Social Referencing and Social Appraisal: Commentary on the Clément and Dukes (2016) and Walle et al. (2016) articles

We welcome the attention paid in these two articles to social aspects of the emotion process. Social referencing and social appraisal are by definition processes that entail the presence (usually physical, sometimes imagined) of one or more others. Despite the fact that this is the typical set of circumstances in which emotions are experienced (Shaver, Wu, & Schwartz, 1992; Wallbott & Scherer, 1986), this inherently social setting is not one that features prominently in the psychological research on emotion (Fischer & van Kleef, 2010).

There is much that we like in each article and several points that could be pursued, but here we focus on the similarities and differences between social appraisal and social referencing, which is the main point of disagreement between the two sets of authors. Whereas Walle and colleagues (2017) regard social referencing and social appraisal as “coterminous,” Clément and Dukes (2017) regard social referencing as a special case of social appraisal. Each position is defensible.

A first point worth noting is that there is general agreement between the two sets of authors about what social appraisals are: the appraisal by an individual of others’ emotional reactions to a stimulus or event, which then informs the individual’s own emotions and behaviors. We agree with Walle et al.’s (2017) position that social referencing and social appraisal are “functionally equivalent” in the sense that in both cases an emotional response is shaped by witnessing one or more others’ appraisals of the same event. Moreover, these authors have a case when they argue that theorists have tended to use the two terms interchangeably. They point out that “social appraisal” was used by Campos and Stenberg (1981) in one of the earliest publications on what came to be referred to as social referencing. Similarly, social appraisal theorists (ourselves included; Manstead & Fischer, 2001) often refer to the social referencing literature in explaining what they mean by social appraisal.

We also endorse Walle et al.’s (2017) implicit argument that we should avoid making fine distinctions between concepts and constructs that are at some level equivalent. However, we also agree with Clément and Dukes’s (2017) argument that we should avoid using the same terms to refer to underlying processes that differ in some important respect(s). In the view of the latter authors, a criterial attribute of social referencing is that the “knower” (prototypically a caregiver) engages in “ostensive” communication, intentionally communicating to the uncertain child whether it is safe or unsafe to cross to the deep side of the cliff, or to engage in interaction with a stranger. They propose that this is a special case of a more general process of social appraisal, which in turn is a subtype of a still more general process of what they call “affective social learning.” In their view the key difference between social referencing and other forms of social appraisal is whether the knower seeks to inform or influence the learner. In social referencing the knower intends to influence the learner. In “affective observation,” on the other hand, the learner simply observes the knower interacting with a person or object, and makes evaluative inferences about the latter on the basis of the knower’s expressive behavior. Both social referencing and affective observation are instances of what Clément and Dukes call “social appraisal.”

The question is whether this distinction between social referencing and affective observation is one that is scientifically useful. In answering this question, it is worth considering the two central panels of Figure 2 in the Clément and Dukes article (2017), because these capture what these authors regard as the key process differences between the two constructs. In one case, social referencing, the learner interacts with a stimulus that gives rise to uncertainty, or even anxiety. The knower monitors this interaction and deliberately modulates it by engaging in ostensive communication. In the case of affective observation, the knower interacts with a stimulus that generates emotionally expressive behavior in the knower without any explicit intention to communicate this to the learner; the learner observes this interaction and draws an inference about the emotional qualities of the stimulus.

This distinction is, we think, conceptually clear and is likely to be important in understanding the role played by emotional expressions in social learning. Take a child who interacts for the first time with another child who belongs to a different ethnic group. The first child may be uncertain and look to his or her caregiver. The latter’s smiles encourage the child to interact with the outgroup child. The adult thereby actively teaches the child a “lesson” about the emotional value of outgroup members. Contrast this with a situation in which a child witnesses his or her caregiver interacting warily with an adult who is a member of a different ethnic group, and making a negative expression behind the other adult’s back when the interaction is over. Here the adult unwittingly teaches the child a lesson about the emotional value of outgroup members. These two sets of circumstances differ in significant respects (not least in terms of the implications for interventions).

However, if we were to focus on the value of this distinction for understanding how emotional experience and behavior are influenced by others, we might arrive at a different conclusion. To understand the way in which social influences operate on the emotion process, it makes little difference whether the learner’s emotional response to a stimulus is shaped by being actively steered by expressive behavior that is deliberately communicated by a knower, or by witnessing the knower interacting with the stimulus and apparently enjoying (or disliking) the experience. Either way, the learner’s emotional response to the stimulus is shaped by inferences made about the knower’s appraisals.

A final point is that we believe that there is another potentially important difference between social referencing and social appraisal that is not explicitly mentioned in either article, namely the circumstances under which these phenomena take place. Social referencing is likely to be used in situations that are novel and the individual is seen to need others’ appraisals to help him or her to disambiguate the situation. In the classic social referencing context, the knower’s appraisal becomes the child’s appraisal. In principle, social appraisal applies to a broader set of circumstances, having the potential to shape both initial appraisals of a stimulus and subsequent reappraisals. Thus passengers who are initially calm during a turbulent flight may become anxious when they hear the screams of a flight attendant, and students who are mildly disappointed with how their essays have been graded may become angry when exposed to fellow students’ appraisals of unfairness. It is not that others’ appraisals of the event are adopted because the passengers or students are uncertain about how to appraise it; rather, others’ appraisals serve to modulate an initial appraisal. Thus social referencing is likely to occur early in the emotion process, whereas social appraisal could occur at virtually any stage. This is a subtle but potentially important difference when it comes to identifying the processes that are triggered when we are exposed to others’ emotions.