# Hilbert's fifth problem

Template:Mergefrom Hilbert's fifth problem is the fifth mathematical problem from the problem list publicized in 1900 by mathematician David Hilbert, and concerns the characterization of Lie groups. The theory of Lie groups describes continuous symmetry in mathematics; its importance there and in theoretical physics (for example quark theory) grew steadily in the twentieth century. In rough terms, Lie group theory is the common ground of group theory and the theory of topological manifolds. The question Hilbert asked was an acute one of making this precise: is there any difference if a restriction to smooth manifolds is imposed?

The expected answer was in the negative (the classical groups, the most central examples in Lie group theory, are smooth manifolds). This was eventually confirmed in the early 1950s. Since the precise notion of "manifold" was not available to Hilbert, there is room for some debate about the formulation of the problem in contemporary mathematical language.

## Classic formulation

A formulation that was accepted for a long period was that the question was to characterize Lie groups as the topological groups that were also topological manifolds. In terms closer to those that Hilbert would have used, near the identity element Template:Mvar of the group Template:Mvar in question, there is an open set Template:Mvar in Euclidean space containing Template:Mvar, and on some open subset Template:Mvar of Template:Mvar there is a continuous mapping

F : V × VU

that satisfies the group axioms where those are defined. This much is a fragment of a typical locally Euclidean topological group. The problem is then to show that Template:Mvar is a smooth function near Template:Mvar (since topological groups are homogeneous spaces, they look the same everywhere as they do near Template:Mvar).

Another way to put this is that the possible differentiability class of Template:Mvar does not matter: the group axioms collapse the whole C k gamut.

## Solution

The first major result was that of John von Neumann in 1933, for compact groups. The locally compact abelian group case was solved in 1934 by Lev Pontryagin. The final resolution, at least in this interpretation of what Hilbert meant, came with the work of Andrew Gleason, Deane Montgomery and Leo Zippin in the 1950s.

In 1953, Hidehiko Yamabe obtained the final answer to Hilbert’s Fifth Problem:

If a connected locally compact group Template:Mvar is a projective limit of a sequence of Lie groups, and if Template:Mvar "has no small subgroups" (a condition defined below), then Template:Mvar is a Lie group.

However, the question is still debated since in the literature there have been other such claims, largely based on different interpretations of Hilbert's statement of the problem given by various researchers.

More generally, every locally compact, almost connected group is the projective limit of a Lie group. If we consider a general locally compact group Template:Mvar and the connected component of the identity G0, we have a group extension

G0GG/G0.

As a totally disconnected group G/G0 has an open compact subgroup, and the pullback G′ of such an open compact subgroup is an open, almost connected subgroup of Template:Mvar. In this way, we have a smooth structure on Template:Mvar, since it is homeomorphic to (G′ × G′ )/G0, where G′/G0 is a discrete set.

## Alternate formulation

Another view is that Template:Mvar ought to be treated as a transformation group, rather than abstractly. This leads to the formulation of the Hilbert–Smith conjecture, unresolved Template:As of.

## No small subgroups

An important condition in the theory is no small subgroups. A topological group Template:Mvar, or a partial piece of a group like Template:Mvar above, is said to have no small subgroups if there is a neighbourhood Template:Mvar of Template:Mvar containing no subgroup bigger than {e}. For example the circle group satisfies the condition, while the [[p-adic integers|Template:Mvar-adic integers]] Zp as additive group does not, because Template:Mvar will contain the subgroups: pkZp, for all large integers Template:Mvar. This gives an idea of what the difficulty is like in the problem. In the Hilbert–Smith conjecture case it is a matter of a known reduction to whether Zp can act faithfully on a closed manifold. Gleason, Montgomery and Zippin characterized Lie groups amongst locally compact groups, as those having no small subgroups.

## Infinite dimensions

Researchers have also considered Hilbert's fifth problem without supposing finite dimensionality. The last chapter of Benyamini and Lindenstrauss discuss the thesis of Per Enflo, on Hilbert's fifth problem without compactness.