A Historical View of Context

Posted: January 12th, 2006 | No Comments »

In A Historical View of Context, Early draft of paper in J. CSCW 13(3), 223-247, August 2004, Matthew Chalmers a number of approaches of context-aware systems design, emphasizing on the way to reflect the historical aspect of context and how to make good use of the past to support ongoing user activity. He also pragmatically reasses of the notion of invisibility or disappearance that often characterises ubiquitous computing and opens doors to my research interest:

It seems more difficult to accept Weiser’s ideal of ‘invisible’ technology as an achievable ideal, as we have to accept that a system will be, and should be, used in an more ready–to–hand way occasionally.

Definition of context differ in CSCW and ubicomp, and I am stuck between both trying to make both fields bridge or even synthesis:

CSCW focuses on intersubjective aspects of context, constructed in and through the dynamic of each individual’s social interaction, and defends against reductionism and objectification. In contrast, context–aware and ubiquitous computing often concentrate on computational representations of context that span and combine many senses and media—rather than the social construction of context in interaction.

The problem with the ubicomp perspective is that it tends to emphasize objective features that can be tracked and recorded relatively easily, and to de-emphasize or avoid aspects of the user experience such as subjectivity perceived features and the way past experience of similar contexts may influence current activity – issues which are central concerns of CSCW.[...] One key issue has been how systems can represent work and its context without over-formalising, over-simplifying and over-objectifying it.

Workflow-like representation of activity have been brought into context-aware computing (Activity-Based Comuting (Christensen 2002; Bardram 2003), “task driven computing” (Garlan 2002). However, such representations of activity have a potential danger of becoming too formal:

“that their design is predicated entirely by formal procedures—ignoring (and even damaging) the informal practice” (Bardram 1997)

Dourish’s approach to combine CSCW and context-aware systems, the “embodied interaction”:

A contrasting approach to combining CSCW and context–aware systems emphasises socially–constructed situated action, and is also inspired by some of the foundational work of ubiquitous computing. The embodied interaction perspective on HCI (Dourish 2001) binds together CSCW and context–aware computing issues in presenting everyday human interaction as non–rationalising, intersubjective and bodily activity. [...] More recently, however, Dourish applied this embodied interaction perspective to the notion of context and context–awareness more practically (Dourish 2004). This paper focuses on a question highly relevant to our focus here: “how can sensor technologies allow computational systems to be sensitive to the settings in which they are used, so that, as we move from one physical or social setting to another, our computational devices can be attuned to these variations?

Dourish suggests that the field’s ideals of combining the social and the technical computing have not yet been achieved: “turning social observation into technical design seems to be problematic” and “these two positions are incompatible”. This is something that I would like to investigate in my thesis.

Dourish points out 3 design principle to allow forms of practice to emerge and evolve, rather than requiring users to fix their work and their information to predefined patterns:

  • Systems should display their own internal state and configuration to users “to make continual determinations of the potential consequences of their actions and their opportunities to reconfigure or realign the technologies through which they are conducting their actions
  • System’s internal structure becomes a resource fo the work of adaptation and contextualisation
  • Interfaces should offer “direct experience of the structure by which information is organized”

The problems with these design principles is that we have to be selective to reveal details of the system (we cannot present and let manipulate every details). There is a need for a degree of reduction and objectification, due to formal representational schemes of programs and databases, and finite capacities for storage, communication and calculation:

We must take a pragmatic stance if we are to design the finite and formal representation that constitute context-aware and CSCW systems. Embodied interaction is a good exemplar of research in CSCW and contet-aware computing that begins to bridge between useful practices and strong theory.

Make at tool invisible or ready-to-hand (Heidegger) through accommodation and appropriation

In time, this process of accommodation and appropriation lets one focus on the use of the tool, and not on the tool in itself, thus making the tool ‘disappear’ as Weiser later discussed. [...] Disappearance happens through the process of coupling and contextualization i.e. the circle of interpretation, action and experience that weaves together both ready–to–hand and present–at–hand uses of a tool by people over time. [...] No tool or system can always be invisible, and perhaps should not, as there are times when one cannot “focus on the task not the tool” because the task is the tool.

Weiser was clear that it was not the technology in itself that made for ubicomp. Instead he suggested that we should aim for and support accommodation and appropriation of computing into everyday life. One of my aims is to understand this accommodation and appropiration process in the real-world, and uncontrolled ubiquitous environments.

Reference I must read:
Abowd, G., Mynatt, E., and Rodden, T. (2002): The Human Experience, IEEE Pervasive Computing, Jan-Mar, 48-57.