How Might We Collect and Archive Multiple Temporalities? — An Experimental Practice through “Reflection Notes” and the “Generative Archive”

During the three-day fieldwork of CDW 2025, our aim was to shift perspective from clock time (Chronos), which has underpinned modern society, toward plural temporalities (Kairos) that emerge within relationships.
In conventional program design, it is common to consolidate experiences into a single output through group discussion at the end of a fieldwork, or for a curator or editor to synthesize the experience into a coherent narrative. However, relying solely on such approaches risks losing the plurality embedded in participants’ cultural backgrounds and social positions (positionality)—including subtle bodily rhythms, affective resonances, and faint senses of discomfort that arise in the field.
We hypothesized that plural temporalities are not defined top-down but emerge bottom-up. Furthermore, by referencing what others felt and how they responded, participants might gain new insights and experience transformation themselves. From these hypotheses, the design of our tools began.

As an experimental attempt, we developed two tools:
・“Reflection Notes”, using a LINE Business Account and
・“Generative Archive”, an interactive archive built from data collected via LINE and processed using NotebookLM

We constructed an interactive archive that allows users to explore texts and images left as “individual reflections,” which were then incorporated into generative AI systems for dialogical engagement.

“Reflection Notes” is an introspective tool that utilizes LINE—an app familiar to participants—allowing them to casually input reflections from their smartphones during gaps in the fieldwork schedule. There are no correct answers; participants are encouraged to “write without overthinking.” The tool serves as a container to preserve each person’s sense of “time” in their own words.

We designed four guiding prompts:
・Situation: Describe the place or context visited.
・Event: Reflect on what occurred there.
・Affect: Address the emotions or resonances that emerged.
・Insight: Consider what realizations arose from the experience.

Using the LINE Business Account, when a participant taps one of the menu buttons, an automated response displays the relevant question along with input examples.

LINE Business Accounts support image uploads and voice input, enabling participants to output sensory impressions in forms other than text. Multiple project members can manage and edit the account, and user inputs can be downloaded as CSV data.

When collected records are edited, they risk being reabsorbed into the perspective of whoever “summarizes” them, thereby undermining the preservation of plurality. To address this, we constructed the Generative Archive.
Text and image data entered via LINE were downloaded, and statements were extracted in a way that prevents identification of individuals. To link reflections from different participants who experienced the same place or event, we inferred contextual elements—such as location, experience, or visual focus—from metadata including input timestamps and textual content.
These processed data were made readable in NotebookLM, forming the basis of the Generative Archive. In preparing the dataset, we also utilized other AI tools such as ChatGPT and Cursor.

This archive is not an AI designed to provide definitive answers. Rather, it is a tool that allows fieldwork participants to reference perspectives other than their own—voices that walked the same field—and to view different reflections side by side. Our aim was to preserve the coexistence of multiple temporal sensibilities without erasing individual voices.

The Generative Archive is available here ​(accessible until the end of December 2026)