Psychology of Language

Orthographic

changing one letter

Lexical approach

you see word and it goes through orthographic input lexicon - contains memory of written and learnt words

sub-lexical approach

use grapheme phoneme conversion - bypasses the lexical system and uses your knowledge of phonemes and graphemes to decode word

context effects

memory is aided by being in the physical location where encoding took place - influence of environment

Inter-stimulus interval (ISI)

time between the offset of the prime and the onset of the target

stimulus onset asynchrony (SOA)

the time between the onset of the prime and the onset of the target

What are phonological neighbours?

words that can be created by changing one phoneme of a word (e.g., gate, geit)

sentence-level context effects

processing words is easier when they appear in a semantically congruent sentence

What is an example of a congruent sentence?

to see the pandas, mike went to the zoo

What is an example of a incongruent sentence?

to see the popstars, mike went to the zoo

the magic moment

Baltoa (1990) - point in time when person recognises a word and accesses its meaning

what is a common assumption among linguists

the longer the time taken to process a written word = the more difficult the word is to process

reading techniques (with no secondary task)

Measure how long people usually spend looking at a word when reading (by eye-tracking or subject-paced).

Naming tasks

measure how long people take to start saying a word (Reading aloud; picture naming).

Lexical decision tasks

Measure how long people take to decide whether or not a string of letters is a word.

Semantic categorisation tasks

measure how long people take to categorise a word ("cat" is an animal).

What happened in Reicher's (1969) study into parallel processing in visual sensory memory?

Participants were to identify a single letter that appeared in one of four possible locations. The letter appeared either alone, in a four-letter word, or in a four-letter non-word. All stimuli were masked, making it difficult to identify them.

According to Reicher (1969), if we process the letters in a word in a serial fashion (one letter at a time)

it should be more difficult to identify a letter in a word than that letter by itself.

According to Reicher (1969), if we process the letters in a word in parallel (all the letters at once),

it should be just as easy to identify a letter in a word as it is to identify that letter by itself.

In Reicher's (1969) study, the non word condition allowed him to see what?

there was something unique about processing a letter in a word vs. that letter in a mixed group of letters.

What did Reicher (1969) find in his word superiority effect study?

Participants recognised those in the word condition and non-word conditions more than in the isolated condition

One popular attempt to explain the Word Superiority Effect is

the interactive activation model of McClelland & Rumelhart (1981).

What is the Interactive Activation model?

They suggest that both feature processing and word processing provide additional support to letter processing. Therefore, when a letter appears in a word, the letter is recognized more easily than when the letter appears alone or in a nonword.

Word superiority effect shows

we don't recognise words by processing each letter in turn

Syllables are

units of organisation of a sequence of speech sounds

Do syllables have a functional role in visual word recognition?

Prinzmetal et al (1986) said syllables defined by purely phonological criteria did not affect feature intergration

What did Prinzmetal et al (1986) do in their study?

Tested a method for determining the units of analysis used by the visual system in word perception and explored the nature of these units in 5 experiments, using 60 university students.

Procedure of Prinzmetal et al (1986) study

Participants were asked to report the colour of the letter in the next word, for example: report colour of 'D' --> VODKA

What were the results of Prinzmetal et al's (1986) study?

When syllables were defined by orthographic constraints or by morphological boundaries, they were functional units in the visual analysis of words and wordlike stimuli.

morphemes are the

smallest units of language that have meaning

Derivational morphology

derivational morphemes change the meanings of words by applying derivations (dark + ness = darkness)

inflectional morphology

The alteration of words to make new grammatical forms

inflectional morphemes don't change the meanings of words by applying inflections (cat + s = cats)

Compounding

multiple words are connected to make new compound words (tea + pot = teapot).

What happened in Taft & Foster's (1975, 1976) lexical decision task?

Readers first decompose morphologically-complex words into morphemes. Then, they access the entry for the root morpheme in the lexicon (decompositional account).

What were the results of Taft and Foster's (1975, 1976) study

Rejecting nonwords with real words in them (e.g., 'dustworth') took longer than nonwords with no real words (e.g., 'mowdflisk').

Readers recognise the real words embedded in non-words.

Also provides support for morphological decomposition in visual word recognition.

What happened in Marslen-Wilson et al's (1994) cross-modal priming task (with lexical decision)?

Questioned how derivational morphemes are represented in the lexicon and investigated the lexical entry for morphologically complex words in English.

Aim of Marslen-Wilson et al (1994)'s study?

To find out whether the lexical entry for derivationally suffixed and prefixed words is morphologically structured and how this relates to the semantic and phonological transparency of the surface relationship between stem and affix.

Procedure of Marslen-Wilson (1994) study

Conditions: relationship between prime and target;

(1) morphologically & semantically related: happiness - happy - 'ness' - a morpheme to make a noun

(2) morphologically related, but semantically unrelated: apartment - apart - 'ment' - a morpheme to make a noun

(3) morphologically & semantically unrelated: tinsel - tin - 'sel' - not a morpheme

All conds had a control cond with an unrelated prime

Results of Marslen wilson et al (1994) study

Results: Priming effect (difference between expt and control) - (1) > (2) = (3).

Only 'semantically-transparent' morphemes are decomposed.

What are frequency effects?

Commonly used words are easier to recognise and are responded to more quickly than less commonly used words.

What did Whaley (1978) show regarding the lexical decision task?

frequency is the single most important factor in determining the speed of responding in the lexical decision task

What are neighbourhood effects?

Some words have a large number of other words that look like them (e.g., "mine" has "pine", "line", "mane", among others), whereas other words of similar frequency have few that look like them.

Andrews (1997) concluded that

neighbourhood SIZE has more effect than neighbourhood FREQUENCY

Orthographic neighbourhood effects: Size

• Coleheart et al (1977): lexical decision task

• Andrews (1989): lexical decision task

Orthographic neighbourhood effects: Frequency

• Grainger et al (1989): lexical decision task

• Grainger (1990): lexical decision and naming task

• Sears, Hino, & Lupker (1995): Lexical decision task & naming task

Phonological neighbourhood effects: Size

• Yates, Locker & Simpson (2004): Lexical decision task

Concreteness/Imageability Effects

High-imageable words are easier to process than low-imageable words.

Semantic effects

• Lexical Decision Tasks (James, 1975; de Groot, 1989; Kroll & Merves, 1986)

• Naming (Bleasdale, 1987; Balota, Cortese, Sergent-Marshall, Spieler & Yap, 2004; Strain, Patterson, & Seidenberg, 1995)

• Strain et al. (1995): Imageability interacts with word frequency and spelling-to-sound consistency in naming: low-frequency words with inconsistent spelling-to-sound mapping produced the largest imageability effects: e.g., suave (low imageability, low frequency, irregular spelling) vs. sword (high imageability, low frequency, irregular spelling)

What are priming effects?

presenting material before the word to which a response has to be made.

What is the most common paradigm of priming effects?

involves presenting one word prior to the target word to which a response (such as naming or lexical decision) has to be made.

What is the stimulus-onset asyncrhony (SOA)?

The time between when the prime is first presented (its onset) and the start of the target

What can we learn about manipulating the relation between the prime and target words, and by varying the SOA?

we can learn a great deal about visual word recognition

What is the basic priming paradigm in orthographic priming effects?

the relationship between prime and target will be manipulated. Usually, there is an experimental condition (prime-target related) and a control (baseline) condition (primte-target unrelated).

What happens in masked priming during orthographic priming effects?

mask usually either forward or backwards, prime usually presented briefly (e.g., 50ms).

An advantage of this is that the prime will become almost invisible, and subjects will be usually unconscious about the prime. Thus, it can prevent subjects to have a strategy about the primes.

What has been found using masked priming in orthographic priming?

Priming effects were found for repetition priming and orthographic priming vs. the baseline

In lexical decision, larger repetition priming effects were found for word repetitionthan non-word repetition, although non-word repetition still elicited priming effects compared with the baseline.

A liner relationship was found between the duration of the prime presentation and the magnitude of the priming effect

What did Evett and Hymphreys (1981) do in regards to phonological priming effects?

They used masked priming with word recognition tasks.

Meyer and Schvaneveldt (1971) provided a demonstration of what is one of the most robust and important findings about word recognition. What did they show?

the identification of a word is made easier if it is immediately preceded by a word related in meaning.

How did Meyer & Schvaneveldt (1971) use to show that the identification of a word is made easier if it is immedately preceded by a word related in meaning?

Used a lexical decision task - but the effect can be found with different magnitudes of effect, across many tasks and is not limited to visual word recognition.

For example, we were faster to say that "doctor" is a word if it is preceded by the word "nurse" than if it was preceded by a word unrelated in meaning, like "butter", or presented in isolation.

The first word might speed up recognition of the second word, in which case we talk of

facilitation

Sometimes the prime slows down the identification of the target, in which case we talk of

inhibition

What were the results of Meyer & Schvaneveldt's (1971) lexical decision task?

RTs were shorter when the prime and target were semantically related (e.g., doctor - NURSE) than they didn't (e.g., bread - NURSE).

--> Classic semantic priming effects

Possible mechanisms of semantic priming?

Semantic feature overlaps: Features shared by both words are activated.

Word associations: Within the lexical network, semantically related words are often closely connected, and activations spread.

Two words are 'associated' if one is

produced in relation to the other in a word association task.

Semantically related words are often also associated

some associated word pairs are not necessarily semantically related (e.g., spider - web) Also, some semantically-related pairs are not necessarily associated (e.g., dolphin - whale)

Associative vs semantic priming effects

• Lupker (1984): Naming task

• Shelton & Martin (1992): Lexical decision task

What are sentence-level context effects?

Processing words is easier when they appear in a semantically congruent sentence.

What were the three different conditions in Hess, Foss and Carroll's (1995) naming task?

1) globally related, but locally unrelated (discourse-level context)

2) globally unrelated, but locally related (sentence-level context)

3) control

What were the results of Hess, Foss and Carroll's (1995) naming task?

o Results: Naming latencies: (1) < (2) = (3) Discourse-level influence on word recognition.

purpose of IAC model

was to account for word context effects on letter identification

What kind of model is the interactive activation and competition (IAC) model?

A connectionist

There is an input level of _______, a level where units _________ and an output level ____________.

visual feature units, correspond to individual letters, where each unit corresponds to a word

each unit is connected to...

each unit in the level immediately before and after it

each unit connection is either

excitatory (positive) if appropriate and inhibitory (negative) if inappropriate

Example of excitatory connection

the letter "T" would excite the word units "TAKE" and "TASK" in the level above it but would inhibit "CAKE" and "CASK".

Excitatory connections made the destination units ______ while inhibitory connections make them _______

more active, less active

when looking at interactions WITHIN the units

each node in the same unit is connected with each other and all connections are inhibitory

when looking at interactions ACROSS the units

each unit is connected to the one immediately before and after it and each connection is either excitatory or inhibitory

There is often no ______ between spoken words - in written words you have one between words.

'break' (pause)

assimilation: phonemes can be influenced by

some acoustic properties of their neighbours (due to co-articulation)

speaker variability: no two speakers

produce the same phonemes in the same way due to differences in shape and size of vocal tracts and accents etc

what is the segmentation problem?

words are often uttered without a clear break between them

what are segmentation strategies?

possible word constraint and metrical segmentation strategy

what is possible word constraint strategy?

Speech is segmented in such a way that there is no syllable left unattached to words.

what is metrical segmentation strategy?

Subjects inserted word boundaries before stressed syllables and eliminated weak ones (stress-based segmentation)

What is voice onset time?

The delay between the release of the contraction of the airstream and the start of the vocal cord vibration

What is the McGurk effect?

a perceptual phenomenon that demonstrates an interaction between hearing and vision in speech perception.

When does the McGurk effect occur?

when the auditory component of one sound is paired with the visual component of another sound, leading to the perception of a third sound

What is semantics?

the study of meaning in language

What is syntax?

The study of word order

What is morphology?

The study of words and word formation

What are pragmatics?

The study of language use

What is phonology?

The study of how sounds are used within a language

Why do we know whether a sentence is grammatical or ungrammatical even though we haven't heard it before?

This is possible because sentences have structure

An important aspect of language is that we can

construct sentences by combining words according to rules

Sentences with the same words

can have different meanings

What does sentences with the same words which have different meanings show us about sentence processing?

It is more than just combining the meaning of words - we must process the sentence structure too

Some sentences are __________ even though they dont make sense

Grammatical

Sentences which do not make sense but are grammatical show us what?

Sentences consist of more than just the meaning of words

Syntax is

independent of meaning (semantics)

Some sentences have more than one interpretation because

there are two different structures - meaning could be dependent on syntax