What is this thing you call "behavior."

I submit that there is nothing useful to be gained from calling some subset of actions or experiences “behaviors.” How do you know something is a behavior, specifically?

The most common answer I’ve heard is that a behavior is observable by others.

To which I reply: What about actions that are not observable by others, such as things you do when nobody is around?

The common answer is: Behaviors are things that are potentially observable by others.

What leads me to ask: What things are not even potentially observable by others?

I welcome your ideas!

1 Like

I wondered what someone who devoted his entire adult life to understanding behavior thought, so I looked up Mr. B.F. Skinner. Here’s the most relevant section from his first book on the topic, The Behavior of Organisms.

It is necessary to begin with a definition. Behavior is only part of the total activity of an organism, and some formal delimitation is called for. The field might be defined historically by appeal to an established interest. As distinct from the other activities of the organism, the phenomena of behavior are held together by a common conspicuousness. Behavior is what an organism is doing or more accurately what it is observed by another organism to be doing. But to say that a given sample of activity falls within the field of behavior simply because it normally comes under observation would misrepresent the significance of this property. It is more to the point to say that behavior is that part of the functioning of an organism which is engaged in acting upon or having commerce with the outside world. The peculiar properties which make behavior a unitary and unique subject matter follow from this definition. It is only because the receptors of other organisms arc the most sensitive parts of the outside world that the appeal to an established interest in what an organism is doing is successful.

By behavior, then, I mean simply the movement of an organism or of its parts in a frame of reference provided by the organism itself or by various external objects or fields of force. It is convenient to speak of this as the action of the organism upon the outside world, and it is often desirable to deal with an effect rather than with the movement itself, as in the case of the production of sounds.

However, this broad definition of “movement” begins to break down as we think about observability (or potential observability) of an action. Where do we break down observability? We can observe minute electrical firings of neurons in response to thoughts, but is thinking behavior?

Upon further research, it appears that some animal behaviorists drawn a line in the sand between organism and organs. That is, if the action is limited to an organ then it is not regarded as a “behavior.”

What I mean here is that, as Levitis et al. note, sweating in response to increasing body temperature is not generally thought of as a behavior per se. But when an animal moves to the shade in response to heat and its own sweating, most ethologists would agree that this is a behavioral response - (link)

1 Like

Fun topic! I’m formally trained in behavior analysis. Actually have a post on the site about some of the reading I did in the field somewhere. Couple questions:

Gary - what field are you approaching your definitions from? Any citations or just ones that you have heard circulated?

Ernesto - the field has evolved quite extensively since Behavior of Organisms was published in 1938. I’ll supply a slightly more recent article that is quite an it longer, but comprehensive “one stop shop” for someone interested in the almost 80 years the science has progressed: https://www.researchgate.net/publication/265124757_Evolving_the_Future_Toward_a_Science_of_Intentional_Change

One thing to note is the authors, the multidisciplinary approach they take, their roots in single subject and group research design, and the scale of projects they have affected change in successfully.

Much as technology and QS practices have changed since 1938, Behavior Analysis has. Looking forward to an informed discussion!

1 Like

Thanks Ryan and Ernesto for getting this going. Ernesto and I come at it with some history of thinking together (from different perspectives); Ryan this thread was prompted by our previous short exchange in the QS Slack. But since we are starting a new forum topic I think we should go “from scratch.” I hope others will join.

My skepticism about the concept of behavior is not limited to a single specialized definition, but is general. That’s why I began with the question: “What is this thing you call behavior.”

Ernesto begins with Skinner, and picks out a good, typical passage in which Skinner uses dubious but scientific-sounded vocabulary, such as “the receptors of other organisms” in a circular argument that conspicuously fails to do what he wants it to do. (That is, supply a definition that makes behavior “unitary and unique subject matter.”)

He says that we know something is a behavior because it acts on the outside world, where the outside world is identified specifically with observations by others, who recognize that a behavior has taken place because it has acted upon them. In other words, we recognize a behavior because a behavior is the kind of thing that we recognize. Oy.

Ryan’s citation to Wilson-Hayes-Biglan-Embry’s paper puts us on different terrain altogether. Here, thoughts are understood to behaviors, because thoughts are the outputs of an evolutionary processes that drives change more generally. Without touching on the details of their argument, which strikes me as, mainly, a strained attempt to rescue certain useful techniques that developed in the context of behavioral and cognitive psychology from the conceptual nonsense associated with their origin, I’ll just note that, in contrast to Skinner, they have given up on limiting the term “behavior” to some unique and specifiable domain. If behavior can both thoughts and unconscious mental processes, as well as intentional acts, unintentional actions, and ambiguous movements whose status as intentional is uncertain, what kinds of things are not behavior?

I think you have to look at this practically. There are 3 things that can be internally observed by a person - feelings, cognitions, and behaviors. To another person body language and behavior convey the first, speech and behavior conveys the second, and behaviors communicate themselves. What causes these 3 things is more complex, but includes memory, history, current environment, genetics, personality, disease, and other factors.

Thanks Gary.

Ernesto’s citation may be circular in how it was referenced, but it’s important to remember that was Skinner’s 36 publication out of about 311. As I said before, 80 years have passed. So that’s not really a worthwhile conversation unless it’s just to talk about historically. I’ve read Skinner’s work, start to finish. His system changed quite drastically throughout his publications. I’ll agree it wasn’t perfect, but it’s being misinterpreted here if we are saying it’s an accurate representation of the field. There’s over 30 forms of behaviorism currently. There’s a Philosophical worldview that the article I linked stands from, functional contextualism, which informs the conceptual foundations of the science of behavior. Are you familiar with Pepper’s World Hypotheses? Maybe we can use that as a starting ground. It’s the “base” of the article and is required to understand their central points - including the concept of behavior.

I think we can critique the term when we first are clear what conceptual system we are using. If that includes behavior analysis, then we must agree on which philosophical assumptions we are subscribing to - which will help us lock in which behavioral “lens” we are viewing the world through.

I think it’s also important to remember that any scientific endeavor can be judged off of the pragmatic value of its findings and the technology and processes it produces. Are you familiar with the influence behavior analysis has produced over the years? I’d be curious which projects you think of when you think about the field. Seems your quite knowledgeable! :slight_smile:

Marc - interesting points. What’s the field or conceptual system you’re referencing? Or is it also just your opinion?

Hi Ryan,

In a field as diverse as “behaviorism” it is easy to get lost in citations, so I propose that we do our best to express our ideas directly, citing sources where useful, but not using them as a substitute for arguments. The reason for this is that if we allow mere citations to substitute for arguments, we will not be able to proceed in this forum, because we will have to pause every time a citation is offered to go read it, make our own judgment as to its relevance to this particular discussion, and then restate it’s argument in order to reply. That’s not going to work.

As to Pepper, I disagree with you that Pepper is the philosophical basis of a critical reformation of behaviorism on which later work can be securely founded. But to explain why, we’d first have to agree that a close look is useful to this discussion, and I think you should justify why you believe this to be the case, if you’d like to take us in that direction.

I agree that the value of a branch of science should be determined by its fruits, and I accept that there is some value in the world of people who think of themselves as behaviorists; however, not a large amount, and mostly useful in spite of the theoretical overconfidence and misleading intuitions supplied by the behaviorist framework. I’d be glad to go in the direction of examples, or take the lead from you.

Wondering if there were yet other ways to approach this (from other disciplines/perspectives). For instance, if the ‘self’ is a complex system, then the emergent and evolving patterns arising from the interactions with its environment could be seen as behaviors? Or alternatively, the Function-Behavior-Structure framework by Gero that relates behavior to function and structure (for AI agents, but perhaps applicable to us).

Just thinking aloud :slight_smile:
Any thoughts?!

Thanks Siddharth for the reference to FSB. I’m interested in the way attempts to formalize design processes connects to attempts to more general formalizations of human experience. The way I read this history, behaviorism takes its inspiration from Ebbinghaus, Pavlov, and other early scientific work, which revealed machine-like regularity in certain dimensions of human/animal life. An overconfident BF Skinner proposes the extension of these models to vastly larger domains. His research program fails, but not before exerting a great deal of influence, including over the designers of machines and supposedly predictable “social machines.” However the behaviorist framework fails even when it comes to machines, because undervalues what in FSB is called “the situation.” This leads to a great deal of elaborate reformulation, which has the advantage of being more useful, because it incorporates much of what the reformers have learned from experience.

The disadvantage, however, is that the new formulations have a somewhat arbitrary quality, as more and more specialized vocabulary is brought in to describe a bunch of widely varying insights. I will spare you citations, but you can easily find many, many variants on the FBS framework. When you come to chose which is useful, you’ll end up choosing your conceptual tool in the expert but non-scientific way an experienced builder might choose what hammer to use: what feels right for the job. To the extent that the word behavior makes us imagine that there is a scientific discipline that has discovered how to predict and control human beings (in any of the ways we care most about) it is misleading and gets in the way.

Thanks Gary, some great food for thought. (And thanks for the succinct overview of behaviourist history!) Agree on all counts, this does seem to be the application of (yet another) reductionist framework onto a temporal, complex and dynamic ‘subject’ (- behaviours, and also the self), leading to inadequate answers that need to be reformulated/propped-up constantly. Or like you’ve said, a misleading myth of the ‘universal key’. In any case, the sciences of prediction and control seem to be failing when it comes to most natural or artificial complex systems, be it the weather or financial markets or the next AI wave, not just human behaviour. Not sure if there’s a workaround or solution for any of this (- the inherent limits of knowledge?!) but I do wonder what’s next :slight_smile:

Alright I’ll take the last few threads and try to reply to each segment:

Gary: “In a field as diverse as “behaviorism” it is easy to get lost in citations, so I propose that we do our best to express our ideas directly, citing sources where useful, but not using them as a substitute for arguments. The reason for this is that if we allow mere citations to substitute for arguments, we will not be able to proceed in this forum, because we will have to pause every time a citation is offered to go read it, make our own judgment as to its relevance to this particular discussion, and then restate it’s argument in order to reply. That’s not going to work.”

Sounds good - so long as we are ok with this being a non-informed discussion around the literature and current state of the field of behaviorism. Any critiques are invalid since we aren’t discussing with the most current information. I understand that it’s not easy to read into the literature one is citing, but that’s how this sort of stuff works IMHO. To each their own :slight_smile: This is likely my last post since it’s heading down that route.

Gary: “As to Pepper, I disagree with you that Pepper is the philosophical basis of a critical reformation of behaviorism on which later work can be securely founded. But to explain why, we’d first have to agree that a close look is useful to this discussion, and I think you should justify why you believe this to be the case, if you’d like to take us in that direction.”

The article that I linked above describes the link to Pepper, including the original references of how it was tied to Pepper’s Worldviews in the mid-80s. There’s a pretty prolific research line starts with somewhere around 120 RCTs in the last decade or so. Small in relation to other sciences, but it’s accelerating quite quickly in relation to the research in the 70-80s. If we aren’t reading into any of this and just going off our perception of the field, then again I think I can’t really answer your question directly.

Gary: “I agree that the value of a branch of science should be determined by its fruits, and I accept that there is some value in the world of people who think of themselves as behaviorists; however, not a large amount, and mostly useful in spite of the theoretical overconfidence and misleading intuitions supplied by the behaviorist framework. I’d be glad to go in the direction of examples, or take the lead from you.”

This is a valid critique of the field if you’re coming from anything prior to the 1970s. There’s an immense number of projects demonstrating it’s success (strongest automated reading and reading comprehension curriculum on the market, landmine and TB detection, the only WHO approved process for handling Ebola outbreak, a hands-on approach to working with cultural practices in Sierra Leone to halt the Ebola outbreak, a mandated healthy eating program in every school in Ireland that helps kids like to eat their fruit and veg with a 2 week program that follows up at over 70% at recommended daily amounts, most every assay for identifying the effects of drugs was developed by behavior analysis, the most effective systemic teaching approach identified through Project Follow Through ($1B study from 70-90s), money-back guarantees of teaching kids 2 grade years per year through Precision Teaching, selection process for astronauts… I could go on…) and it’s ironically for the very reasons you reference as it’s weaknesses that these programs succeed. Your reference to machines and behaviorism show that the literature you’ve been in contact with was on the mechanistic side, which was dispelled anywhere from 2-9 decades ago in Behvior Analysis depending on which of the behaviorisms that you look at (remember, Skinner’s was the most famous of over 30).

If you open up to the idea of revisiting the progress that’s been made the last 50-80 years then I’m glad to reciprocate the effort and further discuss these things. I’d appreciate being open to what’s happened and not coming to the approach so stuck in the past and highly misunderstood behaviorism that has been taught the past 50 years.

To all: FBS is interesting - it’s missing a couple key distinctions that have been shown to be quite fruitful in conceptualizations of human performance. Israel Goldiamond and J. R. Kantor’s systems - glad to talk with whomever is interested about the weaknesses of FBS and how the others are stronger. I’ll take that off the forum, since it doesn’t seem that my approach is quite so welcomed here :slight_smile: ryanodonnell23@gmail.com or 775.482.4112

Off to build an empire off behavior analysis! Hope everyone has a good time finishing up this thread and in their personal/professional efforts.

RyanO

Hi everybody. I’m traveling and must keep this brief, but I want to reply to Ryan specifically before returning to this conversation next week. Ryan: Your point of view and experience are welcome! In asking you to make arguments directly rather than allowing citations to substitute, I’m simply asking to keep this conversations open to people who come from a variety of disciplines. While it is common and often necessarily for disciplinary scientists to resist engaging with people who don’t share their own specialization, it’s not always necessary. If scientists from the much older and much more technical fields of chemistry, physics, and biology can discuss methodological and philosophical issues with outsiders, then I think you shouldn’t be insulted by my request to keep the discussion open to non-specialists. I note that you have said goodbye, but I wanted to be clear that there was no offense intended.

Realizing I am quite late to this interesting thread, but I’d like to add that there are connotations in ordinary, lay usage that add cultural and social valences which also matter. In ordinary usage, there seems to be a difference in connotation when we talk about animal behavior and human behavior. Animals appear to be framed as always already working on instinct, as a function of biology, and yet when applied to humans it appears to have a connotation of “that which is subject to change”. The use of the term seems to anticipate a need for intervention in some way. We can see this most clearly when views on what is/isn’t a “behavior” changes. Gay people’s sex used to be (homophobicly) framed as a “behavior”, as a thing someone did and not who they are. Now many people speak in very different terms. The non-technical usages also strike me as assuming a individualistic sort of mind/body/immediate environs loop, which backgrounds societal structures and cultural beliefs. When fatness is a “behavior” problem, we’re usually not talking about food systems, but a moralizing self-control problem. In a technical way, of course, most of us know better, but the use of the word focuses discussion on individual solutions to non-individual problems. Similarly, when I lived in Russia, I was able to feel culturally untranslatable emotions (I think, anyway. They were certainly new to me). Was I “behaving” or just being? Or perhaps just learning the language, which then became available to me without any action per se? Maybe I was only “behaving” in relation to the contrast with English speaking friends wondering what the heck I was doing, because they saw other ways of conducting themselves.

At a personal level, I’ll say that as a woman in the tech field, I definitely do not appreciate it when my style of interactions being called a “behavior” which needs to be adapted in order to fit in. When I hear the word, it’s like there’s a weapon that I’m on the wrong side of.