A regularization of quantum field Hamiltonians with the aid of \(p\)-adic numbers (Q1383777)

It is well known that in quantum field theories, computations of observable effects often lead to physically meaningless infinite values. For example, we can write down Maxwell's equations of classical electrodynamics for a system of electrons, and get meaningful results if we substitute the measured values of the electron's mass \(m_e\) and charge \(q_e\); however, a natural quantum analogue of these equations (quantum electrodynamics QED) leads to infinities. Physicists know several regularization techniques which enable us to avoid (at least some of) these infinities: for example, in QED, not only observable quantities are infinite, but also observable values of the electron's mass and charge turn out to be infinite; if we reformulate the QED equations in terms of the observable mass and charge (as opposed to the original values \(m_e\) and \(q_e\)), we get finite (and experimentally correct) predictions. There are numerous regulations techniques; some of these techniques are mathematically justified, but many are still heuristic methods, in which we subtract, divide, and do other algebraic operations with infinite values which do not have a mathematical justification. It is, therefore, desirable to find a general mathematical formalism which can consistently formalize the physicists' heuristic operations with infinite values. One possibility of such a formalization is given by non-standard analysis (NSA), in which all the theory is extended from the field \(R\) of real numbers to its non-Archimedean extension \(R^*\). The NSA approach provides a theoretical consistency, but since it is, unfortunately, not algorithmic, it does not help to actually compute anything. To overcome the non-constructive nature of NSA, the authors propose a new, more constructive physically motivated approach. Its basis idea is that real numbers are only an approximation to reality. In real life, in each measurement, we get only finitely many distinct results, results which are usually represented by rational numbers; for example, if we measure a current with an accuracy of 0.1A, and the instrument's scale goes from \(-\)5A to 5A, we get possible values \(-5.0,-4.9,\ldots,0,\ldots,4.9,5.0\). If we apply more and more accurate measuring instruments, we get a sequence of rational numbers whose limit is the actual value. If we use a standard metric \(\rho(a,b)=| a-b| \) to describe this limit, we get real numbers; the fact that we get infinities means that we may have a sequence of rational measurement results which does not have a limit in this sense, so it makes sense to try metrics \(\rho(a,b)=\| a-b\| \) for other valuations \(\| .\| \) on the field of all rational numbers. Such valuations are all known: in addition to \(| .| \), we have \(p\)-adic norms. So, the authors consider a \(p\)-adic version of quantum field theory, in which real numbers are replaced with \(p\)-adic ones. They show that for several reasonable Hamiltonians, we indeed get meaningful results (instead of physically meaningless infinities). One mathematical reason for this regularization is that Hamiltonians with singular potentials (which are too singular to be consistently described in \(R\)-based Hilbert space) become bounded operators in a \(p\)-adic version \(L^2_p\) of a Hilbert space \(L^2\).