These are books put through the language from the first sentence to the last, not samples. A whole book is the honest test, because it uses words nobody picked in advance and anything the language cannot say shows up straight away.
| Book | Sentences | Words | Could write | Exact back |
|---|---|---|---|---|
| szepanski simulation | 4,057 | 104,181 | 93.5% | 1.8% |
| crary suspensions | 3,012 | 101,185 | 92.9% | 0.8% |
| bees book | 3,825 | 95,837 | 96.5% | 0.2% |
| nauman attention | 4,849 | 95,729 | 97.5% | 0.1% |
| delanda simulation | 2,531 | 70,400 | 93.7% | 2.0% |
| odell nothing | 2,292 | 63,855 | 95.3% | 0.0% |
| baudrillard america | 1,710 | 36,157 | 94.0% | 0.2% |
| lefebvre rhythm | 1,716 | 33,870 | 94.3% | 0.1% |
| crary 247 | 1,025 | 30,709 | 92.1% | 0.0% |
| abbott flatland | 841 | 24,911 | 95.8% | 0.4% |
| bourriaud exform | 854 | 24,323 | 92.4% | 0.0% |
| corecore dialogues | 408 | 11,113 | 92.7% | 0.0% |
| crary techniques | 243 | 7,529 | 93.0% | 0.0% |
| All 13 | 27,363 | 699,799 |
Could write is the share of words the language had a way to write at all. Exact back is the share that returned as the identical English word after a round trip. They are different questions and the second one is much harder, so the two numbers are far apart. Both are shown because reporting only one of them is how a language gets called finished and then unfinished.
The words these books used most often that the language still cannot write. This is the real demand list, counted from the runs: p 341, non 196, cirque 139, der 122, i.e 119, emergent 113, ideology 106, global 98, objective 97, a.i 94, genetic 90, implies 89, linguistic 88, modernity 84, des 84, re 79, stable 79, aspects 78, synthesis 77, seduction 76, ing 76, stimulus 73, stimuli 71, automata 70, positive 67, evolution 65, de 64, subjectivity 64, equilibrium 62, courtesy 59, singularity 59, evolutionary 59, apparent 58, theoretical 58, cognitive 57, eventually 57, pre 56, anti 55, singularities 55, und 55.