Language 3 Flashcards
Model of speech production
Message -> Syntax -> Morphemes -> Phonemes
Sapir-Whorf hypothesis
The specifics of language change the content of thought
Linguistic determinism
Strong version of Sapir-Whorf hypothesis.
States that language determines how we think.
In other words this means that people who speak a different language view the world differently than us.
Linguistic relativism
Weak version of Sapir-Whorf hypothesis.
States that there are not absolute differences in cognition between languages but relative change in cognition depending on language.
Does the amount of colour terms influence colour perception
According to Berlin and Kay (1969) there is a constant underlying hierarchy for colour perception.
The Dani people from new Guinea were able to remember colours for which they have no words.
These studies are evidence against linguistic determinism
Thierry et al (2009) study on greek vs english speakers distinguishing light and dark blue
Greek people have two different terms for light and dark blue.
In this experiment the mismatch negativity for green is the same for greek and english subjects.
However the MMN for blue is greater for greek than english subjects.
This experiment is evidence for weak version of Sapir-Whorf hypothesis.
2 stages of lexical access
- Lexical selection: LEMMA, information on syntactic and semantic properties
- Word retrieval: LEXEME, morpheme stage, information about phonetic properties
Word frequency effect
It’s easier to pronounce high frequency words.
Wingfield (1986) showed that simply matching a word to a picture has no frequency effect. This is evidence that frequency information is stored at the lexeme level.
Homophone
1 sound form (lexeme) with two meanings (lemma’s)
Jescheniak & Levelt study on low and high frequency lemma’s and lexemes
Low frequency words with a high frequency homophone are fast, as if it is a high frequency word.
Connections between words at lexeme level influence language production.
Semantic and phonetic information are stored seperately from each other.
Effects of semantic and phonological distractors
Semantic distractor has an early effect.
Phonological distractor has a late effect.
Difficult to say the word when you have a semantic distractor, only with early SOA (stimulus onset asynchronies).
Easier to say the word with a phonological distractor, only with late SOA.
This suggests lexical access is sequential two step process: first lemma then lexeme
Lexical bias
Tempting people in the lab to make speech errors: phoneme switch, saying barndoor instead of darn bore
30% speech errors if the switch results in real words, only 10% if switch leads to non-words.
Dell’s model of speech production
Interactive activation in speech production where lower levels affect higher levels.
Parallel with TRACE model from speech perception.
This explains why there are more speech errors in phoneme switches if they result in real words.
Speech rate affecting speech errors
Fast speech causes slightly more non-word than word outcome speech errors
Slow speech has more lexical bias, meaning more errors that lead to word outcome.
Monitoring while speaking
Auditory target map is prediction of what you intend to say
Auditory state map is what you actually say
Auditory error map is the difference between the two (planum temporale)
