Spatium Novum

Spatium Novum means "a new space". It is where I write about models: how they get built, what they leave out, and what happens when someone trusts them anyway.

My background is quantitative research. I rebuild models of complicated systems from their papers and data, then test where their assumptions stop earning trust. The subjects range from commodity trading and finance to machine learning, pharmacology, economics and cognitive science. The method—not a single subject—is what holds the work together.

The line on the front page revisits Richard Levins's 1966 remark that our truth is the intersection of independent lies. In The Epistemic Value of Independent Lies: False Analogies and Equivocations, Margherita Harris gives it the necessary qualification: agreement among models does not by itself confirm a claim about the world. Models may share assumptions and errors, and agreement inside "model land" still needs a reason to count as knowledge outside it.

Anchor essays go up roughly every two weeks, with shorter reading notes in between. If something here is useful, or wrong, I want to hear about it.

Denis Joly
notes@spatium-novum.com · LinkedIn · GitHub · RSS · Privacy