How to explore: drag to rotate · scroll to zoom · hover dots or lines · click for detail · toggle panel ↗
The dissertation tests whether Soviet-era talking points on famine commemoration transform predictably when they cross register boundaries - secret memo → embassy/Novosti → domestic news → diaspora/scholarly print - and whether that transformation is dated and measurable (Montreal 1983 as anchor, post-1983 explosion as qualitative rewire). These 3D models make Lotman’s spatial vocabulary falsifiable: they read from the same open JSON as Figures 1-12.
Two phrase-network maps stacked vertically. Lower plane = memos before 1983 (only two early files). Upper plane = memos after 1983 (fourteen files). Each dot = a wording-habit cluster; each line = those habits appear in the same memo. Red lines = co-occurrences that did not exist before Montreal - new internal wiring. Grey vertical pillars = topics that continue across the rupture at the same horizontal position.
Before 1983, three topics co-occur in a tight triangle - operational vocabulary, OUN actor references, scare quotes. After 1983, new nodes appear («artificial famine», hostile «gathering») and new edges link them. The institution did not merely write more memos; it rewired which phrases travel together when commemoration proceeds despite counter-measures. That is Lotman’s explosion as a dated, countable claim.
Lotman’s ‘explosion’ names qualitative reshaping, not louder word counts. After Montreal 1983 the memo phrase network gains new nodes («artificial famine», hostile «gathering») and new edges between them - the institution rewires which wording habits co-occur when the outside story ‘wins.’ The lower plane is the pre-rupture bundle; red upper-plane edges are co-occurrences that did not exist before.
Connections on the upper plane are internal to the fond - they do not prove outward press changed line-by-line. They prove the secret register’s phrase ecology bifurcated after the 1983 anchor. Pair this figure with L1 (outward relay) and L2 (operator skew) for the full argument: boundary non-equivalence at the shell edge and qualitative rewire inside the archive after the outside story wins.
Δ topics = 4 · Δ links = 16 · new topics: artificial_famine, hostile_gathering, pedagogy, symposium_recode
Try it: click the red line between «Quoted artificial famine» and «Hostile gathering» on the upper plane - read “What this connection tells us” and “Thesis significance.” Compare to the lower-plane triangle where that edge did not exist.
| Period | Topic A | Topic B | Shared memos | Status |
|---|---|---|---|---|
| Pre-1983 | Operational measures | OUN / nationalist actor | 1 | carried |
| Pre-1983 | Operational measures | Scare quotes | 1 | carried |
| Pre-1983 | OUN / nationalist actor | Scare quotes | 2 | carried |
| Post-1983 | Quoted «artificial famine» | Hostile «gathering» | 3 | NEW |
| Post-1983 | Quoted «artificial famine» | Operational measures | 5 | NEW |
| Post-1983 | Quoted «artificial famine» | OUN / nationalist actor | 5 | NEW |
| Post-1983 | Quoted «artificial famine» | Scare quotes | 6 | NEW |
| Post-1983 | Quoted «artificial famine» | Symposium recode | 4 | NEW |
| Post-1983 | Hostile «gathering» | Operational measures | 4 | NEW |
| Post-1983 | Hostile «gathering» | OUN / nationalist actor | 5 | NEW |
| Post-1983 | Hostile «gathering» | Scare quotes | 5 | NEW |
| Post-1983 | Hostile «gathering» | Symposium recode | 4 | NEW |
| Post-1983 | Operational measures | OUN / nationalist actor | 12 | carried |
| Post-1983 | Operational measures | Pedagogy / counter-propaganda | 4 | NEW |
| Post-1983 | Operational measures | Scare quotes | 13 | carried |
| Post-1983 | Operational measures | Symposium recode | 6 | NEW |
| Post-1983 | OUN / nationalist actor | Pedagogy / counter-propaganda | 4 | NEW |
| Post-1983 | OUN / nationalist actor | Scare quotes | 13 | carried |
| Post-1983 | OUN / nationalist actor | Symposium recode | 7 | NEW |
| Post-1983 | Pedagogy / counter-propaganda | Scare quotes | 4 | NEW |
| Post-1983 | Pedagogy / counter-propaganda | Symposium recode | 1 | NEW |
| Post-1983 | Scare quotes | Symposium recode | 7 | NEW |
Edge weight = memos where both phrase families appear. Source: bc_combined_experiments.json → c_explosion_rewire.
Each panel opens one step on the way from the memos to the two networks, with the equation it uses.
For two phrase families a and b, the weight is the number of memos in a period that contain both. 1[a ∈ m] is 1 if family a appears in memo m and 0 if not. Before 1983 there are only two memos, and after 1983 there are fourteen.
A line is red if the pair occurs together after 1983 and never before. Before 1983 the network has three topics and three lines. After it has seven topics and nineteen lines: three carried over and sixteen new. That gives the summary under the figure, Δ topics = 4 and Δ links = 16. The table under the figure lists every line with its count.
Lines get thicker with the number of shared memos, up to a cap. The two periods sit on separate planes. Dot positions on each plane are stored in the data. The lower plane is a simple triangle, because all three early topics occur in the same early memos (1976 to 1977). On the upper plane, topics that already existed keep their left-right position and new ones are placed near the topics they co-occur with. Grey pillars join a topic's two positions. Position on a plane is a layout choice, not a measurement.
Colours: gold for lines on the lower plane, blue for lines carried over on the upper plane, red for new lines.
Play before and after moves the view down to the before-1983 plane, holds while the caption describes it, then rises to the after-1983 plane and settles back where it started, about fifteen seconds in all. Only the camera moves; the network is fixed. Turn the model spins the two planes slowly. Nothing starts on its own.
The counts come from bc_combined_experiments.json under c_explosion_rewire, as the page notes under its table. The page is generated in the thesis lab by build_lotman_3d_visualizations.py and copied here on sync. This section is kept in explain/fig-lotman-explosion-3d.html in the site repo and added back on every sync.
Before 1983 the memos stick to three themes that turn up together. After the symposium went ahead despite the KGB's countermeasures, four new themes appear, among them a quoted "artificial famine" and the hostile "gathering", and sixteen new pairings link them to the old ones. The memos didn't just get more numerous. Their way of talking rewired.
Lotman's explosion is a sudden change in what can be said, not more of the same said louder. This figure is the dissertation's way of dating and counting such a change inside one institution: the outside world's story won a round in 1983, and the secret register reorganised itself in response. The before side rests on only two memos, so the journal treats this as a candidate result until more early memos are found. Read it alongside L1, which shows what crossed outward, and L2, which counts the wording habits.
The most exciting idea on the site and the most starved of data. Two memos before 1983 are too few to call anything a rewiring with confidence, since one extra early memo could move many of the "new" pairings onto the before side. That's a limit of the archive, not a flaw in the idea, and the journal already treats the result as a candidate.
What would help: more pre-1983 memos from the same fond, so the two periods are closer in size; a comparison that adjusts for the number of memos, for instance by drawing many two-memo samples from the after period and counting how often that many new pairings appear by chance; and a check that the operators from the 1955 manual persist while the neighbourhoods change, which the colloquium notes suggest as a slow clock. If the rewiring survives those, it becomes the dissertation's clearest dated instance of Lotman's explosion.