Optimistic and Adverse Reinforcement by Jack Michael: A Misconstrued Article

Optimistic and Adverse Reinforcement by Jack Michael: A Misconstrued Article

Thanks to the readers who helped me with this paper. Any errors are my very own.

Some terminology in habits science is notoriously exhausting to get one’s head round. One among these phrases is unfavourable reinforcement. Not solely is that this studying course of itself a problem to know, however the terminology itself is counterintuitive. Habits scientists specialise in coaching, educating, and studying, so naturally, if a time period from their very own area journeys individuals up, they’re going to analyze the issue. The terminology for unfavourable reinforcement has already been modified as soon as, within the Fifties to early Sixties. There was extra dialogue since then. This publish is concerning the article that began the newer dialogue, and the way it’s typically misunderstood within the animal coaching group.

In 1975, psychologist Dr. Jack Michael revealed an article named, “Optimistic and unfavourable reinforcement, a distinction that’s now not obligatory; or a greater strategy to discuss dangerous issues” within the journal Behaviorism.

This journal article is extensively mischaracterized, in my view. It’s generally quoted by individuals who use aversives in coaching and search to attenuate that when discussing or defending their strategies. And positively, the title sounds very promising for simply that objective. However provided that you ignore the final phrase about “dangerous issues.”

Some individuals declare the article says that the excellence between the educational processes of unfavourable reinforcement and constructive reinforcement doesn’t exist or is immaterial. They are saying that the distinction between constructive and unfavourable reinforcement is blurred and may’t at all times be decided. Some say that Dr. Michael dismisses all of the potential causes for sustaining a distinction between the 2. That is false (see web page 43 within the paper).

Michael’s paper facilities on higher methods to make descriptions of and determinations concerning the contingent processes of operant studying. The declare that Michael states that there’s little distinction between constructive and unfavourable reinforcement is fake. This declare misrepresents each the main focus and the conclusions of the article. Notice once more the final a part of the title: “A greater strategy to discuss dangerous issues.” 

Within the article, Michael asks whether or not we have to make the excellence between what we name constructive and unfavourable reinforcement. His closing reply is sure, that we’d like the excellence. He concludes, “We have to make the excellence to be able to have a reputation for the dangerous issues in our world.” (web page 43)

Dr. Michael is worried about terminology on two fronts:

  1. He desires to eliminate constructive/unfavourable and current/take away within the descriptions for several types of reinforcement.
  2. He desires to discover a higher nomenclature to point when an aversive is concerned.

He proposes an answer, which I’ll describe beneath.

There are 4 main elements to the paper: a historical past of the utilization of the phrases for reinforcement and punishment, a critique of the present terminology, a bit that explores whether or not we’d like the excellence or not (his reply: sure), and a proposed answer. I’ll summarize every briefly. The next 4 sections are headed with the subtitles used within the paper.

1. A Temporary Historical past of the Distinction Between Optimistic and Adverse Reinforcement

This part includes 75% of the paper and is dedicated to a retrospective of the utilization of the phrases for reinforcement and punishment, beginning with Skinner in 1938. As some individuals know, what Skinner initially known as “unfavourable reinforcement” is what we now name punishment.

A textbook revealed in 1950 by Keller and Schoenfeld (1950) used totally different terminology, and in 1953 Skinner reversed his utilization of the phrases in his personal textbook, defining them as we all know them in the present day. There was a interval of transition—Michael mentions that it needed to have been particularly powerful for the scholars who attended programs on the similar time that employed totally different textbooks—and by the Sixties Skinner’s revised utilization, what we use in the present day, was in widespread use.

I’m not going into element right here, however Dr. Michael did. Eight and a half of the eleven pages of the article are devoted to the modifications in definitions and usages of the phrases and the resultant confusion. This can be a main focus of the article and a serious a part of his criticism of the usage of “constructive” and “unfavourable” with regard to reinforcement.

2. What’s Incorrect with the Current Utilization?

On this part, Michael says, “Since 1953 there should have been hundreds of man-hours spent within the try to stop the learner of behavioral terminology from equating [negative reinforcement] with punishment…”

Regardless that I’m not credentialed in that area, I do know what he means. I’ve spent many hours myself determining the processes of operant studying, with bonus time on unfavourable reinforcement, and plenty of hours as effectively attempting to move on my primary understanding to others. Reinforcement, punishment, and the plusses and minuses may be complicated, particularly since virtually all of the phrases used produce other meanings or widespread metaphorical makes use of.

Michael goes on to explain one other downside that features semantics, on this often-quoted part:

One other issue with present usages is that the important distinction between constructive and unfavourable reinforcement relies upon upon with the ability to distinguish stimulus modifications that are shows from these that are removals or withdrawals, and these latter phrases will not be very passable descriptions of modifications. The circumstances below which we tend to say “current” actually appear to vary from these the place we are saying “take away” in vernacular utilization, however a few of these variations are irrelevant to a science of habits, and there are a selection of circumstances the place the excellence just isn’t simply made.

Michael, 1975, p. 40

Notice that he’s not saying that it’s troublesome to detect the variations between aversive and appetitive stimuli. The difficulty he objects to is the usage of the terminology of presenting and eradicating stimuli.

…In different phrases, from the viewpoint of the behaving organism shows and removals are each merely sorts of environmental modifications. In the event that they differ, the distinction should not be primarily based upon the variables controlling the one that causes the change.

Michael, 1975, p. 40

This part deserves cautious studying. His main objections to the ideas of “presenting” and “eradicating” are that they focus unnecessarily on the actions of the setting or a 3rd celebration and that they’ve societal and linguistic baggage (e.g., he mentions that elimination can sound unfavourable). He says what is actually essential to the topic organism is solely that one thing modified, and it’s the viewpoint of the topic that we must be involved about describing. We don’t want to speak about including or eradicating stimuli, we have to describe dangerous modifications and good modifications from the standpoint of the topic. (Michael makes use of the terminology of “dangerous” and “good” all through the article, which can be a deviation from commonplace apply.)

In different phrases, it seems that “current” and “take away” are abbreviations that may generally stand rather than a extra full description of each the pre-change and  post-change situation. The abbreviation is normally potential within the case of unconditioned reinforcements, though even right here it should at all times be potential to deduce the traits of each pre- and post-change situations if we’re to suggest behavioral significance.

Michael, 1975, p. 41

An attention-grabbing level: he states “current” and “take away” are incomplete descriptions. He’s not arguing to disregard the character of the circumstances the organism finds itself in. He’s arguing towards shorthand. He’s arguing that we have to describe the state of the setting and the character of the change extra precisely to be able to decide the educational course of in play.

There’s rather more on this part about shows and removals not being particular sufficient for scientific utilization, and it’s on this part one actually will get a way of Dr. Michael’s issues.

He additionally addresses an argument that has been happening for a very long time in habits science. It goes like this:

You possibly can’t inform the distinction between constructive and unfavourable reinforcement should you practice utilizing meals since you don’t know if you’re including meals or eradicating starvation.

Varied individuals

Consultants within the area talk about this query earnestly and with goodwill. However additionally, you will see it glibly thrown into arguments by trainers who search to masks their use of aversives (it’s is a favourite amongst drive trainers). I humbly supply my very own research of this query, however right here’s a shock. Dr. Michael addresses this very scenario in his paper.

Once we say that we current a meals pellet to the rat the listener can at all times assume that the pre-change situation is one during which no meals is offered. Let’s imagine that we take away the “no-food” situation, however then the behaviorally essential facet of the change would stay to be described. Once we say that we terminate a 50 volt electrical shock, the following “no-shock” situation can usually go with out additional description, but when it have been described alone little info could be offered.

Michael, 1975, p. 41

He’s saying that just one description is often correct in a selected scenario as a result of the opposite one fails to explain essential elements of the scenario. Once more, he’s arguing that we have to analyze the reinforcement scenario with details about the setting earlier than and after the change, not by specializing in one stimulus and whether or not somebody “offered” or “eliminated” it. What is occurring from the animal’s viewpoint? Is it a “good” change or a “dangerous” change, and does it contain a nasty factor (aversive)?

3. Why Do We Hassle?

On this part of the paper, Michael examines potential variations between unfavourable and constructive reinforcement and discusses whether or not every explicit facet may or must be the rationale we have to make a distinction.

As we discover ourselves making use of behavioral evaluation to increasingly advanced human conditions we discover it more and more troublesome to differentiate between presenting and eradicating, or we discover an growing variety of conditions that appear to contain each. A reasonably widespread response to this case is to keep away from making the excellence, and easily confer with the related environmental change as “reinforcement,” with out making an attempt to find out whether or not a constructive reinforcer is being offered or a unfavourable eliminated. One would possibly effectively ask, then, why we hassle making the excellence even in these circumstances the place it may well simply be made.

Michael, 1975, p. 41

In case your objective is to usually decrease fallout of the usage of unfavourable reinforcement, you possibly can cherrypick the above paragraph with out persevering with and make it seem like Michael is saying the excellence between constructive and unfavourable reinforcement is pointless. Quite the opposite, that is the part the place he particularly rejects that interpretation. He considers 4 causes for making the excellence. He reductions the primary three as follows.

  1. Are the (behavioral) strengthening results of R+ and R- totally different? He solutions that they don’t seem to be any extra totally different than the variations between totally different types of R+.
  2. Do R+ and R-  contain totally different physiological constructions or processes? He doesn’t assume attempting to make this distinction is a good suggestion in view of the altering area, however he leaves room for future analysis. This text was revealed in 1975, earlier than a lot of the present discoveries that confirmed precisely that: that totally different physiological processes are possible concerned (Total, 2013, p. 69). 
  3. Ought to we maintain the present terminology in order to warn individuals solely to make use of “constructive,” not “unfavourable”? He seems to be asking whether or not we should always truly attraction to the double which means of constructive. Once more he solutions no, that we shouldn’t base a scientific definition on a social distinction. (Notice that in utilizing the time period “social distinction” he’s referring to the phrases “constructive” and “unfavourable,” to not the precise studying processes.)

So he rejects three causes for making the excellence between constructive and unfavourable reinforcement. Then, in a bit that’s not often quoted, he goes on to reply his authentic query within the affirmative, saying that we do want a strategy to distinguish the distinction. He says:

The layman steadily finds it essential to establish an environmental occasion or situation as one which he doesn’t like, which he makes an attempt to flee, or keep away from. He might confer with such an occasion as “dangerous” (with out the ethical implications of this time period), “undesirable,” “unfavorable,” and many others., and he additionally has “punishment” to make use of as a distinction with “reward.” A science of habits additionally wants a manner of figuring out such occasions.

Michael, 1975, p. 42

And eventually:

We have to make the excellence to be able to have a reputation for the dangerous issues in our world…

Michael, 1975, p. 43

He’s arguing that we’d like the excellence between what’s at the moment known as unfavourable and constructive reinforcement in order to have the ability to specify when a “dangerous factor” is concerned. So it’s incongruous that this paper is cited in assist of arguments to blur and erase the usage of aversives.

4. The Answer

Michael spends a lot time specializing in complicated terminology within the paper that it’s unusual he doesn’t dedicate more room to creating his answer clear.

However here’s what he wrote.

So, the answer to our terminological downside is to confer with the great issues as reinforcers and reinforcement and name the dangerous issues punishers and punishment. One set of phrases refers to modifications which have a strengthening impact on the previous habits; the opposite to modifications which have a weakening impact. The excellence between two sorts of reinforcement, primarily based in flip upon the excellence between presentation and elimination merely may be dropped.

Michael, 1975, p. 44, daring added by Eileen

The final sentence of that citation will also be taken out of context in a deceptive manner. A hasty reader, or one with an agenda, can declare Michael is saying that there isn’t a distinction between the educational processes we name R+ and R-. However he has already stated that we have to specify when there’s a dangerous factor concerned. He’s arguing to not base the excellence on the terminology of presentation and elimination. 

Lastly he writes:

The arguments set forth above satisfied me about 6 years in the past to cease making the excellence between unfavourable and constructive reinforcement and to confer with the dangerous issues as punishers and punishment.

Michael, 1975, p. 44

That’s the manner he achieves his objectives of eliminating the terminology of shows and removals and discovering a greater strategy to describe the “dangerous issues.”

It’s a disgrace that Dr. Michael doesn’t give some examples of making use of his terminology. However I might counsel a few examples, following his lead. Each of those are what we’d now name unfavourable reinforcement.

  1. In a shock experiment with the objective of accelerating habits, the educational course of might be known as utilizing reinforcement with the punisher of shock.
  2. In an escape protocol the place an animal’s habits is strengthened by giving them extra distance from a scary factor, the educational course of concerned might be known as utilizing reinforcement with the punisher of a feared stimulus.

It appears clunky at first, however when you notice a nasty factor (“punisher”) may be concerned in reinforcement in just one manner, escape/avoidance, it falls into place.

Dr. Michael makes it clear that we have to stipulate when there’s a dangerous factor included as a part of the educational course of. He additionally states that what we name unfavourable reinforcement features a dangerous factor, and presents a cogent argument that the variations between what we at the moment name R+ and R- are essential and are distinct from one another in real-life conditions.


In 2013 (sure, I’ve been engaged on this publish for seven years), I attempted to contact Dr. Michael to ask for some examples of how he utilized his terminology: how and when he made the excellence {that a} punisher was concerned. I reached his spouse, who stated he was not in a position to talk about such issues any longer resulting from dementia. He handed away this 12 months: November 13, 2020.

Dr. Michael’s paper prompted a number of others in the identical vein, questioning the terminology of “constructive” and “unfavourable” with regard to reinforcement. In my studying, the arguments had a number of the similar taste however weren’t precisely the identical. I’ve included these articles within the references beneath. My arguments above apply to Michael’s 1975 article alone.

These papers normally get a footnote in habits science textbooks, however the usual nomenclature hasn’t modified to replicate the concepts put forth, which Michael himself later famous (Michael, 2005). I just lately heard a habits analyst being interviewed in a podcast voice an analogous concern with “shows and removals.” She talked about that in her work it’s most essential to watch whether or not habits is below aversive or appetitive management, and people are the classifications she makes use of.

And allow me one second of editorializing: I don’t know any trainers who don’t use unfavourable reinforcement. Even the kindly act of letting an animal go away or take a break from a troublesome process signifies that R- is a deliberate a part of a coaching plan. Most of us would agree that permitting escape is much less intrusive than flooding, however we additionally strive mightily to coach with sufficient talent that the animal doesn’t wish to go away within the first place. So my goal right here is to not preach purity, though I attempt to keep away from the usage of R- in each potential manner. My argument is with people who find themselves disingenuous about their use, and who cherrypick quotes from this paper to aim to obfuscate the contingent processes of operant studying.

Relaxation in peace, Dr. Michael, and I hope my efforts right here have finished this well-known paper justice.


Baron, A., & Galizio, M. (2005). Optimistic and unfavourable reinforcement: Ought to the excellence be preserved?. The Habits Analyst28(2), 85-98. 

Baron, A., & Galizio, M. (2006). The excellence between constructive and unfavourable reinforcement: Use with care. The Habits Analyst29(1), 141-151. 

Chase, P. N. (2006). Educating the excellence between constructive and unfavourable reinforcement. The Habits Analyst29(1), 113. 

Iwata, B. A. (2006). On the excellence between constructive and unfavourable reinforcement. The habits analyst29(1), 121. 

Keller, F. S., & Schoenfeld, W. N. (1950). Ideas of psychology: A scientific textual content within the science of habits.

Lattal, Okay. A., & Lattal, A. D. (2006). And But…: Additional feedback on distinguishing constructive and unfavourable reinforcement. The Habits Analyst29(1), 129.

Michael, J. (2006). Touch upon Baron and Galizio (2005). The Habits Analyst29(1), 117. 

Michael, J. (1975). Optimistic and unfavourable reinforcement, a distinction that’s now not obligatory; or a greater strategy to discuss dangerous issues. Behaviorism3(1), 33-44.

Nakajima, S. (2006). Hypothesis and Express Identification as Judgmental Requirements for Optimistic or Adverse Reinforcement: A Touch upon. The Habits Analyst29(2), 269. 

Total, Okay. (2013). Handbook of Scientific Behavioral Drugs for Canine and Cats. Elsevier Well being Sciences.

Sidman The Distinction Between Optimistic and Adverse Reinforcement: Some Further Issues

Skinner, B. F. (1938). The habits of organisms: an experimental evaluation.

Photograph Credit score

Skinner Field diagram credit score Wikimedia Commons.

Copyright 2019 Eileen Anderson

Source link