mixedmath

There is no reason why amateurs can't make worthwhile contributions to research mathematics. But there is also no reason why every mathematician must listen to any particular amateur.^{1 1}This is much more true now that it's easier to write mathlike documents with help from an LLM. I don't have time to offer personal advice or guidance to everyone who contacts me.^{2 2}If I have sent you to this page, it likely means that I have also given you a boilerplate response. You've probably asked me for arxiv endorsement and I said no. For some reason I don't think mathematicians will value your work. But much of the fun of doing math is the doing of math and not the sharing of math. I applaud those who try to take a more scientific, quantified, statistical, or mathematical view to their interests. But often I want to say the same basic thing, so I say it here.

Other Advice

Firstly, there is already useful advice. I really like Henry Cohn's advice, though this is aimed a bit more at people who think they've solved famous problems.^{3 3}This is also more useful in the LLM era. It is flatly easier to write plausible-but-incorrect mathematics than it is to refute it. Mathematicians are not prepared for this. Go read it. It's great!

The arXiv is not a blog

The arXiv is a preprint server. It provides a valuable service to the community and science advances faster and more openly because of it. As a preprint server, it hosts preprints, typically articles in pre-publication format, often soon-to-be submitted to a journal.

The arXiv is not a blog. Sometimes established researchers write informal or opinion pieces with the intention of communicating specific ideas to others and put these on the arxiv. This is not common.

A note or article that contains only a short list of observations (or worse, code snippets and graphs with implicit observations) could make a reasonable blog post. I like it when people write ongoing notes. Working with the garage door open is a great way to implicitly invite attention! (But it does not guarantee attention, nor does it inherently deserve attention).

If you ask me to endorse something that I think would make a better blog post, I'll decline and say so.

Writing a math blog is straightforward! You can share short math notes with only a url. Github pages are free and allow MathJax for beautiful math presentation. Any platform that allows hosting static HTML is enough, for example see this simple math website. Wordpress has been around forever and is adequate and easy.

Articles are for the Community

Preprints on the arXiv are each deliberate communications to the community. There are established conventions for writing in the community that make it easier for researchers to understand and appreciate the work. These conventions are not taught and are somewhat fluid, but a good way to start to appreciate what they are is to read some mathematical writing on the arXiv.

One responsibility of a good article is to help situate it within the literature. How does it relate to known work? What does it build on? What's new, what's old, and how does it fit together?

If you are trying to submit an article that doesn't contextualize the work, then I will assume that you are trying to make the Reader fill in this context for you; and I presume that the Reader probably has an infinite list of things to do and doesn't want to waste time.

Similarly, if you are asking for endorsement on an article that doesn't cite anything within the last 100 years, then I'll assume that you haven't read the last 100 years of literature and thus have probably missed all the relevant context. It is interesting to go through the exercise of reproving some mathematical work with possibly different methods, but this sort of exercise is often a better blog post (or homework problem in an upper level math class) than an article.

A good article tells a story. The article is a method of communication, and making the story as digestable as possible is the clearest form of communication.

Before trying to submit to the arXiv, get feedback from friends or collaborators. If you're trying to get your first feedback by submitting to the arXiv, then you're probably not ready to submit to the arXiv. If you can't get anyone to read your paper now, what makes you think that anyone will want to read it after it appears on the arXiv?

More Basic Advice

I end with a short list of comments.

1. If you want someone to endorse your article, then you have to send the article. If you do not send the article, then the only valid response is no.
2. Spamming many mathematicians with endorsement links does not increase endorsement chances. Requested endorsers can flag requests as spam or spam-like. The mechanics of how this works aren't public (I think), but one should guess that this tanks the submission.
3. If your work makes use of LLM output, you should say so. LLMs are powerful and can help in research, but LLMs are currently completely incapable of writing mathematics well. If I get the sense that the article is half-baked LLM output, then I will flag the endorsement request as LLM spam.

4. Consider asking your mathematical friends, teachers, or professors for advice. If they are capable of providing endorsement, then they are probably the correct avenue. If they are capable of connecting you with someone who can provide endorsement and they are established mathematicians, then that is also a reasonable avenue. I have wide research interests, but I typically only consider endorsements in number theory or arithmetic geometry.^{4 4}I have never been asked for CS endorsement and have nothing to say there. I don't make time to endorse in algebra, algebraic geometry, analysis, or pure computational math.

5. If your best bet for endorsement is to cold email someone you don't know, then you should try to make the whole process as easy as possible for them. This includes respecting their time. I rarely make time for cold emails on endorsement; when I do, there will be at most one email exchange.

6. Offering constructive criticism is very demanding. The endorsement process is probably not the time for this. If you really want constructive criticism, one attainable route is to pay a graduate student^{5 5}In practice, this could feel like tutoring. To a certain extent it is. When I was a graduate student I would give high-level tutoring for 80 dollars an hour. I suspect one could get reasonable feedback for a few hundred dollars by describing this offer to the department manager or secretary of any PhD granting math institution. A less expensive first step now is probably to ask an LLM for flaws in your article. for it. If money is no obstruction, you could instead hire a mathematician as a consultant. I don't think this is a good return on investment.

Leave a comment

Comment via email