Competence is where the problem lives. The practitioner most likely to reproduce structural power without noticing is the one who is best at reading the room.
A firefighter entered a burning row house and felt, before he could name why, that something was wrong. The heat was coming from the floor, not the walls. The room was quiet where it should have been loud. He ordered his crew out. Seconds later, the floor collapsed into an unknown basement fire below. Gary Klein, documenting this moment in Sources of Power, named what happened: the expert's pattern library ran the scene and flagged a mismatch before conscious thought arrived to witness it. The practitioner's read had preceded his reasoning by a significant margin. That gap is this issue's subject — and its warning.
The firefighter's case is Gary Klein's central exhibit in the Recognition-Primed Decision (RPD) model — documented across dozens of hours with fireground commanders, military decision-makers, and emergency physicians. The finding is consistent: experts in high-stakes environments do not reason through options sequentially. They pattern-match. They run the current scene against a library of prior situations, and when the match is strong enough, they act. When the match is anomalous — something feels wrong that can't yet be named — they pause. The pause itself is information. The expert is not deciding. They are reading.
Klein's firefighter felt the anomaly in his body before he located it analytically: the unusual floor heat, the atypical silence. The read was faster than the name. This is not a malfunction. In Klein's account, it is precisely what expertise produces — a perceptual system trained to extract signal from complex environments at a speed that deliberative cognition cannot match.
Research on expert intuition — Gary Klein's work on recognition-primed decision-making, and the broader expertise literature — describes how seasoned practitioners in relational fields recognize a pattern before they can put it into words. The read arrives faster than deliberate analysis, below the threshold at which conscious reasoning operates. The expert's read of "this group is ready to go deeper" or "this participant is being resistant" or "the energy is shifting" is not a reasoned conclusion. It is a perceptual event. It arrives the way the firefighter's unease arrived: faster than the name for it.
This is an understudied vulnerability in professional self-assessment: expert confidence can correlate negatively with recognition of automatic-processing errors. The more experienced the practitioner, the more certain they are that their reads are accurate — and the less likely they are to notice when the read is reproducing something other than what is actually in the room.
This is the entry point for Freire. If expertise operates below conscious awareness — if the practitioner's read arrives before reasoning can examine it — then the question is not whether the read is fast. The question is what the read was trained on. Pattern recognition is not neutral. It is a library. And every library was curated by someone, somewhere, in conditions that were not neutral either.
The argument this issue makes requires three distinct bodies of work, each of which establishes a different layer of the same structural problem. None of them is sufficient alone. Together, they close the case.
Klein's RPD model establishes the mechanism: expert decision-making in high-stakes environments relies on rapid situational pattern-matching against a library of prior experience, not deliberate option comparison. Speed is the feature, not the bug — until the situation does not match what the library was trained on. Klein's contribution is to name the structural vulnerability precisely: the RPD model works when the current situation belongs to the same population of cases that built the library. When the situation is structurally different — when the room contains dynamics the practitioner's formation never indexed — the expert reads a false pattern at high speed and acts on it with high confidence. The firefighter's library flagged the anomaly. But not every anomaly is flagged. The ones that are invisible to the library pass through the pattern-recognition system undetected. The practitioner does not feel uncertain. They feel competent.
Freire's distinction between naïve and critical consciousness — in Pedagogy of the Oppressed, Chapter 3 — establishes that naïve consciousness does not perceive conditions as conditions. It perceives them as natural facts. The participant who redirects to process failures is "resistant." The participant who challenges the frame is "avoiding accountability." These reads are not reported as interpretations. They arrive as perceptions. Freire's practitioner-facing implication — rarely quoted — is that the educator's own consciousness is the first site of transformation required. The practitioner's perceptual apparatus, not only the content they deliver, is the primary object of critical inquiry. The expert who has accumulated thousands of facilitated hours has not become more critically conscious by accumulating them. They have become more fluent at a set of reads whose structural assumptions have been reinforced, not examined, by repetition.
Goffman's account of impression management in The Presentation of Self in Everyday Life supplies the layer Klein and Freire do not: why the practitioner does not pause. The expert's competence is not only cognitive — it is performed. The facilitated room is a performance space in which the practitioner continuously maintains a coherent front. Critical self-examination — especially public, in-session examination — threatens front maintenance. To say "I'm noticing that my read of this participant might be reproducing the room's existing power structure" is not just epistemically difficult. It is socially costly. It introduces uncertainty where certainty is expected. It surfaces the possibility that the expert is not the most accurate reader in the room. Goffman explains why the practitioner who feels uncertain does not name it: the performance structure of expertise is itself a barrier to the slow-down that critical consciousness requires.
Kahneman's framework in Thinking, Fast and Slow names the operating mode underlying Klein's RPD: System 1 — fast, associative, pattern-driven, and largely invisible to conscious monitoring. System 2 — deliberate, effortful, sequential — engages only when System 1 encounters a mismatch strong enough to trip an interruption signal. Kahneman and Klein collaborated on a 2009 paper, "Conditions for Intuitive Expertise," that identifies precisely when System 1 expert intuition is reliable: environments with regular structure, sufficient practice, and rapid unambiguous feedback. The facilitated room violates all three. Structure varies across organizations, demographics, and power configurations. Feedback is delayed, filtered through evaluation forms that measure satisfaction rather than accuracy, and systematically distorted by social politeness. Kahneman's contribution to this issue's argument: expert confidence in a read is not evidence of its accuracy. System 1 is confident by design. It does not experience the difference between "pattern-matched correctly" and "pattern-matched incorrectly" until the floor collapses. The practitioner who feels certain is not better calibrated than the practitioner who feels uncertain. They have simply stopped questioning a library that has been reinforced long enough to feel like perception.
Sports coaching introduced the slow-motion review to correct a specific problem: the expert's real-time read of a play is faster than the play. By the time the coach has formed a conscious judgment about what happened, the moment has been overwritten by the next one. The replay at 25% speed does not give the coach new information. It gives them the same information with enough time to examine what was always there.
Surgical training uses the same method. The attending physician's intraoperative decisions arrive faster than verbal narration can track them. The teaching value is not in the outcome — it is in the frame-by-frame decomposition of the judgment that produced it. What was seen. What was weighted. What was not registered at all.
The film strips below run the same facilitation moment at two speeds. Left strip: the automated read — what the practitioner perceived, labeled, and acted on in real time. Right strip: the structural frame excluded by that read — what was present in the room but not indexed by the pattern library. These are not two different interpretations of the moment. They are the same moment, examined at different processing speeds.
Participant redirects to "process failures." Pattern library fires: defensive framing.
Read confirmed by posture: arms crossed, voice tightened. Signal: resistance to accountability.
Practitioner redirects group. Frames forward movement. Participant falls silent. Read: integration.
Session closes. Evaluation forms positive. Library reinforced: the redirect worked.
Participant is only non-direct-report in the room. "Process failures" is the accurate diagnosis she raised six months ago and was told to drop.
Arms crossed, voice tightened: body-state of someone without structural safety in the room. Not resistance. Self-protection.
Participant silent: not integrated. Suppressed. The redirect centered the senior leader's frame. Structurally least safe person removed from conversation.
Library reinforced incorrectly: "redirect worked" because the room's existing power structure was reproduced without friction. That is not success.
The automated read arrived faster than deliberate thought. The excluded frame was present in the room the entire time.
The slow-motion review does not add new information. It adds processing time.
The concept this issue names is not a failure of skill. The practitioner above may be excellent at their work by every conventional measure. The concept is that the RPD library is a historical object — built in specific rooms, with specific bodies, under specific conditions of professional formation. When the current situation contains structural dynamics that the library never indexed, the expert reads the situation accurately within the library's frame. The library's frame is the problem.
Freire's naïve consciousness names this: perceiving conditions as facts rather than as historical, political, and changeable constructions. The expert practitioner's version of naïve consciousness is not born of inexperience. It is born of fluency. The read has been reinforced so many times that it does not feel like a read. It feels like perception.
The argument this issue makes generates predictable resistance in experienced practitioners. That resistance is worth taking seriously — not because it is wrong, but because the form it takes is itself a data point about the problem.
No. Klein's account of the RPD model is an account of expertise working correctly within its domain. The firefighter's fast read saved lives. The argument is not that pattern recognition is unreliable. The argument is that pattern recognition is a function of training data. When the training data is comprehensive — when the library includes the full range of structural dynamics the practitioner will encounter — the fast read is accurate. When the library has systematic gaps — rooms it never observed, bodies it never centered, power configurations it never indexed — the fast read is accurate within the library and wrong in the room. The issue is calibration, not capability.
True, and not what this issue proposes. The slow-motion review is not an in-session practice. It is a preparation and debrief practice. The firefighter did not run a slow-motion review while the floor was collapsing. The review happens afterward: what did the library flag, what did it miss, what would a different pattern library have read differently? The practitioner who builds a slow-motion review practice over time is not slowing down in the room. They are recalibrating the library that runs before the room begins.
Reflective practice, as most practitioners execute it, asks: what did I do, and what would I do differently? The slow-motion review asks a structurally prior question: what did my pattern library exclude from the read, and why? These are different inquiries. The first is an audit of action. The second is an audit of perception. Goffman's contribution here is decisive: the performance pressure of expertise makes the second inquiry socially costly in ways the first is not. Saying "I would redirect differently next time" maintains the front of competence. Saying "my read of that participant reproduced the room's power structure" does not. The resistance to framing this as DEI work is worth noting: it is exactly the kind of automated front-maintenance Goffman describes.
Consciousness — from the Latin conscientia, knowing together, shared knowledge. Critical — from the Greek krinein, to separate, to judge, to discern. Critical consciousness is not simply awareness. It is the capacity to separate what is perceived from the conditions that shaped the perception — to see the frame as a frame, not as reality itself.
Freire developed this concept in the context of popular education in northeastern Brazil in the 1950s and 1960s. The conditions he was working against were specific: a peasant class that perceived their poverty as natural, their illiteracy as personal failure, their exclusion as the inevitable shape of the world. Naïve consciousness was not stupidity. It was the internalized epistemic structure of a social order that benefited from its subjects not perceiving it as a social order.
The expert practitioner's problem is a different instantiation of the same structure. The professional formation system — graduate training, clinical supervision, certification processes, continuing education markets — is not a neutral library. It was built in institutions shaped by specific racial, economic, and gender formations. The rooms that built the RPD library were not a representative sample of all possible rooms. They were the rooms practitioners were trained in, credentialed in, observed in. The bodies that were centered in those rooms shaped what the library learned to read as signal and what it learned to read as noise.
This is not a claim that experienced practitioners are malicious or that their expertise is worthless. It is a claim that the library is a historical artifact. Every library is. The question is not whether the practitioner's pattern library was shaped by structural conditions — it was, inevitably, because all pattern libraries are built somewhere. The question is whether the practitioner knows this, and whether their practice includes systematic mechanisms for auditing it.
The slow-motion review is one such mechanism. Not the only one. But a specific one — suited to the exact vulnerability the fast read names. The library fires at a speed that outpaces reflective examination. The review slows the tape down after the fact, not to undo the judgment, but to see what the frame excluded. That is the practice. That is the discipline. That is what distinguishes critical consciousness from its well-intentioned imitation.
Two separate questions arise from this moment. They are easy to collapse into one, but they should not be.
What does it look like to repair that moment — not just in the relationship, but in the practitioner-client system? Reaching out to the participant after the session is the minimum. The more significant repair is with the client (the senior leader or the team): naming, at the appropriate moment, that the session may have reproduced a structural dynamic worth examining. This is not comfortable. It risks the engagement. It is the repair that the slow-motion review, honestly run, requires. The practitioner who only does the interpersonal repair and not the structural one has addressed the symptom without touching the mechanism.
What would have had to be different in your preparation — not your facilitation skill, but your structural awareness — for the read not to happen in the first place? This is the more important question. It is not about what you would do differently in the moment of the read. The read arrives before intention can intervene. It is about what the pre-session brief would have had to contain: the organizational map, the power relationships, the history of who has been heard and who has been redirected. A practitioner who enters a post-mortem with that structural brief has a library that can catch the anomaly. A practitioner who enters with only facilitation skill has the fast read uninterrupted by structural awareness. Preparation is not logistics. It is library calibration.
Even with a full structural brief, the fast read may still fire. Klein's fireground commanders received the library update through years of debriefs, slow-motion reviews, and AAR (After Action Review) processes — not through a single pre-session read. The library recalibrates slowly. A practitioner who has run one slow-motion review after one session has begun, not completed, the work. What this scenario requires in the long term is a consistent debrief practice in which structural exclusions are named, documented, and fed back into the library over time. Not because the practitioner is broken. Because the library was built in conditions that did not index everything the room contains.
Freire's argument — that the educator's own consciousness is the first site of transformation required — is not a call to self-flagellation. It is a technical claim about the sequence of the work. The practitioner cannot support structural transformation in a room they are inadvertently stabilizing. The slow-motion review is not a guilt practice. It is a calibration practice. The instrument is not broken. The reading protocol needs work.