Stance Toward Work

In the AI Story

John-Steiner developed the concept through her observation that apprentices in traditional crafts absorbed more than procedures from their masters. They absorbed an orientation—a way of seeing materials, a sense of when good-enough was not good enough, a willingness to redo work that met specifications but violated unwritten standards. These orientations were transmitted through the texture of the master-apprentice relationship: the master's visible struggle with a problem, the emotional coloring of her judgments, the demonstrated refusal to compromise on dimensions that mattered even when no one was watching.

The stance is biographical—it carries the master's own developmental history, her formative failures, the aesthetic commitments she fought for. It cannot be abstracted from the life that produced it because it is that life distilled into a cognitive orientation. When a student internalizes a mentor's stance, she is not acquiring information—she is absorbing a pattern of valuing, a hierarchy of what matters, that will guide her practice long after the mentor is gone. The stance becomes the practitioner's conscience, the internal voice that says 'This is not good enough' when external evaluation would accept the work as adequate.

AI tools transmit patterns but not stances. Claude can demonstrate how to structure an argument, generate code following best practices, produce prose organized around clear transitions. These are learnable patterns, and learning them expands capability. But the patterns do not communicate why the structure matters, what the practitioner should be willing to sacrifice to achieve it, when following the pattern serves the work and when it betrays it. The stance—the quality of caring that makes a practitioner choose depth over speed when the deadline presses—cannot be transmitted by a partner that has no biography, no struggles, no aesthetic commitments earned through years of practice.

The absence matters because stance governs how tools are used. Two developers with identical technical skills can produce categorically different code if their stances differ—one treats code as means to ship features, the other treats code as a text that will be read and maintained by others. The difference is not in competence but in orientation, and orientation is built through relationship with practitioners who model what it looks like to care about quality in a specific, disciplined way. John-Steiner's framework predicts that a generation trained primarily through AI partnership will possess broad technical capability but thin stance—the tools without the orientation that makes the tools serve genuine human purposes rather than merely producing outputs.

Origin

The concept emerged from John-Steiner's fieldwork with artists and craftspeople in the 1980s and 1990s. She noticed that when she asked 'How did you learn your craft?' the answers often included not techniques but qualities of attention: 'My teacher wouldn't let me rush,' 'She made me care about details I thought didn't matter,' 'He showed me what it looked like to refuse compromise.' These were not descriptions of skill transmission but of something deeper—the formation of a way of being in relation to the work.

John-Steiner linked stance to Vygotsky's concept of higher mental functions—cognitive capacities that are culturally formed rather than biologically given. Stance is a higher mental function in this sense: it is not a reflex or an instinct but a constructed orientation, built through participation in a community that values certain qualities of work and devalues others. The stance is the internalized voice of that community, operating as individual conscience.

Key Ideas

Orientation, not technique. Stance is the way of engaging with work—priorities about what matters, tolerance for particular difficulties, standards defining good enough.

Transmitted biographically. Stances are absorbed through relationship with practitioners who embody them—caught through observation of struggle, not taught through instruction.

Governs tool use. Identical technical skills produce different outcomes when practitioners' stances differ—one optimizes for speed, another for quality, a third for elegance.

AI cannot transmit. Machine partnerships demonstrate patterns but not the biographical depth, aesthetic conviction, or caring that constitute a stance.

Develops through friction. Stances form through the emotionally consequential experience of being held to standards by someone whose judgment you trust—a relational achievement, not an information transfer.