How Artefacts Do Leadership: A Ventriloquial Analysis

Literature Review

Method: Ventriloquism and Actor Network Theory

Situating this study within the emerging wave of interest in discursive leadership, in order to analyze the in-situ sociomaterial doing of leadership by human and non-human actants, we use a ventriloquial approach inspired by the Montreal School (Cooren, 2010). This school of organizational communication can be loosely described as consisting of scholars whose research concerns the way in which communicative practices embody the organization, and more especially with how both human and non-human actants play a role in the communicative constitution of the organization.

In this respect the Montreal School partly draws its inspiration from Actor Network Theory (Callon, 1986; Latour, 2005). Key to ANT is the refusal to separate the non-human and the human, thus calling into question the dichotomization of the world into the human, which has agency, and the non-human which are passive objects through which human agency is enacted. Contrary to such a dualist world, ANT supposes that actants, defined as “anyone or anything that might be deemed to make a difference in a situation” (Fairhurst & Cooren, 2009, p. 472), are not necessarily human, rather they are networks of human and/or non-human entities. The classic illustration of this is Latour’s (1999) example of a speed bump—which no doubt inspired Grint’s (2005) later formulation of the way in which traffic lights do leadership. In the example of the speed bump, Latour (1999) argued that when a motorist sees a speed bump she necessarily slows down. Consequently, a non-human actant is active in causing a human actant to do something.

However, as with the example of the traffic light previously cited, it is not the speed bump per se that causes a motorist to slow down, rather it is a network of actants that is invoked and which has agency (see Cooren, 2018). That is to say, it is a plethora of agencies, human and non-human, such as legislation that authorized the speed bump to be built, the workers that built it, the principles of road safety that justified it, and so on that has agency. This agency is enacted through a spokesobject (i.e., the speed bump) which is the visible face of the actor network. The upshot of this is that agency is always hybrid and should always be considered as shared across a plenum of intertwined agencies both human and non-human (Latour, 1999).

Thus, from the perspective of ANT, researchers should not consider the “naked” leader, rather they should investigate the actor networks that constitute the agency of the leader. Thus, leadership is not only about the person, it is also about “how other ‘things’ in our environment lead us, such as architectural designs and layout solutions, physical objects and even city spaces” (Ropo et al., 2015, p. 2). The importance of ANT for leadership research is thus, as Grint (2005, p. 49) demonstrated, that artefacts can indeed lead. However, ANT is not concerned with how actor networks function in naturally-occurring in situ communicative practice. The Montreal School, on the other hand, goes beyond observations about traffic lights and speed bumps and has given ANT an “interactional and organizational twist” (Fairhurst & Cooren, 2009, p. 474), which, therefore, can show how human and non-human actants express and reveal themselves in communication. In doing this, scholars from the Montreal School counter the traditional view of communication as one in which two or more people interact with each other, and they offer a view of communication in which the agency of all the actants, whether human or non-human, is considered.

More specifically, scholars working within the framework established by the Montreal School base their analyses on fine-grained transcripts of naturally-occurring interaction, and their work is thus compatible with a discursive approach to leadership that this study takes. These analyses often use the vocabulary and insights of conversation analysis (Sacks, 1992) to demonstrate exactly how it is that networks of actors (both human and non-human) are made present and so influence the interaction. The advantage of using fine-grained analyses of interaction is that the researcher never leaves the terra firma of an event that actually happened and the researcher can, provided there is sufficient warrant for their claims in the analyses of the transcripts, make observations that are pertinent to wider social theorizing (Caronia, 2018). Observations that are pertinent to wider social theorizing can be made because, taking the linguistic turn in organizational studies seriously, we argue that phenomena such as leadership are not external to interaction, rather they are constructed through the social action that communicative activities perform.

Thus, we argue that since the social action of influencing others to realize group objectives is achieved through communication (cf. Pearce and Conger’s (2003) definition of leadership), it is in the communicative process, made visible in video-recordings and transcripts of naturally-occurring interaction, that researchers can profitably search for leadership. In short, communication is the means by which organizational phenomena such as leadership are established, enacted, and sustained. Thus, whatever claims the researcher makes about whatever phenomenon (such as artefacts doing leadership) must have their warrant in the fine-grained analysis of actual interaction.

Consequently, the approach to analyzing interaction as proposed by the Montreal School becomes an ideal way of analyzing leadership as it occurs in situ and as a way of making visible, and thus analyzable, the interplay between human and non-human actants. This interplay is primarily made visible through the metaphor of ventriloquism, which is defined as denoting “actions through which someone or something makes someone or something else say or do things” (Cooren, 2015, p. 476). It is through such ventriloquial effects that hybrid presences of human and non-humans are made present and relevant to the interaction. However, as Cooren (2010) also pointed out, ventriloquism should not be seen as unidirectional in the sense that it is the human operator who makes the dummy say or do things. Rather, to continue the metaphor, the dummy can also be seen as animating the human ventriloquist.

For instance, during a board meeting, a top manager can ventriloquize what a financial report is supposed to show by making it say that a line of products should be discontinued (in which case the top manager is seen as the ventriloquist, while the report is seen as the dummy). However, this top manager can also be seen, at the same time, as the dummy to the extent that she can be positioned as the means, medium or intermediary by which this financial report ends up having a possible influence on the discussion (in which case, the report is seen as something that ventriloquizes the top manager when she speaks). These effects of ventriloquism are quite important to understand how communication functions as we see that they allow analysts to acknowledge how things such as a financial report can do things and materialize themselves in various discussions.

If the metaphor of ventriloquism allows us to identify how things can be made to say things through various turns of talk (as in the case of the financial report during a board meeting), it also allows us to unveil the same phenomenon even when people do not stage these actions in their own talk. For instance, instead of spelling out what the financial report is supposed to show, the top manager could have simply invited the other participants to read its conclusion, which means that the report would have been here presented as speaking for itself (even if, of course, the top manager can then also be seen as making this report speak for itself). In other words, ventriloquism is also at stake when people invite others to experience something (reading a report, for instance) or even when they demonstrate something, in which case the demonstration is also presented as speaking for itself (as, for instance, when we show what technology can do by demonstrating to others its capabilities, which we will examine later).

A ventriloquial approach thus relies on the premise that the world in all its instantiations speaks to us, not because it articulates sounds as we do when we actually talk, but because interpreting or making sense of this world can always be envisaged as making this world say or do things to us. Instead of reducing sensemaking activities to human beings and human beings only, a ventriloquial approach thus acknowledges the part that-is-made-sense-of plays in these activities. For instance, acknowledging that interpreting or making sense of a number in a financial report relies on our knowledge of accounting and finances does not mean that this number does not say or do anything. On the contrary, it is because of this knowledge that this number can indeed tell us something about the financial situation of a company (Fauré et al., 2019). If the number tells us something, it is not only because we make it say something, but also because it makes us say something.

Similarly, demonstrating the capacity of technology can be interpreted as making this demonstration speak for itself not because the human beings suddenly disappear from the portrait, but because they are capable of understanding what this demonstration is supposed to show or say about the capabilities of this technology. The metaphor of ventriloquism thus allows us not to have to define, once and for all, the absolute origin of action or sensemaking because it precisely problematizes this origin. If the world can indeed be envisaged as speaking for itself, it is not only because we learned to make it speak (for itself), but also because it makes us speak too (Cooren & Caïdor, 2019). Incidentally, this metaphor does not imply that the world suddenly becomes univocal or unambiguous, as we all know of situations where people disagree about what a situation or a document is supposed to tell them. This potential ambiguity and polyvocality, however, means that people might still find ways to finally agree about what this situation tells them. Thus, through drawing on the notion of hybrid presences and through not leaving the terra firma of the analysis of transcripts of naturally-occurring interaction, the metaphor of ventriloquism may prove useful in unraveling exactly how leadership is enacted through the interplay of human and non-human actants.

The Study

Empirical Story

Our case focuses on an overlapping inter-organizational collaboration between three organizations (given the pseudonyms Technica, SquareSoft, and CoolDesign) between September 2012 and October 2013. Technica (based in Canada) was founded in 2011 to produce medical sensors. It had an open and horizontal structure, which co-founders labelled “self-organizing.” Technica had had a series of successes, gathering not only prizes, awards, and funding for its prototypes, but also governmental, academic, and public attention for its business model, self-defined on the website as “open, decentralized, and self-organizing.” Following these accomplishments and increasing local notoriety, along with an open membership orientation that enabled “easy access to participation” (Technica website), Technica experienced exponential growth, from three co-founders in the first semester of 2011 to 120 affiliates by the end of 2013.

As Technica was horizontally structured and affiliation was increasing, around mid-2012, its co-founders demonstrated concerns regarding efficiency and coordination. In this respect one of the co-founders, given the pseudonym Sergi, championed arguments not to compromise the foundational values of horizontality, decentralization, and openness during the growth process. Sergei sought out two inter-organizational collaborations to address the issue. In September 2012, Technica started collaborating with SquareSoft (a company managed by partners given the pseudonyms Joe and Mary, based in Midwest USA) to develop a network planning software. In April 2013, in parallel with the work with SquareSoft, Sergei also started collaborating with the co-founder of CoolDesign (based in West Coast USA), given the pseudonym Brad, on the development of another set of tools. The software is also named CoolDesign (like the company’s name). See Table 1 for a list of dramatis personae.

OPEN IN VIEWER

Table 1.

Dramatis Personae in Inter-Organizational Collaboration.

Organization	Representing member(s)	Activity
Technica	Sergei	Search for software to coordinate Technica’s growth
SquareSoft	Joe and Mary	Development of “Network Coordination” software
CoolDesign	Brad	Development of “CoolDesign” software

Technica’s streams of work with SquareSoft and CoolDesign were happening in parallel, that is, there was limited interaction between SquareSoft and CoolDesign. As this situation progressed, Joe demonstrated concerns that Brad’s work was competing and conflicting with his own work. While Brad was not open to interacting with Joe and Mary directly to discuss their roles, he agreed to a Google Hangout meeting to work out clarifications and determine who would do what and whose software they would use. Figure 1 illustrates this story from Technica’s foundation to the meeting that we analyze in this study.

OPEN IN VIEWER

Figure 1.

Timeline.

Analysis Process

In keeping with Nathues et al.’s (2020) methodological framework for ventriloquial analyses, our analysis consisted of first meticulously transcribing the naturally-occurring interaction. Then we identified both what actions were being performed in the interaction and the invocations (figures/dummies) and animations (vents/spokespeople/spokesobjects) that were enacted in order to perform these actions. We then related the activities that were being performed to the networks of authorship that the participants invoked and made relevant to the interaction. Finally, we selected a “powerfully illustrative” (Nathues et al., 2020, p. 10) vignette that exemplified the networks that were invoked to perform these activities. While traditional forms of conversation analysis—usually identified as a “small d” discourse perspective (Alvesson and Karreman, 2000) —tend to exclusively focus on what people do and how they do what they do (Sacks, 1992), other forms of discourse studies—usually categorized as “big D” Discourse perspectives (Alvesson & Karreman, 2000) —identify the various discursive formations or ideologies that are reproduced through what people say and do (Grant & Hardy, 2004).

In contrast with these two opposite perspectives (small d vs. big D discourse studies), the ventriloquial approach proposes not to choose by acknowledging the presence of multiple co-authors in what people say and do. In keeping with discourse perspectives, the focus is, therefore, on what is done in saying something and how saying something are negotiated in interaction, except that these doings and sayings do not have to be exclusively attributed (by the analysts or the participants) to human beings. Echoing Discourse perspectives, a ventriloquial analysis, therefore, acknowledges that people are spoken and acted upon as much as they speak and act, except that what acts and speaks through what they do and say does not have to be reduced to ideological forms.

Concretely speaking, this perspective tends to focus on what or who is “acting for, with, and through” (Cooren, 2018, p. 412) a given turn of talk or discourse. A typical form of ventriloquial analysis thus consists of identifying what is implicitly or explicitly invoked or evoked in what people say, given that invoking and evoking precisely consists of making someone or something say something (i.e., giving him/her/it a voice, as when an administrative clerk refers to a policy in order to reject a client’s request). But another form of ventriloquial analysis also consists of focusing on what is shown or displayed by participants, given that this type of action typically amounts to letting what is shown or displayed make a difference in an interactional situation. In keeping with the multimodal character of discourse in interaction (Mondada, 2016), a ventriloquial approach thus focuses on the multiple forms of communication (verbal and nonverbal) that constitute interactions: not only what is said, but also what is shown or displayed. This multimodal dimension is what we especially focus on in the episode we selected.

Findings

In the analyses below, we demonstrate how the hybrid presence of Brad and CoolDesign does leadership by influencing the other participants to accept CoolDesign as the software that they will use to achieve the group’s goals of producing medical sensors.

OPEN IN VIEWER

We join the interaction as Brad is manipulating the screen to show what CoolDesign can do. As he is manipulating the software on the screen he is talking the others through what CoolDesign demonstrably seems able to do. For example, he displays a “sort of a work around” (line 4); demonstrates how he will “save this all to new map” (line 11); shows that he “will go ahead and take everything (0.3) in here (0.5) that’s not applicable and remove it from the map” (line 16); and so on. Through this talk and the action of manipulating the screen to show what CoolDesign can do, Brad positions himself and is positioned as the means by which CoolDesign can reveal itself to others.

As pointed out previously, saying that CoolDesign reveals itself through these demonstrations does not mean that Brad does not make any difference to this situation. On the contrary, he is the one who is definitely performing this demonstration. However, making a demonstration precisely consists of trying to show what CoolDesign demonstrably seems capable of doing. Consequently, a hybrid presence of Brad/CoolDesign is made relevant to the interaction. CoolDesign becomes a participant to the extent that someone—here, its designer, Brad—speaks on its behalf and reveals its features to his interlocutors. By showing what CoolDesign is capable of doing, Brad makes an effort to convince the other participants of the capabilities of his software and this allows Brad to let, to some extent, CoolDesign speak for itself.

As in any demonstration, Brad not only tells his interlocutors what his software is capable of doing but also shows it to them. Showing how CoolDesign works also implies, by definition, that what is shown is supposed to make a difference to the situation. Furthermore, if Brad is demonstrating what the software can do, it is also the software he is using at this moment that is supposed to display its capabilities. After all, Mary and Joe are participating in this meeting not to assess Brad’s dexterity but to see what his software can reveal about itself, especially in terms of performance. To this extent we can affirm that Brad’s ambition, at this point, is to let CoolDesign speak for itself, not, of course, because this software would suddenly be capable of literally speaking by articulating words but because its performances are supposed to say something about its features.

Therefore, CoolDesign, as displayed on the screen (see Figure 2), indeed becomes an active participant in presenting a future in which CoolDesign, if Brad can persuade the others to use it, will be part. The hybrid presence of Brad and CoolDesign as ventriloquized through Brad could also be active in influencing the other participants to buy into the use of the software.

OPEN IN VIEWER

Figure 2.

The screen.

We propose that like the traffic light, which Grint (2005) argued has a leadership effect through the formal authority embedded in it, so the hybrid presence of CoolDesign and Brad appears to be meant to influence and lead others through authority. However, on this occasion the source of authority is not legal, rather it is knowledge-based. On the one hand, this knowledge is displayed through Brad’s talk as he explains what CoolDesign will be able to do, yet on the other hand, it is the on-screen manipulation of CoolDesign that displays what it can do. Multiplying the sources of what is being said, which is the essence of ventriloquism, potentially also has important consequences in terms of leadership since it can be a rhetorical technique (Clifton, 2017b) for influencing people to buy into proposals—in this case the proposal to use CoolDesign.

Authority is, therefore, not only about securing obedience by providing a display of knowledge, it is also—and perhaps mostly—a matter of demonstrating how one’s proposals are authorized/authored by a variety of actants that lend weight to the proposal. Authority is enacted because ventriloquizing amounts to multiplying the authors as a way of influencing others. Cooren (2010) indeed noted that the words “author” and “authority” have the same Latin root, auctor, which means the father, origin or source. By having a say in a given discussion, artefacts can make a difference to the way a discussion evolves, for instance by leading the participants to adopt specific courses of action rather than another and so achieve the group’s objectives. These authoritative/ventriloquial moves are particularly interesting to understand how things can be said to do leadership in a specific situation (see also Holm & Fairhurst, 2018).

However, as Holm and Fairhurst (2018) argued, doing leadership is a fluid negotiated process in which authoring may be accepted or resisted so that:

leadership will be located in the struggle over the meanings to be assigned to situations, including whose and which ideas (read, authored acts) are going to prevail vis-a-vis the tasks at hand, the authoritative texts they produce, and the attributional outcomes that follow. (p. 696)

Therefore, Brad and CoolDesign’s authoring of a future in which CoolDesign has its place in the inter-organizational collaboration is not leadership until the proposal is accepted by the others. Brad appears to be aware of this and throughout the talk and through his use of “right,” spoken with a rising intonation (lines 20, 32, 35, and 37), he displays an awareness of the need to solicit agreement and so confirm that the others are following his—or rather the hybrid presence of Brad/CoolDesign’s—lead. In other words, if CoolDesign is supposed to speak for itself, it remains to be seen what it actually says, at least to Joe and Mary.

In this case, whilst nobody interjects, at least Brad displays a need for the other participants to align with and accept his proposal of using CoolDesign. However, we can also see that as Brad and CoolDesign are displaying what the future with CoolDesign might be like, Joe and Mary, who until now have been skeptical of its use (see the empirical story), begin to align with Brad’s vision that CoolDesign should be used. First, this alignment is seen in Joe’s turn in line 9. Prior to this, Brad as he is demonstrating the software states that “what I would actually do here and this is sort of a work around.” As he hesitates (line 8: “er er”), Joe overlaps Brad’s turn in progress and states that he is “good with work arounds” (lines 9–10). This displays that Joe is beginning to be influenced to align with the proposal of using CoolDesign and accept Brad/CoolDesign’s lead. He does this by displaying that he has the abilities that the future with CoolDesign requires and that he can, and implicitly will, participate in such a future.

Similarly, in line 27, Joe states that “yeah this would make sense (.) coz there’s probably more protocols.” This positive assessment of Brad and CoolDesign’s display of what they can do, illustrates that Joe is now beginning to be influenced by the hybrid presence of Brad/CoolDesign, and so he begins to align with an emerging proposal to use CoolDesign that the hybrid presence of Brad and CoolDesign is arguing for. Further, we can also see displays of alignment, albeit minimal, as Joe backchannels agreement: “yeah” (line 30) and “‘kay” (line 42). Mary, who has been silent through Brad’s display of what CoolDesign can do then also displays that she is being influenced by the hybrid presence of Brad/CoolDesign and so aligns with the future that the hybrid presence of Brad and CoolDesign is presenting. In line 43, she aligns fully with CoolDesign and Brad’s extended turn by saying “yeah I like that ability I think we ought to try it and do that so we can have kind of a top-level view that’s very simple.” Significantly, Mary uses the pronoun “we,” which signals that an inclusive group is now relevant rather than the SquareSoft/Joe and Mary and CoolDesign/Brad split that had previously been relevant. While up to this point in the conversation agreement had been sparse and somewhat restrained, when CoolDesign is brought to the fore we see displays of alignment that suggest Joe and Mary have been influenced to accept the use of CoolDesign.

In sum, we argue that leadership, in the form of influencing others to achieve an organizational goal, is not only happening through Brad. Rather, we argue, that the software also manifests itself, playing a key role in confirming Brad’s claims of expertise and convincing Joe that the work SquareSoft and CoolDesign do is complementary rather than duplicative and that, therefore, they should decide to use CoolDesign. In other words, if CoolDesign is supposed to speak for itself, we also learn, through Joe’s and Mary’s reactions, that what it tells them is positive and reassuring, which probably was Brad’s ambition throughout his demonstration. As pointed out before, for this software to speak for itself through this demonstration, it needs, of course, to rely on people’s capacity to understand and evaluate what it actually says (here, Joe and Mary’s experience with such software).

We should also point out that we do not claim that Joe and Mary have had a sudden revelation that CoolDesign can be used. Rather, we are more modestly claiming that aligning and agreeing with the future use of CoolDesign that Brad and CoolDesign are proposing is gradual, and what we are witnessing is part of the process of influencing others to achieve the group’s goals, which is, as Pearce and Conger (2003) argued, the essence of leadership. Further, we also note that Sergei is silent during these exchanges. However, in a later exchange, not analyzed for reason of space, he also signals his agreement and also incrementally moves toward accepting the use of CoolDesign.

Discussions and Conclusions

Returning to the research question: how, then, is it that artefacts do leadership? Modifying Pearce and Conger’s (2003) anthropocentric definition of leadership as “a dynamic, interactive influence process among individuals in groups for which the objective is to lead one another to the achievement of group or organizational goals or both” (p. 1, italics added), we would argue that the influence process is not solely amongst individuals, rather it is hybrid and distributed through networks of human and non-human actants. This influence process is achieved notably through the mobilization of authority that enables Brad to persuade Joe and Mary to agree to use CoolDesign. Authority is achieved interactionally because of the hybrid presence of Brad and CoolDesign, which allows Brad to display his knowledge of CoolDesign and CoolDesign to speak for itself by showing what it is able to do.

The combination of these authoritative sources, we argue, influences Joe and Mary to accept that CoolDesign is a suitable choice of software for the organization. Thus, we argue that, in this case, leadership occurs neither through the artefact alone, nor the person alone. Rather, leadership occurs through the hybrid presence of Brad and CoolDesign. CoolDesign, the artefact itself, could not “do” leadership unless it is presentified in the interaction and made relevant through being ventriloquized by Brad. On the other hand, Brad’s ability to influence the other participants in the choice of software is limited by what the artefact (CoolDesign) can and cannot do.

Further, Brad’s ability to influence Joe and Mary to accept CoolDesign as the software that will be used is constrained by how persuasive CoolDesign is and the extent to which it can, or cannot, persuade others to follow it—or, rather the ventriloquized hybrid presence of Brad and CoolDesign. We, therefore, also note that leadership, as the process of influencing others to achieve the group’s goals, is negotiable and that any leadership move/attempt to influence can either be acquiesced to or resisted. In this case, the analyses reveal how Joe and Mary gradually align with this proposal, accepting the leadership of the hybrid presence of Brad and CoolDesign.

To clarify our position, our claim that CoolDesign speaks for itself—a controversial statement in the context of more than fifty years of social constructivism (Berger & Luckmann, 1966) —has to be understood relationally and never in an absolute way. As pointed out in our analysis, a great deal of work is being done by Brad to demonstrate his software’s capabilities, which means that if CoolDesign happens to speak for itself, it is also because Brad makes it do so, which is the essence of a ventriloquial move. However, one could retort that if Brad makes it speak for itself, it actually means that his software does not act as autonomously as this expression would imply. If we agree that this autonomy is relative (as any form of autonomy is), we believe that this expression is still useful and accurate to mark the difference the software, through its performance, makes in the situation we examined.

What is at stake here is our anti-reductionist position, which consists of never erasing the difference something or someone makes in a given situation. This stance is where we clearly depart from traditional forms of social constructivism (where everything is reduced to people’s action or sensemaking activities) and where we side with a relational perspective (Cooren, 2020; Kuhn et al., 2017). Adopting a relational perspective means that we acknowledge the differences humans and non-humans make in a given situation, knowing that these differences always interact with each other. In the case we studied, we showed, for instance, how displaying the performance of the software appeared to make a key difference in the unfolding of the interaction, a difference that can be reduced neither to what Brad was doing, nor to what the software was doing.

Retorting that the software, through the display of its performances, was in no way speaking for itself would consist of reducing the difference it was making to what Brad was doing. This reduction would amount to saying that the software’s performances did not make any difference, which is obviously not the case, as shown in our analysis. We, of course, recognize that some differences are more obvious than others, which means that “speaking for itself” can always be assessed relationally (by both the analysts and the participants). For instance, if someone asks us what the weather looks like and we simply reply, “Look!,” inviting this person to contemplate the blue and sunny sky we are pointing out to him, we imply, by this invitation, that the (appearance of the) sky somehow speaks for itself, meaning that it tells us what the weather looks like: sunny and blue. But note that even in this case, speaking for itself is not something that the sky exclusively does as it also relies on our experience and knowledge about what good weather consists of.

In the case of Brad, Joe, and Mary’s discussion, speaking for itself is less obvious for CoolDesign because this act relies on the participants’ professional experience (Goodwin, 1994), that is, their capacity to precisely make the performance of this software speak for itself. Reducing this difference to their experience would be as mistaken as reducing this same difference to what CoolDesign exclusively does. It is professional experience that makes CoolDesign speak for itself, but it is also Cool Design’s performance that makes Joe and Mary react the way they do. It is in this oscillation/vacillation that the relational nature of ventriloquism lies. There is no absolute starting point, rather there are relations that more or less unveil the differences that relata (i.e., things that are related by the relation) make in the situation. Without Joe and Mary’s professional experience, the performance of CoolDesign could not speak for itself. However, this performance is still crucial in this equation, as without it, there would not be anything to be assessed by Joe and Mary. Defining Joe and Mary’s expertise or CoolDesign’s performance as the origin of what is happening is nonsensical from a ventriloquial/relational perspective, as it forces us to arbitrarily determine a point of origin. It is this either/or logic that the ventriloquial/relational perspective precisely wants to fight against.

In conclusion, we suggest that by using the concept of ventriloquism as articulated by the Montreal School and by demonstrating how artefacts “do” leadership we add to leadership research in three ways. First, we argue that by adopting a ventriloquial approach to the role of artefacts in the doing of leadership we can make visible, and thus analyzable, how such hybrid presences do leadership as part of in situ practice—something that Clifton et al. (2020) argue is currently relatively lacking in research on the sociomateriality of leadership. We argue that, at least in this case, leadership occurs because the actor network that is operationalized through the hybrid presence of Brad and CoolDesign provides authority for the claim that CoolDesign should be the software used by the group. This authority is based in a display of what CoolDesign, the artefact, when ventriloquized through Brad, can do. Thus, Brad’s proposal becomes more persuasive and, as we see in the analyses, it is able to influence others to buy into the decision to use CoolDesign.

Thus, mobilizing authority is shown to be a key way in which leadership is achieved through the presentification and ventriloquation of artefacts in naturally-occurring interaction. Authority whilst being, on the one hand, an a priori potential, nevertheless has to be enacted in the here-and-now of an interaction for, to borrow a phrase from Garfinkel (1967), another first time. Authority is, therefore, always negotiated, fragile, and fluid, and recent research (e.g., Holm & Fairhurst, 2018; Van De Mieroop et al., 2020) has investigated how authority, as a key vehicle for the doing of leadership, is negotiated in interaction. However, to our knowledge, such research has not shown how artefacts can be an integral part of networks, distributed across a series of actors, that become part of a dynamic, interactive influence process in and through which organizational tasks are achieved (cf. Pearce and Conger’s (2003) definition of shared leadership). Thus, this study responds to Van De Mieroop et al.’s (2020) call for more research that investigates participants’ in situ mobilization of sources of authority to do leadership. Specifically, they (Van De Mieroop et al., 2020), argued that “investigating participants” in situ mobilization of sources of authority other than legitimate authority (such as, e.g., knowledge) may prove to be an interesting area for further research into shared leadership (p. 511, italics added).

Second, our analysis of artefacts doing leadership points to the relevance of distributed concepts of leadership in which we can claim that if leadership is distributed, it follows that it is distributed across not only human but also non-human actants. Indeed, the very existence of hybrid leadership presences (cf. Grint, 2005) points to the fact that leadership is necessarily shared and that the concept of the atomized individual is insufficient to grasp the relational complexities of leadership. Further, responding to the criticisms that distributed leadership has remained largely abstract and is lacking empirical evidence of distributed leadership-in-action (e.g., Clifton, 2017a), this study provides fine-grained analyses that explicate exactly how leadership is distributed within the group—and we move notions of distributed leadership forward by demonstrating how leadership is distributed between both human and non-human actants.

More specifically, as Arvedsen and Hassert (2020) pointed out, recently, a proliferation of leadership studies consider the influence of “things” such as material surroundings, objects, and bodies. However, they (Arvedsen & Hassert, 2020) noted that such studies are often lacking a fine-grained analysis of naturally-occurring interaction that makes visible, and thus analyzable, the role that artefacts and other material phenomena play in the doing of leadership as part of in situ social practice. Thus, this investigation responds to calls by researchers, such as Clifton et al. (2020), who ask for more work that uses the fine-grained analysis of video-recorded naturally-occurring workplace interaction that enables researchers to capture and systematically consider the use of artefacts as part of the multimodal interactional accomplishment of leadership.

Third, through locating the doing of leadership in the analysis of transcripts of naturally-occurring interaction in which the sociomaterial practices of leadership are made observable, we join calls by researchers, such as Fairhurst and Connaughton (2014) and Tourish (2014), who have argued for a more communicative approach to leadership studies. Such an approach may, by unravelling the (sociomaterial) practices of organizational players, reveal specific processes through which leadership is achieved in the terra firma of interaction. An interactional approach to leadership is thus less susceptible to putting forward reifications by informants who may report on what they say they do, rather than what they actually do, or who may reinterpret what they do in terms of fashionable discourses of leadership (Alvesson & Sveningsson, 2003). Further, the findings of this research also add to recent work that takes a discursive approach to leadership, which, whilst having provided significant insights into how leadership is achieved through talk in human interaction, has to date generally overlooked how non-human actants also communicate and so play a role in leadership processes.

Finally, we recognize that the analyses presented here provide a snapshot of an artefact, notably software, doing leadership in a particular situation. Moreover, we acknowledge that the case we studied could be called an easy test case. Testing the idea that leadership is distributed across individuals and nonhumans by analyzing an online decision-making business meeting among tech-savvy representatives from different firms seems like gathering evidence to fit a predetermined outcome (which amounts to a sort of confirmation bias). After all, we explain that this collaboration valued horizontality, self-organizing, decentralization, and openness. We believe that this is definitely a methodological issue, but we also think that this analysis allows us to demonstrate the value of acknowledging this phenomenon before future researchers do the challenging work of tracing the boundary conditions of hybrid distribution of leadership. It would, therefore, be interesting to expand this research by investigating other situations and contexts in which other artefacts and hybrid presences are involved in the doing of distributed leadership.