Models of spoken word identification

Question 1

What is a scientific model?

Answer 1

A representation (diagram/verbal description, computer program etc.) of an empirical phenomenon that captures the fundamental characteristics of that phenomenon. A model simplifies and substitutes for the phenomenon.

Question 2

Why do we need cognitive models?

Answer 2

Cog models needed to
• See if it is a good fit of the real phenomenon (accurately predicts empirical data collected from observations and experiments)
• Provide evidence about the validity of the assumptions about the structure and organisation of the mind instantiated in the model

Question 3

What is the Logogen Model (Morton, 1970)?

Answer 3

• The same lexical (logogen) system is accessed by both spoken and written words
• Bidirectional links between the logogen and cog system
• Separate logogens (evidence-collecting devices) for each word are activated in parallel according to match with input - Simultaneously collects evidence from words accessed outside, to see whether or not current knowledge matches with the input
• Evidence can come from visual, auditory or cognitive sources
• When a logogen’s evidence exceeds a threshold it fires –> word is identified

Question 4

According to the Logogen model, how does frequency and context affect word identification?

Answer 4

Frequency: Higher frequency = lower threshold –> quicker word identification
- Threshold permanently lowered every time the logogen is activated

Context: Provides environmental evidence for a word –> quicker identification
- Easier to identify a word in context of a sentence than in isolation

Question 5

What are the problems with the Logogen model?

Answer 5

Does not specify how the logogens accumulate the evidence
Does not show how the system ensures that the correct word is first to reach threshold
The system may never be able to recognise low freq words if context is similar to a high freq word

Question 6

What does the Cohort Model suggest about how spoken word are identified?

Answer 6

Speech signal is processed sequentially
Early in the presentation of a word, all the words that start with presented sounds are activated - “Word initial cohort”
As more perceptual evidence becomes available, words are eliminated from the cohort until only one left –> successful word identification