On Sun,  2 Nov 2008 15:08:59 -0800 (PST), Annette Taylor wrote:
[snip]
> Mike: OK, a new book on my amazon wishlist: Artificial Dreams. 
> Interestingly, it has no reader reviews, 

Expect APA's online book review journal PsycCritiques to have a
review in the near future.  Some obscure Tipster wrote it.

> and is a bit pricey. Hmm, maybe I should get it from the 
> library :) 

Or you could get it in paperback.  On Amazon the paperback
version is $24.99. Similar price at Barnes & Noble online
but if you have a B&N membership, it's $22.49. At Buy.com
it's $19.99.

>How do you find time to keep up with all the literature?

Oh, it just seems to come my way.

> I knew I could count on tips folks to get a discussion going; I will follow 
> along. 
> 
> However, I think I still stand by my current thinking (but am willing to 
> change) 
> that for the junior/senior level undergraduate what I want to teach them is 
> more 
> strongly focused on our incomplete knowledge developed from behavioral 
> studies, rather than spending too much time talking about imaging correlates 
> of 
> incomplete knowledge. Does that make sense?

I'd suggest that focusing on experiments which have been replicated
many times over and produce consistent data is more important than
the theories used to explain them because theories are (a) tentative
explanations which will be modified or abandoned as new research
identify the flaws/limitations of those explanations and (b) given that
much of the cognitive neuroscience research is relatively new, one
should wait for a good number of replications before investing too
much meaning or significance in them (e.g., on pages 50-51 of
Bruce Goldstein's 2nd ed Cog Psych text, he referes to research by
Quiroga et al 2005 that suggests the existence of "Jennifer Aniston"
or "Halle Berry" detector cells based on the "grandmother cell"
idea which is covered earlier on pages 40-42;  I don't know what
is actually going on Quiroga's situation but I think that it might be
irresponsible of Goldstein to present such research because it is
so recent and the theoretical explanation, I think, is wanting).

With respect to point (a) above, one can use the following
to illustrate the role of method/data relative to theory:

(1)  Cermak & Craik's (1970) text "Levels of Processing in 
Human Memory" provides a number of chapters that identify
why the original explanation (i.e., sensory memories = fragile,
semantic memories=strong) is wrong and other explanations
have to be considered (I'm somehat partial to Douglas
Nelson's chapter).  So, we can readily replicate a levels of
processing effect but what is its explanation?  J.J. Jenkins
chapter which presents the "tetrahedral model of memory"
might be one thing to present in order to direct thinking about
the study of human memory.

(2)  I assume that people still do Shepard & Metzler type
studies of mental rotation because these are available on
opl.apa.org and other websites.  But, though these results
are consistent with an interpretation that visual mental images
have psychological reality (i.e.,  mental representations 
have pictorial properties which can be mentally transformed
such as rotation).  However, back in the 1970s there was 
a significant debate between people who supported
an "analog" view (i.e., mental representations preserve the
properties of the objects they represent; visual mental images
have the 2-D properties of 2-D physical images) or a propositional
view (i.e., "mentalese", the language of cognition is purely abstract
and mental imagery is an epiphenomenal by-produce of more
fundamental abstract processing).  Zenon Pylyshyn was and still
is associated with the latter perspective.  However, John R.
Anderson (1978) in Psych Rev argued that there was no real way to
decide whether mental representations had sensory properties or
purely abstract representations since abstract representations could
have all of the info contain in an analog/sensory representation).
The re-emergence of this debate in the early 2000s by Kosslyn
and Pylyshyn (see their exchange in the journal "Trends in Cognitive
Science") shows that old arguments die hard and though Kosslyn
would like to use a neuroscientific explanation (i.e., visual mental
imagery activate the areas of the brain that appear to be involved
in visual perception), he also has to concede that, as Pylyshyn notes,
the neuroimaging results have note been consistently supportive.

(3) Finally, I like to use the Psych Review article by Tien Ming Chang
(1986) "Semantic Memory: Facts and Models" which lays out the
empirical results about semantic memory know to that time as well as
the list of theoretical models developed to explain the results.  I think
it provides an interesting example of how to systematically look at the
relationship between research results and their theoretical explanations
and how to decide when certain theoretical positions should be
abandoned in favor of theories that are more consistent with the empirical
research.

I'll shut up now.

-Mike Palij
New York University
[EMAIL PROTECTED] 


---
To make changes to your subscription contact:

Bill Southerly ([EMAIL PROTECTED])

Reply via email to