Erving Goffman on AI: Resurrecting a Seminal Sociological Voice

May 02, 2025

In education's unfolding drama, artificial intelligence stands as potential villain or hero. The same technologies that can monitor, sort, and control can also liberate, enlighten, and transform. As the great Erving Goffman might observe, these divergent futures hinge on how we structure what he called the "interaction order" (the implicit rules governing face-to-face interaction) [3].

Below, two possible scenarios are presented, each interlaced with Goffmanesque phrases: one where AI extends institutional power, the other where it amplifies human potential.

AI at Education's Crossroads

The classroom has always been a stage where identities form through ritualized performances. Teachers and students enact what Goffman termed "frontstage behaviors" (deliberate self-presentations in public settings) [1], carefully presenting versions of themselves appropriate to educational contexts.

Into this theatrical setting, AI now steps—not as a neutral tool but as an active irritant in what Goffman would call the "definition of the situation" (how participants establish a shared understanding of social context) [1], altering how learning happens and who controls it.

Which future awaits us depends not on the technology itself but on the social architecture we build around it. Will AI become what Goffman might term a "total institution in code" (an all-encompassing system that regulates all aspects of daily life) [2]?

Or will it serve as what he might describe as a "backstage collaborator" (an assistant in the private preparation space for public performances) [1], helping teachers and students access richer possibilities while preserving their autonomy in educational performances?

Future One: The Algorithmic Institution

In this troubling future, Goffman's concepts of surveillance and institutional control dominate through omnipresent monitoring systems where every classroom becomes what he would identify as a "focused gathering" (a group of people engaged in sustained attention to a common activity) [3]. Facial recognition cameras track student attention while keystroke analysis evaluates writing process rather than just final products.

What Goffman called "batch processing" (institutional handling of people in standardized groups) [2] becomes automated and inescapable. Students who fidget excessively get flagged for possible ADHD, despite what Goffman might recognize as normal "role distance" (separating oneself from the expectations of a social role) [4], behaviors where individuals naturally struggle to distance themselves from institutional expectations.

Those whose writing patterns match "underperforming" profiles receive remedial content without human consultation, trapping them in algorithmic categories that reflect Goffman's concern with how institutions create fixed identities for their subjects.

Performance Under Pressure

Teachers find themselves engaged in what Goffman would term "impression management" [1] directed at algorithmic evaluators rather than human observers. Their classroom performances, once guided by professional judgment and student needs, now conform to metrics that AI systems can easily measure—speaking time, question frequency, vocabulary level—while neglecting aspects of teaching that defy quantification.

Students, similarly constrained, develop what Goffman called "secondary adjustments" [2]. They learn to maintain eye contact with screens even when mentally elsewhere, knowing that attention-tracking software interprets this as "engagement." They craft essays using phrases algorithms reward rather than exploring authentic ideas. What Goffman termed the "mortification of self" [2] (systematic stripping away of a sense of self) occurs gradually as students internalize algorithmic expectations.

Ritualized Compliance

The rich interaction rituals Goffman documented become mechanized under algorithmic observation. Class discussions follow structured templates designed to generate measurable participation metrics. Group work proceeds according to prescribed patterns that facilitate automatic evaluation. Spontaneous learning moments—what Goffman might call "flooding out" (momentary breaking of frame when emotions overwhelm social expectations) [5]—become rare anomalies in a system optimized for predictability.

Administrative decisions reflect what Goffman termed "cooling out the mark" (helping people accept disappointment without challenging the system) [6]. Students denied opportunities based on predictive analytics receive automated messaging explaining their "personalized learning pathway," disguising limited options as customized education. Teachers facing performance concerns receive "growth plans" generated without human involvement, exemplifying what Goffman described as how institutions reframe control as support.

Future Two: The Augmented Learning Community

In this promising alternative, AI functions as what Goffman might call an "ancillary team member" [1] in educational performances—supporting rather than directing human actors. Instead of constant evaluation, AI tools become available resources that students can activate when needed, preserving what Goffman termed "territories of the self" (personal spaces and possessions that individuals claim as their own) [3]—those personal spaces essential for identity development.

Teachers employ AI assistants for what Goffman would recognize as effective "backstage preparation" [1]—generating diverse examples, identifying potential misconceptions, suggesting discussion prompts—while maintaining complete authority over classroom interactions. These tools expand possibilities without constraining professional judgment, exemplifying what Goffman described as how supporting resources can enhance rather than undermine performances.

Students discover what Goffman might call "collaborative knowledge scaffolding" through AI study devices that adapt to individual curiosities rather than standardized pathways. A student wrestling with meiosis might explore the concept through different explanatory metaphors, visualizations that align with personal interests, or connections to topics previously understood—creating what Goffman would term "keying" (transforming an activity from one frame of reference to another) [5].

Intellectual Amplification

Learning analytics in this future serve what Goffman might identify as "reflective practice" rather than evaluative control. Students review personalized insights about their learning patterns, e.g., moments of engagement, conceptual connections, growth trajectories, developing metacognitive awareness of what Goffman called "interaction orders" [3] in learning acts as growth space in their educational experiences.

AI tutors function as what Goffman might term "friendly strangers" (individuals who offer interaction without ongoing social obligation) [3], offering the benefit of fresh perspectives without social judgment. When students struggle with complex ideas, these systems present alternative explanations, relevant examples, or clarifying questions, engaging in what Goffman could call "remedial exchanges" [3] that preserve intellectual dignity while facilitating understanding.

Language models become what Goffman would recognize as "cultural bridges," helping students access knowledge across linguistic and conceptual boundaries. A student researching Chinese history can interact with primary sources through translation assistance that preserves cultural nuances, experiencing what Goffman described as "frame alignment" (achieving shared understanding across different perspectives) [5] where different worldviews become mutually intelligible.

Empowered Performance

Classroom discussions expand through what Goffman might call "augmented interaction rituals" where AI systems suggest relevant resources without interrupting human dialogue. When conversation touches on climate science, for instance, visualization tools make abstract concepts tangible while discussion prompts help students connect scientific principles to lived experience—exemplifying what Goffman termed "keying" [5].

Assessment becomes what Goffman would identify as "performance" rather than standardized evaluation. Students demonstrate understanding through projects where AI tools help them express ideas in multiple modalities—visual, textual, interactive—while human teachers evaluate holistic understanding. This approach honors what Goffman called "face" (the positive social value a person claims through self-presentation) [3].

Teachers engage in what Goffman termed "teamwork" [1] with AI systems that handle routine tasks while freeing human attention for complex interpersonal work. Grading assistance, content curation, and administrative clerical support allow teachers to focus on what Goffman recognized as the core of education, the same core which resonates in persistent expressions of AI angst over its potential loss: Those delicate human exchanges where knowledge becomes meaningful within social contexts.

Humans at the Controls

These divergent futures represent not the inevitable consequences of technology but the fruits of deliberate choices about institutional design. As Goffman consistently emphasized throughout his work, social, human-created structures shape individual experiences far more profoundly than technical capabilities alone. His dramaturgical framework reminds us that education has always been performance—the question now is who writes the script and directs the action.

The algorithmic classroom can extend institutional control or expand human possibility; it can standardize learning experiences or diversify them. What Goffman would surely insist upon is that these outcomes depend not on artificial intelligence itself but on the very human decisions about how it integrates into education's existing interaction order.

The responsibility falls to educators, policymakers, technologists, parents, and students themselves to design systems that enhance rather than diminish human learning and enlightenment. To face this moment in history and safeguard the miracle of education requires precisely the sociological imagination Goffman championed—the ability to see beyond technical, predictable, mechanical specifications for human social life to a better social reality expanded through tools.

The future we choose will reflect our deepest values about learning, autonomy, and human development. In Goffman's terms, we face a choice about the "definition of the situation" [1] in tomorrow's classrooms. Will we define education as efficient information delivery best managed by algorithms? Or as a profound human exchange where technology serves as one resource among many for expanding minds and connecting hearts? Or will we take to the sidelines and leave it to chance?

The stage is set. The choice is ours.

Learning to Read, Reading to Learn

Discussion about this post