The Internal Lexicon

李如婷 李 (Cynthia)
白班晢明 (Benjamin)

Introduction

● Internal Lexicon:
○ Representation of words in permanent memory.
○ When the word is found, the properties we
associate with the word become available for use.

● Lexical Access:
○ The process by which we activate the meanings of
a word.
○ Example of "Elephant" (pp. 103-4)
Dimensions of Word Knowledge

● What does it mean to know a word?

● Phonological Knowledge:
○ Pronunciation.
○ Homophones (e.g., ate/eight).
○ Tip of Tongue Phenomenon (TOT) (p. 104).

TOT

● We can't retrieve the right word.

● We can remember how it sounds.
● Retrieve and reject similar sounding words.
● Example:
○ Sextant
○ Secant
Dimensions of Word Knowledge

● Syntactic Knowledge:
○ Syntactic category (part of speech).
○ Content words (open-class) vs. Function words
(closed-class).

● Agrammatism: (\(ˈ)ā-ˈgram-ə-ˌtiz-əm\)
○ Disorder.
○ Omit closed-class words, but maintain open-class
words.
○ Shows distinction in our internal lexicon between
words.

Dimensions of Word Knowledge

● Morphological Knowledge:
○ Free morpheme vs. bound morphemes.
○ Inflectional morpheme vs. Derivational morpheme

● Semantic Knowledge:
○ Sense and reference.
 ○ Synonyms.
○ Coordination.
○ Hypernymy vs. Hyponymy.
○ Meronymy.

The Word Association Test

● Used to find the sense relations.

● Idea:
○ Say a word out loud to someone.
○ That person says the first word that comes to mind.

● Four types of semantic relations (p.108):

○ Taxonomic relations.
○ Meronyms.
○ Attributive relations.
○ Functional relations.

Dimensions of Word Knowledge

● Denotation:
○ Objective or dictionary meaning of a word.
○ Six parts of the definition can be found. (p. 108):
(a) Phonological (b) Orthographic (spelling)
(c) Syntactic (d) Semantic
(e) Morphological (f) Other (etymology)

● Connotation:
○ Certain aspects of meaning beyond that which it
explicitly names or describes.
○ Example: bachelor vs. spinster. (p. 110)

Organization of the Internal Lexicon

● Concept of Semantic Network:

○ Interconnected elements (nodes).
○ Connected by relationships between one another.
● Models:
○ Hierarchical Network Model.
○ Spreading Activation Model.
● Hierarchical Network Models (p. 111):
○ Nodes stand above/below other nodes.
○ Appears to be a waste of memory space to store
information that is highly unlikely to ever be used.
○ 歇洛克 福爾摩斯
· (Sherlock Holmes).

"You see," he explained, "I consider that a man's brain

originally is like a little empty attic, and you have to
stock it with such furniture as you choose. A fool takes
in all the lumber of every sort that he comes across, so
that the knowledge which might be useful to him gets
crowded out, or at best is jumbled up with a lot of other
things so that he has a difficulty in laying his hands
upon it. Now the skillful workman is very careful indeed
as to what he takes into his brain-attic. He will have
nothing but the tools which may help him in doing his
work, but of these he has a large assortment, and all in
the most perfect order. It is a mistake to think that that
little room has elastic walls and can distend to any
extent. Depend upon it there comes a time when for
every addition of knowledge you forget something that
you knew before. It is of the highest importance,
therefore, not to have useless facts elbowing out the
useful ones.

Organization of the Internal Lexicon

● Cognitive economy principle:

○ Space available for the storing of semantic
information was limited.
○ Information is only stored at one place and that
place is the highest possible node.
● Semantic verification task (Table 5.2 p.
113).
● Category-size effect:
○ An A is a B An A has a B
○ The higher B (related to A) in hierarchy the longer
th ti t T F

Organization of the Internal Lexicon

 ● Typicality effect:
○ Items that are more typical of a given subordinate
take less time to verify
○ Opposite for false statements. (e.g., p. 114; #9-12).

● Basic-level terms:
○ level in hierarchy were the most distinguishing
features are assigned.
○ Terms that children learn first:
■ Cat/Dog
■ Mom/Dad

Organization of the Internal Lexicon

● Spreading Activation Models (p. 115):

○ Retain the network, but do not follow strict
hierarchical structure.
○ The organization is closer to a web of
interconnecting nodes.
○ Retrieval process:
■ Spreading activation.
■ Closer related concepts are more likely to be
activated than distant ones.

Bock & Levelt 3 level model (p. 117)

 ● Conceptual Level:
○ Top level.
○ Concepts.
● Lemma Level:
○ Middle level.
○ Syntactic aspects of word knowledge.
● Lexeme (Sound) Level:
○ Bottom level.
○ Phonological properties of a word.

Lexical Access

Search Model (p. 118):

● Lexicon is independent of other language
processing systems.
● Divisions:
○ Orthographic (spelling) Properties.
○ Phonetic Properties.
○ NOT directly influenced by syntactic or semantic
factors.
● Higher frequency words are searched
before lower-frequency words.
Lexical Access

Logogen Model (p. 119):

● Logogen = Morpheme's various attributes:
○ Semantic, Phonology, etc.
● Logogen activated:
○ Sensory input (spelling, sound).
○ Contextual information (sentence structure).
● Logogen is like a scoreboard counter:
Sensory & Contextual increase the same
counter.

Lexical Access

Cohort Model:
● Auditory word recognition.
○ Humans recognize spoken words rapidly.
○ Listeners are sensitive to the recognition point of a
word.

● Word recognition occurs in three stages.

Stages of Word Recognition

● Acoustic-phonetic analysis:
○ Bottom-up.
○ Many possible candidates.
○ Word-initial cohort.
● One member of the cohort is selected for
further analysis.
○ Sensitive to multiple sources of information.
○ Initial candidates are progressively eliminated:
■ Contexts of a spoken sentence.
■ More phonological information comes in.

Stages of Word Recognition

● The selected lexical item is integrated into

the ongoing semantic and syntactic context.

● Cohort model captures the best of the

search and logogen models.
Variables That Influence Lexical
Access
● Word Frequency:
○ Phoneme monitoring (p. 120).
○ Lexical decision task (Table 5.3 p. 121).
● Phonological Variables (Sounds):
○ Stress and intonation.
○ Substitution of similar sounding words.
● Syntactic Category:
○ Open-class words (word frequency effect)
○ Closed-class words (no word frequency effect)

Variables That Influence Lexical

Access
● Morphological Complexity:
○ Independent vs. Combined Storage.
○ Govern + ment vs. Government

○ Suffix: Easier
■ -ment (Government).
■ -ence (Existence)
 ■ -ion (Decision) Harder

Variables That Influence Lexical

Access
● Semantic priming:
○ Word presented earlier activates another,
semantically related word.

● Two phrases:
○ (1) Priming stimulus is present:
■ No response to the prime is required or
recorded.
■ Little interest.
○ (2) Second stimulus (the target) is presented:
■ Some response to it and time taken to make this
response is recorded.

Lexical Ambiguity

● Single word with more than one meaning.

● Significant property of language.
○ Vital that any model of language processing must
deal with these words.
● Ambiguity:
○ Spoken (righting vs. writing)
 ○ Written (letter = mail or ABC)
● Even with strong prior contexts we active
the multiple meanings of Ambiguous words.
○ Logogen threshold helps to figure out the correct
meaning of the word.

Lexical Ambiguity (p. 124)

● Do ambiguous words have more than one

representation in the lexicon?

● Do we consider multiple meanings of

ambiguous words when we hear or see one?

● How might the sentence context influence

how lexically ambiguous words are
processed?

Appraising Models of Lexical Access

● All models deal with:

● Word Frequency Effect:
○ Search: Higher the frequency = Higher the node.
○ Logogen: Higher the frequency = Lower the threshold.
 ○ Cohort: Higher the frequency = Higher probability.
● Semantic Priming.
○ Search: Each word generates a list of words that might
come next.
○ Logogen: Lowering the threshold that are related to a
prime.

Why Cohort Model is Best

● It is better position to explain the full range

of factors that influence lexical access.

● It is more explicit about the time course of

spoken word recognition.

● Better able to explain how sounds in

different positions within the word may affect
recognition.

Jeopardy!

● American game show.

● Answers are in the form of a question:
 CLUE:
"It is a popular street food in Taiwan with a
strong smell."

"What is stinky tofu?"

http://www.youtube.com/watch?v=vCVfWjlMSa4

Discussion

● This is an example of lexical ambiguity.

 Question:

What caused the man to give the

wrong answer?

Discussion

Hoe = Garden tool.

Ho = Slang for prostitute.

Rake = Garden Tool. (high frequency)

Rake = A man who leads a wild life with women
and alcohol. ((very) low frequency)
Thank You
Questions
&
Comments