Production and Aphasia

Question 1

Why is studying production (rather than comprehension) of language so complex?

Answer 1

you have to account for…

1. recruiting and deploying the linguistic resources appropriate for communicating one’s thoughts
2. being able to think in ways that are readily converted into language (“thinking for speaking”)
3. bringing to mind words/morphemes that are suitable for conveying the thought
4. arranging the words or morphemes in the way the language requires so as to mean to the listener what the speaker means to say
5. energizing the arrangements into sound forms through the actions of the speech musculature

Question 2

What are the basic processes in production?

Answer 2

1. conceptualization
2. formulation (word selection + sound processing)
3. articulation

Question 3

What type of evidence can we use to justify the basic processes in speech production?

Answer 3

1. speech errors (slips)
2. tip of the tongue
3. experiments

Question 4

What do speech errors tell us about the language processing system?

Answer 4

By knowing which slips are possible and which are not, we can constrain the theories of production. Slips are not random, and models of speech productions need to account for these regularities in slips.

Question 5

What are some types of slips?

Answer 5

The classic Freudian Slip

• • arises from concurrent (or opposing) action of two different intentions
• • the intended meaning + disturbing intention = speech errors
• • the psycholinguistic approach however assumes that the mechanics of slips can be studied linguistically without reference to their motivation

Spoonerisms
– tend to switch similar segments (phonemes, morphemes, words, etc.)

Question 6

What’s the point of looking at speech errors?

Answer 6

They can teach us how the system breaks down, which will tell us something about how it works.

• • errors aren’t random
• • look for regularities in the patterns of errors

Question 7

What do we learn from shift errors (“a maniac for weekends” vs. “a weekend for maniacs”)

Answer 7

This suggests that speech is planned!

• • there is accommodation to the phonological environment (ie. the plural being pronounced /z/ instead of /s/)
• • tells us that there is order of processing (selection of morpheme – error – application of phonological rule)

Question 8

What can we learn from prosody in errors?

Answer 8

Stress can shift while intonation stays intact. This suggests that prosody is planned independently of segmental phonology!

Question 9

What can we learn from substitution errors (“bat a tog” vs. “pat a dog”?)

Answer 9

In this instance it’s a matter of swapped voicing (a shift of the phonetic feature) which implies that phonetic features are psychologically real units of speech production.

Question 10

What are some regularities in speech errors?

Answer 10

1. consonant-vowel rule: consonants never exchange for vowels, and vice versa.
   - - suggests that vowels and consonants are separate units in the planning of the phonological form of an utterance
2. errors produce LEGAL non-words
   - - suggests that we use phonotactic rules in production
3. lexical bias effect
   - - speech errors are more likely to result in real words than non-words
4. category constraint
   - - when words are substituted or exchanged they typically substitute for a word of the same grammatical class

Question 11

What is Levelt’s Model (1989)?

Answer 11

1. conceptualization
   - - conceptualize what we wish to communicate
2. formulation
   - - formulate the thought into a linguistic plan
3. articulation
   - - execute the plan through the biological speech system
4. self-monitoring
   - - monitor the speech to check whether it is what we intent to say and how we intend to say it

Question 12

How does the Levelt model break down?

Answer 12

CONCEPTUAL PREPARATION (lexical concepts)

The formulation structure:

FORMULATION contains:

1. lexical selection (lemmas)
2. morphological encoding (morphemes)
3. phonological encoding/syllabification (phonological words)
4. phonetic encoding (phonological gesture score)

ARTICULATION (sound wave)

Question 13

What type of errors can be accounted for by Levelt’s Model?

Answer 13

1. semantic substitution (“dog” for “cat”; at conceptual preparation and lexical selection)
2. morphological errors (“productful” for “productive”; at lexical selection and morphological encoding)
3. sound exchanges (“fig beet” for “big feet”; at phonological encoding/syllabification and phonetic encoding; more common after being primed)

Question 14

What is the difference between a lexeme and a lemma?

Answer 14

Think ToT states, where you can retrieve the lemma without the lexeme.
– know the meaning but cannot produce the word
(so meaning = lemma and lexeme = word produced?)

Question 15

What are some things that happen during a ToT experience?

Answer 15

people. ..
1. accurately predict whether they will come up with the correct word soon
2. report the correct number of syllables
3. accurately report the first phoneme
4. may be aware of the meaning but can’t access the form
5. report word-specific grammatical information (ie. syntactic class, gender)
6. are more accurate about the beginning and end phonemes than the middle
7. report words that sound like the target
8. have more ToT’s for less frequenct words
9. resolve about 40% of ToT’s within a few seconds to a few minutes

Question 16

What type of self-repairs are there?

Answer 16

1. instant repairs (replace with the correct word)
2. anticipatory repairs (the speaker retracts back to some point prior to the error)
3. fresh starts (just start over)

Question 17

What is the relationship between monitoring and repairs?

Answer 17

Self-repairs happen after an error, while self-monitoring helps to prevent overt errors (pre-output monitoring).
– when speakers make an error they often replace the error with the correct word with no delay/nearly no delay

BIG THING: since speech planning takes time, the plan for the correction must be undertaken as the error is being produced; therefore the error must be detected before it is spoken!!
– internal loop

Question 18

What is the difference between picture naming and picture recognition?

Answer 18

People activate different concepts at about the same speed. Picture recognition speed is constant, regardless of word frequency, whereas picture naming is dependent on word frequency.

Concepts compete for selection (the picture-word interference)

Question 19

What is the lexical bias effect?

Answer 19

It is how words in the lexicon influence sound substitutions.
– sound substitutions resulting in words are more likely than those resulting in non-words

Question 20

What is the positional constraint?

Answer 20

For sound exchanges, sounds switches occur at the same place (onset with onset, coda with coda)

Question 21

What is the category constraint?

Answer 21

How word exchange errors usually occur with categories (“my piano plays the girlfriend”)

Question 22

What is evidence for interactivity in the Levelt (?) model?

Answer 22

• • lexical bias effect
• • category and positional constraints
• • mixed errors (word substitutions sound like and have similar meaning to the target; interpreted as evidence for phonological-lemma feedback during lemma selection)