Re: Difference between cognition and perception?

[Christian Kaernbach <mailinglist@xxxxxxxxxxxx>]

Woojay Jeon asked:

> I am wondering if anyone can clarify the exact difference between
> "cognition" and "perception", at least in terms of acoustics, and also
> provide some examples illustrating the difference?
I guess there is no consensus on clear-cut definitions among
researchers. You'll find textbooks that give such definitions because
they are expected to do so, but that would not mean that everybody would
agree. To my students I give a simplistic criterion: as soon as memory
is involved, it's cognition.

You asked for examples. If we study the just noticeable difference (jnd)
in pitch as a function of some manipulations of the stimulus (say,
duration), this would be clearly a perceptual study. I guess nobody
would call this cognitive research. The experiment (let's call it
Experiment 1) would be performed as a two-interval forced-choice (2IFC)
procedure ("in which interval the sound was higher in pitch?").

Now let's make a slight modification of Experiment 1: Let's increase the
pause between the two intervals, say 6 s instead if 0.5 s. In most
literature this (Experiment 2) would now be called S1-S2 paradigm (2IFC
and S1-S2 are in principal both independent of the length of the pause,
but the focus is different). We could now manipulate the pause (e.g.
silence versus tones versus ....), and we would study how pitch is
memorized. Most researchers would agree that this is now cognitive
research. The consensus on calling Exp 2 "cognitive" is, however, a
little bit smaller than the consensus on calling Exp 1 "perceptual".
Please note that Exp 1 would not work if there were no memory: in a 2IFC
paradigm we need memory in order to compare across intervals. Memory is,
however, not in the focus of the research, but just a means to help
understanding processes that are not related to memory.


Christophe Pallier suggested:
> My opinion is that "perception" and "cognition" are not very useful
> concepts in information processing models.
Oh, let's make it a pun and turn this sentence to:
       My opinion is that information processing models
       are not very useful concepts
       in the study of perception and cognition.
;-)

Best regards,
Christian

--
Christian Kaernbach
Institut fuer Allgemeine Psychologie
Universitaet Leipzig
Germany
www.kaernbach.de