We’re back at it as part of Semester 2 of the Master of Applied Cybernetics 3A Institute. The Build course from last semester continues and our questions course is replaced by a course focused on practice. Each fortnight we’ll be exploring component’s of 3A Institute 3As and 3Is — Assurance, Autonomy, Agency, Intent, Interfaces and Indicators (read more on this here).
During the first fortnight we were faced with this provocation from Churchman (1970): if a change in a system is to be called an improvement, then reflecting on the boundary of analysis of that system is crucial.
A range of resources were explored including:
- Sathe (2011)’s narrative on the implementation of Universal Identification (UID) as a mechanism for improvements in the empowerment of India’s population
- Hau’ofa (1993)’s framing of his home as ‘Oceania’ as an improvement on the belittlement and neocolonialism associated with ‘the Pacific’
- Jain et al (2010) who discusses in detail how their computer software improves on previous methods of randomising patrol scheduling and deters terrorism.
Defining a boundary (singular) may not be as useful (or possible) as defining the different types of boundaries that authors employ (implicitly or explicitly). A few types of boundaries are explored below.
Geographic Boundaries
Jain et al (2010) provides the most ring-bound geographic boundary where a software assistant is described in the specific context of LAX airport and then expanded to US-based flight paths. Moving the boundary of application to a non-US context, where, for example, the airport security context is different, the nature of targets is different and the assumption of an individual leader / follower frame may not match the cultural context, would likely change its position as a ‘success story’.
Similarly, focusing on a particular country context is a boundary drawn for Sathe (2011); in this case, India. Changing the boundary to focus in on the diversity of India’s numerous states and how differences in geography, demographics, access points, existing social, information and other infrastructure etc impacts implementation may change the story of ‘improvement’, ‘success’ and ‘desirability’ of the UID.
Hau’ofa (1993) also explores geographic boundaries, describing how the colonial influence of naming his home ‘the Pacific’ (focused on land) was part of an overall strategy of belittlement and ‘othering’. Hau’ofa (1993) reflects on how notions of size and importance change when considering his home as ‘Oceania’ which includes the oceans, trade routes and the collective whole of neighbouring communities.
Stakeholder perspective boundaries
Both Sathe (2011) and Jain et al (2010) focus their articles on the perspective of the ‘doer’ -the designer of the algorithms and the designer of the UID implementation, Nilekani, respectively. There are no interviews or perspectives of other stakeholders for example those responsible for translating the plan into action such as the security patrol at LAX, the ‘recruiters’ in different Indian communities or ‘the beneficiaries’ of the UID, India’s poor. The inclusion of other stakeholder voices would widen the boundary and in doing so introduce different lenses and potentially conflicting stories. This could then muddy the author’s narratives of ‘success’.
For Hau’ofa, the perspective is more personal, it’s written in the first person in a narrative style. A boundary in many ways is Hau’ofa’s own lived experience. Changing the boundary to include the voices of other generations, particularly the youth that are referred to in the article anecdotally could influence what an improvement may look like.
Time boundaries
Every report and piece of research is representative of a snapshot in time. This is a boundary that is drawn implicitly or explicitly. In the case of Sathe (2011), the choice to focus on implementation excludes the period of time in the leadup to the decision that was made on IUD in the first place, and does not allow for stories of impact (or harm) as they will occur post implementation. For Hau’ofa (1993), he highlights how stories taken in the time of his ancestors would have shown a very different interpretation of his home compared to contemporary narratives, even though the substance (the islands and oceans) are the exact same. Similarly, the story of empowerment could look quite different in a future decade where his story is superseded by younger generations.
As Churchman (1970) suggests, success inextricably linked to the boundaries drawn. Boundary-drawing from my previous world would be called ‘setting metrics or KPIs (Key Performance Indicators). When you think about it, metrics of also assume (or make explicit) a particular period of time. In its own way, defining metrics is a boundary-forming exercise: it gives specificity over what to focus on, what to care about. And in doing so, defines success.
What are the implications of this when building cyber-physical systems of the future? Can we begin to look at boundaries as ways of framing intent and/or making intent tangible and explicit? What do those boundaries do or say for those ‘on the outside’ and how do you manage that as a practitioner? Which types of boundaries are consciously decided upon and which ‘choose you’? Do boundaries stay the same over time? If not, what influences them changing?
References:
C. West Churchman (1970). Operations research as a profession. Management Science, 17, B37–53.
Manish Jain et al (2010). Software Assistants for Randomized Patrol Planning for the LAX Airport Police and the Federal Air Marshal Service. Interfaces, 40(4), 267–290. DOI: 10.1287/inte.1100.0505
Epili Hau‘ofa (1993). Our Sea of Islands. In A New Oceania: Rediscovering Our Sea of Islands, ed. E. Waddell, V. Naidu, and E. Hau‘ofa, 2–16. Suva, Fiji: School of Social and Economic Development, University of the South Pacific.
Vijay Sathe (2011). The World’s Most Ambitious ID Project: India’s Project Aadhaar, Academic Paper