Models of linguistic planning

Question 1

SERIAL MODELS

Garrett’s model

Three level of processing:

Answer 1

Three levels of processing:
- Message level: conceptualization
- Sentence level: formulation process (lemma)
-Functional level: syntactic-semantic representation
-Position level: phonological representation
- Articulatory or motor level: articulatory processes

Question 2

SERIAL MODELS

Garrett’s model

Language processing happens in successive stages:

Answer 2

1. Message planning: conceptual content
2. Selection of lemmas and assignment to slots in the syntactic frames
3. retrieval of word (or morpheme) forms and assignment of sounds
4. The specified sound representation leads to a motor plan for articulation

Question 3

SERIAL MODELS

Formulating messages:
Experiment from Zenzi Griffin and Kay Bock

Answer 3

All groups showed same eye movement patterns, looked at both of the characters and settled on the character that eventually appeared as the subject of the sentence, then they moved to the other.

Conclusion: eye movements provide strong evidence that we analyze the full scene and identify the relationship between characters before beginning to speak.
Well organized plan appeared to be in place before word retrieval began.

Question 4

SERIAL MODELS

Formulating messages:
When message planning and word retrieval overlap
Experiment: Gleitman

Answer 4

Cross at centre of the screen. Black square appears quickly on top of one character. After, image appears.

When attention was drawn to dog in scene: dog is following man
When attention drawn to man: man was running away from the dog

The harder the message planning, the more likely that word retrieval or even speaking will begin before you finish formulating your message.

Question 5

SERIAL MODELS

Structuring sentences:

Answer 5

• Choice of a particular syntactic structure might be guided by subtle shadings of meaning or emphasis
• Passives put more emphasis on the end state of and action than on the action itself
• Easiest syntax in response to pressures of time and memory:
  -Time pressure
  -Accessibility: use one we have just used
  -Memory pressures

Question 6

SERIAL MODELS

Putting the sounds in words

Answer 6

sound-based speech errors show lexical bias, errors that result on real words are statistically waiting to happen

Question 7

SERIAL MODEL

Serial model of language production

Answer 7

• Activation flow in one direction, from lemma to sound
• Unidirectional links between the lemma and sound levels.
• Activated sounds don’t feed back to the lemma level

Question 8

CASCADE MODEL OF WORD PRODUCTION

Answer 8

• Later stages of word production don’t need to wait until earlier ones have been fully resolved, but can be initiated while earlier stages are still in progress
• Bidirectional links between the lema and sound levels, which allows activation from the sound level to flow back up to influence the activation of lemma candidates