Liberman et al. (1957) demonstrated categorical perception of phonemes. What does this mean?
A: Listeners can detect infinitely small acoustic differences between sounds
B: Phonemes are perceived as discrete categories despite continuous acoustic variation
C: Speech perception depends entirely on visual lip-reading cues
D: Different languages perceive the same sounds identically
Correct: Phonemes are perceived as discrete categories despite continuous acoustic variation
Categorical perception means we hear sounds as belonging to one phoneme category or another — not as varying along a continuum. Voice-onset time (VOT) can shift gradually across a recording, yet listeners hear a sharp boundary between /b/ and /p/. This boundary effect is robust and, to some degree, universal, though experience with one's native language shapes exactly where the boundary falls.
In the McGurk effect (McGurk & MacDonald, 1976), participants who hear the sound /ba/ while watching lip movements for /ga/ typically perceive /da/. What does this demonstrate?
A: Visual information overrides auditory information in speech perception
B: Speech perception is a purely auditory process unaffected by visual cues
C: Speech perception is multimodal — auditory and visual signals are integrated
D: Adults cannot use lip-reading to improve speech comprehension
Correct: Speech perception is multimodal — auditory and visual signals are integrated
The McGurk effect shows that speech perception is inherently multimodal. The brain does not simply receive an auditory phoneme and look it up — it integrates acoustic and visual articulation information, often producing a percept (/da/) that matches neither input alone. The effect is robust, automatic, and persists even when observers know it is an illusion.
In Marslen-Wilson's (1987) Cohort Model of spoken word recognition, which process determines when a word is uniquely identified?
A: When the word's frequency in the language exceeds a threshold
B: When the listener has heard the entire word including final phoneme
C: At the "uniqueness point" — where only one word in the lexicon matches the input so far
D: When the word activates semantic associations in long-term memory
Correct: At the "uniqueness point" — where only one word in the lexicon matches the input so far
The Cohort Model proposes that as each phoneme is heard, the listener activates all words starting with that sound (the "cohort"). Members are progressively eliminated as more phonemes arrive, until only one candidate remains — the uniqueness point. This predicts that words can be recognised before they end, which has been confirmed with gating tasks: people correctly identify many words before the final phoneme.
Meyer & Schvaneveldt (1971) found that participants were faster to recognise "nurse" as a word when it was preceded by "doctor" rather than "bread". This phenomenon is called:
A: Repetition suppression
B: Semantic priming
C: Phonological bootstrapping
D: Lexical satiation
Correct: Semantic priming
Semantic priming is the facilitation in lexical access caused by prior exposure to a semantically related word. Seeing or hearing "doctor" activates a cluster of associated concepts in the mental lexicon, reducing the time needed to recognise related words. Priming effects are used to map the structure of semantic memory and reveal how meaning is organised in the brain.
The sentence "The horse raced past the barn fell" causes comprehension difficulty because:
A: It contains a word the reader has never encountered
B: The parser initially attaches "raced past the barn" as the main verb, then must reanalyse when "fell" arrives
C: It violates Grice's maxim of quantity by being unnecessarily long
D: It uses passive voice, which is harder for the brain to process than active voice
Correct: The parser initially attaches "raced past the barn" as the main verb, then must reanalyse when "fell" arrives
This is a classic garden-path sentence. The parser initially treats "raced" as the main verb, but "fell" at the end forces a reanalysis: "raced past the barn" is a reduced relative clause (the horse that was raced past the barn fell). The garden-path effect reveals that syntactic parsing is incremental and initially commits to the simplest reading — minimal attachment — often needing costly revision.
According to Grice's (1975) cooperative principle, which maxim is violated by a referee who writes only "Mr X's command of English is excellent, and his attendance has been regular" in a letter for a philosophy position?
A: Quality — the referee is lying
B: Quantity — the referee is providing too little relevant information
C: Relation — the content is completely off-topic
D: Manner — the letter is written unclearly
Correct: Quantity — the referee is providing too little relevant information
Grice's maxim of quantity states one should be as informative as required for the purpose. A philosophy reference that only praises English and attendance is conspicuously uninformative about philosophical ability — and by violating quantity, it implicates (without stating) that there is nothing good to say philosophically. This is conversational implicature: what is communicated goes beyond what is literally said.
Baddeley (1986) proposed a phonological loop within working memory. What is its primary function in language processing?
A: Translating written text into speech output
B: Storing phonological representations temporarily to support sentence comprehension and learning new words
C: Suppressing irrelevant semantic associations during word recognition
D: Integrating prosodic cues with syntactic structure
Correct: Storing phonological representations temporarily to support sentence comprehension and learning new words
The phonological loop consists of a phonological store (holding ~2 seconds of speech-based information) and an articulatory rehearsal process that refreshes it. It is critical for maintaining word sequences during sentence processing — especially long or complex sentences — and for learning new phonological forms. The word-length effect confirms its role: longer words are harder to hold than shorter ones.
In the dual-route model of reading (Coltheart, 1978), the non-lexical (phonological) route converts print to sound via:
A: Whole-word look-up in a stored visual lexicon
B: Grapheme-to-phoneme correspondence rules applied serially
C: Semantic mediation — accessing meaning before pronunciation
D: Prosodic pattern matching with stored speech templates
Correct: Grapheme-to-phoneme correspondence rules applied serially
The dual-route model proposes two pathways for reading aloud. The lexical route addresses familiar words directly via stored visual word forms. The non-lexical route converts spelling to sound through grapheme-to-phoneme correspondence (GPC) rules — it can read regular words and nonwords (e.g., "blorp") but fails on irregular words. Patients with surface dyslexia rely exclusively on GPC rules; those with phonological dyslexia can only use the lexical route.
Language Processing
Liberman et al. (1957) demonstrated categorical perception of phonemes. What does this mean?
About this quiz
How does the brain turn sounds and symbols into meaning? This quiz covers speech perception, word recognition, syntactic parsing, semantics, pragmatics, and the role of working memory in real-time language comprehension.