The Phenomenon of Speaking-Time Hierarchies

We live in an interdependent world. Addressing societal and environmental problems requires interdependent group efforts.

The need to come together in efforts to respond to the COVID-19 pandemic and the climate crisis provides two stark examples, but public consciousness of a host of other problems is increasing. This can be observed in daily conversations and discussions we have with family members, workmates, friends, and people we meet on the street; it manifests in the discussions taking place among government administrators and corporate board members, in media groups, think tanks, online discussion groups, and broader social media exchanges.

But the extent to which group discussion leads to solutions that reflect egalitarian knowledge sharing and collective intelligence often depends on very simple things, including the balance of speaking time among group members. Who dominates the conversation? This important question was the focus of an interesting study by Khademi and colleagues (2020).

One might assume that interdependent tasks and problems that require coordinating inputs from individual group members would prompt less dominance, less of a steep hierarchy in terms of distribution of speaking time, and a more egalitarian approach to working together in general. This seems optimal, particularly if each group member has unique knowledge that can contribute to collective problem-solving.

But knowledge is often hidden from view: Group members don’t necessarily know what other people know. And people are not prompted by default to adopt an egalitarian approach to knowledge sharing in groups; they simply begin talking and working together, and some voices emerge as more dominant. Interdependent work is also complex, because individuals with unique knowledge must coordinate their activities. This is difficult in practice as more coordination implies more complexity in the group dynamics.

As noted by Khademi and colleagues (2020), when a task or problem is more complex, the chance of conflict increases. Hierarchy emerges as a solution, potentially reducing group conflicts by offering a control mechanism to resolve conflicts (Keltner, Van Kleef, Chen, & Kraus, 2008; Cantimur, Rink, & van der Vegt, 2016), which may support better performance (Ronay, Greenaway, Anicich, & Galinsky, 2012).

Of course, a hierarchy that functions to reduce conflict is different from a hierarchy that simply dominates the airwaves. And in a situation where unshared knowledge needs to be shared in order for collective knowledge and problem-solving capacity to grow, if some people dominate speaking time, knowledge sharing will be reduced across the group.

The hierarchies that emerge in so-called leaderless groups are interesting: Dominant voices emerge and indeed people may become "leaders" over time simply by dominating speaking time. This reflects an interesting group dynamic as there is no selection of the dominant member(s) or group leader(s)—for example, based on their specific role (e.g., managing, facilitating, coordinating group activities) or competencies (e.g., specialised knowledge, skill, or experience in addressing related problems). As Khademi and colleagues note, it’s important to understand how hierarchies emerge and under which conditions they might become steeper or flatter.

Khademi and colleagues hypothesised based on previous research that one way to prompt an increase in knowledge sharing and a more egalitarian distribution of speaking time among group members is to provide feedback on speaking time. They examined this hypothesis experimentally by working with 135 groups, each of which included 3 people engaged in discussion as part of a personnel selection task.

In the experiment, groups were randomly assigned to one of two speaking feedback conditions, in which they were provided either with (1) feedback on total group speaking time, or (2) feedback on their individual speaking time and the speaking time of each group member. This was achieved using a visual display linked to real-time audio inputs.

Intelligence Essential Reads

What DNA Can—and Can't—Tell Us About Intelligence

How Do You Know What You Know?

Information on four job applicants (all seeking employment as a pilot) was provided to each group member. Based on the information provided, only one applicant was the best candidate for the job (i.e., based on the balance of positive and negative attributes).

Importantly, in addition to the experimental manipulation of speaking feedback, groups were also assigned to one of two task conditions—a high or low interdependent task condition—based on the requirement to share information.

In the high interdependent task condition, group members each had six pieces of information about each candidate, two pieces of information held by them alone (i.e., in the resumes they were provided with), and four pieces of information known to all group members. They would need to exchange their uniquely held information to identify the best candidate.

In the low interdependent task condition, each group member had all the information needed to identify the best candidate; they all had the same information. As such, the requirement to share unique information was not a feature of their discussions.

Regardless of their task condition, group members were required to memorize applicant information prior to group discussion. This made salient the importance of speaking from memory rather than referring to written notes. After a brief warm-up exercise in which group members introduced themselves, the groups engaged in discussion as their speaking time was recorded.

As expected, some group members dominated the discussions early on, while others spent less time speaking. As such, a "hierarchy" of speaking times that was more or less steep emerged in each group.

However, among those who initially dominated discussions, their dominance decreased over time in the condition in which they were provided with real-time feedback on their individual speaking time alongside the speaking time of other group members. In other words, the steepness of the hierarchy was reduced when each group member could see how much time they spent speaking.

Khademi and colleagues also found that a steeper hierarchy of speaking times emerged in the high interdependent task condition in which the exchange of unique information was central to the task of identifying the best candidate for the job. While dominant speakers emerged in both task conditions, the steeper hierarchy in the interdependent task condition was largely explained by the fact that less-dominant speakers decreased their participation over time by speaking less.

These are interesting findings. We all appreciate how speaking time provides group members with an opportunity to share their knowledge and influence others. It’s central to a collaborative and democratic ethos that many of us value, and it requires an egalitarian approach to dialogue and discussion among group members that somehow moderates the influence of dominant voices.

The most dominant voices are not necessarily the most knowledgeable, experienced, or skilled. Providing speaking time feedback to members of working groups is one way to facilitate more balanced and equal contributions.

While we might assume that more balanced and equal contributions will arise more naturally when people come together to address interdependent problems, like many of the societal problems we currently face, the results of the study by Khademi and colleagues warn against this assumption. In reality, steeper dominance hierarchies may emerge when a group is faced with more complex, interdependent tasks that require more group coordination. We need to be conscious of this and work to facilitate groups as best we can.

LinkedIn image: fizkes/Shutterstock