Mathematical phantoms
The field with one element or characteristic one ($\mathbb{F_1}$ or $\mathbb{F_{un}}$) is a mathematical phantom, which can defined as a beast who clearly (i.e. within the current mathematical framework) does not exist, but there are many pointers in a direction that it should.
Are there other examples of mathematical phantoms ?
$\endgroup$ 31 Answer
$\begingroup$I agree with Dietrich (in the comments) that the meaning of "mathematical phantom" is not entirely clear. In particular, the demarcation from merely very fruitful abstractions is blurry; perhaps useful criteria are:
- There should be a statement which was held to be obviously true before the discovery of the phantom, but which is false in view of the new concept.
- It should have required a nontrivial effort to make it precise (to "help it come into being").
- It should have great explanatory power and vast consequences (Gavin Wraith: "which [...] obtrudes its effects so convincingly that one is forced to concede a broader notion of existence", sorry for linking to http).
Phantoms in this stricter sense could include:
- The irrational numbers. (Running counter to the basic tenet "all is number" by the Pythagorean school, where by "number" they meant "rational number".)
- The complex numbers. (The idea of a number squaring to $-1$ was surely held to be obviously nonsensically.)
- The $p$-adic numbers.
- Actual infinity, together with the flexible notion of sets we have nowadays (vastly surpassing recursive subsets of $\mathbb{N}$) and the axiom of choice.
- Sobolev function spaces.
- Infinitesimal numbers.
Phantoms in a broader sense (where I can't think of any held-to-be-obviously-true statement falsified by them) could include:
- Symmetries of zeros of polynomials, or more generally groups.
- The field with one element.
- Ideals in number theory.
- Motives.
- $\infty$-categories.
- Toposes. (Generalizing and unifying various cohomology theories.)
- Nonclassical logics. (Born in the foundational crisis, nowadays with lots of applications in mainstream mathematics.)