Lecture 4 - Spoken Word Production

Question 1

What does MLU stand for?

Answer 1

Mean length of utterance

Question 2

Why would MLU be preferred to number of words?

Answer 2

MLU gives a better reflection of the understanding of grammar by the child.

‘He drop’ vs ‘He dropped’

Both have 2 words, but the second one has 3 morphemes.

Question 3

What is the significance of a child reaching a specific MLU? And what is the MLU for which this is significant?

Answer 3

If a child has a mean length of utterance between 2.5 and 3.5 morphemes, they seem to be at the stage of grammar acquisition.

Question 4

What is telegraphic speech?

Answer 4

First utterances which do not include any function words, determiners, prepositions or inflections.

It does have syntactic structure and follows word order rules of the language used.

Question 5

What did Berko-Gleason (1958) demonstrate about rule application in children?

Answer 5

Children aged between 5 and a half and 7 years could articulate plural versions of nonsense words, and could produce the correct phonological form.

Question 6

What are the 5 different speech errors?

Answer 6

• Word blends
• Word exchange
• Sound exchange
• Anticipation
• Perseverance

Question 7

How common are speech errors?

Answer 7

One every 2,000 words.

Question 8

What are the two types of speech errors?

Answer 8

Word errors: not restricted by distance and are always of the same type (verb for verb, noun for noun). Happen early.

Sound errors: Close together and can cross word type. Happen later.

Question 9

What does Garrett’s model of speech production propose?

Answer 9

3 stages of speech production:

• Conceptualisation/message generation
• Formulation
• Articulation

3 stages of formulation:

• Functional level
• Positional level
• Sound level

Question 10

Describe each level of formulation according to Garrett’s model.

Answer 10

Functional level - lexical selection. Choosing the verb, noun, adjective, etc that you want to talk about.

Positional level - grammatical encoding, word order.

Sound level - encoding how the words are going to be articulated.

Question 11

Where do word errors occur in Garrett’s model?

Answer 11

Functional level of formation.

Question 12

Where do sound errors occur in Garrett’s model?

Answer 12

Sound level of formation.

Question 13

What is lexicalisation?

Answer 13

Lexical retrieval - finding the right word.

Question 14

What are the two stages of lexicalisation?

Answer 14

• Lemma: meaning and some syntactic information (e.g. noun or verb)
• Lexeme: form/phonology - sound.

Question 15

Describe evidence supporting the idea that lexicalisation has two distinct stages.

Answer 15

Wheeldon & Monsell, (1992); Monsell et al., (1992).

Found that prior exposure to words facilitated recall for related pictures of the words.

However, effect was only present if languages were the same and if the prior exposure was not for a homophone (i.e being exposed to boot, in the context of a car and then seeing a picture of a boot that you walk in)

Question 16

Repetition priming increases the strength between what?

Answer 16

Lemmas and lexemes

Question 17

Describe evidence from Schriefers et al., (1990) that supports the two stages of lexicalisation.

Answer 17

Picture-word interference task in which PPS had to name a picture while ignoring an auditorily presented target word (dog).

Found that semantically related words (fox) presented before the presentation of the picture inhibited recall speed.

Phonologically related words (doll) presented after the picture presentation facilitated recall.

Question 18

What is discrete lexicalisation?

Answer 18

A single item is selected based semantic information, then it’s phonological form is retrieved.

Question 19

How many types of interactive lexicalisation are there, and what are they?

Answer 19

2 - cascading & feedback

Question 20

What is cascading lexicalisation?

Answer 20

Activation flows to the form lexicon before a single lemma has been detected. Multiple word forms are activated.

Question 21

What is feedback lexicalisation?

Answer 21

Activation cascades down and then feeds back to the level above, resulting in lemma activation of related word forms.

Question 22

What did Levelt et al., (1991) investigate about cascaded priming?

Answer 22

Findings were against cascaded priming.

PPS had to name pictures, and completed a lexical decision task (LDT) between presentation and naming.

The words presented for the LDT were investigated to see if they had priming effects on later trials.

Question 23

What did Levelt et al., (1991) find about cascaded priming?

Answer 23

Priming:

• occurred semantically, at the lemma level (goat).
• occurred phonologically, at the lexeme level (sheet)
• did not occur with ‘goal’, suggesting that phonological activation of words semantically related to target words does not occur.

Question 24

What evidence supports cascaded priming?

Answer 24

Peterson and Savoy (1998) - mediated priming for near synonyms. Couch primes sofa, which primes soda.

Question 25

What is the lexical bias effect?

Answer 25

Sound level speech errors result in words more often than by chance.

Suggests that some feedback may occur, in that semantic systems check phonology before they are articulated.

Question 26

What does the output monitor supposedly do and how would it explain the lexical bias effect?

Answer 26

Monitor words to be articulated. We are more likely to miss errors in real words than errors involving pseudo words.