Controlled Codes
Every motif occurrence in ACIM Atlas is classified along six analytical dimensions. These are called controlled codes because each one uses a small, fixed set of allowed values—not free-form descriptions, but a constrained vocabulary that can be compared across occurrences, filtered, and used for pattern detection.
The codes capture how the Text is using a motif in a given section. Not just that the motif is present (which salience measures), but what register it is operating in. A section might treat forgiveness in a metaphysical, non-dual frame, using a corrective tone, focused on the learner’s own mind. Another section might treat forgiveness practically, using a narrative tone, focused on the learner’s relationship with a brother. Both are forgiveness. The codes distinguish what kind of treatment is happening.
The six dimensions are assigned during Pass 1 as part of the section-level annotation. They appear on the site in Research mode, grouped into three pairs, and they also feed the plain-language section profile (“Understanding This Section”) that Pass 4 produces for every section page.
Approach
The first pair describes the teaching’s orientation—at what level and how abstractly the motif is presented.
Level
What plane of the Course’s teaching is foregrounded.
| Code | Label on site | What it means |
|---|---|---|
| L1 | Metaphysical | Non-dual framing. The teaching operates at the level of reality versus illusion, beyond perception, beyond the dream. God, Christ, Heaven, and the nature of what is real are foregrounded. |
| L2 | Practical | Within-the-dream framing. The teaching operates at the level of forgiveness practice, choosing again, healing perception, and undoing projection—the work the learner does within experience. |
| MIXED | Bridge | An explicit bridge between L1 and L2. The section connects the metaphysical frame to practical application, or moves between the two. |
| META | About the teaching | The Course is teaching about its own method, language, or pedagogy—explaining why it teaches the way it does, or clarifying its own levels. |
This distinction reflects a core feature of the ACIM Text: it operates on multiple levels simultaneously. Recognizing whether a section is speaking about ultimate reality (L1) or about what the learner can practice within the dream (L2) is one of the most important skills in studying the Course, and one of the most common sources of confusion. The level code makes this dimension visible.
Abstraction
How abstract or concrete the presentation is—a separate question from level.
| Code | Label on site | What it means |
|---|---|---|
| concrete | Concrete | Situational, worldly examples are foregrounded—specific scenarios, relationships, or experiences the learner might recognize. |
| psychological | Psychological | Mind dynamics within experience are foregrounded—fear, guilt, projection, denial, attraction, resistance. The teaching is about inner processes. |
| metaphysical | Metaphysical | Non-dual reality/illusion framing is foregrounded. Abstract principles about the nature of what is real. |
| bridge | Bridge | Links concrete or psychological material with metaphysical framing—the section moves between levels of abstraction or uses one to illuminate the other. |
Abstraction and level are related but not identical. A section can be L2 (practical, within the dream) and metaphysical in abstraction (discussing the nature of perception in abstract terms). Another can be L1 (non-dual) and psychological (describing the mind’s resistance to accepting reality). The two codes together give a more precise picture than either alone.
Voice
The second pair describes the tone and interpretive frame—how the teaching is delivered and whose interpretation is foregrounded.
Tone
The dominant voice or rhetorical style of the motif’s treatment.
| Code | Label on site | What it means |
|---|---|---|
| expository | Expository | Straightforward explanation. The section defines, describes, or explains the motif directly. |
| corrective | Corrective | Undoing or refuting ego logic. The section is actively dismantling a misperception, reversing an assumption, or issuing a directive correction. |
| poetic | Poetic | Elevated, evocative, imagery-forward. The section uses language that is more lyrical or metaphorical than instructional. |
| narrative | Narrative | Story, analogy, or dialogue is the primary vehicle. The teaching is delivered through a scenario or extended comparison rather than direct instruction. |
| meta | Meta | Teaching about teaching. The section discusses the Course’s own method or explains why it approaches a topic the way it does. |
Tone matters because the Course uses different rhetorical strategies for different purposes. A corrective passage that dismantles an ego defense is doing different work than an expository passage that calmly defines a concept, even when both address the same motif. The tone code makes this visible.
Frame
Which interpretive lens is foregrounded—the ego’s or the Holy Spirit’s.
| Code | Label on site | What it means |
|---|---|---|
| ego | Ego | The ego’s interpretation is foregrounded. The section is describing, exposing, or dismantling how the ego sees things. |
| holy_spirit | Holy Spirit | The Holy Spirit’s correction is foregrounded. The section is presenting the reinterpretation, the alternative, the undoing. |
| mixed | Mixed | Both frames are explicitly contrasted or the section pivots between them. The teaching is built on the juxtaposition. |
Many sections of the Text work by presenting the ego’s interpretation and then offering the Holy Spirit’s correction. The frame code indicates which side of that contrast is dominant, or whether the contrast itself is the point.
Focus
The third pair describes where the teaching is directed—what relational context and what site of change.
Relational focus
The primary relational target where the teaching is applied.
| Code | Label on site | What it means |
|---|---|---|
| self | Self | Intrapersonal. The teaching is directed at the learner’s own mind, perception, or choice. |
| brother | Brother | Interpersonal. The teaching foregrounds another person or a relationship dynamic—how the learner sees, judges, or forgives a brother. |
| sonship | Sonship | Collective. The teaching addresses the shared identity of the Sonship—“we” as joined mind, beyond individual relationships. |
| god | God | The relationship with God, Heaven, or Knowledge is foregrounded. |
| mixed | Mixed | Multiple relational contexts are equally active or explicitly contrasted within the section. |
Agency
Where the “doing” is located—the apparent site of change that the teaching points toward.
| Code | Label on site | What it means |
|---|---|---|
| behavior | Behavior | External action is emphasized. The teaching references what the learner does in the world, even if only to reframe it. |
| mind_choice | Mind choice | Inner decision is emphasized. The teaching points to choosing a different teacher, shifting perception, or making a decision at the level of mind. |
| beyond | Beyond | The teaching points past doing and choosing into being, knowledge, or timelessness—beyond the learner’s sense of personal agency. |
Agency is a particularly useful dimension for studying the Course because one of its central teachings is the distinction between behavioral change (which is form) and mind-level change (which is content). The agency code tracks which side of that distinction the section is emphasizing.
The “unknown” value
Every code dimension allows a value of UNKNOWN. This is assigned when the evidence does not support a confident classification—the section’s treatment of the motif is ambiguous, the signals point in multiple directions without a clear dominant, or the occurrence is too faint (typically an echo) to classify reliably.
UNKNOWN is not a failure. It is a statement of epistemic honesty. A forced classification on ambiguous evidence would be less accurate than acknowledging the ambiguity.
How the codes are used
On the site, the six codes appear in Research mode as three labeled groups—Approach (Level + Abstraction), Voice (Tone + Frame), Focus (Relational + Agency). Each code displays as its human-readable label, not its raw value. Every section page also carries a plain-language interpretation of the codes in the “Understanding This Section” block on the section header.
Behind the scenes, the codes serve several functions beyond display:
- Trail-level change detection. When Pass 2 evaluates whether an occurrence represents a shift in the motif’s treatment, changes in controlled codes are one of the signals it looks for. A motif that shifts from L1/metaphysical to L2/concrete between two occurrences is showing genuine development, even if both occurrences have similar salience.
- Editorial input. The Pass 4 editorial layer uses the codes as input when writing discussions and identifying common confusions. A long discussion for an L1/corrective occurrence reads differently than one for an L2/expository occurrence.
- Section profile generation. The section-level “Understanding This Section” voice/approach/focus fields are written by interpreting the controlled code distribution across all motif occurrences in the section.
A note on precision
Each occurrence gets one value per code dimension. This is a simplification. Many sections of the Text are genuinely multi-tonal, multi-level, or multi-focused—a single passage might begin expository and shift to poetic, or foreground both self and brother in alternation.
The codes capture the dominant register for each dimension—the best single label for how the motif is used in this section. Where the section genuinely operates on two levels equally, the bridge or mixed values exist for that purpose. But the system does not attempt to model every nuance of a section’s rhetorical complexity. It provides a consistent, comparable feature set, not an exhaustive description.
This is a trade-off the project makes consciously. A richer coding system (multiple values per dimension, weighted blends, paragraph-level granularity) would capture more nuance but would be harder to score consistently across the full corpus and harder to use for comparison. The single-value approach prioritizes consistency and comparability, which are the qualities that make longitudinal tracking possible.
For what the corpus-wide distributions reveal about how the Text teaches, see the Findings page.