Announcing the Ronald Jensen Nachlass Project

In memory of Ronald Jensen, and in service of the notes he left behind. August 31st, 2026

Ronald Jensen passed away almost a year ago. He left behind an enormous amount of mathematics. Beyond what he published, a lot of his results only ever existed in his own handwriting and were never typeset.

These notes are generally very readable (barring the odd scanner artifact here and there) and mostly freely available on his website.

The Ronald Jensen Nachlass Project aims to transcribe all of them into beautiful LaTeX. This is not to replace the originals, but rather to complement them and to provide them in an alternative format which is searchable and easier to print.

The first manuscript is already done: "On the Subcompleteness of some Variants of Namba Forcing".

How to contribute

With modern tools, transcribing a single note is a manageable amount of work. All of them together is a different story, so the project needs your help to succeed. Please go ahead and shoot me an email if you want to help! Let me give you the details on how this might look, and then I'll explain how the first set of notes was transcribed.

Before you do anything, please get in touch with me, so we can avoid transcribing the same notes twice.

Let me know which notes you wish to transcribe. You can either do everything on your own or request a preliminary .tex file produced by Claude from me. You can then upload the corrected version to a shared Overleaf project.

You can also contribute by making your students do it. The next time you want a student of yours to read one of Jensen's notes carefully, let them produce the transcription. This should actually be a pretty good exercise.

Let me stress that you do not need to be an Inner Model Theorist to help! Jensen was a very well-versed set theorist who proved many important results in e.g. pure forcing theory. Only half of the notes focus on advanced inner model theory, with the rest covering different topics.

My workflow

As much as I like to criticize our AI overlords, I am happy to admit that Anthropic's Claude was extremely helpful and that this kind of project would not be feasible without such tools.

I first produced a preliminary LaTeX transcription with the help of Claude. Specifically, I used the Claude plugin for Visual Studio Code, which allows Claude to do everything on its own: downloading the handwritten notes, organizing them into folders, creating .tex files and compiling them into PDFs.

Even without having told Claude to do so, it is clear that Claude does not just try to parse the scans word by word, but that it understands the mathematics and uses the context to guess if necessary. The upshot is that if Claude gets something wrong, it is very consistent in getting this wrong and can be told to change this specific thing throughout the whole document.

I would estimate that the very first version was already 95% accurate to the handwritten notes.

After notifying it about some of the consistent mistakes, it was probably 98% accurate. As is the nature of such things, most of the actual work is in the remaining 2 percent.

The PDF needs to be fully proofread against the scanned pages and the remaining issues have to be corrected (either by hand or by telling Claude to do so).

The result is a .tex file which is highly accurate to the originals. The only intentional changes are some fixed typos. I also added some more details to the references. Other than this, Jensen's notation, numbering and formatting are faithfully reflected in the transcription.

Final Thoughts

My hope is that enough volunteers step forward so that the project can be completed in, say, the next two years.

I have heard about some rumors that there are even more scans of notes of Jensen, which could also become part of this project at some point.

For now, I will host the transcriptions on my website. Once a critical mass of transcriptions are done, I will try to find a better home for them.