure formant frequencies and analyze speech characteristics. The source-filter theory explains that speech sounds result from the interaction between the glottal source (vocal fold vibration) and the vocal tract filter (which shapes the sound through resonance).

Acoustic classification uses formants (resonator frequencies) to characterize sounds. Vocalic sounds have formants in integer ratios; non-vocalic sounds do not. Additional characteristics include consonantal force, pitch, compactness/diffuseness, diphthongal/non-diphthongal, bemoled/non-bemoled, interrupted/non-interrupted, sharp/non-sharp, and voicing. Voice characteristics arise from vocal fold vibration: pitch depends on fold length/thickness/tension, timbre from overtones in resonators, loudness from amplitude. A phoneme is an abstract unit capable of distinguishing words, realized through allophones. Transcription methods include phonetic (one sign per sound, IPA-based) and phonemic (same sign for same phoneme regardless of allophones).

Digital communication creates permanent records that can be accessed and analyzed. Unsended messages and digital communications may reveal important information about relationships and intentions. Understanding digital evidence requires examining patterns in communication, identifying inconsistencies, and recognizing what information may have been intentionally withheld. Physical evidence collection involves systematic searching, proper documentation, and preservation of items for analysis. Investigators must identify relevant evidence, document its location and condition, and ensure items are properly collected and stored. The process requires attention to detail, understanding what constitutes evidence, and recognizing how physical items can reveal important information about events and circumstances.

Digital forensics distinguishes between logical and physical evidence collection: Logical collection involves copying data that has been interpreted by a file system or operating system, meaning the data is presented in a human-readable format with metadata (such as file names, creation dates, and folder structures). Physical collection involves copying raw data at the bit level without interpretation, capturing exactly what exists on the storage medium regardless of how it is organized or interpreted by the system. Logical collection is useful for analyzing file contents and metadata, while physical collection preserves the complete raw data structure including deleted files and unallocated space.

This segment covers the collection and analysis of digital and physical evidence in criminal investigations. Digital evidence, including cell phone data, can provide crucial information such as location data, communication records, and application usage. Physical evidence should be properly collected, documented, and preserved according to proper protocols to maintain chain of custody. Investigators should document the location and condition of evidence when discovered and seek corroboration through multiple sources.

This segment covers two critical aspects of criminal investigation: digital evidence collection and physical evidence processing. For digital evidence, officers obtain warrants for telecommunications data (pen register warrants) and social media subscriber information, serving them through company portals and downloading data to secure forensic computers. For physical evidence, officers follow chain of custody protocols when discovering items during searches, including documenting findings, consulting supervisors, securing items in temporary storage, and properly booking items into evidence. The process ensures evidence integrity from discovery through storage.

This segment covers evidence collection and digital forensics: (1) Systematic search of crime scenes including offices and living spaces, (2) Preservation of physical evidence (bottle with message, diskette), (3) Digital evidence analysis including password recovery attempts, (4) Correlation of physical and digital evidence, and (5) Understanding the importance of evidence chain of custody. The dialogue demonstrates how investigators systematically search for clues and use technology to access protected information.