LingA stands for “Linguistic Annotation”. It is based on the Annotation Graph model[1]. Labels are represented as arcs connecting nodes in an acyclic directed graph.
LingAs represent the annotations in ALingA. Because this format is chiefly about the alignment of annotations, the nodes in a LingA are meant to be eventually bound, or anchored, to a specific point, or granule, in some signal manifold, and hence to the underlying manifold.
What is notable about LingAs is their unrestricted connectivity. See the diagram below for an example.
havs (hear and view speech) provides a GUI to manipulate such structures.
[1] Steven Bird & Mark Liberman (2001). A formal framework for linguistic annotation. Speech Communication 33, 23-60. http://www.arXiv.org/abs/cs/0010033.