Language II Flashcards
What are phonemes?
elementary (smallest) units of sound. For example ‘rice’ and ‘lice’ differ from each-other by just one phoneme
What are morphemes?
elementary (smallest) units of meanings in language. For example, dog-s (two morphemes)
What is phonology?
Rules governing the sound of words and parts of words
What is syntax?
Rules governing word order and meaning resulting in sentences
What are semantics?
meanings of words and sentences.
What are pragmatics?
the use of language as a function of context and social rules
At it’s most basic level what can we think of language as?
we can think of language as a system of symbols and rules that enable us to communicate:
- Symbols: written and spoken words, signs
- Rules: specify how sentences are formed
What is the average speech rate?
150 words per minute
What does where speech errors occur often reveal?
discrete points in the language production system
What was Altmann & Kamide’s (1999) experiment about speech perception?
- If you hear a speech fragment ‘the child ate the…’ whilst looking at a picture containing an apple chair, and cat there is good evidence that you’ll look at the picture of the apple in anticipation of what is likely to come next in the sentence (Altmann & Kamide, 1999)
- Influence of top-down information
What can be said to be the three stages of speech production?
• Conceptualization
- Think of something to say
• Formulation
- Find a way to express your idea given the specific language tools
• Articulation
- Physical action of moving your muscles to produce speech
What is the WEAVER ++ theory of speech production? (Levelt et al., 1999)
- Idea
- Zoom in on the appropriate lexical item in the mental lexicon (lemma level)
- Retrieve a word’s morphemic code
- Retrieve a word’s phonological code
- Syllabify the word & access the corresponding articulatory gesture
- Move the muscles and produce speech waves
- No flow of information back up
- Serial set of processes – feed-forward information flow system
- Lemma is an intermediate stage between activating an idea and the speech sound
- The link from meaning to sound (lexicalisation) is not a one stage process
- You will be self-monitoring while you are speaking (still don’t always capture errors)
What is the spreading activation model of speech production? (Dell, 1986)
- Network in your head with different nodes representing different things
E.g. a set of nodes for semantics, a set of nodes for words, a set of nodes for phonemes
Different words may share some of the same semantic and phoneme nodes
Connection between phonemes, words and semantics
Almost like a neural network - Interactive information flow (from meaning to sound & sound to meaning)
- Lemma is an intermediate stage between activating an idea and the speech sound
- Lexicalisation is not a one stage process
What is a TOTs (tip of the tongue state)?
- A tip of the tongue state is what happens when the meaning clusters light up, but the sound clusters don’t activate completely, because the signal in your brain takes a detour instead of following the right path
- You can remember a lot of information about the word but not the word itself
What is supporting evidence for the models of speech production?
- Tip-of-the-tongue state (TOTs)
- Picture-word interference
- Speech errors
What are some features of TOTs?
- people can retrieve the first phoneme of the target word, the number of syllables of the target word, and the gender of the target word
- Most of the TOTs occur with low-frequency words (Harley & Bown, 1998)
What happens with picture-word interference?
- Semantic relatedness slows down picture naming (e.g. you would be slower to name a picture of an apple if it has the word ‘lemon’ written in it)
Lexicalisation is not a one stage process - Phonological relatedness speeds up picture naming (e.g. if you see a picture of a house with the word mouse in it you would name it as a ‘house’ more easily
- Arieh & Algom (2002)
What are the different types of speech errors?
- semantic error (where is my tennis bat instead of racquet)
- word exchange (my chair looks empty without my room instead of my room looks empty without my chair)
- sound-exchange (spoonerisms) (go shake a tower instead of go take a shower)
Why study speech errors?
- We gain insights into how the complex cognitive system involved in speech production works when it actually breaks
- We understand how information flows between different mental processes during speech production
- We understand how much planning speakers do when they speak
What two effects show that speech errors are not random
- Lexical bias effect
‘Fig beet’ instead of ‘big feet’ more often than ‘hig borse’ instead of ‘big horse’ (Baars, Motley & MackKay, 1975)
You are more likely to produce speech errors within your language - Mixed-error effect
Let’s start’ but not ‘let’s begin’ instead of ‘let’s stop’ (Dell, 1986)
Which model of speech production is most supported by the evidence?
The spreading activation model - evidence suggests that speech production is an interactive process
What is coarticulation and what is good and bad about it?
- The pronunciation of a phoneme by a speaker depends on the preceding and following phonemes. The /b/ phonemes in ‘bill’, ‘ball’, ‘able’, ‘rub’ are acoustically different. This actually helps speech perception.
- The word ‘job’ is likely to be misperceived if the ‘jo’ part was recorded from the word ‘jog’ (which some splicing)
- Bad news:
No one-to-one relationship between acoustic signal and phonemes - Good news:
Allows for prediction and makes perception faster
What is the Motor theory of speech perception?
- Listeners perceive spoken words by reproducing the movements of the speaker’s vocal tract rather than by identifying the sound patterns that speech generates
- Feed-forward
- The motor system is involved in both speech production and perception
There is a bridge between production and perception
What is thought about the motor theory and mirror neurones?
- Activation of mirror neurones (close to motor cortex)
- Think that mirror neurones may be activated when listen to speech so ‘you mimic it’
What provides support for the motor theory?
- McGurk effect
- Categorical Speech perception
- fMRI and TMR
What is the McGurk effect?
- An illusion occurs when the auditory component of one sound is paired with the visual component of another sound, leading to the perception of a third sound
- We integrate non-acoustic information into what we hear
- Our auditory and visual information processing is merged
What is categorical speech perception?
- Raizada & Poldrack (2007)
- Artificial 1 to 10 sound continuum where 1 sounded as /b/ and 10 as /d/
- Speakers did not perceive gradual change
- They suddenly switched from perceiving one sound to perceiving the other
- Better discrimination of pairs of sounds when sounds were across the crucial point
- Shows that sounds belong to specific categories – not sensitive to subtle differences
What do the fMRI and TMS show about speech perception and the motor cortex?
- fMRI
Listening to speech and imagining you are speaking activates speech motor cortex - TMS
Motor TMS (disrupt motor cortex) decreases phoneme discrimination
What is the TRACE model of speech perception?
- Connectionist ‘neural model’
- Interactive process
- Very much based on the fact that you often hear what you are expecting to hear (top down)
- Three different levels of representation:
Word units
Phoneme units
Feature units (acoustic sounds)
1. Nodes influence each other creating patterns of activation
2. Interaction of bottom-up and top-down processes
What provides support to the TRACE theory?
- Lexical identification shift aka Ganong effect
When listening to sounds ranging from /dash/to/tash/ phonemes were assigned to words rather than non-words (Ganong, 1980) - Phonemic restoration effect
When listening to ‘it was found that the *eel was on the shoe’, participants heard ‘heel’
When listening to ‘it was found that the *eel was on the orange,’ participants heard ‘peel (Warren & Warren, 1970)
