Chapter 4 – What Is Really Inside?

Reading 4.6 – Do feelings require verbal behavior?

Readings 4.2 and 4.3 (among others) address the behavior analytic approach to feelings. Those readings emphasize a distinction between statements about feelings based on their origins. Some feelings statements are rooted in actual physiological events that the individual can detect and, with the assistance of a verbal community, learn to label and describe as feelings (e.g., “I feel sick to my stomach”). Others are not apparently tied to any detectable physiological events (e.g., “I feel like going to a movie”).

That distinction is straightforward, but neither reading fully examines the role of verbal behavior in how we define feelings. Is it useful to refer to feelings when there are indeed detectable physiological events but no verbal repertoire that allows them to be recognized? This question arises in the case of babies or other individuals who have not acquired a verbal repertoire and non-human species such as dogs.

Dogs offer a great example

Let’s take our faithful furbabies as an example for discussion, in part because it is so tempting to characterize a dog’s behavior as embodying feelings of some sort, such as happiness, guilt, confusion, eagerness, jealousy, and so forth. Consider a dog’s reaction to its owner coming home after being gone for a good while. It jumps around, wags it tail vigorously, and licks the owner’s face – all with an open (smiling) mouth. In vernacular dialect, we say that the dog is happy, implying that it is experiencing the feeling of happiness.

To pursue our question, let’s suppose that there are real physiological events occurring that are related to the owner coming in the door. Let’s also suppose that the dog is wired so that it could detect these events if contingencies were arranged to establish recognizing or responding to them. We certainly don’t know whether those assumptions are true, but let’s start with those possibilities.

We do know, however, that dogs do not have a verbal repertoire. Unlike their owners, they have not been taught to respond to private sensations in any particular way. That is, they do not recognize them in any way or distinguish them from other sensations. How do we know this? Doing so would require a learning history teaching differential responding to one sensation versus another, and dogs have no verbal community that arranges the necessary instructional contingencies. (However, it is certainly possible to arrange contrived contingencies under controlled conditions that allow us to determine a dog’s ability to detect certain private events, such as particular odors.) The sensations merely occur and play whatever causal role a dog’s biology dictates. Perhaps in this example they increase the likelihood of jumping, tail wagging, face licking, and so forth, although we must also acknowledge the contribution of a learning history based on the consequences of such behavior. (Dogs may not jump up because they have been taught to keep all four feet on the floor.) Furthermore, we should acknowledge the possibility that such behavior may be fully learned in the absence of any physiological goings on.

Be careful about “experiencing”

Our verbal community has taught us to call those behaviors emitted in such an environmental circumstance “happiness” or “feeling happy,” but is the dog feeling happy in the same sense that a verbal human might recognize or label their own behavior? Is the dog experiencing a particular physiological feeling under these circumstances? The term “experiencing” is risky in this context because it implies recognition, identification, or awareness of the sensations. For instance, humans are not generally able to detect their blood pressure. Although our blood pressure affects our behavior in some ways, we are usually unaware of it and have not acquired an ability to even crudely identify changes in blood pressure, in spite of effects they may have. Are we “experiencing” our blood pressure? The vernacular term implies awareness, which is more than we should be willing to say.

Chapter 5 points out that awareness requires a verbal community to teach responses to events. We say we are aware of something only if we are able to respond to it with the benefit of a verbal repertoire. Recognizing, identifying, labeling, and so forth can only result from a learning history involving a verbal community. In the absence of such a history, although physiological events may occur that might be detectable in principle, we would have no means that would enable us to respond to them in ways we call being aware of them.

Because dogs lack such a history, particular physiological events merely occur and have their effects on behavior, if any. The dog’s behavior with regard to such events does not involve any awareness or recognition. To repeat an example used throughout the book, this is the same situation as when we are daydreaming while driving and are unaware of visual and other sensations that are nevertheless affecting our driving behavior. We do not say we are “experiencing” these stimuli, which makes the point about the overly broad implications of this term.

How should we talk about feelings?

The question posed in this reading asks what role verbal behavior should play in our definition of feelings. Is it appropriate to view feelings as no more than raw, detectable, physiological events occurring in an organism even though it does not learn to detect and respond to them because the necessary contingencies have not been arranged? Alternatively, should we reserve the term for physiological events that verbal humans have learned to recognize or otherwise respond to differentially? Neither position is right or wrong. The fundamental issue is about our verbal behavior as behavior analysts. Which of these two definitions is more consistent with our conceptual framework and useful in our research agenda? Or, as long as we appreciate the role of verbal behavior, can we have it both ways?

Although the physiological events assumed to be detectable by the organism may be the same in these two circumstances, it is clear that the option of a learning history that establishes the behavior of recognizing, identifying, detecting, being aware of or otherwise responding to a sensation leads to different ways of talking about feelings. With this learning history mediated by a verbal community, we would talk about feelings only with regard to verbal humans. Lacking this history, we would agree that babies, dogs, and other non-verbal organisms would not have learned to be aware of the same physiological events, although they would still have whatever effects the organism’s biology dictated, and would not be described as having feelings.

A dog responding to its owner coming home may be jumping about and so forth because the associated physiological events facilitate such behavior, although we should not ignore the concurrent impact of direct reinforcement on the behavior, which could even fully explain its occurrence. However, the dog would have no means for recognizing or responding to the events. They would merely occur and have their effects on behavior, if any. Assuming for the sake of argument the same hypothetical physiological events, a typical human would greet a long-lost friend with species appropriate behavior but would recognize and label their feelings as happiness. Should we say that the dog is “feeling” in the same way as the verbal human?

If the canine example is not satisfying, consider the familiar feeling we have each had of being hungry. The physiological basis for internal events corresponding to how long it has been since we last ate is clear, and we know what hunger pangs “feel” like. However, what if you were completely non-verbal and had never learned to recognize these sensations in any way? Even that is likely too difficult to imagine, so consider this situation with babies. Sufficient food deprivation may cause crying, but should we say in our technical dialect that the baby is “feeling hungry?” It has acquired no distinctive means of responding to hunger pangs other than what its biology naturally produces. Furthermore, it cannot respond differently in ways learned from its verbal community to different kinds of sensations. Are the hunger pangs themselves what we should consider feelings or should we reserve the term for learned responses to hunger pangs?

Again, this entire argument is fundamentally about how we should talk about our definition of feelings. The conclusion that we should say that an organism can have a certain feeling only if related private physiological events can be recognized in some way seems consistent with human experience and common parlance, not that that those are appropriate standards. The alternative position is that we should acknowledge that physiological events that are detectable in principle constitute feelings in and of themselves, even if the organism cannot recognize and differentiate among them because they have no means of acquiring that skill. Perhaps the most enlightened position is that both verbal and non-verbal organisms experience sensations as private physiological events but only verbal humans learn to respond to them as feelings.