By Todd A. Ward, PhD, BCBA-D
Founding Editor, bSci21.org
The distinction between positive and negative reinforcement is useless! That was the conclusion published by Baron and Galizio in The Behavior Analyst. Though Jack Michael was the first to abhor the distinction in his famously titled article “Positive and Negative Reinforcement, a Distinction that is No Longer Necessary; or a Better Way to Talk about Bad Things”, Baron and Galizio revisited the issue decades later.
We are all familiar with the the standard definition of positive and negative reinforcers as stimuli which increase the likelihood of the behavior that produced their presentation or removal. Sounds pretty cut and dry, right? I mean, who could argue against the validity of such a clear and obvious distinction?
The authors restated Michael’s original argument as follows: positive and negative reinforcement refer not to the presentation or removal of a stimulus, but to changes in stimulus condition from one moment to the next. The presentation and removal of stimuli are always present simultaneously.
Baron and Galizio cite an example wherein an animal kept in a cold chamber will be more likely to press a lever if the response is met with heat. A clear case of positive reinforcement, right? After all, heat was added to the environment and the response that produced it became more likely. Not so fast, says the authors. The same behavior also produced the removal of cold air — negative reinforcement. A presentation, and removal, of stimuli co-occurring in the same event.
However, textbooks retain the distinction between positive and negative reinforcement and the distinction is ingrained in every behavior analyst as part of ABA 101. Moreover, the main categories of functional analysis retain the distinction at its very core with terms such as “escape from demands,” and “attention-maintained behavior.” To the latter, Baron and Galizio ask “Is it better to speak of the consequence as increased attention or as relief from loneliness? As escape from an aversive task or as access to an alternative activity?”
What do you think?
Let us know in the comments below. But before you do, be sure to check out the full article, which has many more details not discussed here.
Also be sure to subscribe to bSci21 via email to receive the latest articles directly to your inbox!
Todd A. Ward, PhD, BCBA-D is President of bSci21 Media, LLC, which owns bSci21.org and BAQuarterly.com. Todd serves as an Associate Editor of the Journal of Organizational Behavior Management and as an editorial board member for Behavior and Social Issues. He has worked as a behavior analyst in day centers, residential providers, homes, and schools, and served as the director of Behavior Analysis Online at the University of North Texas. Todd’s areas of expertise include writing, entrepreneurship, Acceptance & Commitment Therapy, Instructional Design, Organizational Behavior Management, and ABA therapy. Todd can be reached at [email protected].
What do you think of Goldiamond’s response? http://philpapers.org/rec/GOLASA-4
From a functional contextualist perspective it doesn’t really matter if what is truly happening in either positive or negative reinforcement procedures. If managed properly, the presentation or removal of an environmental event actually does lead to an increase in a defined behavior. It all comes down to understanding why we tact such events as observers and why we produce certain intraverbals as speakers or writers.
I agree that any change in the state of the environment is simply producing a different state. An analysis of the parameters that can be applied to measuring how State 1 is different (or not different) from State 2) is the source of a plethora of generative verbal behavior that reflects the influence of multiple proximal and distal contextual cues and contingencies. If this level of complexity of an analysis is warranted to “understand” what is going on in positive and negative reinforcement procedures from an academic perspective is one thing, if it is needed for the management of contingencies to effect significant clinical change in an applied setting is another. What “works” in one setting may not “work” in another.
I prefer Timberlake’s analysis, where positive reinforcement involves arranging a contingency between two behaviors – lever pressing and eating. Food doesn’t reinforce unless the organism can eat it. Deprivation (what we typically call the EO), in this case is a reduction in baseline levels of responding – so the organism eats at a lower rate than normal. Thus in our traditional example, access to eating reinforces lever pressing. Food delivery means nothing outside of that context. When considered in this light, a fundamental distinction between positive and negative reinforcement remains. In negative reinforcement, no contingency is arranged between behaviors specifically. At least I can’t think of how it would at the moment. Just my two cents.
Interesting thoughts! I don’t quite understand how negative reinforcement isn’t a contingency though.
Oh I wasn’t indicating that negative reinforcement isn’t a contingency. I was saying that even given timberlake’s framework, the distinction between positive and negative reinforcement is still useful. That they’re different kinds of contingencies. At first glance positive reinforcement would a behavior-behavior contingency relation, while negative reinforcement would be a behavior-stimulus condition contingency relation. These are my off the cuff thoughts. I think there’s plenty of room for debate.
bsci what he meant about no contigency in neg. ref. is that untill the aversive stimuli is present at the environment the organism doesn.t do anything. He only acts when the stimuli (surrogate EO) is affecting him. In pos. ref. though there’s an EO that alters SD’s effectiveness and it evokes organism’s response. so in pos. ref. there are more interlaced contigencies in force.
I think it is an important distinction because of the differences in performance by the organism. Look at Jesus Rosales-Ruiz’s work on the “poisoned cue”. http://reachingtheanimalmind.com/pdfs/ch_09/ch_09_pdf_05.pdf
Skinner (1938) initially used “negative reinforcement” to refer to the decrease in rate of lever pressing after an electric shock was applied contingent on the response; a process we would all label punishment today. In that sense positive or negative referred to the direction of rate change (up or down) and reinforcement simply meant strengthening that rate change. I kinda wish we stuck with that definition. Makes more sense to me.