How is knowledge coded for language

What are the main steps in identifying a written word?

Fixate word → identify letters → encode letter order → compare with lexicon → retrieve meaning → compute pronunciation

Why must readers directly fixate on words?

High-acuity vision is limited to the fovea

What is the fovea?

The small retinal area responsible for high-acuity and colour vision

How does the fovea compare with the blind spot?

The fovea is about 10 times smaller

Why is speed reading mostly a myth?

The retina can only take in a limited amount of text per fixation

Why do long words often need more than one fixation?

They exceed the useful high-acuity visual span

Do we identify words mainly by whole-word shape?

No, we identify words through component letters

What does recognising an unusual visual version of “yacht” show?

Word recognition depends on identifying letters, not overall visual shape

What are abstract letter codes?

Letter representations that treat visually different forms like A and a as the same letter

What evidence supports abstract letter codes?

Masked priming between different cases, such as kiss/KISS or read/READ

What is masked priming?

A brief prime influences target recognition, often without conscious awareness

What indicates priming has occurred?

Faster responses in related than unrelated cognition

Why must the visual system encode letter order?

Words can contain the same letters in different orders such as DOG and GOD

Do only external letters matter for word recognition?

No, internal letter order also matters

Give an example showing internal letters matter

form vs from

What is context-dependent letter coding?

Letter identity is coded together with position

What is conjunctive coding

Binding a letter to its specific position in a word

Which model uses position-specific letter detectors

The Interactive Activation model

In the IA model, how are SALT and LAST distinguished?

They activate different position-specific detectors: S1 A2 L3 T4 vs L1 A2 S3 T4

What is a limitation of strict context-dependent coding?

It struggles to recognised familiar letter strings in novel contexts

What is context-independent letter coding?

Letters are represented independently of their positions

How can position be represented in context-independent coding?

By temporarily binding letters to positions in short-term memory

In context-independent coding, how are SALT and LAST represented initially?

By the same set of letter detectors

Why can context-independent coding help recognise CAT in TREECAT?

CATs letters can be activated even in a novel larger context

How can context-dependent and context-independent theories be distinguished?

By testing which words or primes are treated as similar

Under context-dependent coding, are form/from and form/film similarly similar?

Yes, because position-specific overlap is similar

Under context-independent coding, why are form and from more similar than form and film?

They contain the same letters, just transposed

What is transposed-letter priming?

Faster recognition when prime and target differ by swapped letters, e.g. talbe → TABLE

What is double-letter substitution priming?

A prime differs from target by replacing two letters, e.g. tacme → TABLE

What does context-independent coding predict for table-TABLE vs tacme-TABLE?

More priming for table-TABLE

What does context-dependent coding predict for table-TABLE vs tacme-TABLE?

Similar priming

What does evidence generally support?

More support for context-independent coding

How is Geon theory similar to letter coding

Objects are built from abstract parts, like words from abstract letters

In Geon theory, how are objects recognised?

Abstract parts are combined in short-term memory

What is the mental lexicon?

The mental dictionary of previously learned words

What happens after encoding the letter string?

It is compared with the lexicon and the best-matching word is selected

What role does competition play in word identification?

Similar words compete to be recognised

What is lateral inhibition?

Activated word candidates suppress competing candidates

In the IA model, how is a word activated?

To the extent that it contains the correct letters in the correct positions

Why might TRAP activate TRIP?

They share 3 out of 4 letters in the same positions

How can lexical competition be tested?

By comparing priming from word vs nonword primes

What is the prediction if words compete?

Nonword primes would produce more facilitation than word primes

Why should a nonword prime like “trup” facilitate TRAP?

It is similar to TRAP but does not strongly compete as a real word

Why might a word like “trip” inhibit TRAP?

It activates a real competing lexical entry

What is the baseline condition in this example?

An unrelated word prime like cart before TRAP

What is the prime lexicality effect?

Nonword similar primes facilitate recognition, while word similar primes can inhibit it

What does inhibitory priming strongly suggest?

Lexical competition

What happens after a word is selected from the lexicon?

Its meaning is retrieved

What are two ways orthographic information can activate meaning?

Cascaded processing and discrete processing

What is cascaded processing?

Later processing begins before earlier processing is complete

What is discrete processing?

Processing at one level finishes before information passes to the next level

In cascaded processing, can losing word candidates briefly activate meaning?

yes

In discrete processing, what activates semantic processing?

Only the winning word candidate

How can cascaded processing be tested?

By testing whether embedded words activate meanings

Example of an embedded word in CATCH?

CAT

Why might “Is CATCH an animal”? be slower than is “Is CATCH a body part”?

CAT is an animal, causing semantic conflict

What does slower response to “Is CATCH an animal” suggest?

The embedded word CAT briefly activated its meaning

What is an interference effect?

Extra reaction time caused by conflicting semantic activation

Does cascaded processing also occur in spoken word identification?

Yes

Does cascaded processing also occur in object identification?

Yes

What is the final step in word identification for reading aloud?

Compute pronunciation

English words are often divided into what two types?

Regular and irregular

What is a regular word?

A word whose spelling-to-sound mapping follows common rules

What is an irregular word?

A word whose pronunciation does not follow regular spelling rules

What does the Dual Route model claim?

There are two ways to read aloud: lexical and sub-lexical routes

What is the lexical route?

Reading via stored word knowledge in the lexicon

What is the sub-lexical route?

Reading by spelling-to-sound conversion rules

Which route is needed for irregular words?

Lexical route

Why must irregular words use the lexical route?

Rule-based pronunciation would give the wrong sound?

Which route is needed for nonwords?

Sub-lexical route

Why must non-words use the sub-lexical route?

They are not stored in the lexicon

Which route can regular words use?

Either lexical or sub-lexical route