Lecture 9 - 03-18-13
Chapter 8; Concepts and Generic Knowledge
prototypes and typicality effects
difficulties with categorizing via resemblance
concepts as theories
when we have a concept like dog or chair, we take it for granted that this should be something
easily understood. Defining these concepts are more complex than our first glance. These are
the building blocks of all the information that we know. It's the building blocks of our general
knowledge. We start by knowing what the dog or chair is.
dog: mammal with 4 legs, barks, wags its tail. There's examples that are exceptions to this rule.
If you saw a 3 legged dog, you would still know that it's a dog, or a dog that doesn't bark. If you
can always have an exception to the definition, then how are we defining these terms in which
we believe are so simplistic?
This philosopher, Ludwig Wittgenstein had this idea that simple concepts have no definition. If
you try to define them, there's no real way of defining them. Example: game, played by children,
for fun, some type of rules, involves multiple people, done during leisure time; we can think of
exceptions to what a game is.
Gambling is also a game. Playing games for fun can also be professional sports. There are rules
in some games but not something like Lego. There are things that involves multiple people, but
solitaire is a game but it's individual. Competition could be a possibility but tea party is also an
The way that we can start to think about this, rather than saying that there are definitions, we
can think of family resemblance. Guy with the beard, glasses, etc., that's the ideal member of
the family. There are also atypical members. Maybe we do look like a brother, sister, mother or
father. Maybe same thing for your cousins, may have overlapping features.
Dog probably has 4 legs, barks and wags its tail. Creatures without these feature are unlikely to
be a dog. There's an ideal member that is a dog with these traits and there are members of the
concept dog that don't necessarily fit that in which we still recognize as a dog but is not likely to
The more characteristic feature an object has, the more likely we are to believe it is part of the
category. If we see a random animal run by, may have heard a bark, 4 legs, probably a dog. How
many of these characteristics actually fit the correct description of a dog?
Prototypes and typicality effects
prototypes We have prototypes, ideal members are prototypes of that concept. Prototypical dog is German
sheppard, other dogs are less prototypical yet they still fit a lot of that prototype in which you
can generalize to. That prototype possesses all features when we think of a dog for instance.
This differs amongst individuals and cultures. We can say that having this prototype is based on
what we mostly see often. If you grow up in a particular part of the word with specificities, it'll
be different than other parts of the world. What your prototype is will be different. You're
averaging out all the ones that you basically see. If you're living in Canada, you have specific
prototypes. Exposure to more examples of that house changes your general idea of what your
ideal house would be.
We can say that prototypes have fuzzy boundaries. Because of the way that they are set, there's
no clear boundaries in which a certain thing is defined as a prototype or not. If you were looking
at the red squares, which are the best red, studies have shown that people have similar
believes. It seems odd to do this, the fact is that people can still can do that task. There's
something different about as being a best example of a red despite all the squares are all
typically red. In the text, it talks about odd and even numbers. You can have a number and it fits
into one of those categories. Let's say you're given a bunch of numbers and to have to judge
which are "evener", people can still rate them on different levels and so there's something
about this that kind of brings the point that maybe there's better examples of the prototypes in
other exemplars that fit into that concept.
Most of these items are somewhat implicit. It's more or less than when you see these things
growing up, it builds a conceptual base of a web of information in your head and becomes a
type of activity in which you distinguish.
Robins are birds, penguins are birds, people are slower at defining the latter. You have to take a
second think whether or not it fits your prototype of a bird. It's not that you consciously thought
about the fit. When people respond faster to something like this, you're saying that the example
that fits has more characteristics in common with the prototype that the individual has.
These typicality effects can also be seen when we give participants production tasks. Example:
name as many birds as you can. They'll start out with items that are more typical with their
experience with birds (more common) and as you move down the list, they become more and
Does this picture show you a phone? The first one is an iphone, if you have more experience
with that, it becomes quicker. The latter is the older phone, you can still recognize that but is a
slower response, it's more temporal because this generation is more used to the typical phone
that resembles something like the iphone. It fits that particular individual prototype in which is
shaped by where you are, and your experiences.
We also talk about privilege memberships. Category members are more privileged if they're
closer to your prototypes. The chart is about the fruits and birds. Rank these on how well/bird-
like/fruit-like they are. Apples and pears and grapes and strawberries are rated quite high, quite
typical. Pumpkins are less fruit like. Same thing with birds, bats, wings, size of a bird. If we're asked to write a sentence about birds, a lot of times people will say that there are two
birds in the tree, which is fine. Penguins are birds, robins are birds. When we fill it in, it's better
to use something closer to the prototypes. Which is why Robin will fit into the sentence.
Penguins won't because they're not the prototypical bird.
Three fish, which is more attractive? The first fish is founded more attractive consistently,
something that is closer to the prototype, more attractive; the latter two don't necessarily fit
what we think of as a fish. Typicality also influences judgments about attractiveness.
Just as certain members of a category are privileged, things that fit that category more are
privileged, rated as more fruit like or bird like, we can also say that there are certain types of
categories that are more privileged. In this case, the picture is a chair. That's a likely initial
response. Some people say furniture which are too general, upholstered armchair which is too
specific. The category itself is privileged as the middle ground.
Basic level categories; single word i.e. chair, it's the default for basic level and it's easier to
explain why it belongs to a particular category. The category fits into the middle of broad and
specific. Children tend to use these basic levels. We can say that our knowledge of basic
categories are concepts that fit our prototypes.
Exemplar (check the table for exemplars on the ppt.)
When we think of typicality as something that is typical, the prototype is the average of the
category itself. And the exemplar is encountered more often, those things can have a great deal
of overlap. Sometimes it isn't so much because you've learned about them elsewhere. Graded
membership for the prototype is based on how similar or less similar it is from our average. The
exemplar is basically more often, more times that we've encountered it like seeing a chair often.
For music, you can think of both of these. Let's say you're used to a certain genre of music, you
can identify it's music when it's from another country. The prototype theory says that if it's less
similar compared to the music you listen to, it's less privileged. For parents when they haven't
heard it before, it's less fitting to the prototype. For exemplars, it's because you have
encountered a lot less. Illustration fits into both. Ideal fruit (apple) versus less ideal (fig);