Lecture 11 - 04-01-13
Chapter 10 Visual Knowledge
long-term visual memory
the diversity of knowledge
Can you picture this...
"picture a puppy rolling in the snow..."
On a day to day basis, we do a lot of visual imagery. How many windows do we have in our
condo/house? We can picture it. How do we have this picture that isn't really a picture, how do
we bring this up? They're done on a regular basis. Another is: was so and so in class the other
day? We don't know that information right off the bat but we can picture the class. Another is:
buying new clothes, we can hold the jeans/sweater in front of us and visualize how we look in
Francis Galton (1883)
Ask them to visualize an object and then ask them to report as much information as they could.
Most people could do something like that. What they reported was very different, very broad
range of self reports - detailed images, cats, colours of the eyes, nails trimmed, etc. Others were
very basic in terms of what they were picturing. The question is: self reports/introspections
aren't always the best articulate way of explaining what they're looking at. Is it the case that the
reporting is off? Or the case that what they're imagining in their mind is different? Detail versus
Another type of studies were looked at. These kinds differ because they add some objectivity to
what people are picturing. We call these chronometric studies. We ask participants to
manipulate the image. Do something with it that should take a different amount of time based
on how they're imagining this image to be and then we time it. It's a more objective type of
view. Observe how long the manipulation takes.
Kosslyn (1976): premise was this: if you think about a cat, generally the things that come up
about the animal or what you would normally talk about in how it differentiates from other
animals are claws and whiskers; heads and ears may not the first pieces of information that
comes to mind. When you have to draw a picture of the cat, you usually draw the head first, not
a defining feature, that's the biggest object that you would draw to begin the process in terms
of a picture. Defining feature is not prominent pictures versus non-defining features but
prominent pictures. They had participants answer whether the cat had x or y. Subjects were
quicker at identifying the cat has a head when asked first to visualize a cat. It fits more with the
way in which people would draw an image. People are actually looking a picture in their head.
The reverse was also done, just to think about the cat. Quicker at defining these defining
Image-scanning procedure. Kosslyn et al. (1978)
A map is shown, asked to memorize the map. Memorize it to the extent that they can draw it from memory. The question then is posed to them, in this case; there's a straw hut and a well.
You already memorized the picture. Imagine a little dot from straw hut to the well and tell me
when you get there. And then they would say now tell me from straw hut to the tree, and then
to the rocks at the top. The distance between the two objects, as the distance increases, more
time is taken. More than likely, people are actually picturing this map and thinking about this
dot travelling across it, just as if the map was in front of them, and their eyes follow the
trajectory. The results show that the effects are consistent/robust. Imagined distance
correspond to real distance. Follow up studies have been done: follow up from A to B in your
house, etc. The further away they're traveling, the longer it takes. Something similar is imagine
that we're asked to picture a car sitting here, and then a mouse sitting next to the car. We are
then asked if the mouse has whiskers. We would be able to confirm or deny but it may take a
little bit of time whereas if we are asked a mouse next to a quarter and asked if the mouse has
whiskers, we are quicker at identifying. The quarter is more proportional to the mouse in size as
opposed to the car. We must zoom-in in the former example just like we're looking closer at the
picture. The latter, we're in the same frame as the detail so we are quicker to respond. Supports
that mental pictures/images are there. Another study is the mental-rotation task (Friedhoff). Are
these two objects the same? Generally we would be looking at the speed at identifying if they're
the same. We would rotate either clockwise or counter clockwise. In the latter set of objects, we
must rotate to a larger degree. The results of these studies show that it takes subjects longer to
identify these latter ones. It correlates with the amount of time of spinning the actual object. It's
not just two dimensional but something to do with three dimensional. The greater the spin or
the angle needed to spin, the longer it takes for people to do it. It's less like pictures but more
Mental images are pictures or sculptures. There's also the counter argument that you have
some expectation for the researcher. Most people want to please the researcher and get the
right results. The majority of people want to do that. When people are asked the question: how
long does it take you to get here or rotate this, there's some characteristic being demanded of
them that they try to fulfill. This is not the case though. Even if you tell them what they're there
for, they still take the same amount of time.
If we're asked to picture what something sounds and looks like, auditory versus visual, there are
perceptual mechanisms in place that seem to interfere with each other if the modality is the
same. Why would we want to look at something like this? It was mentioned that the brain is
very efficient in how it utilizes the resources that are there. This is a way to begin to look at
whether there is a competition of resources based on what they're asked to visualize and what
they're asked to respond to. There should be some conflict if they're using the same resources
(visualize and visual task), if not, it should be easy. This research looked at auditory versus visual.
There is interference when modalities are the same. There is overlap in picturing something in
mind's eye and seeing something physically in front of you.
We would say that imagery can interfere with perception (mismatching). It can also facilitate
perception. One way to think about this is that if your task of visualization is congruent with
what you're asked to picture in mind's eye, visual stimulus in picturing something, and it's
similar, it actually helps it, it's a facilitation effect. I.e. if primed first, it helps you respond to that
picture as a stimulus. These things have been looked at in other types of populations. We talked about neglect when
people have neglect syndrome and don't pay attention to one side of their world. The images
that are visualized are similarly neglected as the ones that they experience in real life. One
example here is that someone who had neglect syndrome, they're asked to think about the
town they live in, sitting in the front/main street of the town, and label the buildings. E.g. Yonge
and Dundas, told to label the buildings left and right. The patients with neglect, they give labels
of buildings on the right and they neglect the left side. They have people imagine they're facing
the other way. Now they can label left and ignore right.
We're saying that mental images are pictures. From the studies, it looks like they're using the
same parts of the brain. The next question is that: the imagery and perception, are they
equivalent, is there functional equivalence? If the dots are not in your visual periphery, it takes
more effort to see if the dots are close in proximity (in the study). These studies looked at the
visual acuity of the dots, people imagine these two dots. Shown then imagine. What people
imagine are very similar to what they see in real life.
People who have been blind since birth also demonstrate the same effects in mental-rotation or
image-scanning tasks, with response time being proportional to the distance travelled. Thus we
need to distinguish between visual imagery and spatial imagery. Spatial imagery may be based
in movement or body imagery, or it may be abstract and not tied to any one sense.
Vivid imagers versus non-imagers
Recall the puppy and snow. There's a good number of people who claim that they can't actually
see images in their mind's eye. So when we're asked a question as such, people would
understand but they don't see a puppy, an actual dog there. We started to call these people
vivid imagers or non-imagers.
Eidetic memory or photographic memory. In autism (Rain Man), this is possible. There's also the
hypothesis that 5% of kids has E.M. aka P.M. In all the things we talked about in memory, people
who have good memory are better at certain things like processing things deeply i.e. creating
mnemonics around it. There's a few cases where in this Alice in Wonderland picture, kids shown
this for 30s, and are asked how many stripes on cats back, how many pedals on flower, they can
provide you good detail. There isn't much note about why this is the case. It's not something
that is very common at all. With the rarity of it and lack of research, it's something that we know
exists and it comes to the question that - if we figured what occurred, would it be something
that we want to instill in ourselves or our