Michael Caruso made two points about a much-maligned McMaster 
psychology department webpage which I'll deal with in separate e-
mails.  Here's his second point, which I'll consider first:
 
> In the second [McMaster] graphic he [the author]
> explains that when a negative reinforcer is
> present following the response, this is punishment and decreases the
> strength of the behavior (no error here).  The graphic also says that
> the removal of a negative reinforcer following a behavior increases
> its strength (no error here either).  He calls this escape, where most
> intro texts would use the term "negative reinforcement", but escape
> strikes me as an acceptable, if less common, term.
> 
> What exactly are the errors in the graphics?
> 

I believe the error is that the second graphic mixes definition with 
empirical observation. Historically, negative reinforcement was first 
defined as punishment (application weakens response ). Then it 
switched to its current usage (removal strengthens response). If a 
negative reinforcer is _defined_ as a stimulus whose removal 
strengthens response, then it turns out to be an empirical 
observation that such stimuli, when _applied_, often act as 
punishers.

A general question is  to ask whether stimuli which are negative 
reinforcers (when removed) will _always_ be punishers (when applied). 
Certainly, that's my impression. But through many editions of his 
textbook the eminent behavior modifier Alan Kazdin has maintained 
there are exceptions to this rule. For example, in his latest 6th 
edition of _Behavior Modification in Applied Settings_ (2001), he 
says in the legend to Figure 2-1 (p. 57):

"The figure implies that a particular event that can negatively 
reinforce behavior can also be used to suppress (punish) some other 
response that it follows. Although this is usually true, many 
exceptions exist. It is not necessarily the case that the same event 
whose removal negatively reinforces a behavior will suppress a 
behavior when it is presented or vice versa".

Long ago, I queried him on his "many exceptions", and got a confusing 
and (in my opinion) unsatisfactory reply. But whether he's right or 
not, the conclusion remains: if you _define_ negative reinforcement_ 
as increasing response strength by removal, then you should not also 
be defining it as decreasing response strength by presentation. That 
should be left to empirical observation.

But the short answer to what's wrong with the graphics is that 
students are confused enough by our current definition of negative 
reinforcement without adding to their misery by stating it in a way 
which can only confuse further. Current consensus is that a negative 
reinforcer should never be defined as "a stimulus that reduces the 
probability of any response it follows" ( which is the what the 
legend accompanying the first McMaster graphic tells us), not unless 
we want to totally confuse students and have them flunk the question 
on the GRE.  However, it may be empirically observed to have that 
effect on many, if not all occasions.

Now on to the first point in my next e-mail (q.v.)...

Stephen
___________________________________________________
Stephen L. Black, Ph.D.            tel:  (819) 822-9600 ext 2470
Department of Psychology         fax:  (819) 822-9661
Bishop's  University           e-mail: [EMAIL PROTECTED]
Lennoxville, QC  J1M 1Z7
Canada

Dept web page at http://www.ubishops.ca/ccc/div/soc/psy
TIPS discussion list for psychology teachers at
 http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips    
_______________________________________________


---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to