1 Introduction
In the recent years, there has been an increasing interest in the so-called Swampland pro-
gram [1] (reviewed in [2–6]), which aims at drawing the boundaries between the low energy
effective theories than can be consistently coupled to Quantum Gravity and those that cannot.
These boundaries are typically formulated as a series of conjectures that constrain the low
energy effective theories, whose arguments range from general black hole physics to explicit
String Theory constructions. One of the best established conjectures at the moment is the
Swampland Distance Conjecture [7], which predicts the presence of an infinite tower of states
becoming exponentially light with the distance as one moves far away in the moduli space of
the theory. The nature of the possible massless towers has been analysed in [8], where it was
conjectured that it must always come from either a KK tower (or dual versions), or from the
excitations of an emergent critical string becoming tensionless (this is known as the Emergent
String Conjecture).
On the other hand, massless vectors at a given point in moduli space are a signature of a
larger gauge group, and all massless states arrange into representations of this group. Thus,
exploring specific points (or more generally, hypersufaces) in moduli space allows to examine
the symmetry enhancements associated to this phenomenon. In particular, a detailed know-
ledge of all the possible enhancements is essential to determine whether all consistent low
energy vacua in Quantum Gravity can be obtained from String Theory (i.e. string universal-
ity). Progress along this direction has been recently made by systematically studying toroidal
compactifications of the heterotic/CHL string [9–12] and F-theory on K3 surfaces [13], as
well as by applying general Swampland arguments to supergravity vacua [14–17] . In this
context, a natural question arises, namely what is the interplay between these symmetry en-
hancements and infinite distance boundaries of the moduli space, specially in relation to the
Distance Conjecture. Such points have been recently studied from this perspective in the
circle compactification of the heterotic string [18] and in 8d theories realised from F-theory
on K3 surfaces [19–21], where it was shown that in these limits one gets an infinite tower of
massless states that realize an affine algebra. Particularly, in the decompactification limit of
the heterotic string on the circle the KK tower combines with the E8×E8or Spin(32)/Z2
gauge group to give rise to the affine version of their algebras, respectively, with the U(1) of
the circle direction playing the role of imaginary root.
In this paper, we study the boundaries of the moduli space of toroidal compactifications of
the heterotic string, which correspond to (duals of) different decompactification limits. Despite
being a very simple playground, the toroidally-compactified heterotic theory proves to be rich
enough to describe all the possible decompactifications and the related affine algebras that are
realised in the corresponding moduli space. We show that when decompactifying kdimensions
from 10 −dto a point in the 10 −d+kdimensional theory with gauge algebra AL×U(1)d−k
R,
we obtain the k-loop version of AL×U(1)d−k
R. The central extension for the left/right moving
algebra is the holomorphic/antiholomorphic part of the current associated to the kdirections
that are decompactified.
1