Week 8 Lecture 8 - the speaking brain 1 Flashcards
What studies show that animals have language?
Washoe:
- learned about 200 manual signs
- evidence of overgeneralisations (e.g. using “hurt” for tattoo), and the combining of words for unfamiliar objects (e.g. “water
bird” for duck)
Koko:
- 1,000 signs, which she was able to
combine in complex ways
Kanzi:
- 200 symbol “lexigrams”
What does the Chinese room problem suggest about semantics (Searle, 1980)?
- Syntax doesn’t suffice for semantics
- In order to really use language, you have to understand the meanings or the thought contents
Where is the neural bases of sematic memory?
temporal lobes
What is sematic memory?
Semantic memory represents our conceptual knowledge of the world
How can we measure semantics/meaning?
by measuring N400s (ERPs)
Kutas & Hillyard (1980) had participants read sentences on the screen presented word-by-word
What were the 2 conditions?
- Semantically appropriate ending: ’He took a sip from the glass’
- semantically inappropriate ending: ’He
took a sip from the transmitter’
Kutas & Hillyard (1980) had participants read sentences on the screen presented word-by-word
What were the results and conclusions?
- Participants anticipated the last word of
the sentence - Incorrect endings – mismatch between
expectation and the actually presented
word - Mismatch elicited a larger N400
component
Parise and Csibra, 2012 used the N400 to investigate the development of semantic
representations
What was found?
- 9-month-old infants detect the mismatch between an object appearing from behind an occluder and a preceding label
- The label primes the object that
should appear - Infants understand the meaning of some words
What is a concept (semantic memory)
- represent classes of things, events, or ideas
- unite things, qualities, and occurrences on the basis of a similarity of characteristics
How are concepts organised (semantic memory)?
in a network
What does the activation of a concept lead to?
activation of associated concepts
What evidence supports the spreading activation of concepts?
- N400 was modulated by the semantic connection between the expected and presented word
- Semantically related words elicited
smaller N400 compared to semantically unrelated words
What do all models of semantic memory propose?
All models propose that concepts are
comprised of a constellation of constituent features
What are features (semantic memory)?
- Properties of or facts about a concept
- E.g. a lion is an animal; it has 4 legs; it is
carnivore
How are features linked together (semantic memory)?
via a network
What are amodal representations?
- features are represented as abstract knowledge
- not tied to sensory or motor information
- they are independent of input or output modality
What was semantic memory traditionally assumed to be? Why?
- traditionally assumed to be amodal
Reason: - semantic memory can be accessed from
multiple kinds of sensory input
What is the hierarchal model of sematic memory (Collins & Quinlan, 1969)?
- Concepts are organized in a hierarchy
- Features are associated with the concept at the appropriate level
- Concepts are represented as abstract
knowledge = amodal model
What reaction time studies provide evidence for Collins & Quinlan (1969) hierarchal model?
- Fastest: Canary is yellow
- Slowest: Canary is an animal
- canary as an animal is further away in the hierarchy and so takes longer
What evidence is there for Collins & Quinlan (1969) hierarchal model of semantic memory from neuroscience?
- fMRI study of naming and categorization (Rogers, et al., 2006)
Superordinate and subordinate information has different neural substrates:
- Subordinate-level processing (specific):
– e.g. robin –> activates anterior temporal pole
- Superordinate-level processing (general)
– e.g. animal –> activates posterior temporal lobes
What is the problem with amodal representations?
- Each word needs to be defined by other words
- If using a lexicon, it is impossible to understand learn new words/concepts without understanding some
words/concepts in advance - (the symbol grounding problem)
What is a possible solution to the symbol grounding problem?
Grounded concepts/features
What are grounded concepts/features?
- concepts not defined in terms of each other, but in terms of our experiences and interactions with the world
- Concepts of “green” and “kick” are linked to sensory and motor experiences rather than abstract/amodal representations
- Concept representations are distributed over several distinct brain areas
What is the Fully-grounded model of Allport (1985)?
- Different features of a concept are
represented in different information
channels (modalities) - These are the same channels the features were acquired through
- Representations of features belonging to
the same concept are connected
E.g. telephone:
* Auditory regions (how it sound)
* Visual regions (how it looks)
* Action related regions (how to use it)
How does priming provide evidence for grounded cognition?
- Hear: “The ranger saw the eagle in the
sky” - Picture naming: participants are faster in
naming the flying eagle than the stationary eagle
How do action words provide evidence for grounded cognition?
- Processing words such as “lick”, “kick” and “pick” activates body-based
neural representations
Naming pictures of animals and tools elicit
category-related activity where?
in the ventral occipito-temporal cortex
True or false?
obeject representations are limited to a discrete area
False
they are widespread and overlapping (=grounded cognition)
True or false
category-related activations reflect the
retrieval of information about category-specific features and attributes
true
Category-related patterns of activation was found for images animals and tools in occipital and posterior temporal lobes
When were similar patterns also found?
Similar patterns of category-related
activity occurred when subjects read the
names of or answered questions about
animals and tools (Chao et al., 1999)
Are category-specific deficits limited to the visual domain?
- no
- deficits also for defining words
Different categories rely on different types
of characteristics
Give 2 examples of this
Animals:
- sensory characteristics are more
important: how they look and/or sound
- Animals usually have eyes, mouth, self-initiated movements
Tools/man-made objects:
- functional characteristics are important
- relationship between shape and function
Categories emerge based on what?
out experiences with their features
Is category specificity innate?
At least some categories are hardwired:
category specificity in the brain is innate
(Caramazza & Shelton, 1998)
- Animals
- Plants
- Conspecifics (other humans)
- Tools
What evidence is there that category specificity is innate?
- Prenatal face processing (Reid et al, 2017)
- Newborns prefer biological motion (Simion, Regolin, & Bulf, 2008)
What evidence based on congenital blindness suggests that visual functional brain organisation is innate?
- Visual stimuli elicit category-specific
activity in the ventral-temporal cortex
(VTC) in sighted individuals (faces,
scenes, body parts, and objects) - Natural sounds representing different
categories elicit similar discriminatory
responses in VTC in individuals who
were blind since birth - If visual functional brain organization
does not rely on visual input then it must
be innate
A study involving those who are congenitally blind concluded that if visual functional brain organization
does not rely on visual input then it must
be innate
What is a contradiction to this?
what if these representations are not
strictly visual, but shape-based – blind
people can acquire this via touch
A case study by Farah and Rabinowitz (2003) studied Adam
Where did Adam have brain damage?
damage to the posterior brain areas at the age of 1 day
A case study by Farah and Rabinowitz (2003) studied Adam
What were Adam’s impairments?
Selective impairment in knowledge of living things:
- naming of pictures of living things
- retrieving verbal information about living things:
1. questions probing visual knowledge (Does the dog has four legs?)
2. questions probing nonvisual knowledge (Is chicken served in restaurants?)
A case study by Farah and Rabinowitz (2003) studied Adam
What could Adam do?
Could name and answer questions about inanimate objects
A case study by Farah and Rabinowitz (2003) studied Adam
What was concluded?
- brain damage sustained too early for experience to have contributed to the organisation of semantic memory
- innate basis for the living–nonliving distinction in semantic memory
