The influences of facial expression and conducting gesture on college musicians' perceptions of choral conductor and ensemble expressivity

Method

Results

We began our analysis by comparing both experimental pairings between Conductor A and Conductor B. Finding no significant differences between conductors, we combined participants' ratings by conducting condition. Thus, there were overall ratings of “expressive gesture/static face” and “static gesture/expressive face” for each of our two dependent variables, conductor expressivity and ensemble expressivity. All data related to the practice conductor were disregarded. Using a paired-samples t test, we found no significant difference between participants' ratings of conductor expressivity for excerpts that were conducted with an expressive face/static gesture (M = 4.99, SD = 1.67) compared to those with a static face/expressive gesture (M = 5.08, SD = 1.84); t(155) = 0.479, p = .63. Similar results were found for participants' ratings of ensemble expressivity, with excerpts that were conducted with an expressive face/static gesture (M = 5.91, SD = 1.58) being rated almost identically to those featuring a static face/expressive gesture (M = 5.87, SD = 1.69); t(155) = 0.275, p = .78.

To analyze responses to the three open-ended questions, we independently reviewed participants' responses, assigned codes, and combined codes into themes (Creswell, 2007; Gibbs, 2007). Once descriptive coding (Saldaña, 2016) was complete, with comments manually summarized using short words or phrases for categorization, we exchanged the list of themes with one another. To address trustworthiness, the primary author analyzed and coded participants' data, then we all discussed emergent themes together until consensus was reached (Denzin & Lincoln, 2005), thus minimizing overlapping and redundant categories. Saldaña (2016) explained that “some researchers may wish to transform their qualitative data and/or codes into quantitative representations” to “corroborate quantitative data” (p. 86). Because we were interested in a hierarchical understanding of participants' perceptions of expressivity, we counted the number of instances that participants wrote about each category. Interrater reliability, using the formula [agreements/total observations], was 85.71% for Question 1, 86.96% for Question 2, and 92.31% for Question 3. Reliability exceeded the acceptability threshold of 80% proposed by Madsen and Madsen (2016).

Five categories emerged in participants' responses: conducting gesture, facial expression, musical elements, conductor investment, and other. Conducting gestures referred to movements and positions of the arms, hands, and body, as well as the mechanics of conducting. The category of facial expression encompassed all features of the face: eyes, eyebrows, forehead, cheeks, and mouth. We included elements of style, dynamics, musicality, phrasing, ensemble togetherness, tone quality, and the selections themselves in the musical elements category. Conductor investment comments related to the passion, desire, interest, energy, and emotion of the conductor, including how well they connected with the ensemble. The “other” category included other personal preferences of the participants or comments regarding undefined expressiveness. Many participants wrote multiple comments for each question, and we coded each separately to capture the various perceptions.

For the first question, “What factors impacted your ratings of the conductors and their ensembles?” participants wrote 358 total comments. The largest number of comments were in the facial expression category (130; 36.31%), followed by conducting gesture (123; 34.35%), musical elements (64; 17.87%), conductor investment (30; 8.38%), and other (11; 3.07%). Some of the positive written comments included “clarity of conducting, connection to the ensemble, and general show of interest in the music”; “if what they were doing made the ensemble more musical”; “The main factor for the conductors was the range in size, variety in shape, and amount of stylistic indications in the conducting pattern,” and “the conductors' use of multiple arms and their facial expressions.” Negative comments included “None of the conductors showed facial expression and the only parts of their body that were moving were their arms and hands”; “I did not notice any discernable differences in the ensembles, but when the conductor appears more expressive, the ensembles do, too”; and “Sometimes some conductors almost act as a metronome/robot.”

Participants wrote 231 total comments for the second question, “What was the biggest difference among the videos?” The largest number of comments were regarding conducting gesture (95; 41.12%), then facial expression (81; 35.06%), conductor investment (26; 11.25%), musical elements (15; 6.49%), and other (14; 6.06%). Some of their comments included “The conductors had different body language”; “Conductor 2 did not seem interested at all in the music”; “Facial expression in video 3 seemed more lively”; and “The second conductor was smiling, looking at the ensemble and cueing, while the others seemed more stoic and unengaged.”

For the final question, “What do you consider to be the most important factor when assessing the expressivity of the conductor?” our participants' comments followed a somewhat different pattern. Of the 230 total comments, most referenced facial expression (90; 39.13%), then conducting gesture (62; 26.95%), conductor investment (45; 19.56%), musical elements (17; 7.39%), and other (16; 6.95%). These included “animated face”; “if the conductor is passionate”; “how well their motions matched the shaping of the music”; and “apparent engagement with the music.”

Discussion

We investigated whether the presence of either an expressive face or expressive gesture—in the absence of the other—would perhaps sway participants' ratings of conductor or ensemble expressivity. The results of this study indicated no significant differences between college musicians' ratings of either dependent variable for excerpts that were conducted with an expressive face/static gesture versus those that were conducted with a static face/expressive gesture. As evidenced by participants' scores being clustered near the midpoint of our scale (i.e., 4.99 vs. 5.08 for conductor expressivity; 5.81 vs. 5.87 for ensemble expressivity), it is likely that the absence of one aspect of conductor expressivity (expressive gesture or facial expression) made it difficult for participants to rate conductor or ensemble expressivity highly, thus resulting in suppressed overall ratings.

Our findings are different than those of Wöllner (2008), whose participants rated the expressivity of a conductor's face alone significantly higher than either arms-only or whole-body conditions. It should be noted, however, that we intentionally designed this study to set up a comparison between conductors demonstrating either expressive gesture or expressive facial expression alone and not the combination of these two factors. Still, observers' general preference for globally expressive conducting over globally unexpressive conducting seems well-established in the research literature, as we have reviewed. ¹ Future researchers might continue efforts in separating and combining various discrete conducting behaviors in the hopes of pursuing a more detailed understanding of the components of expressive conducting and their effects upon observers.

Our qualitative analyses appeared to confirm our quantitative results. Participants commented frequently on both conducting gestures and facial expressions of the conductors. While slightly more of the participants' comments referenced conducting gestures for the first open-ended question, with 130 facial expression comments and 123 conducting gestures comments, the trend was reversed for the second question, with 95 conducting gestures comments and 81 facial expression comments. There were more comments regarding facial expression for the third question related to the most important factor for evaluating conductor expressivity (90) than comments about conducting gesture (62). It was clear that participants perceived the contrasts in our independent variables, as the majority of comments (176 of 231) in answer to the question “What was the biggest difference among the videos?” pertained to gestures and facial expression. Therefore, it is reasonable to assume that, while understanding how the videos differed, participants' quantitative responses were reflective of their value systems when assessing expressivity.

While responses to the first open-ended question provided us with data regarding factors that impacted participants' ratings of the videos in general, we included the third open-ended question to specifically target a single factor that participants deemed most important when considering conductor expressivity. Conducting gesture was mentioned frequently (26.95%) but was second only to facial expression (39.13%).

For all three questions, many participants commented on the investment of the conductor (101 comments of 819), and that it was an important element affecting their perceptions. We were curious about how our participants determined that the conductor was or was not invested in the music and the ensemble, given that no conductor was seen communicating verbally. It is possible that these were ideas that might be translated into observable behaviors of facial expression, conducting gesture, congruence between musical and gestural/facial affect, or other factors. Passion, interest, and the connection between a conductor and the ensemble are elements that merit continued study, ² both in choral contexts as we have done here, and with instrumental ensembles as well, using a greater variety of musical literature.

Importantly, many of our participants' comments were not value-laden; rather, they reflected a summative statement of their perceptions. We were careful not to categorize their responses as positive or negative. Similarly, very few participants differentiated between the ensemble's expressivity and the conductor's expressivity. We encourage future researchers to explore these elements more richly, perhaps through interviews and focus groups, to delve more deeply into participants' perceptions and with pointed questions unveiling more nuance.

Our data show that both facial expression and gesture are important elements of the overall construct of expressivity. Like Johnson et al.'s (2003) participants, ours referenced facial expression as a preferred mode of nonverbal communication, and its presence was noted. As referenced earlier, the persistent ratings near the mid-point of the scale for both neutral/expressive pairings suggest that the absence of either one is detrimental to perceptions of conductor or choral ensemble expressivity. Still, we should note that although our participants saw conductors from the visual perspective of ensemble members, this was not a live ensemble setting; they did not know the conductors they viewed on screen, and they did not sing in response to what they saw. Importantly, neither did the conductors see or respond to the simulated ensemble. These considerations may help explain the somewhat disparate findings in prior research that demonstrate expressive facial conditions were preferred (Johnson et al., 2003; Van Weelden, 2002), but not uniformly so among all observers (Whitaker, 2011). Moreover, such facial expression is not always used even among expert teachers (Sherrill, 1986), nor is highly expressive gestural movement always seen among experts (Price & Chang, 2005). It seems best to understand these elements as tools in an expert conductor's repertoire, used judiciously toward specific ends (cf. Byo & Austin, 1994).

These findings also suggest that both elements (face and gesture) are inextricably linked when participants consider expressivity. Indeed, we noted that our volunteer conductors found it quite challenging to suppress only one expressive modality at a time. Expressive gesture is a universal component of undergraduate conducting coursework (Romines, 2003); our results highlight the necessity for teaching both gesture and facial expression as part of a comprehensive conductor preparation program. When considering the implications for teacher education, the findings of this study lend credence to the idea that the whole body communicates expressivity. Rather than conceiving of the gesture as the single most important element of conducting, it is clear that facial expression also plays a large role in how others perceive a conductor's expressivity. Future researchers may choose to investigate conducting curricula with these variables in mind.

There are many expressive behaviors that can be used and demonstrated by a conductor. Within that range of choices, a conductor must determine which ones are most appropriate when demonstrating the character of the music. The capacity and procedure to learn these skills remain as areas of continued inquiry. Even if the most important variables in ensemble conducting are difficult to discern and isolate, continued research on these matters is warranted.

Supplemental Material