Final _ chapter 9 Flashcards
Conceptual Knowledge
- Knowledge that enables us to recognize objects and events
- and to make inferences about their properties
This knowledge exists in the form of concepts
Concepts
- Allows people to discriminate members of a categorie from non-members.
- A Mental representation of a class or individual.
What is the difference between a concept and a category?
A person’s idea about what should and shouldn’t go in a category is their concept of that category.
- A concept refers to a mentally possessed idea or notion
- A category refers to a set of entities that are grouped together.
Categories
- Facilitate the storage and retrieval of information.
- supplies a principle of organization by which new information can be banked efficiently in memory.
- Provides rules and road maps to accessing information.
We categories to organize concepts
Categorization in use
- You recognize a perceived input
- find its category in your internal mind map
- You are then able to focus your energy on specifying what’s special about that particular input.
Relates each experienced entity to an extant representation
Definitional Approach
How are objects placed into categories?
- Classical view
- Deciding whether something is a member of a category by determining whether a particular object meets the definition of the category
[Cons]
- Works well for well defined concepts [geometric shapes],
- Not so will for complex ones.
[E.g.]
- Different objects with drastically different features can be considered chairs.
Ludwig Wittgentein [1953]
Noted the problem of the definitional approach
and proposed the idea of family resemblance
- Referes to the idea that things in a particular category resemble one another in a number of ways.
Allows for some variation within a category
Prototype Approach
How are objects placed into categories?
Membership in a category is determined by compairing the object to a prototype that represents the category
A ‘typical’ member of the category
Eleanor Rosh [1973]
Proposed that the ‘typical’ prototype is based on an average of members of a category that are **commonly **experienced.
- Thus the prototype is not an actual member of the category, but an average representation of the category.
Typicality
[High typicality]
- Means that a category member closely resembles the category prototype.
- A ‘typical’ member of the category
[Low typicality]
- Means that the category member does not closely resemble a typical member of the category.
Rosh et Al. [1975]
Experiment
Investigated the basic nature of categories and quantified the idea of typicality.
- Presented participants with a category title [e.g. bird] and a list of about 50 members of the category.
- Participants were tasked to rate the typicality on a 7-point scale [1-high and 7-low/ is not a member at all]
[Results]
- [1.18 rating for sparrow] reflects the fact that most people consider a sparrow a good example
- [4.53 rating for penguin]
- [6.15 rating for bats] reflect they are not considered good examples.
Typicality and family resemblance
Prototype effects
High typicality category members have a higher family resemblance
[In a study done by Rosch and Mervis]
- Good examples of a category share many attributes with other members of this category
- While those that were rated as bad examples didn’t.
- They concluded that there is a strong relationship between family resemblance and prototypicality.
Sentence Verification Technique
Method
-Participants are presented with statements and are asked to answer
- yes is they think the statement is true
- No if they think it isn’t
Typicality Effect
Prototype effects
Ability to judge highly prototypical objects more rapidly
[In a study done by Smith et al 1974]
- Used the sentence verification technique to determine how fast people could answer questions about an object’s category.
- Highly typical category members were responded to faster
- E.g ‘Is an orange a fruit?’ faster than ‘Is a papaya a fruit?’
Highly typical category members are primed more effectively
Naming
Prototype effects
When participants are asked to list as many objects in a category as possible
- They tend to list the most prototypical members first
[E.g] sparow would be named before penguin
Priming
Prototype effects
Occurs when presentation of one stimuli facilitates the response to another stimuli that usually follows closely in time.
[In Rosch’s experiment]
- Participants first hear the prime [name of a color]
- 2 secs later they saw a pair of colors side by side
- They we asked to indicate whether the two colors were the same or different
[results]
- When participants hear the word green, they judged two patches of primary green as being the same more rapidly than two patched of light green
- The prime facilitated response to a stimuli if it contained some of the information needed to respond to the stimuli.
Name the 4 effects of prototypicality.
- Family resembance
- Typicality
- Naming
- Priming
Examplar Approach
How are objects placed into categories?
- Suggest a category is defined based on a set of strored examplars
- Membership in a category is determined by comparing a target to stored members of the category
Takes into account the wide variation among items in a category
Examplars
Actual members of the category that a person has encounted in the past.
How does the examplar approach compair to the prototype approach?
Like the prototype approach, involves determining whether an object is similar to others
But, instead of an average..
- The standard for the examplar approach involves many examples
- each called an examplar
The examplar approach explains the typicality effect
- By proposing that objects that are like more of the examplars are classified faster
Similar to the idea of family resemblance as well
Advantages of the examplar approach
- By using real examples, it can more easily take into account atypical cases [flightless birds]
- The ability to take into account individual cases, means that no information that might be useful later is discarded
- Requires only that we remember varying examples
Exemplars and prototypes working together
It has been proposed that as we initially learn about a category, we may average exemplars into a prototype, but as we learn, individual exemplars become stronger.
- Early in learning we would be poor at taking exceptions into account
- later on exemplars for these cases would be added to the category
- Both produce our rich store of conceptual knowledge allowing each kind of knowledge to explain the tasks that are most suited for it.
[E.g]
- we know generally what cats are [prototype]
- But we know our own specific cat the best [exemplar]
Hierachical Organization
kind of organization in which larger more general categories are divided into smaller more specific categories creating a number of levels of categories.
Colour Categories
After a color spectrum classification task
[Regier et al. 2007]
- Concluded that there are universal properties that guide the development of colour categories in different languages
- However there is a variation from language to language
- Suggested that warmer colours are more useful and more likely to be named
Rosch’s Approach
Levels of categories
Noticed that there are different levels, and that when people use categories, they tend to focus one one of these levels
- Superordinate level [Global]
- Basic level
- Subordinate level [Specific]
Proposed that the basic level is psychologically special because..
- Going above it results in a large loss of information
- Going below it results in little gain of information
Naming things
Demonstration
Rosch et al. 1976
- Showed participants a category label, then after a bried delay
- Asked participants to name as quickly as they could objects presented infront of them
- The results showed that they accomplished this task faster when naming basic level categories rather than global level categories
How Knowledge can affect categorization
The ‘optimal’ level that people tend to focus on ultimatly changes depending on individual knowledge.
- In a series of experiments, researchers conducted a categorical naming task with participants who knew from as little about the category to participants who were experts
- The results varied drastically from the ‘optimum’ basic level response
- For the experts, their responses fell under the specific level
Semantic Network Approach
Network models of categorization
Collin & Quillian [1969]
Proposes that concepts are arranged in networks
Network consists of:
-** Nodes**
- Represent a category or concept
- placed in the network so that related concepts are conneted
- A number of properties are indicated for each concept
-Links
- Connect nodes together
- indicate how they are related to each other in the mind using a hierarchical model [ [top]global to [bottom]specific]
Access knowledge by going upwards
Cognitive Economy
Semantic Network
Way of storing shared properties just once at a high-level node for efficiency.
[Cons]
- Does not account for individual differences at lower level nodes.
- Collins & Quillian added exceptions at lower nodes to try and account for that.
similar to prototype issue of over generalization.
Evidence
Semantic Network
The time it takes for a person to retrieve information about a concept should be determined by the distance that must be traveled through the network [From bottom up]
Sentence verification Method:
- It should take longer to answer ‘yes’ to the statement [a canary is an animal] than [a canary is a bird]
- Collins & Quillian tested this by measuring the reaction times to statements that involved traversing different distances in the network.
- Prediction was confirmed.
Spreading Activation
Semantic Network
Activity that spread out along any link that is connected to an activated node
- Activating the canary to bird pathway activates additional concepts that are connected to ‘bird’
- Additional concepts then become primed and can be retrieved more easily from memory
Lexical Decision Task
Method
Participants read stimuli [some which are words and others non-words]
- Their task is to indicate as quickly as possible whether each entry is a word or non-word
Can spreading activation influence priming?
Semantic Network
- Meyer & Schvaneveldt [1971]
- Used a variation of the lexical task
- Participants were tasked to press [as quicly as possible] ‘yes’ when both items were words or ‘no’ when at least one of the words was a non-word
[Results]
- Key variable was the association between the pairs of real words
- In some trials the words were closely associated and in some weakly associated
- The reaction time was faster when the pair present was closely associated
[Conclusion]
Proposed that the results were due to retrieaving one word from memory triggering a spread of activation to other near by locations in a network.
Criticism of the semantic network approach
- The theory cannot explain the typicality effect, does not account for typical words being retrived faster.
- Question the concept of cognitive economy since people may in fact store specific properties of concepts.
[Rips et al. 1973]
- sentence verification task
- ‘a pig is an animal’ was verified more quickly than ‘a pig is a mammal’
- Based on the semantic network, there’s a shorter distance between mammal and pig than between animal and pig.
Connectionist Approach
Network models of categorization
McClelland and Rumelhart [1986]
Approach to creating computer models for representing cognitive processes
- Parallel distributed processing [designed to represent concepts]
- Proposed that concepts are represented by activity that is distributed across a network
Units
- Inspired by neurons found in the brain
- Concepts and their properties are represeted in the network by the pattern of cross activity
- can be activated by stimuli from the environement or from other units within the network
Lines
- connections that transfer information between units
- Roughly equivalent to axons in the brain
Input Units
Connectionist Approach
Units activated by stimuli from the environment
-Send signals to hidden units, which sends signals to output units
Connection Weight
Connectionist Approach
Determines how signals sent from one unit either increase or decrease the activity of the next unit
- These weights correspond to what happens at a synapse that transmits signals from one neuron to another
- [High connection weight] result in a strong tendency to excite the next unit
- [Lower connection weight] cause less excitation
- [Negative connection weight] can decrease excitation or inhibit activation of the receiving unit
What does the activation of a unit in a network depend on?
Connectionist Approach
- The signal that originates in the input units
- The connection weights throughout the network
Basic principle of connectionism
A stimulus presented to the input units is represented by the pattern of activity that is distributed across the other units.
Connectionist approach compaired to Semantic approach in representation
- Connectionist networks indicate properties by activity in the attribute units and by the pattern of activity in the representation and hidden units
- While the Semantic [hierarchical] network represents these properties at the network’s nodes
- With the connectionist network, a concept is represented by a pattern of activity in ALL the units in the network [some stronger some weaker]
Training a Network
Connectionist Approach
For the network to operate properly, the connection weights have to be adjusted so that activating a concept unit and a relation unit activates only the corresponding property unit
- Acheived by a learning process
Back Propagation
Connectionist Approach
Learning process that occurs when the erroneous responses in the property units cause an error signal to be sent back through the hidden and representation units
- The error signal provides information about how the connection weights should be adjusted so that the correct property units will be activated
Graceful Degradation
Connectionist Approach
- Because information in the network is distributed across many units, damage to the system does not completly disrupt its operation.
- Disruption of performance occurs only gradually as parts of the system are damaged
Generalized Learning
Connectionist Approach
- Because similar concepts have similar patterns
- training a system to recognize the properties of one concept also provides information about other related concepts
Sensory-Functional Hypothesis
How concepts are represented in the brain
Warrington and Shalice [1984]
- Reported on patients with category-specific impairements
- [Patient K.C & patient E.W]
- Patients were able to identify non animals, but impaired ability to identify animals.
[Hypothesis]
Our ability to differentiate living things and artifacts depends on a memory system that distinguishes sensory attributes and a system that distinguishes functions.
- Living things are more likely to be distinguished their by sensory features.
- Artifacts [non-living] are more likely to be distinguished by their functional features.
Flaws in the sensory-functional hypothesis
Artifacts are not a single homogeneous category, many effects of brain damage can’t be explained by the simple distintion between sensory and function.
[Lambon Ralf et al. 1998]
- Studied a patient who had a sensory deficit
- She performed poorly on perceptual tests
- Yet was better at identifying animals than artifacts
[Hoffman and Lambon Ralf]
- Studied patients with poor comprehension for small artifacts [tools..
- But better knowledge for larger artifacts [vehicles..
The Multiple-Factor Approach
How concepts are represented in the brain
Distributed representation
- Though it is true that members of specific categories share similarities, it is also clear that we need to used multple features in grouping them.
[Hoffman and Lambon Ralf 2013]
- Had participants rate a large number of items with regards to different features.
- [7] being strongly associated to the feature, [1] being not at all.
[Results]
- Animals were highly associated with motion and color while artifacts were highly associated with performed actions {much like S-F]
- Intrestingly, mechanical devices such as machines, vehicles, and musical instruments overlapped with both sensory and function
Crowding
Animals tend to share many properties, in contrast artifacts share fewer
[Hypothesis]
Pateints dont really have a category-specific impairment, they have difficulty recognizing living things because they share similar features making it harder to distingish them.
Semantic Category Approach
How concepts are represented in the brain
Proposes that there are specific neural circuits in the brain for some specific categories.
[Hutch et al. 2016]
- Created a neural map based on spoken language.
- Had participants listen to more than two hours of strories from “the Moth Radio” while scanning their brains
[Results]
- While this approach focuses on areas of the brain that are specialized to respond to specific stimuli.
- They found that the brain’s reponse to items from a particular category is distributed over a number of different cortical areas
Embodied Approach
How concepts are represented in the brain
States that out knowledge of concepts is based on reactivation of sensory[size, shape,color] and motor[carrying out action] processes that occur when we interact with the object.
[Hauk et al. 2004]
Measured participant’s brain activity using FMRI under two conditions:
1. As participants moved their right/left foot, index finger, or tongue.
2. As participants read ‘action words’ such as kick, pick, or lick.
[results]
Show areas of the cortex activated by the actual movements and by reading the action words.
Link between perception and motor responses
Semantic Somatotopy
Correspondence between words related to specific parts of the body and the location of brain activity .
Flaws in the embodied approach
According to the embodies approach, a person who has trouble producing actions associated with objects should have trouble recognizing the objects.
[Garcea et al. 2013]
- Tested patient AA
- Suffered a stroke that affected his ability to produce actions associated with various objects.[how to use them]
Concluded that the sensory motor association is not necessary for recognizing objects.
Also approach isn’t well suited to explain abstract concepts
The Hub & Spoke Model
How concepts are represented in the brain
Different areas of the brain process different kinds of features and the Anterior Temporal Lobe serves as a hub that integrates them all together.
- Damage to this area caused a general loss of knowledge
- Semantic dementia
[Evidence]
- Damage to one specialized area [the spokes] can cause specific deficits
- Damage to the ATL [the hub] causes general deficits
Transcranial Magnetic Stimulation [TMS]
Method
Temporary disruption of cortical function by applying a pulsating magnetic field using a stimulation coil.
- If a particular behavior is disrupted by the pulses, researcher conclude that the disrupted area is involved in that behavior.
Difference between hub and spoke function
[Pobric et al. 2010]
- Using TMS
- Participants were presented with pictures of living things and artifacts and were asked to name them as quickly as they could.
- Procedure was repeated while TMS was applied to either the ATL or to an area in the parietal lobe[normally active when manipulating an object]
[Results]
- Parietal inactivation - lowed reaction times for artifacts[highly manipulatable objects {tools…, but not furniture] but not for living-things
- ATL inactivation - caused the same effect for both living-things and all artifacts
- Supports the idea of a hub with general functions, and spokes with more specific functions
Flaws in the hub and spoke model
- It has been suggested that other structures may be ‘hubs’
- or that the most important way concepts are presented is not by ‘hubs’ but by the pattern of connections formed between the ‘spokes’
