The lab's signature method · Extending Rivera & MacTavish (2017) into wave-four research
Configurational Probes, structured visibility for configurations.
A probe is a research instrument that surfaces what people, things, and contexts enact when prompted in a specific way. The lab inherits the cultural-probes tradition (Gaver, Dunne & Pacenti 1999) through the research-through-provocation lineage (Rivera & MacTavish 2017) and extends it for the wave-four moment, when the unit of analysis is no longer the user but the configuration.
Foundation: Rivera & MacTavish 2017, Time / Space / InformationExtension: 3 × 3 attribute framework for wave-four configurationsTwo versions: Light (3 attributes) · Full (9 attributes)
§ 01Lineage
The method's three-generation lineage.
Configurational probes do not emerge from nowhere. They sit at the end of a specific research-through-design tradition that runs from Bill Gaver's cultural probes (1999) through the research-through-provocation framework the lab's director and Tom MacTavish published in 2017. Each generation inherits something and breaks something. The break that matters now is the move from interaction to configuration, and the framework evolves accordingly.
Wave 3 · Origin
Cultural Probes
Gaver, Dunne & Pacenti (1999)
Evocative artifacts deployed into people's everyday lives to surface fragmentary, atmospheric data that wave-two methods could not reach. The third-wave methodological signature.
→
Wave 3 · Extension
Research through Provocation
Rivera & MacTavish (2017) · The Design Journal
A structured prototyping tool using interaction attributes of time, space, and information. Provotypes deliberately displace one attribute to surface tacit assumptions. Validated across three case studies.
→
Wave 4 · Successor
Configurational Probes
This lab (in development) · Builds on prior work
The 3×3 framework extends the wave-three method into wave-four conditions. Where provotypes probe interactions, configurational probes probe configurations, preserving the original three attributes and adding ontological and institutional layers.
§ 02Foundation
The wave-three foundation, preserved.
The 2017 paper proposes three interaction attributes, drawn from Lim et al.'s Interaction Gestalt (2007): Time (Scarcity ↔ Abundance), Space (Intimate ↔ Public), and Information (Tailored ↔ Generic). Each attribute is a subjective continuum. Each is measured at the level of how a person encounters a designed artifact. The framework remains valid for what wave-three research could see, and the lab's wave-four extension preserves it intact as the first row of the new 3×3 structure.
Source paper
Research through provocation: a structured prototyping tool using interaction attributes of time, space and information.
Jaime Rivera & Tom MacTavish (2017), published in The Design Journal (Routledge / Taylor & Francis), 20:sup1, S3996–S4008. Presented at the 12th European Academy of Design Conference, Sapienza University of Rome. The paper proposes the three subjective continua, a four-step process for structured controlled reflexivity prototyping, and validates the approach across three case studies in health, knowledge management, and policy engagement.
DOI: 10.1080/14606925.2017.1352902 · Open Access · Cited in The Design Journal
Time
Scarcity
Rush multitasking
Limited attention
Partial attention
Single focus activity
Abundance
Space
Intimate
Portable personal devices
Carry-on devices
Medium size devices
Large digital surfaces
Public
Information
Tailored
Unique individual
Particular group
General group
Non-specific collective
Generic
The three-position model · Real situation / Provotype / Final prototype · Rivera & MacTavish 2017, Figure 4
Real situation · the control group
Provotype · the experimental displacement
Final prototype · the design outcome
§ 03The 3 × 3 evolution
What changes in wave four, and what the framework grows to hold.
The 2017 framework studies an interaction: a designed encounter between a pre-formed user and an artifact, measurable along three attributes that describe how the encounter unfolds. Wave-four conditions dissolve the pre-formed user. The unit of analysis becomes the configuration, an arrangement in which humans, technologies, and contexts no longer hold still as separable entities. The three interaction attributes still apply (every configuration has time, space, and information dimensions), but they are no longer sufficient. Two new attribute categories are needed: one to describe what the configuration is ontologically, and one to describe how it sits institutionally.
The structural move
From 1 × 3 to 3 × 3. The wave-three framework had one row of three attributes, interaction attributes describing how the encounter unfolds. The wave-four framework preserves that row exactly and adds two more rows: ontological attributes (what kind of arrangement the configuration is, anchored in STS) and institutional attributes (how the configuration is embedded, anchored in organizational theory). The result is nine attributes across three theoretical lenses, with the original three intact as the framework's HCI foundation.
Lens · Category
Attribute 1
Attribute 2
Attribute 3
HCI · Wave 3
Interaction attributes
"How does the encounter unfold?"
Time
Scarcity ↔ Abundance
The subjective experience of time during the interaction. From rush-multitasking through single-focus activity.
Rivera & MacTavish 2017 · Lim 2007 · Kronz 1997
Space
Intimate ↔ Public
The spatial register of the encounter. From portable personal devices through large digital surfaces.
Rivera & MacTavish 2017 · Hornecker & Buur 2006
Information
Tailored ↔ Generic
The specificity of the content delivered. From unique-individual through non-specific-collective.
Rivera & MacTavish 2017 · White, Clark & Moore 2010
STS · Wave 4
Ontological attributes
"What kind of arrangement is this?"
Agency
Directive ↔ Distributed
Where action originates in the configuration. From single-agent direction through distributed across humans, AI, and context.
Barad 2007 (agential cut) · Suchman 2007
Inscription
Determined ↔ Open
How scripted the configuration is by its design. From rigid-determined through open-to-reinterpretation.
Akrich 1992 · Latour 1992
Visibility
Opaque ↔ Diffracted
How legible the configuration's workings are to its participants. From black-boxed through openly readable.
How interdependent the configuration's components are. From loosely-coupled (autonomous) through tightly-coupled (lock-step).
Weick 1976 · Perrow 1984
Logic
Singular ↔ Hybrid
How many institutional logics are operating simultaneously. From a single dominant logic through actively hybrid arrangements.
Thornton & Ocasio 1999 · Friedland & Alford
Adaptation
Static ↔ Reconfigurable
How much the configuration can change in response to environment. From static through continuously reconfigurable.
Teece 2007 · March 1991
Reading the grid
The three attributes marked LIGHT (Time, Agency, and Coupling) are the load-bearing attribute in each row. Together they make up the Configurational Probes Light version. All nine attributes together make up the Full version. The next section explains both.
§ 04Light & Full
Two versions of the probe, for different research depths.
A diagnostic instrument needs to scale with the engagement. The lab deploys configurational probes in two versions. A light version for initial scans, teaching, and rapid diagnostics. A full version for deep research, doctoral-level empirical work, and comprehensive organizational X-rays. Both versions read the same configuration through the same framework. The difference is depth, not kind.
Version 01 · Light
Configurational Probes, Light.
Three attributes, one per lens. A quick X-ray.
HCI
TimeScarcity ↔ Abundance
STS
AgencyDirective ↔ Distributed
ORG
CouplingLoose ↔ Tight
Best for
Initial diagnostics in a new engagement
Teaching the framework to non-specialist audiences
Quick scans across multiple sites or teams
Workshop sessions where time is limited
Journalistic or expository explainers of an AI configuration
Version 02 · Full
Configurational Probes, Full.
All nine attributes, three per lens. A complete diagnostic X-ray.
HCI
Time · Space · InformationThe full Rivera & MacTavish (2017) framework
Coupling · Logic · AdaptationWeick, Thornton & Ocasio, Teece, March
Best for
Deep organizational diagnostics over weeks or months
Empirical research feeding the 3 × 3 Executive Dashboard
Doctoral-level fieldwork on specific configurations
Cross-sector comparative studies
The full lab engagement model with guest researchers
§ 05Process
From four steps to six, evolving the process for configurations.
The 2017 paper specified a four-step process for structured controlled reflexivity prototyping: map the real situation, select fixed parameters, generate the provotype displacement, guide the final prototype. The wave-four extension adds two steps that the original could not anticipate, because they presuppose the configurational unit: scoping the configuration (deciding what is in and out of the arrangement under study) and re-X-raying (re-running the probe over time, because the configuration does not hold still).
Wave 3 · Original
The four-step provotype process.
Rivera & MacTavish 2017 · The Design Journal
01
Map the real situation. Plot the current interaction across Time, Space, and Information. This is the control group.
02
Identify fixed parameters. Select which attribute values cannot change in this project. They constrain the design space.
03
Generate the provotype. Build a low-fidelity prototype that deliberately displaces one of the changeable attributes, the more extreme, the more surfacing power.
04
Guide the final prototype. Use the provotype's findings to direct the design toward the final, deployable prototype.
Wave 4 · Extended
The six-step configurational probe process.
This lab · Configurational Probes v1.0 · In development
01
Scope the configuration. Decide what is in and out of the arrangement under study, which humans, which technologies, which institutional context. The agential cut is part of the method, not prior to it.
02
Map the current X-ray. Plot the configuration across the three attribute categories, interaction, ontological, institutional. Light version: three attributes. Full version: nine.
03
Identify fixed parameters. Which attribute values are non-negotiable in this engagement? They constrain the probe's design space.
04
Deploy the probe. Build the structured prompt that displaces a chosen attribute. Unlike the provotype, the probe is deployed inside the configuration, not against a user.
05
Read what the configuration enacted. The data is the configuration's response, what got produced, what got refused, what got reshaped, what got newly visible.
06
Re-X-ray. Configurations do not hold still. The probe is re-run on a cadence appropriate to the engagement, diagnostic visibility is a continuous practice, not a one-shot deliverable.
§ 06Worked example
SkyWords at Chicago City Hall, read through both frameworks.
To make the evolution concrete, the framework reads one of the dissertation's three case studies (SkyWords, the civic engagement installation at Chicago City Hall) through both the wave-three and wave-four versions of the framework. The wave-three reading is what the original paper offered. The wave-four reading is what configurational probes would surface today.
The original project
SkyWords, a civic engagement machine.
In 2013, an IIT/ID team led by Braun, Rivera, and others installed a site-specific civic-engagement installation on the ground floor of Chicago City Hall for ten days. Participants played a game (inflating and destroying balloons) that gathered policy preferences about the city. The team's initial assumption (an interactive kiosk) was displaced by five different provotypes deployed in the building's lobby. The final design adapted to the rush behavior of people in transit through the lobby, using generic information they could process quickly.
Braun, L., Rivera, J., Mello, J., Hindi, K., Lin, L., Patel, K., & Mathew, A. (2013). SkyWords: An Engagement Machine at Chicago City Hall. CHI '13 Extended Abstracts, 2839–2840.
Wave-three reading · The original paper
SkyWords through interaction attributes.
Time
Real situation: scarcity, people rushing through the lobby. Final design: confirmed scarcity, prototype adapted to rush behavior.
Space
Real situation: public, the lobby of City Hall. Provotypes tested how the installation could occupy that public space without requiring sustained attention.
Information
Initial assumption: tailored (an interactive kiosk with specific questions). Provotypes revealed this was wrong; final design moved to generic information processable in transit.
What the wave-three framework surfaced: the displacement of one interaction attribute (Information, from tailored to generic) drove the design pivot. The provotype process generated useful empirical data about participant behavior. The framework worked exactly as it was designed to work.
Wave-four reading · What configurational probes would surface today
SkyWords through the 3 × 3 framework.
HCI · Time / Space / Information
Same readings as the original, the wave-three framework remains valid for what it measures. Scarcity, public space, generic information.
STS · Agency / Inscription / Visibility
Agency:distributed, the configuration includes participants, designers, the City Hall building, the policy questions, the game mechanic. None acts alone. Inscription:open, the game mechanic invited reinterpretation; people made it their own. Visibility:opaque, participants did not see how their answers became policy input.
ORG · Coupling / Logic / Adaptation
Coupling:loose, the installation and the city's policy process were not tightly bound; data flowed in but was not committed-to. Logic:hybrid, civic-engagement logic and entertainment logic operated simultaneously. Adaptation:reconfigurable, the installation moved across five provotype iterations during ten days.
What the wave-four framework would surface that the original could not: the visibility gap (participants did not know what happened to their data) was a configurational property the interaction-attributes framework had no instrument to measure. The agency distribution (the building, the game, the policy office all had agency) was also invisible. A configurational probe today would specifically test these, and might design a final prototype that closed the visibility gap, or that re-tightened the coupling between participant input and policy outcome.
What the evolution shows
The wave-three framework worked for what wave-three could see. The wave-four extension does not invalidate it, it adds the questions wave three could not yet ask: about agency distribution, about inscription openness, about visibility, about institutional coupling. SkyWords becomes a richer case when read through both frameworks. The same is true for every other configuration the lab will eventually study.
The Configurations Lab · An independent research practice