ARIN6906 Chap.1 Emergence, Mediation and Distributed Agency
Emergence, Mediation and Distributed Agency
The opening block establishes the method used throughout the unit: treat technology as a technical and social system that reorganises perception, agency, power and reality. Extension and capture name the paired movement by which a device enlarges capacity while creating new dependencies, and distributed agency prevents responsibility from being assigned to one visible object alone.
What this chapter covers
- 01
Emergence as systemic change rather than novelty
- 02
McLuhan and technology as extension
- 03
Capture through platforms and institutions
- 04
Actor-network theory and delegated action
- 05
Exteriorisation and reciprocal transformation
- 06
Responsibility across human and nonhuman actors
Worked example · free
Where agency sits in a campus wristband
- +1These four rehearsal points are not an official University mark allocation. Define the system boundary around readers, identity records, occupancy data, platform rules and institutional policy.
- +1Explain how integrated access and navigation extend student capacity.
- +1Show how movement becomes a persistent record and recommendations frame available space.
- +1Allocate responsibility among university, vendor, staff and affected students, then state a condition for acceptable use.
Key terms
- Technological mediation
- The process through which technical arrangements shape perception and action rather than neutrally transmitting intention.
- Distributed agency
- Action produced through relations among people, artefacts, platforms and institutions.
- Extension
- An enlarged practical or perceptual capacity created through a technical system.
- Capture
- The redirection of behaviour, data or dependence into an institutional or commercial arrangement.
- Delegation
- The transfer of a task or prescription into an artefact, procedure or nonhuman actor.
Emergence, Mediation and Distributed Agency FAQ
What makes a technology emerging in this unit?
Novelty alone is insufficient. The useful test is whether capabilities, uses, supporting institutions and social consequences are forming quickly enough to disturb established arrangements. A strong answer also identifies uncertainty and avoids treating promotional momentum as demonstrated technical capacity.
How are extension and capture related?
They describe two sides of one mechanism. A navigation system may extend spatial access while capturing routes, attention and dependence within a platform. The analytical task is to show how the same feature produces both capacity and constraint, then judge the conditions under which the trade is acceptable.
Why does distributed agency matter?
It stops analysis at neither the user nor the machine. A consequential action can depend on designers, data, software, operators, owners and institutions. Mapping those relations clarifies where practical control and responsibility remain, especially when the visible artefact appears autonomous.
How can I identify capture in an everyday object?
Look for the arrangement that receives data, defines defaults or makes continued use dependent on a service. Capture is not limited to surveillance. A device can capture attention, maintenance, identity or future choices. Pair the constraint with the capacity gained so the analysis explains one technical relation rather than two disconnected lists.
What is a good distributed-agency conclusion?
A useful conclusion allocates different forms of control instead of saying responsibility is shared by everyone. Name who designed the prescription, who authorised deployment, who operates the system and who can repair the harm. Then state which actor had the practical power to change the arrangement before the outcome occurred.
Why begin with ordinary technologies?
Doors, maps and access systems make delegation visible because their operations are easier to describe than those of a large model. Once you can show how a mundane artefact prescribes conduct, the same method transfers to AI agents and autonomous machines without relying on dramatic claims about machine intention.
Assessment move
Create a four-column note for every case: capacity extended, behaviour captured, actors connected and responsibility retained. Practise on ordinary objects before moving to AI or robots, because mundane cases make the causal method easier to see. In assessed work, name the technical operation before using abstract terms such as agency or control, and finish with a condition that follows from your map.
Build one diagram with the user on one side and the institution on the other, then place devices, platforms and data between them. Write a verb on every connection. This forces the analysis to specify whether an actor senses, classifies, authorises, pays, repairs or resists.
Rehearse a short counterargument that treats the device as a neutral tool, then answer it by showing how design and infrastructure shape the available choices. When preparing a presentation, introduce the visible technology quickly and spend more time on the relations that produce its consequences. For Q&A preparation, identify the boundary of your evidence and one plausible alternative explanation.
A disciplined answer can concede that users retain choice while demonstrating how the conditions of choice have been reorganised. A second revision pass should test what your map excludes. Redraw the same case three times: first around the user and device, next around the service and its data, and finally around the institution and affected public.
Record which harms, benefits and decision-makers appear only after the boundary moves. This exercise stops a narrow interface description from standing in for the whole system. It also produces useful answers to challenge questions, because you can explain why a particular boundary is appropriate without pretending it is natural.
For debate preparation, assign a teammate to defend the narrowest boundary and another to defend the widest, then resolve the disagreement by returning to the precise proposition. Keep a short evidence ledger beside the diagrams so that every claim about power or dependence points to an observable rule, ownership relation, default, data flow or avenue of refusal.
