Orientation of Roman camps and forts in Britannia: reconsidering Alan Richardson's work

At the time of establishing a new settlement, Romans performed a number of rituals in order to

choose the right place and orientation. These rituals were also present in military settlements where,

despite the practical mentality, the role of religion was relevant (Andrés Hurtado, 2002). Although

some ancient scholars such as Hyginus Gromaticus mentioned the necessity to orientate the towns

according to the path of the sun, others as Vegetius indicated that the chief determinants of camp

orientation were the direction of the march or the enemy position, in which case any preferred

orientation would appear (null hypothesis). Alan Richardson (2005) presented a number of orientations

measured by a protractor on small-scale published plans of Roman military camps and forts in Britain

and elsewhere and proposed that, since some angles are more common than others, the null hypothesis

should be rejected. However, the use of statistics done by Richardson is not accepted by some authors

such as Peterson (Peterson 2007, 103-107), who thinks that the Xi-squared test is not correctly used and

leads to wrong conclusions, and Salt (http://alunsalt.com/2007/02/17/vidi-7/), who proposes the use of

a binomial distribution test and considers it a simpler method which works well with small samples.

In this talk, we will discuss the results of our analysis of Richardson's data. We have analyzed 93

Roman camps in Britain. We have complemented the data acquisition by estimating the angular heights

in the horizon at each site by using a reconstruction of the horizon given by a Digital Terrain Model

(http://www.heywhatsthat.com/), thus also considering the local topography. We have applied a

different statistical method where we compare between our results and a set of homogeneous

distributions. Besides, a preliminary result of a diachronic study of the camps and forts will be

presented. We tend to confirm Richardson?s thesis as a random distribution of the orientations could be

discarded. In fact, there are some concentrations around certain angles, which correspond to sunrise

and sunset positions in particular dates during the Roman warfare season, even if we make a distinction

between camps at the north and south of the Hadrian?s Wall, the main frontier of the Roman Empire in

Britannia. These results apparently show that a solar orientation custom was presumably followed at

the time of laying out a camp, and that some dates would be preferred, as has been found in previous

works on the Roman world (see e.g. González-García, Rodríguez-Antón, Belmonte 2014; Magli 2008).