Periodic Table of Immersion Techniques

Columns are channels: КЛ1 — worldbuilding; КЛ2 — sensory simulation; КЛ3 — tempo and scenic presence; КЛ4 — information gaps; КЛ5 — access to consciousness; КЛ6 — voice and address; КЛ7 — attention and structure; КЛ8 — foregrounding and formal play. Header colors identify families: I — КЛ1–3, II — КЛ4/7, III — КЛ5/6, IV — КЛ8. Column width reflects the number of cards, not effect strength.

Rows are temporal scales: МК — phrase, ЭП — scene, МАК — chapter/book, ПЕР — variable. Position means channel and scale, not quality rank; order within a cell is not a scale. Of 256 cards, 240 occupy the grid; 3 meta-operators (modulating realization) and 13 configurations (combinations) are separate below. Their vectors are a project model, not measured additive effects.

Search by ID or name and filters dim nonmatching grid cards, without recalculating the lower sections. Click a card for its definition, goal, proposed effect, examples, sources and cross-references. Below are 9 preselected pairs (8 numeric φ associations and one qualitative “mirror”; click to highlight a pair), configuration formulas and profiles of 8 authors. This is not an arbitrary live correlation calculator; a profile aggregates assessments, not reader psychology.

Try this: search for П61 (Cyrillic first letter), open “Omission / Iceberg theory” and compare the definition with its examples. Its position is КЛ4 × ПЕР, vector П↑ В↓. Clear the search and locate the П61 + П62 association: the highlight lets you compare omission with information gaps. Inspect manual and machine strips separately: do observations agree, and on what evidence base? Association does not establish causation.

Manual layer: columns represent 104 authors, assessed from author studies, not every book they wrote. ОЧ/Ч/Р mean very frequent/frequent/rare under the project rubric; “—” means “not used” in the assessed material; “н/д” means no data. These are ordinal assessments, not word counts. “Manual” names the author layer: part of the world-corpus expansion was imported from arbitrated machine summaries (*М2); some portraits remain labeled “machine draft”. This does not guarantee a manual reading of every text. Core 93 / Sparse 11 partitions authors by coverage. The original methodology uses >60%; after corpus expansion, Core-author inclusion uses an absolute threshold of ≥100 assessed cells (core_split.py, decision 12.6). This is not an author ranking; there is no separate Core filter in this interface.

Machine book layer 🤖: 155 book columns, not 155 authors, containing draft aggregates of pipeline findings. Its “—” means no confirmed machine finding, not absence of the technique from the book. “н/д” and a missing manual row are not zero. Book-level context, translation and chunking constrain detection. Do not confuse this layer with catalog section T.

Four color modes: valence — Д (story-world motivation), С (both modes), Э (exposed medium/aesthetic distance): a motivation summary, not an effect measurement. Gray means no tag in the current 80-dual-card layer, not “ineffective”. “П vector” shows the sign of the П axis, not all axes; “weight · authors” is the proportion of manual carriers out of 104; “coverage · books” is the proportion of book columns with findings out of 155. ☢ means fewer than 5 manual carriers or a row outside the manual matrix, not danger or poor quality.

Vector П/Г/А/М/В (P/G/A/M/V): immersion / relation to character / authorial presence / world authentication / continuity of experience. ↑ strengthens, ↓ weakens or reverses, ↕ depends on realization; an omitted axis is neutral by model convention, not an experimentally proven zero. LDI: TR — translation-resistant, LS — language-specific, INT — intermediate. RU/EN/FR changes presentation, not the corpus; Cyrillic IDs and codes remain stable. Hover over card strips to see the author/book and assessment.

Data snapshot: 2026-10-26. The table summarizes the research technique catalog, classification, and assessments from the manual and machine corpora. Coverage is uneven; machine assessments remain separate from manual ones and do not establish causation.