Decoding Memory: A Behavioral Perspective

Co-Authored by Angel Monegro

University of Massachusetts Lowell, Class of 2019

Acknowledgement: Many thanks to Mirela Cengher for her helpful feedback on an earlier draft of this blog post.

“Whatever the event does leave behind it in the organism, it isn’t the memory”

-Ludwig Wittgenstein

One of the assumptions of a radical behaviorist worldview is that all human activity is impacted by past events. This is so much so that we sometimes take the activity of recalling the past for granted. Recollection, however, is a remarkable behavioral feat. The prevailing cognitive theoretical framework for memory in psychology is the storage and retrieval metaphor. Alternatively, the behavioral perspective of remembering is conceptualized as thinking about the past (a private behavior, but still behavior), or as responding to past stimuli and/or behavior. Studies in delayed matching to sample (DMTS) procedures have clearly demonstrated this phenomenon.

What is Delayed Matching to Sample (DMTS)?

Image by from Pixabay

In DMTS, a sample stimulus (e.g., a picture, an object, a sound) is presented briefly and then removed. Then, after a brief temporal distance, two or more comparison stimuli (e.g., two related pictures, objects, sounds) are presented, and if the correct comparison is selected, the selection is reinforced. There are similarities between temporal and physical (spatial) distances. That is, the greater they are, the more difficult it tends to be for learners to respond to three- or four-term contingencies. For example, in Verbal Behavior, Skinner gave the example of showing a child a watch and asking What is it? with the watch still present, and then covering and asking the question after increasingly longer intervals. In this case, the watch is the sample stimulus, and the question What is it? is the comparison stimulus. The interval doesn’t have to be stretched much for the child to stop emitting the response “watch.” Non-human animals (namely pigeons) have been trained to respond correctly to a sample stimulus on a specific retention interval for several sessions. The protocol used in previous studies explicitly trains for better performance at increasing temporal distances; this is a reasonable non-human parallel to a behavior chain produced by popular nursery rhymes that we will describe below.

There are also parallels with our daily experience with memory. That is, as more time passes, we become less accurate in our recollection of the event. However, humans can remember things that happened several minutes, hours, days, and years following an event. This is possible because our verbal communities arrange a complex set of contingencies that put their members under certain types of stimulus control—what we commonly call culture. Verbal communities shape critical repertoires from an early age so that individual members can respond to increasingly complex demands over time. One such practice in early education is the learning of nursery rhymes.

Nursery Rhymes

Image by Tom Burgess from Pixabay

When you observe a preschool or kindergarten classroom in the United States (and in other countries worldwide), singing will likely be on the daily schedule. I (Angel) happen to remember almost none of the nursery rhymes I sang as a small child. Perhaps most adults don’t remember these unless they work with young children themselves. The one nursery rhyme I do remember, a Spanish song by the title of A Mi Burro, has taken on a new light after my first read of Skinner’s Verbal Behavior. Nursery rhymes are not just a way of passing time. They are practices selected at the community level to shape important repertoires in developing children. While I may only remember one song, I think that I owe my basic (generic) remembering repertoire in some parts to that song and many others like it!

The structure of some nursery rhymes is strikingly like DMTS procedures; they require the singer to remember a given verbal stimulus at increasingly longer intervals of time. These are memory exercises, much like math exercises, meant to establish much more complex repertoires that will be useful later in life. A similar song that may be familiar in English verbal communities with the same fundamental structure and function is The Green Grass Grows All Around. This song begins by identifying an object—a hole. It is introduced as follows:


There was a hole. (There was a hole.)

In the middle of the ground. (In the middle of the ground.)

The prettiest hole. (The prettiest hole.)

That you ever did see. (That you ever did see.)

And the green grass grows all around, all around.

And the green grass grows all around….

Photo by Andrew Griswold on Unsplash

Think of the final verse, “The green grass grows all around, all around…” as the terminal response of the exercise. After this, another object is added—a tree. Now the singer is required to name both objects in the order of most to least recently added, as well as to name their spatial relation. (Well, the tree in the hole, and the hole in the ground…), and the terminal response is emitted again. The song goes on like this until six objects in total are named in the same sequence. The song is a way to teach, among other things, a behavior chain by requiring the singer to remember each of the previously named objects from an increasing temporal distance created by the other parts of the song and the increment in the number of objects named. For the singer that sings “from memory” each verbal operant acquires intraverbal control over the next, bridging the distance.

A Behavioral Interpretation

Photo by Allef Vinicius on Unsplash

A behavioral analysis of nursery rhymes as shaping procedures for operant classes from which memory is inferred reveals the true nature of remembering. That is, remembering might not involve responding to the past at all. Instead, it is perhaps more accurate to say that remembering is responding to present circumstances in ways that have been reinforced in the past. We reconstruct past events with our verbal behavior, to an extent, and then respond to that reconstruction as required by the present variables. For example, we can engage in intraverbal responding to “remember” nursery rhymes, visual imagining to “remember” what we did this weekend, or emit echoics and self-echoics to help bridge the temporal gap between sample and comparison stimuli when we repeat the five items we need to purchase at the grocery store. (see Palmer (1991) for a more eloquent account of a behavioral interpretation of memory).

This formulation is like the cognitive concept of mental representations, except that the representations (read: comparison stimuli) are not in our minds, but in the environment. Then, a verbal performance, improvised or highly structured, as in a song, does not come from a specific place, as the storage and retrieval metaphor would suggest, but is the result of history and present stimuli. This analysis also demonstrates that an exploration of early childhood will be required for a full account of our most complex behavior. It may reveal why we remember so little of our early years, important as they were.

If you are interested in learning more about this topic, including how we can teach recall of past events to young children, we encourage you to visit the links listed below.