A key aspect of the project’s
work will be the elucidation of the logistical infrastructures
which evolved in the different parts of the medieval world from
the period of the later Roman empire until the twelfth century.
Communications in general but especially road systems are
fundamental; likewise, the supportive capacity of the lands in
which armies were based or through which they passed, will be a
basic feature of the research. The issue of the relative
availability of resources, including the relationship between
size of population, levels of agrarian and other production, and
the size of armies and their supporting structures within and
between medieval societies will also be key aspects. These
are subjects which remain almost entirely neglected within the
three fields referred to; while virtually no work has been
carried out to place the social logistics of the early
Byzantine, Frankish and Islamic worlds in the broader
comparative context which they clearly need if the direction and
tendency of their different development is to be properly
understood. Although several of these aspects have
occasionally been the subject of scholarly attention, they have
also been the subject of unproven assumptions which have rarely,
and in most cases never, been tested against the evidence.
Armies are, by definition, amongst
the simplest of human social groups. As agents, armies are
deterministic and demonstrate highly directed behaviour. Any
model of the impact of an army essentially revolves around
movement and provisioning. Armies move within parameters
dictated by their military goals, usually by moving from one
offensive action to the next, in a condition of relative
safety. Except under duress their progress is a balanced
procedure offsetting required speed and acknowledging the
necessity of maintaining the body of troops in battle-ready
physical condition " that is fed as adequately as possible.
Armies therefore are goal-seeking and endeavour to be
self-sustaining in an efficient manner. In this, armies
sidestep many problems associated with the mathematical
modelling of complex human societies and approximate more
closely the behaviour of predators or, perhaps, meta-entities
such as ant colonies.
Ecology provides a strong set
of models for such behaviour. The most fruitful of these is
probably “Optimal Foraging Theory”. This procedure finds its
origin in mathematical biology. Optimal Foraging Theory
essentially asserts that natural selection favours animals whose
behavioural strategies maximise their net energy intake per unit
time spent foraging. The prediction is that an animal strikes a
balance between two contrasting strategies: spending a long time
(i.e. using more energy) searching for highly 'profitable' food
items, or devoting minimal time (i.e. using less energy) to more
common but less profitable food items. Various factors can
cause animals to deviate from optimal foraging. For example,
the risk of predation may force the animal to select less
profitable food items in a relatively safe location, rather than
opting for the energetically most efficient feeding strategy.
Optimality models therefore
share certain basic elements:
An actor
who chooses or exhibits alternative strategies or states
A strategy set defining the range of options available to
the actor generally taken to be a maximising strategy
A currency in which the costs and benefits of
alternatives are measured in reproductive success, mating
frequency, survival frequency, harvest rate, or in the case of
armies simple maintenance.
A set of constraints that determine the feasible
strategies and payoffs associated with each.
Two
categories of elements in optimisation models should also be
noted:
Those subject to choice by actor
Those not subject to choice by actor - constraints.
Extrinsic constraints: features of social and natural
environment
Intrinsic constraints: abilities/phenotype (behavioural,
cognitive) requirements: (physiological, nutritional)
Although armies fit well into
the general parameters of classic optimal foraging theory, such
models are rarely applied in such a simplistic manner and
modified foraging theory accepts the reality of important
cognitive constraints including lack of perfect knowledge on
which to base decisions. Early armies, in particular, were
probably characterised by endemic ignorance of local conditions
when groups were involved in long-distance actions. As
extensive lines of supply and inadequate logistics demanded
foraging activity, this suggests that ancient campaigns should
approximate modified optimal foraging behaviour.
It should be noted that
biologically deterministic foraging theory has a considerable
history in archaeology and anthropology. However, such models
have generally been restricted to relatively simple social
groupings such as hunter-gatherers or farming societies whose
subsistence strategies may be theorised by site catchment
theory. However, there has been considerable criticism of even
these applications, as even the simplest human societies may
possess complex stochastic behavioural patterns that are not
well described by such simple models. More recently, works by
archaeologists such as Steve Mithen have used optimality models
for archaeology that incorporate observed behavioural variation
in diet and food acquisition strategies and model behavioural
decision making among hunters and gatherers. These are known as
“prey choice” (or diet breadth) and “patch choice” models and
are designed to predict the food items the forager will attempt
to exploit (handle) and those it will ignore in favour of
continued search for more preferred food, thus incorporating a
decision component within the model.
Least Cost
Routes may be derived from “Cost Surfaces”. Such maps define
the energy cost of traversing a landscape, derived from a number
of variables - the most efficient route of the line which
carries the least cumulative cost
Although optimal foraging
theory appears to offer great potential in understanding the
behaviour of armies, there are serious practical and technical
points that must be considered. Until recently, many models
were highly theoretical with minimal competence to be
implemented. Foraging models demand a very detailed knowledge
of the environment as well as a technical base to permit
analysis and modelling. Archaeological application was always
hampered at any level because of our imperfect knowledge of past
settlement distributions and inadequate environmental data.
Moreover, there were few technologies that could adequately
represent 4 dimensional models (topography or X, Y, Z values
plus time/movement).
Since the initial application
of optimal foraging theory within the archaeological domain,
Mediterranean studies have seen an exponential increase in the
amount of available settlement and quality of environmental
data. Major landscape survey projects and the application of
environmental methodologies have begun to provide a detailed
landscape and settlement context for some regions (including
approximations of settlement systems and landscape
productivity). Within parameters, these areas provide an
appropriate backdrop for models that approximate military
behaviour.
A related point is the
development of computer technologies that permit the
manipulation and visualisation of complex, spatially referenced
geographic and mathematical data in complex situations. GISs
are at the forefront of this development although virtual
reality and a variety of visual technologies are relevant to the
modelling of such behaviour.
*********************
It
is not that no studies have been devoted to these questions; but
they are often deeply flawed. A number of articles have
appeared on various aspects of one or another of the themes
outlined above, in which the conclusions reached are at the
least questionable - this applies particularly, for example, to
the issue of the size of early medieval armies and their
logistical support (where the latter is taken into account at
all), and to the organisational inheritance of the late Roman
world, especially where the early Frankish armies are
concerned. Constant argument over the same sparse references
in a few medieval texts gets us no farther in our efforts to
understand such matters, and results which are produced even by
the most competent and wide-ranging scholars generally remain
almost entirely subjective and unfalsifiable, because not
testable. Trying to determine the parameters within which the
agricultural output of a region at a given time can be
quantified, along with the producing and surplus population it
can support, are essential elements in any attempt to resolve
this and related issues. The project proposed will therefore
seek to place these questions within a framework set both by the
broader issues raised here and by the various categories of
evidence, both medieval and non-medieval, through which some
answers might be offered.
Back to Logistics Home Page
