Book Review: Lost in Math by Hossenfelder

When it comes to the philosophy of science, not many publications are relevant to modern practice. Let’s take math. The current literature still talks about platonism. Look harder and you might find the rise of non-Euclidean geometry or other breakthroughs like cardinality. In short, the bulk of mathematical philosophy still consists of math that’s hundreds of years old. While these topics are still important, I find it much more interesting to look at the new philosophical issues present in modern mathematics and science.

That’s why I was delighted to find Lost in Math by Sabine Hossenfelder, who is also the author of a popular physics blog called Backreaction.

It’s written by a physicist on the philosophy and sociology of foundational physics. The author studies the influence of beauty in the development of science, and reflects on the conflict that beauty has with the scientific method. This is the modern stuff.

This book consists of conversations with fellow physicists on popular theories like string theory and other ‘theory of everythings’. According to the author, many aspects of current theories are untestable, and yet are developed because physicists are so attracted to the mathematical beauty of the theories. For this beauty, they are willing to forgo the key elements of the scientific method.

Basically, Hossenfelder is concerned with biases, and the ‘beauty bias’ in particular. The problem, according to the author, is that beauty is causing physicists to deviate from the tried-and-true scientific method as a guide for research development. That in turn erodes the cohesion of the discipline.

Being a mathematician, I was perhaps less troubled by the ‘beauty bias’ than I should have been. Physics and mathematics now are quite different fields of study, but at one time the two communities were much more intertwined. So, I do not find it disturbing that physicists are taking the time to pursue ideas that may in fact be too devoid of science to be called science.

But, there are two other problems that are touched upon briefly in this book that I found more interesting and I thought closer to the true problem. The first is that some of these theories that may not be ‘true science’ are being passed off as more applicable to science than they are. This is a general phenomenon in all of math and science: the pressure to find links and applications to other areas when there may be none. This can be seen in biology with grants mentioning (far-fetched) applications to cancer or in mathematics with claims that one area may suddenly unveil new understanding of another. This happens at the paper-writing and research level, as well as at the university administration level, where it is becoming more common to quantify the research of the faculty with useless metrics. In short, part of the problem is portraying one type of research as another.

The second problem, which is think is far more pertinent, is the general scientific research machine whereby grants are awarded by established researchers and prestige is awarded by publications. Basically, the problem is that this method of scientific progress works well for some areas and poorly for others.

In the context of Hossenfelder’s book, I think how theoretical physics is supported is not so good. The author explicitly mentions this as well: certain areas can be pursued beyond their value to the rest of the scientific community, and that’s simply because the few established researchers can support those fields through grant awarding and accepting papers.


Regardless of which areas are suffering the worst from various biases, the author brings up some very important issues that need more study in the philosophy of science. This is especially important in fields that have reached such a level of maturity that human biases or even human limitations begin to erode their cohesion.

Personally, I don’t believe that this means any field is doomed. Rather, many areas of study deserve scrutiny to find the best strategy for future development. Just as individual students need guidance in order to proceed to the level where they can be independent researchers, so too do entire communities need guidance in order to proceed past certain waypoints in the development of an entire discipline. When it comes to entire communities, we of course don’t have more experienced mentors that can help us. Rather, we need to devote resources and time to introspection so that we can guide ourselves. This would be much better than just proceeding blindly with the old systems of the past.

We should take Lost in Math very seriously as one of the few published books daring to examine the modern machine of scientific progress.


Leave a comment

Fields marked with * are required. LaTeX snippets may be entered by surrounding them with single dollar signs. Use double dollar signs for display equations.