The Human Aperture: Sovereign Learning in the Age of Analytics v.02

In evaluating the 2024 EDUCAUSE Horizon Report: Teaching and Learning Edition, one technology and one trend intersect directly: Evolving Data and Learning Analytics Capabilities, and the report's growing concern around cybersecurity, data privacy, and behavioral monitoring.

When an LMS measures time-on-page, click cadence, or paused video to infer "engagement," it commits a foundational error: assuming a digital representation can capture, or govern, human intellectual growth. Learning is a relational, often nonlinear process — confusion, backtracking, and quiet incubation are a needed contemplation space and part of how understanding actually forms, not noise sitting on top of it. A dashboard built to reward continuous, legible activity has no way to register the value of the student who goes still for ten minutes to think. As I've argued elsewhere, representation never inherits sovereignty over the thing it represents — a telemetry stream is not the learner, and treating it as an adequate stand-in for judgment is where the harm starts.

Two primary socio-technical forces drive this tension:

Institutional Demand for Predictive Oversight (Institutional/Economic Force): Facing enrollment pressure and demands to prove retention, universities deploy centralized analytics to make student behavior standardized and "legible." As political anthropologist James C. Scott observed, centralized state systems attempt to flatten complex, organic local reality into simplified data points so they can measure and manage it—a dynamic Philip Agre formalizes in his capture models. When an institution reduces human learning to trackable metrics, its gaze shifts from supporting the learner to enforcing administrative legibility, making genuine understanding secondary to system compliance.

The Rise of Data Sovereignty & Edge Architectures (Social/Technological Force): Concurrently, there is a growing push toward data minimization and learner agency. Drawing on Sylvia Wynter’s framing of humans as Homo narrans — storytelling beings who narrate themselves into meaning rather than existing as fixed data — forcing students into compulsory tracking flattens exactly the capacity education is supposed to develop. The counter-force is edge-based, localized processing — architectures that protect privacy by design and honor student agency.

Figure 1: Structural Agency Checkpoints & System Architecture — Operationalizing the Agency Ledger, Reversibility, and Authority Impact Statements (AIS) to ensure digital representation never inherits sovereignty over the learner.

A workable alternative doesn't require abandoning analytics, just changing what they're allowed to keep. An Agency Ledger approach — a transparent accounting of what a system gains, what a student surrenders, and how fully any of it can be revoked — would treat consent as ongoing rather than a one-time checkbox at enrollment. Paired with edge-based processing, where analysis happens locally and only aggregate, non-identifying signals leave the device, institutions could still measure what they need for course design without building a permanent behavioral record of every student's private struggle to learn.

The right question for a learning platform isn't how much it can see.

It's how little it needs to see in order to actually help gain knowledge.

References

Agre, P. E. (1994). Surveillance and capture: Two models of privacy. The Information Society, 10(2), 101–127. https://doi.org/10.1080/01972243.1994.9960162

EDUCAUSE. (2024). 2024 EDUCAUSE Horizon Report: Teaching and Learning Edition[cite: 1]. EDUCAUSE Library[cite: 1]. https://library.educause.edu/resources/2024/5/2024-educause-horizon-report-teaching-and-learning-edition[cite: 1]

Scott, J. C. (1998). Seeing like a state: How certain schemes to improve the human condition have failed. Yale University Press.

Wynter, S. (2003). Unsettling the coloniality of being/power/truth/freedom: Towards the human, after man, its overrepresentation—An argument. CR: The New Centennial Review, 3(3), 257–337.