Internet Archaeol. 16. Mlekuz. Modelling past sounds in GIS

4.2 Modelling past sounds in GIS

The first step in modelling past soundscapes is to model sound propagation across the landscape.

The propagation of sound over any given landscape is an extremely complex process, influenced by almost every aspect of the environment (Smith et al. 1996, 62-67; Truax 1999 [sound propagation]). Nearly all atmospheric factors, such as air absorption, wind and temperature gradients, air humidity etc. contribute to this process. Equally important are surface effects, such as absorption, reflection and diffraction due to the surface properties (topography, barriers, texture) of the landscape.

The modelling of sound propagation over a landscape is computationally extremely difficult, as it depends on a range of variables, which are, at best, ill defined.

If these variables are omitted, predicted sound levels can turn out very different to real ones. These effects are particularly important where sound is propagating over distances greater than a few hundred metres. It is very difficult to imagine the existence of a comprehensive paleoenvironmental record describing exact past states of the atmosphere on a very fine-grained time-scale.

A reconstruction of past sonic environments, a sequence of sound profiles in a space at a specific moment, is therefore impossible.

However, this does not mean that it is impossible to study past soundscapes. If our interest lies in the way soundscapes changed through the constant construction and re-interpretation of soundscape features, especially by soundmarks then only stable or long-term properties of a landscape (e.g. topography) are important. Unstable and transitory properties (e.g. state of the atmosphere) can safely be ignored as they do not contribute to the long-term structure of sound profiles. Study of past soundscapes becomes study of past soundmarks and their role in social life.