More tools · The understanding engine · working name
Noogeny
Each -ogeny site describes its field by the choices things are made of. Noogeny reads all of them as one kind of knowledge: thirteen facets of understanding, each with a test. It shows every site where its understanding is thin.
Van Riel Online Services started it on 27 October 2026. It runs privately for now, and reads the sites without changing them.
| Facet | Stylogeny | Morphogeny | Lumogeny | Quotogeny | Glossogeny | |
|---|---|---|---|---|---|---|
| What it is | Form | built | partly | built | built | built |
| Structure | built | not yet | built | not yet | not yet | |
| Names | partly | not yet | partly | partly | built | |
| Tinbergen’s four | Mechanism | not yet | built | partly | not yet | partly |
| Making | partly | built | built | not yet | built | |
| Function | built | not yet | partly | built | built | |
| Descent | built | not yet | built | built | not yet | |
| Where it lives | Context | built | partly | partly | built | partly |
| Reception | not yet | not yet | not yet | built | partly | |
| Value | not yet | partly | not yet | not yet | built | |
| How sure | Evidence | built | partly | built | built | partly |
| Possibility | built | partly | built | not yet | not yet | |
| Self-knowledge | built | not yet | partly | built | partly |
Thirteen questions
Calling a poster Swiss answers one question about it. Noogeny asks thirteen, and gives each a test that can fail, because a facet without a test is only a belief.
What it is
- Form
- What does it consist of?Test: it classifies examples it has not seen, and two judges agree.
- Structure
- Which choices go together, and which exclude each other?Test: it fills in hidden choices from the visible ones.
- Names
- What is it called, and by whom?Test: it explains the thing correctly to an expert and to a beginner.
Tinbergen’s four
The biologist Niko Tinbergen set these four out in 1963, in “On aims and methods of ethology”. A behaviour, he wrote, is only understood once you know how it works, how it develops, what it is for and how it evolved. Two of his four are -ogenies already: ontogeny and phylogeny.
- Mechanism
- Which choice causes which effect?Test: it predicts what one change will do, and is right.
- Making
- How does one come into being, step by step?Test: it makes a new one by the process, not by copying a result.
- Function
- What is it for, and what job does it do?Test: it picks the right form for a stated goal.
- Descent
- Where does it come from, and what came from it?Test: it places a new thing in the lineage, and dates and locates it.
Where it lives
- Context
- Who does it belong to, and what does it signal?Test: it knows when the thing must not be used, and why.
- Reception
- How is it read, and how is it misread?Test: it predicts how a given audience will read it.
- Value
- What makes a good one or a weak one?Test: it ranks examples the way experts do.
How sure
- Evidence
- How do we know?Test: it points to the source of any claim on request.
- Possibility
- What doesn’t exist yet but could?Test: it proposes new combinations that hold up when they are made.
- Self-knowledge
- What does the system not know?Test: it says “unknown” rather than guessing, and queues the research.
How it reads the sites
Noogeny reads each site’s data where it lies and never changes it. Every statement becomes a claim: the thing, the facet, the value, its evidence, how sure it is and who made it.
A thing’s understanding is then a profile: which facets are filled, how sure each one is and which tests it passes. Where two facets disagree, that is a finding. Two styles that look alike but share no descent are one look invented twice. The first form test found three candidate pairs, such as the Bollywood poster and the Egyptian film poster.
What the first tests found
27 October 2026. The numbers change as the atlases grow.
- 74%of 608 real works in Stylogeny are matched to their own style first, out of 415. In the top five: 92%. By chance: 0.2%.
- 89%of 237 real works in Lumogeny’s atlas are matched to their own style first, out of 62. In the top five: 98%.
- 92.5%of the time, Jev, Quotogeny’s AI judge, agrees with its researchers on a quote’s traits: it rates the ones they gave above the ones they didn’t. Over 438 quotes.
The first two tests are not blind: the works were chosen and reviewed with the descriptions in view. They show that the descriptions are consistent and distinct, not that they generalise.
The weak spots are the point. For ‘epigraph’, Jev and the researchers agree no better than chance (48%): they mean different things by it. And only 138 of Stylogeny’s 415 styles have a real work as evidence. Most of the rest are still in copyright, so Stylogeny shows them with a labelled AI illustration.
What comes next
- 01
The intervention bench. Morphogeny names a design choice in every control, but nothing checks yet that a control does what its name says. The bench changes one control at a time, keeps the rest, and checks each picture with Stylogeny’s vision check.
- 02
A blind test outside design. Give the engine a field’s descriptions without the experts’ vocabulary, and measure how much of that vocabulary it finds by itself. If it finds most of it, a new field needs settings, not a new site.
- 03
Reception and value for the atlases. How a style reads to different audiences, and what makes a strong or a weak example of it.
The name
From the Greek nous, mind. Noogenesis is the growth of mind and knowledge. Noogeny is a working name.
Questions and corrections: tell us, or write to info@vanriel.eu.