PARADIGMS OF

COGNITIVE PSYCHOLOGY

Unit 1
 What is a paradigm?
• Body of knowledge structured according to what its proponents consider
important and what they do not.

• Paradigms are intellectual frameworks that guide investigators in understanding

phenomena.

Paradigms include the assumptions Paradigms also specify what kinds of

investigators make . methods are apt for an investigation.
In learning about the 4 paradigms, ask yourself the following
questions:

• What assumptions underlie the paradigm?

-There is 1 truth out there
-Objectivity is imp. in finding the truth.
Paradigms of
• What questions or issues does the paradigm emphasize? Cognitve psychology-
- How can we predict and control phenomena? 1. Information
-How can we ensure a bias free search for the truth? processing
2. Connectionist
• What analogies does the paradigm use?
3. Evolutionary
- People are like bone china, handle with care
4. Ecological
-The human is like a machine.

• What research methods and measures does the paradigm favor?

Experiments? Participant observation? Interviews? Surveys?
 Information-Processing Approach

• Dominated in the - Remains

influential today (Atkinson & Shiffrin, 1968).

• Analogy b/w human cognition & computerized

processing of info.
Computer metaphor used, shows an
indebtedness to the fields of engineering
The idea that cognition can be thought of as & communications.

information(what we see, hear, read etc) passing

through a system (our minds). It sees the individual as a processor of
information, in much the same way that a
computer takes in information (INPUT),
follows a program to produce an OUTPUT.
Its assumed that information is processed in & it is stored in while being processed.

Our goal-Determine what these stages and storage places are and how they work.

People’s cognitive abilities can be We have different cognitive capacities—

thought of as “systems” of interrelated attention, memory, language etc.
capacities.
Information-processing theorists try to
find the relationships b/w these
capacities.
 Three store model of memory Boxes represent stores.

Stores- hold information for later use

Arrows represent processes

Processes- Operate on the information or transfer it

from store to store.

( So these are also called “boxes-and-arrows”

models )
Certain processes (e.g., recognition)- Used at the
beginning.
A general-purpose information-processing system Others (e.g.retrieval), have to do with memory
storage.
The use of flowcharts (borrowed from computer sciences)- which
illustrate the sequential flow of information through a system. Others (e.g., reasoning) have to do with putting info.
together in new ways.
Its assumed that-

• Information in our minds is stored symbolically.

(Symbols- Letters, numbers etc.)

• The way it is coded & stored greatly affects how easy it is to

use it later. (e.g., Dual-coding theory)

• People, like computers, are general-purpose symbol

manipulators
Typically, information-
• We perform astonishing cognitive feats by applying only a processing psychologists
few mental operations to symbols. use
experimental &
quasi- experimental
The paradigm is rooted in structuralism- Its followers attempt to identify the techniques in their
basic capacities & processes in cognition. investigations.
Many models in the IP framework have been proposed like Broadbent's Filter Model (1958).
Yet there are some evaluative points to bear in mind. These include:

• The analogy b/w human cognition & computers approach is limited.

(i) Human brain has capacity for extensive parallel processing - computers often rely on
serial processing;
(ii) Humans are influenced in their cognitions by no. of emotional & motivational factors.

• The evidenceis largely based on experiments under controlled, scientific conditions.

Although these lab. experiments are easy to interpret, the data may not be applicable to
real world (i.e., it ).
 Computer - Mind Analogy

The development of computers in the 1950s & 60s influenced psychology & was, in part, • Essentially, a computer cod es (i.e.,
responsible for the cognitive approach becoming the dominant approach in modern changes) information, stores information,
psychology (taking over from Behaviorism) uses information, and produces an output
Basic Assumptions (retrieves info).

(1) info made available by the environment is processed by a series of processing systems • The idea of info. processing was adopted
(e.g. attention, perception, STM); by cognitive psychologists as a model of
(2) these processing systems transform or alter the info. in systematic ways; how human thought works.
(3) the aim of research is to specify processes, structures that underlie cognitive
performance; • The information processing approach
characterizes thinking as the environment
(4) info. processing in humans resembles that in computers. providing input of data, which is then
Information processing models consist of a series of stages, or boxes, which represent transformed by our senses.
stages of processing. Arrows indicate the flow of information from one stage to the next.
• The information can be stored, retrieved
Input processes are concerned with the analysis of the stimuli. and transformed using “mental programs”,
with the results being behavioral
Storage processes cover everything that happens to stimuli internally in the brain and can responses
include coding and manipulation of the stimuli.

Output processes are responsible for preparing an appropriate response to a stimulus.

• Some remarkable cognitive abilities are ones we typically take for granted like ability
to use

• These abilities may seem rather trivial to us.

• But researchers in AI found that it is not so easy to program computers to carry out
these tasks.
• So why can young children do these tasks? The evolutionary theory suggests:

Like other animal minds, the human

The mind has responded to evolutionary
mind is a biological system.
pressures to adapt in response to
physical & social environments
It has evolved over generations as per
encountered by our predecessors
laws of natural selection.
.
 Evolutionary psychologists believe we understand a system best if we
understand the evolutionary pressures on our ancestors.

Cosmides and Tooby (2002): Cosmides & Tooby (2002)-

Some of the most significant issues our ancestors faced involved social issues-
People have special-purpose Creating & enforcing social contracts*. To do this, people must be very good at:
mechanisms (including
cognitive mechanisms) specific
to acertain context or class of (i) Reasoning about costs and benefits
problems (e.g., grammar (ii) Detect cheating in a social exchange.
acquisition, mate acquisition).

• Thus people’s reasoning will be especially enhanced when they are reasoning
about cheating.

*Social contract arguments- Individuals have consented, either explicitly or tacitly, to surrender
some of their freedoms and submit to the authority in exchange for protection of their
remaining rights or maintenance of the social order.
• Overlaps quite a lot with evolutionary approach.

• Comes from both psychologists & anthropologists.

All cognitive activities are shaped bythe

culture and by the context in which they occur.

Lave et al (1984) studied how people used arithmetic in their daily lives.
They followed people on grocery-shopping trips to analyze how people calculate
“best buys.”

They found that people’s methods of calculation varied with the context.

This was surprising, because students in our culture are taught to use the same
specified formulas on all problems of a given type to yield one definite answer.
 A typical third-grade arithmetic A problem, posed & solved by a
problem presented by teachers grocery shopper- How many apples
she should buy for her family for the
week? Lave (1988) pointed out a number of
contrasts b/w the 2 problems:
There’s only about 3-4 apples at home.
Brandi had eight seashells. Nikki I have 4 kids, so at least 2 apiece in 1. The second example has many
had five more. How many seashells the next three days. I have a certain possible answers, unlike the first ,
did the two of them have together? amount of storage space in the fridge, which has one.
so I can’t load it up totally. . . I like an
2. First problem is given/presented-
apple sometimes at lunchtime when I
the other is constructed by the
come home.
problem solver herself.

3. First problem is disconnected from

personal experience & interests.
The other comes out of
daily living.
• The idea of studying cognition in everyday contexts was encouraged by J. J.
Gibson.

• Ulric Neisser, , wrote a book in 1976 aimed at redirecting psychology toward

studying more “realistic” cognitive phenomena.

• We also see the influences of Functionalism & Gestalt school.

This approach would deny the
The functionalists Gestalt psychology usefulness of studying cognition
in artificial circumstances.
Focused on the purposes served by Emphasis on the context surrounding any
cognitive processes, certainly an ecological experience
question Thus it relies less on lab.
experiments or computer
simulations & more on
naturalistic observation + field
studies.
 The Connectionist Approach
Parallel-distributed processing (PDP)

• Beginnings are in early 1980s

• Resulted when researchers from different disciplines began

to explore alternatives to the I.P approach.

The name is derived from models depicting

cognition as a network of connections among
simple & numerous (McClelland, 1988).

Because these units are sometimes compared to neurons- these

are sometimes called
A connectionist network, showing units and their connections
• Units-Black circles, with all the arrows pointing
to them (NODES).

• Each unit represents a particular individual.

• Each unit is connected to other units (shown as

small ellipses) that depict certain information
about individuals (e.g.,professions).

• Thus, if the node for “Joe” is activated, it inhibits

activation of the nodes “Claudia,” “Fred,”
“Frank,” “Harold.”

• “Joe” node activates the top left-hand node in

the center circle, and activates the nodes “male,”
“professor,” “Subaru,” “brie.”

• The activation of these nodes inhibits, or lowers,

the activation of all other nodes in their
respective ellipses
Let’s look at a simple example.

• Imagine that a person tells you to define the concept of a dog.

• When the word reaches your ear, the set of neurons associated with
it automatically activate in your brain.

• The activation of this group of neurons spreads to others it’s

connected to.

• This may include neural patterns related to the words mammal, bark,
fur.

• And this will lead you to define a dog as “a mammal with fur that
barks”.
• Each unit has some level of activation at any particular moment in time.

• The exact level of activation depends on the input to that unit from (i) the
environment (ii) Other units to which it is connected.

• Connections b/w two units have weights, which can be or

Causes one unit to excite units to Causes one unit to inhibit (lower the
which it is connected (raise the level of activation of) connected units)
activation of units)

• When any unit reaches a certain

level of activation, it activates all
the other units to which it has
positively weighted connections.
Properties of connectionist systems
For these systems to work like the human brain seems to, they have to fulfill certain conditions:

Neurons, when activated, influence those they’re connected to. This

can occur by facilitating their activation or by inhibiting it. e.g., ‘Dog’ neurons facilitate those
related to ‘mammals’ but inhibit those related to ‘reptiles’.

Learning & experience affect the connections b/w neurons. Thus, if we see
many dogs that have fur, the connections b/w the neurons related to both concepts will be
strengthened. This is how we create the neural networks that help us process information.

• Parallel processing. Neurons don’t activate one after the other. The activation occurs in
parallel b/w all the neurons. And you can have multiple patterns going on at the same time.
Thus, we can interpret a lot of data simultaneously. However, there is a limit to our capacity.
 Major difference

Information-processing Connectionist
cognition is typically assumed to occur assume that cognitive processes occur in
serially—that is, in discrete stages parallel, many at the same time.

Like IP approach, connectionism Information-processing Connectionist

draws from structuralism an interest Look to computer science look to cognitive neuropsychology and
in the elements of cognitive cognitive neuroscience to
functioning. construct theories & models.
Methods: experimental and quasi- Replicate the findings
However, experimental research of experimental and quasi-experimental
research using computer programs.

Try to provide explanations at a more More concerned with the “subsymbolic”

abstract, symbolic level level: how cognitive processes actually
could be carried out by a brain.
PDP
• Connectionism allows for a wide variety of models that can vary in the no. of units hypothesized, no.
& pattern of connections among units, & connection of units to the environment.

• All connectionist models share the assumption, however, that there is no need to hypothesize a
central processor that directs the flow of information from one process or storage area to another.

• Instead, different patterns of activation account for the various cognitive processes (Dawson, 1998).

• Knowledge is not stored in various storehouses but within connections b/w units.

• Learning occurs when new connective patterns are established that change the weights of
connections between units.
 PDP
• Feldman & Ballard (1982)- argued that this
approach is more consistent with the way the
brain functions than IP approach.

• The brain is made up of many neurons connected

Connectionism, being much
to one another in various complex ways. newer than information
processing, is just beginning to
map out explanations for
• Rumelhart (1989) puts the issue more simply: individual and developmental
“Connectionism seeks to replace the computer differences.
metaphor of the information-processing
framework with a brain metaphor.”
General Points
• Each of these four paradigms makes an important contribution to cognitive psychology & offer complementary
perspectives on how the underlying principles of cognition ought to be investigated.

The connectionist approach-

The information-processing
Focuses on the underlying
paradigm- Focuses researchers
“hardware”—how the global
on the functional aspects of
cognitive processes described by
cognition—what kinds of
an information-processing model
processes are used toward what
are implemented in the human
ends.
brain.

Ecological approach- Stresses

The evolutionary approach- the need to consider the
Centers on questions of how a context of any cognitive
cognitive system or function process to understand more
has evolved over generations. completely how that process
functions in the real world.

• Not all cognitive research fits neatly into one of these paradigms.
• Some research incorporates parts of different paradigms; some fits no paradigm neatly.