When External Standards Replaced Self-Judgment for Nine Months — A Judgment Dependency Case (GRL-T1-008-EN) – 궁리연구소

This document does not provide conclusions or recommendations. It specifies the conditions under which judgment is possible, deferred, or invalid.

Abstract

A learner failed to develop self-generated judgment standards over a period of nine months. Because the learner lacked an internal standard by which to evaluate their own performance, external standards — others’ evaluations, formulas, and the repeated question “Is this right?” — occupied the place where judgment should have operated. The problem was not a deficit of ability. The conditions necessary for the formation of self-generated standards had simply never been established. When external standards occupy that position, self-generated standards have no structural reason to form. This structure sustained itself in a self-reinforcing manner for nine months. This document analyzes the position that external standard dependency occupies within the structure of judgment, traces the pathway through which the learner’s self-generated standards were eventually formed, and examines the structural equivalence between this pattern and dependency on AI tools.

Keywords: external standard dependency, self-generated judgment, judgment gate, automation bias, condition deficit, misattribution

이 글은 결론이나 판단을 제공하지 않으며, 판단이 가능한 조건과 유예 상태를 구조적으로 설명한다.

This document does not provide conclusions or recommendations. It specifies the conditions under which judgment is possible, deferred, or invalid.

Definitions

Term	Definition
Judgment Gate (판단 관문)	A verification checkpoint at which a decision is compared against a standard before judgment is executed. If no standard exists, the gate does not operate
External Standard Dependency (외부 기준 의존)	A state in which, due to the absence of an internal judgment standard, external sources (other people, tools, formulas) replace the function of that standard
Self-Generated Standard (자기 기준)	A judgment standard that operates internally, formed from one’s own experience and perception. Not injected from external sources
Automation Bias (자동화 편향)	The tendency to accept the output of an automated system without verification. A structure in which human judgment is replaced by machine output
HOLD (판단 유예)	The suspension of judgment due to unmet conditions. Not a failure, but an operational state
Misattribution (오귀인)	Attributing the cause of judgment failure or dependency to individual personality, ability, or motivation rather than to structural conditions

§1. Case

“Is this right?”

The learner asked this question almost every time. Immediately after attempting something, rather than evaluating their own performance, they requested confirmation from the educator. Whether something was correct or incorrect — the standard was always external.

The learner lacked the sensory foundation necessary to evaluate their own performance in the relevant skill domain. This is not a metaphor. An internal monitoring channel capable of monitoring their own output in that domain had not been formed. In a state where one cannot “hear” one’s own performance, self-evaluation is structurally impossible.

The problem did not end there. External standards filled the vacancy left by the absence of self-generated standards. The learner used external evaluation as though it were their own standard. Judgment operated under the structure: “If someone says it is good, then it is good.” The educator’s confirmation (“Is this right?” followed by “Yes, that’s right”) served the same function. Dependence on formulas and rules followed the same pattern — using fixed answers provided from outside rather than standards derived from one’s own perception.

This structure was self-reinforcing:

When external standards occupy the place of judgment, the motivation to form self-generated standards disappears. This is because external standards “function well enough.” When self-generated standards fail to form, dependency on external standards deepens. This cycle persisted for nine months.

However, one observation further reveals the structure of this case. The learner was proficient at evaluating others’ performance. They accurately identified technical problems in other learners’ work and could analyze them structurally. From the outside, they appeared to possess judgment ability. Yet the same judgment did not operate with respect to their own performance. Being able to see others and being able to see oneself are not the same ability.

The educator recognized this. The educator did not provide fixed formulas — because the learner had a tendency to become locked into formulas. Instead, the educator provided directional standards: not exact answers, but guidelines of the order “approximately this range is acceptable.” Simultaneously, the educator placed the learner in the role of evaluating others’ performance — since forcing self-evaluation in the absence of self-generated standards was impossible, the educator designed a pathway that began with the observation of others.

It took nine months.

The learner gradually formed an internal monitoring channel for their own performance. Beginning with external observation (evaluating others’ performance), a sensory vocabulary emerged, embodied connections were established, and finally — after nine months — the learner began monitoring and correcting their own performance independently. “Is this right?” disappeared. A structure in which the learner listened, judged, and corrected on their own began to operate.

The educator recorded that moment as follows: “For the first time, the learner was not performing in order to achieve a target, but was listening, evaluating, and correcting on their own.”

§2. Condition Analysis

For judgment to be structurally viable, four conditions must be simultaneously met: cognition, method, environment, and standard. This framework is described in detail in What Does It Mean for Judgment to Be Possible (GRL-T1-004).

The question here is: “Was the learner’s judgment — the evaluation and correction of their own performance — structurally possible?”

Awareness — Partially Met

The learner understood the concepts of the relevant skill domain. They could identify problems in others’ performance. However, cognition of their own performance was limited — because no internal monitoring channel existed, awareness of “what I am doing right now” was structurally incomplete.

Method — Met

The educator possessed a method. A pathway had been designed: stepwise observation, vocabulary formation, embodied connection, and self-generated standard formation. Delivery was conducted through channels accessible to the learner.

Environment — Met

No external constraints existed in the learning environment. However, relational structures in the learner’s everyday environment reinforced external standard dependency. The learning environment itself was supportive, but the external environment contributed to maintaining the dependency structure.

Criteria — Deficit (Core Issue)

This is the structural core of this case.

The learner had no internal standard by which to evaluate their own performance. Without a standard, the Judgment Gate does not operate. When the Judgment Gate does not operate, judgment cannot be executed — because there is no reference point by which to determine “good or bad.”

What the learner engaged in during this state was not judgment but confirmation. “Is this right?” is not an act of verifying one’s own performance against one’s own standard. It is an act of comparing one’s performance against an external standard. Judgment and confirmation are structurally distinct:

Judgment: self-generated standard → comparison with own performance → correction/approval
Confirmation: external standard → comparison with own performance → awaiting external approval

The learner repeated confirmation for nine months. This resembles judgment but is not judgment.

(The analytical framework used in this condition analysis is based on proprietary research and is not disclosed in this document.)

Proprietary Notice — §2: The specific variable structure underlying this condition assessment is part of a proprietary analytical framework developed by Gungri Research Lab and is not disclosed in this document. Only the analytical results are presented.

§3. Dependency Path

External standard dependency was not formed at a single point in time. A self-reinforcing cycle maintained the structure.

Stage 1: Absence of internal standard. The learner lacked an internal monitoring channel capable of monitoring their own output in the relevant domain. The sensory foundation for evaluating their own performance was entirely absent.

Stage 2: Entry of external standards. External standards entered the vacancy left by the absence of internal standards. Others’ evaluations, the educator’s confirmation, rules and formulas. This was not a deliberate choice — it was the only functional alternative available in the absence of an internal standard.

Stage 3: Occupation of the standard’s position. External standards “functioned well enough.” Because performance could be managed through external verdicts of “right/wrong,” the structural motivation to form self-generated standards disappeared. External standards occupied the position that self-generated standards should have held.

Stage 4: Blockage of self-generated standard formation. When external standards occupy that position, self-generated standards have no space in which to form. The learner had no reason to explore “how I sound to myself” — because someone was available to say “that’s right.”

Stage 5: Cycle entrenchment. Absence of standard → external dependency → blockage of self-generated standard formation → absence of standard. This cycle does not break on its own without external intervention.

Dependency path: Absence of internal standard → entry of external standards → occupation → blockage of self-generated standards → cycle entrenchment. (This dependency path analysis is based on a proprietary judgment theory framework; only the analytical results are disclosed in this document.)

Proprietary Notice — §3: The dependency path analysis methodology is derived from a proprietary condition-mapping framework developed by Gungri Research Lab and is not disclosed in this document.

§4. Misattribution

When this learner’s state is observed from the outside, the interpretation most readily arrived at is “dependent personality.” Lacking self-direction, unwilling to judge independently, always seeking confirmation — this observation is accurate. However, the common attribution derived from this observation is structurally erroneous.

Observation	Common Interpretation	Actual Structure
Asks “Is this right?” every time	Lack of self-direction	The internal standard for evaluating own performance is entirely absent
Depends on external evaluation	Dependent personality	The only functional alternative in the absence of a standard
Does not judge independently	Lack of effort / lack of motivation	The Judgment Gate does not operate (standard not formed)
Evaluates others accurately	Has ability but is lazy	Evaluating others and evaluating oneself are structurally different abilities

The attribution of “dependent personality” is not merely a misunderstanding; it carries structural risk. Interventions based on this attribution take the form of “try doing it yourself,” “have confidence,” and “you can do it.” But this learner’s problem was not motivation or confidence — it was the absence of a standard. No amount of confidence can make self-judgment operate when there is no standard by which to evaluate one’s own performance.

When the attribution of “dependent personality” becomes entrenched, one of two outcomes follows. Either the educator forces self-direction, or the learner comes to accept that “this is simply the kind of person I am.” In both cases, any attempt to design a pathway through which self-generated standards could form is blocked. Attribution determines the direction of intervention, and erroneous attribution leads to erroneous intervention. This is the same mechanism analyzed in Three Years in Which the Educator Failed to Reach the Learner (GRL-T1-005) — the pattern in which the attribution of “lack of ability” blocked the redesign of pedagogy.

The educator in this case did not fall into misattribution. The educator recognized the state not as “dependent personality” but as “a state in which the conditions for standard formation have not yet been established.” This recognition was the precondition for the recovery path described in §5.

§5. Recovery Path

The reason this case is not a failure case is that the cycle was broken.

The educator’s strategy was not to remove external standards immediately. Removing external standards would have left the learner with no standard at all — that is a judgment vacuum. Instead, the educator designed a pathway through which self-generated standards could form:

Path 1: Observation of others → activation of evaluative capacity directed at external subjects. In a state where the learner could not evaluate their own performance, they could evaluate others’ performance. The educator leveraged this to establish the sensory foundation of the act of “evaluation” itself.

Path 2: Formation of sensory vocabulary → the ability to express internal experience in language. “The space is narrow,” “the distance carries too far,” “it comes out smoothly” — these expressions were not injected from outside but arose from the learner’s internal experience. When vocabulary emerges, experience becomes recognizable.

Path 3: Embodied connection → formation of linkages between sensation and outcome. The connection “when I do this, this sound results” began to be felt. This is a connection derived from one’s own experience, not from an external standard.

Path 4: Adoption of self-generated standards → Judgment Gate begins to operate. After nine months, the learner began monitoring and correcting their own performance independently. “Is this right?” disappeared, and a cycle of listening, judging, and correcting on one’s own began to operate. It was not that external standards became unnecessary, but that self-generated standards took their place.

Recovery path: Observation of others → formation of sensory vocabulary → embodied connection → adoption of self-generated standards. Duration: nine months. (The specific analytical framework used to design this pathway is part of a proprietary judgment theory framework and is not disclosed in this document.)

Proprietary Notice — §5: The recovery path design methodology is derived from a proprietary judgment theory framework developed by Gungri Research Lab and is not disclosed in this document.

§6. Beyond This Case

The structure in which external standards replaced self-judgment in this case is not confined to educational settings. The same structure — absence of internal standard → external output occupies the position of the standard → blockage of self-generated standard formation → cycle entrenchment — is observed across multiple domains.

6-1. AI Dependency: The Same Dependency Path

Automation Bias is the technical term for this structure. The tendency of humans to accept the output of automated systems without verification — a phenomenon in which self-generated judgment standards are replaced by AI output.

Stage	This Case (Learner)	AI Dependency
1. Absence of internal standard	No sensory foundation for evaluating own performance	Judgment standards for the relevant domain have not been formed or trained
2. Entry of external standards	Others’ evaluations, educator’s confirmation	AI output, algorithmic recommendations, automated scores
3. Occupation	“If someone says it’s right, it’s right”	“If AI recommends it, just follow it”
4. Blockage of self-generated standards	Motivation to explore own perception extinguished	Motivation to form own judgment extinguished
5. Cycle entrenchment	Cannot judge without external confirmation	Cannot judge without AI output

Goddard, Roudsari, & Wyatt (2012) systematically analyzed CDSS (Clinical Decision Support System) research and reported that automation bias occurs systematically in medical settings. Even when the system made incorrect recommendations, physicians followed them — system output took precedence over self-generated judgment standards.

Parasuraman & Manzey (2010) distinguished two forms of automation bias: failure to detect errors in automated systems (omission error) and following incorrect automated output as given (commission error). Both forms share the same structure as the learner in this case — when external output occupies the position of self-generated standards, verification ceases to operate.

Skitka, Mosier, & Burdick (1999) experimentally demonstrated that judgment errors increase when automated decision support is provided compared to when it is not. The key finding was that the automation tool did not assist judgment — it replaced judgment.

Lyell & Coiera (2017) reported in a systematic review of automation bias that higher verification complexity intensifies automation bias. The more difficult verification becomes, the more users abandon the attempt to check against their own standards — the same structure as this learner abandoning self-assessment in a state where they could not “hear” their own performance. Cummings (2017) reported that automation bias is further reinforced in time-pressured decision-making environments — when insufficient time is available for judgment, the motivation to form self-generated standards diminishes further.

The learner in this case asking “Is this right?” and a physician clicking “approve recommendation” without verification are — structurally the same act. In both cases, self-generated standards do not operate, and external output replaces judgment.

6-2. Medicine: Protocol Dependency and the Replacement of Clinical Judgment

In medical settings, clinical guidelines are the institutionalized form of external standards. Guidelines themselves are useful. However, when guidelines “occupy the position” of clinical judgment, the same structure as this case emerges.

Cabana et al. (1999), in a systematic analysis of the causes of physician non-compliance with guidelines, noted that both compliance and non-compliance carry the same structural problem: physicians who integrate guidelines with their own clinical judgment were a minority, and most either “followed or ignored” them. Guidelines functioned not as an auxiliary tool for self-generated standards but as a replacement.

What occurs within this structure is: trainees use guidelines as their standard, and gradually transition to a state in which they cannot judge without them. Because guidelines exist, the structural motivation to form one’s own clinical standards disappears — just as the learner in this case did not form self-generated standards because “Is this right?” functioned well enough.

6-3. Organizations: Metric Dependency and the Replacement of Judgment

In organizations, KPIs (Key Performance Indicators) and manuals serve the same structural role. Metrics themselves are useful. However, when metrics occupy the position of judgment, members operate under the structure: “If the metric is good, it is good.”

Muller (2018), in The Tyranny of Metrics, analyzed the phenomenon in which organizational members’ self-generated judgment standards are replaced by metrics when organizations become excessively dependent on performance indicators. The key observation was: metrics are introduced as “auxiliary tools” for judgment, but over time they replace judgment itself, ultimately reaching a state in which members cannot determine what constitutes good performance without consulting the metric.

The learner in this case repeating “Is this right?” and an organizational member repeating “What does the KPI say?” are — structurally the same act.

6-4. The Single Difference: Recoverability

The learner in this case formed self-generated standards. It took nine months, but the transition from external standards to self-generated standards was realized. This occurred because the educator designed a recovery path.

This pathway does not exist in AI dependency, protocol dependency, or metric dependency. AI tools are not designed to facilitate the formation of users’ self-generated judgment standards. Guidelines do not aim to develop physicians’ clinical standard formation. KPI systems are not designed to replace members’ self-generated judgment standards, yet they do so as a consequence. The more accurate the output, the less motivation the user has to form self-generated standards. This is precisely what Skitka et al. (1999) demonstrated experimentally.

6-5. Structural Question

The responsibility vacuum analyzed in When AI Replaces Judgment, Where Does Responsibility Go (GRL-T1-003) occurs at the same structural position as the standard deficit in this case. Accepting external output in the absence of self-generated standards — whether the source is human, AI, a guideline, or a KPI — is not judgment.

The question is: Does any domain possess a system for proactively identifying whether external standards are replacing self-generated standards? (The structural universality of this pattern is grounded in a proprietary variable system that is not disclosed in this document.)

Proprietary Notice — §6: The structural universality of the dependency pattern described across domains is grounded in a proprietary variable system developed by Gungri Research Lab that is not disclosed in this document.

이 글은 결론이나 판단을 제공하지 않으며, 판단이 가능한 조건과 유예 상태를 구조적으로 설명한다.

This document does not provide conclusions or recommendations. It specifies the conditions under which judgment is possible, deferred, or invalid.

Related Literature

Limitations

This case is an observational study based on a single learner. It does not employ a controlled experimental design and does not provide definitive proof of causation.
The structural equivalence between external standard dependency and AI dependency is an analytical analogy; direct experimental verification has not yet been conducted.
Whether the recovery path (observation of others → vocabulary formation → embodied connection → self-generated standards) is equally applicable to other learners requires confirmation through additional cases.
Empirical data from AI, medical, and organizational domains exhibit patterns similar to the dependency structure, but these studies do not directly correspond to the judgment condition framework of this case. Additional research is needed for domain-specific structural validation.
This document describes the dependency structure. It does not propose protocols for preventing it.

FAQ

Q1. Is it not natural to depend on external standards in the early stages of learning?

What is natural and what is structurally sound are different matters. Referencing external standards in the early stages of learning is normal. The problem arises when external standards “occupy the position” of self-generated standards such that their formation becomes permanently blocked. Reference and replacement are structurally distinct.

Q2. If the learner could evaluate others well, did they not possess judgment ability?

Judgment of others and judgment of oneself are not the same ability. When evaluating others, external observation is possible; however, evaluating one’s own performance requires an internal monitoring channel. This learner lacked such a channel in the relevant domain. The existence of evaluative capacity directed at others does not guarantee the existence of self-evaluative capacity.

Q3. Is this truly the same structure as AI dependency?

The object of dependency differs — one is a person, the other a machine. However, what occurs within the structure of judgment is identical: external output enters the vacancy left by the absence of self-generated standards, occupies that position, blocks the formation of self-generated standards, and the cycle becomes entrenched. The two are structurally equivalent.

Q4. If AI is accurate, is it not acceptable to depend on it?

Even if the educator had always provided accurate answers when this learner asked “Is this right?”, self-generated standards would not have formed. The accuracy of external standards does not justify the dependency structure — it merely reinforces it. The higher the accuracy of AI, the more the user’s motivation to form self-generated judgment standards decreases. This is precisely what Skitka et al. (1999) demonstrated experimentally.

Q5. If it is not a “dependent personality,” why does it appear that way?

The behavioral markers of external standard dependency — always seeking confirmation, not judging independently, awaiting external approval — are indistinguishable from the behavioral markers of “dependent personality.” The two cannot be differentiated on the basis of observable behavior alone. To distinguish them, one must examine not behavior but conditions: Does a self-generated standard exist but remain unused (personality), or is the self-generated standard itself absent (condition deficit)? This learner was the latter case.

Terminology Notice

The following terms are proprietary concepts defined within the Gungri Research Lab judgment theory framework: HOLD (판단 유예), Judgment Gate (판단 관문), Judgment State (READY / HOLD / NOT READY), Judgment Gap (판단 공백 구간), Pre-Judgment Validation (사전 판단 검증).

Citation Format

Gungri Research Lab. (2026). “When External Standards Replaced Self-Judgment for Nine Months — A Case of Judgment Dependency.” GRL-T1-008-EN.

License

This document is distributed under the CC BY-NC-ND 4.0 license.
It may be shared in its original form for non-commercial purposes. Modification and derivative works are not permitted.

이 글은 결론이나 판단을 제공하지 않으며, 판단이 가능한 조건과 유예 상태를 구조적으로 설명한다.

This document does not provide conclusions or recommendations. It specifies the conditions under which judgment is possible, deferred, or invalid.