
1. Introduction
The quantization of supermembranes in eleven-dimensional spacetime is key to a better un-
derstanding of M-theory, which provides a unification of different superstring theories in a
nonperturbative manner. Supermembranes are described by a three-dimensional sigma model
[1], whose quantization is fundamentally distinct from quantizing strings. One of the major
difficulties is that three-dimensional relativistic sigma models are not renormalizable, which
invalidates any perturbative treatment. Nevertheless, it is possible to probe some features
of the supermembrane in the light-cone quantization, at least when a matrix regularization
is employed on the spatial manifold of the worldvolume [2,3]. This regularization restricts
the infinite-dimensional basis on the continuum manifold to a finite-dimensional Lie group,
and the supermembrane is described as a subtle limit of supersymmetric matrix quantum
mechanics [4], closely related to D0-branes in the infinite momentum frame and the Matrix
theory description of M-theory [5]. 1However, a complete understanding of the continuous
spectrum of Matrix theory is still lacking [12,13], which requires a multiparticle interpreta-
tion and thus a second quantization of the membrane [5]. This light-cone treatment is also
limited to special backgrounds, due to the absence of spacetime covariance. Finally, when the
sigma model is coupled to a dynamical worldvolume described by three-dimensional quantum
supergravity, there still exists neither clear quantization technique nor classical expansion for
the membrane [14]. See, e.g., [15,16] for excellent reviews.
According to the Matrix theory description, a supermembrane is unstable and tends to
be dissolved into dynamical bits [12]. 2This suggests that membranes might not be the funda-
mental objects. However, it is still highly valuable to probe M-theory from the worldvolume
perspectives, especially in regard of the lacking of any covariant formalism of supermem-
branes. Since the three-dimensional sigma model is nonrenormalizable, which suggests that
new physics may arise above a particular cutoff. It is therefore natural to ask whether there
exists a renormalizable QFT at our disposal, which can be regarded as an ultra-violet (UV)
completion of the nonrenormalizable membrane theory. If so, there might exist a stable
notion of membranes that only become strongly coupled at low energies, where the Matrix
theory description becomes valid and the physics is captured by emergent multiparticle states.
Unfortunately, no such renormalizable worldvolume theory seems to be available within the
relativistic framework.
It is pioneered in [17] that the relativistic membrane might admit a UV completion
that exhibits nonrelativistic behaviors on the worldvolume: a renormalizable sigma model
describing membranes propagating in spacetime can be constructed if an anisotropy between
the worldvolume space and time is introduced. Furthermore, this sigma model is required
1Another way of obtaining Matrix theory is by considering the discrete light cone quantization (DLCQ)
of M-theory, which is usually defined as a subtle limit of the compactification over a spatial circle [6–9]. An
alternative treatment of the DLCQ of string/M-theory as nonrelativistic string/M-theory is recently revived
in e.g. [10,11].
2This instability of the membrane was resolved in the Matrix theory description by realizing that the Hilbert
space of the matrix quantum mechanics naturally contains multi-particle states of the D0-bits [5].
– 2 –