Cognitive Psychology: Language

Question 1

Three things we need to understand a piece of text

Answer 1

1. We must perceive and identify individual words
2. Process the word in context of the sentence and construct the meaning of it
3. Integrate this sentence within its context

Question 2

What is eye-tracking?

Answer 2

measure how long people actually spend looking at a word when reading (camera monitors your eye movements, longer = more difficult to recognise word)

Question 3

What is a lexical decision task?

Answer 3

measure how long people take to indicate that a strong of letters is a word (longer to say yes = more difficult to recognise word, often used in conjunction with priming)

Question 4

What is a naming task?

Answer 4

measure how long people take to start saying a word (pronounce it as quickly as possible without stuttering or mispronouncing, longer = difficult to recognise word)

Question 5

What are three factors affecting word recognition?

Answer 5

1. Word frequency: commonly used words are recognised more easily than infrequent words (high frequency = teacher, low frequency = armadillo, it takes longer to recognise low frequency words)
2. Predictability: predictable words are recognised more easily than those in neutral or misleading contexts (pp’s read an incomplete sentence and then attempt to recognise a single word)
3. Neighbourhood effects: word identification can be speeded when similar words exist in the language

Question 6

What is the difference between orthographic and phonological neighbourhood effects?

Answer 6

• Orthographic neighbourhood: the number of words that can be formed by changing one letter while maintain letter position ‘Tank, task, rank’
• Phonological neighbourhood: the number of words that can be formed by changing one phoneme of a word ‘gait, bait, get’

Question 7

Morton’s Logogen Model (1969,1979)

Answer 7

• A logogen is a collector of evidence
• When enough evidence is collected, then the threshold is reached, the logogen fires, and the word is recognised
• High frequency words have a lower threshold for firing, so low frequency words take longer (i.e. cat vs cot)
• The cognitive component of the logogen model explains how sentence context can affect recognition and the semantic info from the sentence activates it, lowering their threshold (therefore decrease the amount of info needed to fire the logogen)

Logogens can be thought of as as “word detectors”.

Each logogen has an activation threshold which needs to be met before it fires.

Question 8

Word superiority effect (Task, Result & Conclusion)

Answer 8

• Task: stimulus – mask – forced choice
• Result: 10% improvement in performance with the whole word compared to a single letter
• Conclusion: its easier to identify a letter in the context of a word than in isolation

Question 9

Interactive activation model: McClelland & Rumelhart (1981,1982)

Answer 9

used to explain the word superiority effect
1. Contains feature detectors, which can detect letter features, such as vertical and horizontal lines
2. Letters that contain the activated feature would also become activated
3. The output of the letter detectors would then activate word level detectors which contained these letters
- Types of connections: excitatory and inhibitory
- Connections run in both directions: so that the network tends to evolve towards a state of activation in which everything is consistent
- Cannot account for hie we can recognise words that have been mis-spelled (transposed letter priming = jugde instead of judge)
- These findings can be accounted for in recent models in which letter position is not fixed – Spatial coding model (Davis 2010)

Question 10

Dual-route model: Coltheart et al (2001)

Answer 10

• Direct route: connects the visually presented word to the whole word’s mental representation, used for high-frequency or familiar words
• Phonological route: accesses the mental representations of words by using grapheme-to-phoneme conversion rules, used for reading low-frequency and non-words

Question 11

Difference between phonological and surface dyslexia?

Answer 11

Phonological dyslexia: Difficulty with reading non-words.
Phonological dyslexia assumes a selective deficit in developing the phonological route.

Surface dyslexia: Problems reading irregular words
When presented with an irregular word, readers use the lexical route.
Surface dyslexia assumes a selective deficit in the lexical route.
Results in difficulty in pronouncing irregular words (i.e., ‘pint’/‘have’).

Question 12

What is a syntax?

Answer 12

• the words in a sentence aren’t just strung together one after the other, they are structured into phrases and clauses, can be depicted in tree diagrams
• Each sentenced can be broken down to smaller constituents (nodes), which are connected via branches
• The noun and verb phrases can then be combined to make a sentence
• Syntax refers to the way in which words in a sentence are grouped together

Question 13

What is syntactic ambiguity?

Answer 13

• Where a sentence may have more than one interpretation, given the potential grammatical function of the individual words
• Researchers often investigate how people process ambiguous sentences in order to develop theories of syntactic parsing

Question 14

What are two types of syntactic ambiguity?

Answer 14

• 1. Global ambiguity: ones that remain ambiguous even when you get to the end of the sentence (the spy observed the politician with binoculars)
• 2. Temporal ambiguity: starts of being ambiguous but able to be resolved when you reach the end of the sentence (while anna dressed the baby threw up)

Question 15

Theories of parsing: Garden Path theory (serial model) and what is parsing?

Answer 15

The term ‘parsing’ relates to the analysis of the syntactic or grammatical structure of a sentence. So, if we ‘parse’ a sentence, we are assigning syntactic structure to it.

Only one syntactic structure is initially considered, and the sentence meaning is not involved in the selection of this structure. If the simplest structure is incorrect, then then sentence meaning can influence re-analysis. The simplest structure is chosen, following:
a) Minimal attachment – we would choose to build the tree that has the fewest nodes/ branches
b) Late closure – sometimes both trees have the same number of nodes, so we should interpret the incoming material that was most recent

Question 16

Theories of parsing: Constraint satisfaction (Parallel model)

Answer 16

All relevant sources of information are immediately available to the parser. The initial interpretation of a sentence depends on multiple sources of information: context, plausibility, general world knowledge, verb bias. The different sources of information are known as constraints. The syntactic structure receiving the most support is chosen

Question 17

Sentence meanings and what is an idiom?

Answer 17

• Non-literal means we cannot deprive a meaning by direct composition of the literal meaning of the words as guided by the grammar (metaphors, irony, idiom)
• Idiom: group of words in a fixed order that have a particular meaning that is different from the meaning of each word (kick the bucket = to die)

Question 18

Theories of figurative language processing: Standard pragmatic view

Answer 18

use irony to communicate the opposite of what is said, and the literal meaning is assed first, then reanalysing it as being ironic with the context (serial model)

Question 19

theories of figurative language processing: Graded salience hypothesis

Answer 19

makes different predictions depending on if the ironic comment is familiar or not in the mental lexicon. So, there would be a processing cost for unfamiliar ironies whilst the reanalysis takes place

Question 20

Theories of figurative language processing: Direct access view

Answer 20

literal and ironic language are essentially treated in the same way and so the ironic meaning can be accessed without accessing the literal meaning first. So no additional processing cost for ironic language

Question 21

Context and world knowledge: Hagoort et al (2004)

Answer 21

• Context play a crucial role in determining the meaning of a sentence
• ERP’s (even-related brain potentials) measures the electrical activity when the participant is processing the sentence given to them
• The results suggests that we do a lot of this processing with 400 msec of encountering a word

Question 22

Shallow processing: The Moses illusion

Answer 22

• Moses fit in with the biblical context so we didn’t fully process the sentence (e.g. “how many animals did Moses fit on the ark? But is was Noah!!!!)

Question 23

What are three types of inferences?

Answer 23

• Logical inferences: depend only on the meaning of words (e.g. we can infer that anyone who is a widow is a female)
• Bridging inferences: establish coherence between the current part of the text and the preceding text (e.g. if you have two sentences such as ‘Mary poured the water on the bonfire. The fire went out’, the bridging inference ‘water extinguishes fire’ allows us to connect these two sentences together)
• Elaborative inferences: embellish or add details to the text by making use of our world knowledge Elaborative inferences are, in a sense, optional. That is, they are not required to establish coherence.

Question 24

What is the constructionist approach? and what are some supports and against

Answer 24

• Readers typically construct a relatively complete ‘mental model’ of the situation and events referred to in the text
• Support: Showing two slightly different sentences with the second one adding the inference and the pp’s think both were the same (Bransford et al 1972)
• Against: pp’s were read s story about a dictator called Gerald and a week later they were given a test on memory on Hitler. Pp’s mistakenly ‘recognised’ sentences relevant to Hitler that had not appeared in the original story (Dooling et al 1977)

Question 25

What is the minimalist hypothesis? Macon & Ratcliff (1992)

Answer 25

• Inferences are either automatic or strategic (goal directed)
• Some automatic inferences establish local coherence, and others rely on info explicitly stated in the text
• Strategic inferences are formed in pursuit of the reader’s goals (the number of inferences you make depends on what your goals are when you are reading the text)
• Most elaborative inferences are made at recall

Question 26

Support and against for minimalist hypothesis

Answer 26

• Support: G1 instructed to read for comprehension and G2 were explicitly told to try and anticipate what might happen next. G2 were able to draw more elaborative inferences than G1 and therefore it demonstrates thar the number if inferences are influenced on whether or not they have a goal in mind when they are reading the text (Calvo et al 2006)
• Against: Pp’s took longer to read a sentence describing inconsistent action than one describing consistent action, regardless of if the goal was explicit or implicit (Poyner et al 2003)

Question 27

Event-indexing model: Readers keep track of several dimensions

Answer 27

1. The protagonist – the central character in the present event compared to the previous one
2. Temporality – the relationship between the times at which the present and past events occur
3. Causality – the casual relationship between the current and past event
4. Spatiality – the relationship between the spatial setting of the current and past event
5. Intentionally – the relationship between the character’s goals and the present events

Question 28

Strengths and weaknesses of the event-indexing model

Answer 28

• Support: Pp’s were described a series of events (E1, E2, E3, E4) either in chronological or non-chronological order (flashback wither 5 mins or 3 hours ago) Findings: the duration of the second event influenced the speed in which the first event could be assessed. Conclusion: readers put the four events in the correct chronological order
• Strengths: identifies key processes involved in creating and updating situation models
• Weaknesses: has little to say about the nature of the internal representation that is built by readers and listeners to understand language

Question 29

What are schemas?

Answer 29

• The schemas stored in long term memory include scripts (deal with knowledge about events and consequences of events) and frames (knowledge structures relating to some aspect of the world and consist of structed information and have slots for variable information)

Question 30

What is the experimental simulations approach?

Answer 30

• Proposes that modal simulations, bodily states, and situated action underlie cognition
• When applied to language: the same sensorimotor representations that are activated while interacting with the environment are re-enacted or ‘stimulated’ when reading about a similar experience
• Action: ‘turn down the volume’ can facilitate certain bodily actions
• Perception: sentences may evoke a perceptual stimulation
• Evidence from neuroimaging: (speer 2009) brain areas activated during word-by-word presentation of stories corresponded to those activated by performing the action that is described
• Affect: (Niedenthal 2007) showed that emotion words generated emotion-specific facial activation, as measured by examining the activation. Of certain facial muscles. (Havas 2007) showed that facial posture influences the amount of time the participants took to judge whether the sentence was positive or negative

Question 31

Strengths and weaknesses of the experimental simulations approach

Answer 31

• Strengths: framework for explaining the nature of the internal representation created by readers and listeners in order to understand language
• Weaknesses: it is not known whether the experiential simulations approach applies under ALL circumstances