Studio Artifacts

    A Visual Language for Communicating Cognitive Derivation in Hybrid Human-GenAI Workflows

  Why does higher education still lack a sufficiently expressive visual language for communicating how GenAI systems participate in cognitive workflows?

   Knowing whether GenAI tools were allowed is different from knowing what cognitive work they performed in a cognitive workflow.

  Current disclosure practices often indicate whether GenAI tools were used, but rarely explain what roles those tools played in shaping a workflow. The Cognitive Derivation Stamps provide a visual language for making some of those roles more legible.

   We use “cognitive derivation” to describe how cognitive work takes shape across a human-GenAI workflow: how ideas, judgments, revisions, decisions, and artifacts are generated, transformed, evaluated, stabilized, and claimed.

   Eliott’s comic “Two-Stage Story” illustrates the origins of our discussion. Rather than beginning with the question of which GenAI uses are permitted, we began by asking how human agency shifts when cognitive work is distributed across hybrid human-GenAI workflows (and how to make those shifts more legible).

   Seeking new ways to make hybrid workflows more legible across our institution, the Cognitive Robustness Research Studio’s research assistants collaboratively designed a set of stamps intended to be as visually intuitive as possible. All designs were created by the human team without GenAI assistance.

   After extensive and iterative collaborative design discussions, we are thrilled to release the Cognitive Derivation Stamps.

   The Cognitive Derivation Stamps are organized into three galleries:

1. Broader Umbrella;
2. Ghost Gallery;
3. Coding Gallery.

   Although we initially conceived the stamps as a pedagogical tool, we later saw that our project could extend beyond that original context.

Below, you can:

1. Explore a few examples of how to use the stamps.
2. Read the accompanying PDF to learn more about the Cognitive Derivation Stamps.
3. Download a folder containing one image per stamp, along with its alt text.

   The stamps are licensed under a Creative Commons Attribution–NonCommercial–ShareAlike 4.0 International License.

Cognitive Derivation Stamps in Pedagogical Contexts

1. Using Stamps in Assignments

   In this case, specific Cognitive Derivation Stamps can help to:

1. Provide a shared vocabulary for communicating GenAI roles across instructors, students, and disciplines;
2. Help students encounter more consistent GenAI language across courses;
3. Support institutional conversations about Academic Integrity by making some aspects of cognitive derivation more legible across cases.

   See below for two examples of assignments from different divisions that use the same Cognitive Derivation Stamps.

   The meaning and appropriateness of each stamp should be interpreted in relation to disciplinary practices, assignment goals, and instructor expectations.

   In the examples below, both instructors allow GenAI use only for searching/investigating resources and for obtaining critiques of students’ work. The same stamps can therefore communicate comparable cognitive roles even when the assignments belong to different disciplines.

2. Using Stamps in Course-Level Policies

• In this example, the Cognitive Derivation Stamps are used as part of a course-level GenAI policy. The first set of stamps identifies forms of GenAI participation that are permitted in the course.
• In contrast, a crossed-out “Email: Writing or Summarizing” stamp communicates an explicit exclusion: students are expected to read course-related emails themselves and compose their own messages without GenAI assistance.
• In this way, the stamps can communicate not only where GenAI may participate in a workflow, but also where cognitive work is expected to remain human.

Emphasizes that students should read and write all course-related emails.

Cognitive Derivation Stamps in Article Writing

• In article writing, the Cognitive Derivation Stamps can provide readers with a compact account of where and how GenAI participated in the workflow.
• Rather than simply indicating that GenAI was used, the stamps can identify particular roles it played—for example, brainstorming, search or investigation, or the use represented by the Ghost Gallery: Images stamp.
• This makes selected aspects of the article’s cognitive derivation more legible to readers without implying that the stamps provide a complete account of the workflow.

The three stamps communicate more about the workflow than a generic statement that “GenAI was used.”

Conclusion 

• The Stamps offer a visual entry point into cognitive derivation. In many cases, they may be enough to clarify selected roles that GenAI played (or was expected not to play) in a workflow.
• They are not intended to provide a complete record of cognitive derivation, certify authorship or understanding, or resolve questions of academic integrity. Complex workflows may require multiple stamps, additional context, process documentation, or other forms of evidence.
• The Stamps’ immediate value is especially clear during this transitional period, as many academic and professional practices move from predominantly human-produced artifacts toward hybrid human-GenAI workflows. They can be used to communicate expectations about where GenAI may participate, and retrospectively, to make aspects of its actual participation more legible.
• Looking ahead, as GenAI becomes less visible as a separate tool and more embedded as an ambient environment, our practices for communicating cognitive derivation will need to evolve. The Cognitive Derivation Stamps should therefore be understood as an early and intentionally partial visual language for making distributed cognitive work more legible and discussable, and for supporting an infrastructure of trust without pretending to prove it.

Acknowledgments

In preparing this page, Eliott used GenAI tools in this article in these ways: Criticism, Minor editing, and Alt Text.

How to cite the stamps: Silva, Jenifer, Edgar Romero, Mansi Singh, Paul Reif, and Fernanda Eliott. 2026. “Stamps: Clarifying GenAI Use in Pedagogy and Communication.” OSF. July 15. doi:10.17605/OSF.IO/2CYWR.

How to cite this page: Eliott, Fernanda. “How should society communicate cognitive derivation? A visual language for communicating cognitive derivation in hybrid human-GenAI workflows.” The Cognitive Robustness Research Studio: Studio Artifacts. Grinnell College. https://cognitivestudio.cs.grinnell.edu/artifacts/