Explanation
Speech perception is the active processing through which a continuous and highly variable acoustic signal is experienced as an articulated sequence of sounds, words, and meanings. The ear does not receive neatly separated units: articulations overlap, voices differ, and noise masks parts of the signal. Listeners therefore combine acoustic cues with categories, expectations, context, and linguistic knowledge. This completion makes communication robust, yet it can also cause the same sound to be heard differently under different expectations. The concept presents perception as a regulated interaction between a material signal and an acquired space of interpretation, rather than as the passive copying of an already segmented world.
Movement of thought
1. A continuous stream of sound first reaches hearing without objectively inscribed boundaries between all linguistic units. 2. Acoustic features, rhythm, and prosody activate learned sound and word categories together with contextual expectations. 3. Ambiguous or missing portions are completed through interaction between bottom-up cues and top-down knowledge. 4. A counter-test changes context, voice, or masking noise and asks whether the supposedly heard unit remains stable. 5. What was heard returns to awareness as a constructed linguistic event that nevertheless remains answerable to the signal.
Scientific use case
Psycholinguistics can use identification, discrimination, and priming experiments to examine how categorical boundaries, lexical expectations, and phoneme restoration interact. Clinical research can compare these processes in hearing and language-processing disorders.
Example
In the noisy sentence “She put the vase on the …”, many listeners hear the predictable word “table” although one syllable is masked; another sentence context may complete the same trace differently.
Counterexample
A tone is registered only as loud or soft without being processed as a linguistic unit. This is auditory perception, but not speech perception in the narrower sense.
Distinction
Distinguish speech perception from hearing in general, language comprehension, word recognition, and categorical perception. It spans several processing levels; recognition and categorisation are component processes.
Effect on thought and perception
Speech perception rapidly turns variable acoustic events into manageable linguistic units; its efficiency often conceals the constructive work of completion.
Resonance
The incoming signal meets stored patterns of sound and meaning; only their reciprocal response allows an intelligible voice to emerge from sound.
Cite this entry
Reich, Michael (Hrsg./ed.): „Speech perception“. Atlas der Sprache, AT-DEP6-0006. ATLAS 260, data model 1.6.0. https://atlasoflanguage.com/concept/AT-DEP6-0006The address remains stable; editorial content may evolve. Editorial reading state: 2026-10-15. Include your access date for precise references.
Historical background
Sapir argues that a speaker’s familiar sound system affects which differences seem salient in unfamiliar speech. His account changes the question from whether the ear receives a sound to how that sound is classified as linguistically relevant. For speech perception, this is substantial historical conceptual evidence, but not a complete model of auditory processing. A perceived difference need not distinguish words in the listener’s language, and a difficult foreign contrast is not proof of an anatomical inability to hear it. Sapir’s reported examples require independent verification before language-specific conclusions are extended. The Atlas can use the distinction between acoustic variation and learned categorisation while leaving modern experimental mechanisms open. The historical contribution is a disciplined separation of sensation, linguistic habit and functional contrast.
How conceptual history changes the analysis
Acoustic differences need not map onto linguistic categories, and linguistic categories need not mirror every acoustic difference.
References for the historical section
- Edward Sapir, Language (1921), Gutenberg transcription carrying later printing information · Language III: learned sound categories and the hearing of unfamiliar speech
Selected witness, not a complete genealogy or first English attestation. Contemporary Atlas comparisons are editorial interpretations. Sapir transcription contains later printing information. Historical language examples are attributed reports, not independently verified contemporary descriptions; racial classifications and cultural hierarchies are not adopted.