Analysis: Episode Two

While there are a number of similarities in this and the previous episode in terms of the material-discursive practices observable in Paul and Etienne’s engagement, what is dis- tinctive in this account is the manner in which the computational object comes to the fore as an actant that ‘tells’ humans what to do—analogous to a human leader giving commands to her subordinates.

Along these lines, a notable phenomenon is my own orientation to the context, in particular, the way I myself oriented to the computational object as a means to gather information. Thus, at the beginning of the account and for the�rst three paragraphs of episode two is my own ongoing consultion/relation with the build monitor to gather local information through an engagement not dissimilar to an interview accompanied through ‘close observation’ (Van Maanen1988, p. 19). By closely observing the projections made material through the computational object, I was able to discern, much in the same way I might ask questions of a human, aspects of the project status that suggested its initial phase. A crucial consideration here depends on what McLuhan argues is an important form of literacy (1962, pp. 36-40), where:

branchaconcurrently to the changes being made by the developers on branchb. This creates a situation where the changes from branchamust eventually bemergedwith branchbin order to bring the codebase up-to-date. This does not occur automatically; the developers working in branchbmust merge branchaand ensure that their code still performs correctly. This is normally done at a convenient time or at particular project milestones.

[T]echnology is introduced either from within or from without a culture, and if it gives new stress or ascendancy to one or another of our senses, the ratio among all of our senses is altered. We no longer feel the same, nor do our eyes and ears and other senses remain the same (1962, p. 24).

Thus for McLuhan, that I might be able to extend my senses through a computa- tional object is a phenomenon in and of itself worthy of attention. Here, I characterise my own engagement with the computational object as an accessible relationship made

possible in McLuhan’s terms based on my own literacy with it—one I immediately and without question turned to as part of my own sensemaking of the context (Weick, Sutcli�e et al.2005). It was through that relation I ascertained various project metrics by focus- ing with my eyes and not my ears. This is one of the impacts brought about by modern technologies—a shift from an aural to an optical society (cf. Braidotti2006, p. 204; Har- away1991a, pp. 188-189; McLuhan1962, pp. 28-29). So, although I was there to observe Paul and Etienne, as part of my taking in context, the data shows that my attention went

�rst to the computational object as an informant to the research. I, like my own human informants, am apparently not immune from the ubiquity and background status of com- putational object as sources for guidance, knowledge, and information. This status that computational objects hold as a source orpositionas depicted in�gure2.4on page32for

guidance, knowledge, and information is one also viewed in the literature as a source for leadership (Alvesson2011b, p. 73; Nonaka and Takeuchi2011).

Moving now beyond my own initial attention to the computational object at the start of the episode, I would like to call attention to the periods of material-discursive practice involving remote activity around 3:01 p.m. Despite Paul and Etienne being an embodied human pair situated in the same room, there was nevertheless a simultaneous stream of instant messagings (IMs) between Etienne and remote colleagues that were re- lated to the deployment of the project that Paul and Etienne were working on. The ongoing communication between Etienne and these colleagues was situated within the same visual topography as pair programming, reinforcing the visual locus of attention (Guimbretière

2002, pp. 15-19,2003, p. 53) and the primacy of the screen (Braidotti2006, p. 204; Haraway

1991a, pp. 188-189). Such behaviour, in analysing its contextual detail, may also be under- stood as leadership practice, because Etienne was being asked speci�c questions to vouch for the readiness of certain code to be deployed. Thus, based on hispositionas depicted in

�gure2.4on page32, the deployment team deferred to him for a decision (Useem2010a, p. 510) before they proceeded.

Moreover, the�uidity with which Etienne’s IM exchanges occurred was notable. There was a sense of�ow with which Etienne handled these, one that I argue arises out of the communication system itself being situated within the same computational object as

the pair programming work. Thus, the ‘one-stop shopping’ approach that the computa- tional object o�ers to its human can provide particular a�ordances that impact cognitive and communicative practices including the externalisation of thought (Nickerson et al.

2013), coordination (Zammuto et al. 2007), and increased specialisation (Leonardi 2011, p. 161). Such layered material-discursive practices beautifully illustrate the distributed nature of activity as a collective form of leadership, much in the same way that Hutchins’ study of how a collective of people, artefacts, and machines navigate a submarine (1995a, pp. 355-356).

Indeed, relational (cf. Cunli�e and Eriksen2011; Fletcher2012; Ospina and Uhl- Bien2012), distributed (cf. Bolden2011; Gronn2002; Spillane2006), and processual (cf. Day and Antonakis2013; Koivunen2007; Wood2005) approaches to leadership rely on this same observation that Hutchins makes about cognition; namely that as a phenomenon not directly visible, we must cite it in some way for it to materialise. All of the approaches to leadership I just mentioned implicitly understand that citational process as occurring through social systems, which as can be seen from these episodes, must necessarily include the computational objects entangled as part of material-discursive practice. Therefore, the thesis I want to take forward is that leadership is not simply understood in certain ways, but that it is understoodasleadership precisely because it is cited as such. I argue that

leadership, following Ailon and Kunda, acts as a ‘symbolic resource’ (2003) standing for particular values within the context of a community such as an organisation. Thus, upon interrogation, leadershipalwaysstands for something else; some value or values within

the active social system, for which the word ‘leadership’ is shorthand. In the speci�c case of the deployment team asking Etienne to con�rm certain information, this underscores particularvaluesaround authority and responsibility that align with the leadership dimen-

sions ofpositionandprocessas depicted in�gure2.4on page32.

Then at 4:33 p.m., as Paul and Etienne are working on their next work item, the focal computational object stands mutely before the group but switches its colours from green to red, signifying a problem. Then,in response toit, Paul says ‘Uh oh’, the o�ce director says ‘everybody stop’, laughs, and says ‘Who did it?’ In terms of thetemporal

order of these speech acts, they wereprecededby the computational object signifying the

issue and, in doing so, issuing a command to the group. Therefore, an important question at this juncture is who—or what—is acting as the leader?

One way to answer this is to say that the programmers who wrote the software, the people that built the computational object and the operating system, and all the other components, as well as the management who decided that this approach to operations would be implemented, were part of a sociotechnical assemblage of actants (Latour1999, p. 198). This must include the computational object that enabled this enactment where a

crucial problem was reported and then solved. With this ANT-style analysis, we arrive, at a minimum, to a conclusion where a highly distributed form of leadership is present.

However, I would like to suggest something more radical. I propose that we have a situation where the computational object has become thesourceof leadership command

authority (Grint2005b, p. 1477), where, as Haraway (1991a, pp. 188-196) has suggested, the visual dimension is employed to control a group of people; not necessarily for nefarious means, but to control them nonetheless. In the performance of this material-discursive practice, it is the computational object that silently evaluates and delivers feedback to the the group, analogous to a human leader o�ering performance feedback. And from a tem- poral perspective, it is the computational object that explicitly initiates this chain of events where the group subsequently perceives a ‘problem’ that must be addressed as a priority. The command has been issued and has been received. Thus, in an important sense, I argue that the computational object is playing the role of leader, where it initiates action toward compliance, followed by the evident receipt by humans of authoritative orders on what to do. It is therefore plausible that this material-discursive practice constitutes a form of lead- ership as authority or command (Air Command and Sta�College2005; Benoit-Barne and Cooren2009; Miller2008), one that aligns with theresultorprocessdimensions depicted in�gure2.4on page32.

But there is yet another sense that the computational object in this context em- bodies leadership. In its given formalrole, it issues a command in a language that humans readily understand in regards to a problem that needs to be addressed with some urgency. Therefore, the role it plays is analogous to the one that Larsson and Lunholm’s manager (2013) plays in her leadership interaction with a subordinate, however, in their case, what requires a lengthy conversation between manager and subordinate, where the former must persuade the latter, happens in this episode within the space of eight seconds. Here, the dimensions of computational object’s‘personality’and its implicitpurposeas depicted in

�gure2.4on page32are established by the ways humans respond to it. Thus, in this epis- ode, we see all�ve of the dimensions of leadership: person,result,position,process, and

purpose—all enacted by the computational object.

It is important to recognise, however, that this is not simply a claim about a com- putational object playing a leadership role. Rather, it is a claim regarding shared meanings and values that can be embodied by and/or attributed to humans and other objects within material-discursive practice. For just as leaders cannot lead without followers (Bligh and Riggio2012b; Burke1965; Kahai2012), so too must the humans in the room understand the meaning and rami�cation of the message from the computational object and then play their parts in order for any ‘leadership’ to occur at all. The o�ce manager articulated to and for the group this sensibility in his utterance of ‘Who did it?’ By making this state-

ment, he implies that that someone ‘did it’, verbally establishing the existence of a problem, and by extension, that someone needs to ‘�x it’. Etienne and Paul then demonstrated their recognition of the discursive nature of what just transpired. They understood that the build was broken and they needed to determine whether it was their fault. Whether one accepts that the computational object ‘led’ or not, this episode illustrates an example of direction initiated by a computational object that was subsequently responded to by the human group in kind. In this sense, the phenomenon of ‘leadership’ was performatively produced by the particulars of this material-discursive practice (Barad2007, p. 178).

However, I would be remiss if I did not report that the members did not speak of ‘leadership’ in this episode and did not seem to think about what the computational object did as leadership. For these members, my observations revealed that for them, humans ‘do’ leadership. What is therefore fascinating about this case is that while the computational object may have functionally reproduced many aspects of ‘leadership’ as per my analysis, members did not see it as such. In this sense, I would say that while the possibility for a posthuman form of leadership was present, it was not materialised through subsequent citational material-discursive practice.

In this episode, I have demonstrated possibilities for how leadership practices are enacted bybothhumans and computational objects and then further distributed through

the sca�olding provided by computational objects and their underlying systems. I argue that in this episode we have witnessed a case where it is possible that the computational object played the role of leader in a distributed system of leadership, authority, and re- sponsibility, where the human participants are committed to the rules of the game (1953 / Wittgenstein2009, 6-8e; Lyotard1984). Interestingly, the informants did not see what the build monitor did, nor the impact it had on work�ow as anything other than an instru- mental intercession.