Citter, Carlo 2014. Evaluating the environmental sustainability of settlements through cumulative cost surfaces. Towns and castles in medieval Tuscany.

Evaluating the environmental sustainability of settlements through cumulative cost surfaces. Towns

and castles in medieval Tuscany.

Carlo Citter - University of Siena, Dept. of History and Cultural Heritage – http://unisi.academia.edu/CarloCitter

T04 Environment and Subsistence: Geosphere, Ecosphere and Human Interactions

S023 Concepts and methods for exploring sustainability and resilience in the archaeological record

Theoretical framework: studying the sustainability

Usually, archaeologists of the classical antiquity and the middle ages are concentrated on site they are digging. Actually, each settlement is founded, lives, and dies within a given environment. Thus, it is appreciable that some researchers focus on the container (the

environment), more than on the content (the site).

Among the topics related to the archaeological approach to the environment, the sustainability is increasing its importance. One can easily guess that contemporary concerns are driving our questions to the past. However, this approach opens new

scenarios to study past communities.

Key questions: why, what, how?

What?What do we mean for studying the sustainability of past communities? Usually, we mean to study the environmental sustainability of the resources' exploitation. It is intriguing to apply the concept to the economic and social spheres too. We also stress the importance to study the sustainability of long term routes.

Why?Why should we study the sustainability looking behind us if we are going to face severe challenges in the next decades?Because the past is not all but the lost paradise. We need to study how past communities managed the exploitation of the natural resources to understand, whether they were sustainable or not.

How?How can we study the sustainability?There is a great variety of approaches, which we can appreciate during this session. Here on we shall sketch a further approach based on GIS and, in particular, on the making of cumulative cost surfaces.

Cumulative cost surfaces to postdicting the sustainability

What is a Cumulative Cost Surface?It is a raster map that archaeologists

usually use to evaluate the least cost path between two settlements. Here the case of Roman roads in

SW Tuscany.

We propose a wider meaning of the word cost. We wish to assign each raster's cell a value for the cost of doing

something (move, exploit, and so on)Let's think at the crop production.

The evaluated CCS can suggest that

the territory of the given site cannot produce enough food for the

estimated living community

the territory of the given site can produce enough food for the estimated living community

they fulfilled their needs

they missed to fulfil their needs

WHY?- were they a powerful community?- they failed because of social constrains?- could they easily get it from trade?- did they fail and vanish?

POSTDICTING THE HUMAN CHOICES

Cumulative cost surfaces to postdicting the sustainability

several raster cost weighted surfaces to evaluate environmental and human factors independently

attractors detractors facilitators obstacles

act at a distance act “under our feet”

Antonia Arnoldus-Huyzendveld and I introduced the concept of “attractors” to model a ccs for the least cost path. We tested it for settlement location and for resource exploitation. It seems to give

the archaeologist a more flexible tool.

Case study: high to late medieval Tuscany (11th – 14th c. AD) – the geographical context

From a geographical point of view, Tuscany is roughly composed of three parts

1. The Appennines mountains in the northern border

2. The Arno valley right south of them. It is the third Italian river

3. A series of gentle hills and small valleys, being the Ombrone the major one

Case study: high to late medieval Tuscany (11th – 14th c. AD) – the historical framework

At the end of the 10th century AD the towns of the northern part of the region started growing faster. They were already the leading towns in the Roman period. In the 13th century Pisa was one of the major Mediterranean hubs and Florence was the most populated town in Europe (around 100.000 people).

On the contrary, the southern portion, despite its potential, never developed large towns. The few people was spread into small hilltop fortified villages (castles) and four towns. Siena, though belonging to the hilly side, is considered the most southern of the major northern towns.

Case study: high to late medieval Tuscany (11th – 14th c. AD) – the dataset

Literary sourcesFood production and consumption, trades, resources exploitation, famines, epidemics

Archaeological data

Palaeoenvironmental studies

Historical maps

Digital cartography

Maps of land units and land evaluation are rare. So are LIDAR and RADAR surveys extent. However, we have digital maps of

corine land cover and lithology

Cabrei and, later, cadastres with detailed land use before the industrial and the

green revolution

We almost miss a good dataset of palaeoenvironmental studies for the middle

ages. However, things are improving

Anthropological dataSeveral cemeteries have been

investigated. Good dataset about diet, diseases, and working activities

Food production and consumption, trades, resources exploitation

How could the rather small territory of a growing town like Florence sustain its fast development?

Why did great towns never develop in the southern portion of the region?

We counted 1,554 medieval castles among 2,666 hilltop settlements in Tuscany (top right). However, the total amount of aerial photographs indicating potential hilltop settlements counts 4,234 dots (left). How many of them could be unsuccessful castles? Which were the reasons of their unsustainability? Political struggles? Environment? Other?

Case study: high to late medieval Tuscany (11th – 14th c. AD) – the questions

These two maps are the results of the ASFAT project, a research made 15 years ago by the laboratory of Medieval Archaeology of the University of Siena.

http://archeologiamedievale.unisi.it/miranduolo/territorio/cartografia/cartografia-archeologica/contesto-toscano

Florence: evaluating the needs of a successful town without cumulative cost surface

10th to 12th c. AD Florence's territorial expansion within a radius of 10 miles from the city (left) has been traditionally interpreted as a growing town, that needed a larger political

space to rule upon.

The most profitable land for crop production (light grey) within a 10 miles radius from the city (right) could feed a maximum of 11,000 people. Frederick I favoured the neighbouring town of Fiesole by granting to Florence the uplands, but not the fertile plain (dashed area).

Case study: high to late medieval Tuscany (11th – 14th c. AD) – the evaluations

Sassoforte: where did they live? Questioning models

The survey (top left), the excavation in the castle of Sassoforte, which was the central place, and the land evaluation map (top right) suggest this castle in southern Tuscany had not much land in the surroundings to fulfil the needs of the population living within its territory. On the contrary, the nearby village of Sassofortino (bottom left) seems in the most suitable position.

Sassofortino

Sassoforte


Land evaluation map of the territory of

Sassoforte made by Antonia Arnoldus-

Huyzendveld


We make a ccs to evaluate which is the best position for a village according to the need of:

1) good arable land in the surroundings2) road network3) gentle slope

Sassofortino is mentioned since the 12th century, but we found pottery in the survey dating from the 6th.

We made a ccs evaluating the land suitability for 50%, the proximity to roads 25%, and the gentle slope 25%. The result is a qualitative estimation we grouped into 4 categories (1 very good, 4 very bad).

A 2 km buffer centred on the two sites shows clearly that Sassofortino is in a better position, than Sassoforte.

Though caution is always welcome, can we suggest that the lords planned a walled village (Sassoforte), which turned unsuccessful because the peasants wanted to settle near the arable fields and the roads (Sassofortino)?


Grosseto

geological terrace

drillings

elevationsAntonia Arnoldus Huyzendveld and I worked for years searching for the medieval saltworks of Grosseto. We simulated a DTM for the high middle ages comparing several data (top right). We had no LIDAR survey, but only elevations each 100m.

lagoonal coastline

pote

ntial

stre

am

pour point

basin

We simulated the potential streams to get the potential basins (sort of flat bowls) for salt winning along a coastline of a lagoon opened to the sea (bottom right).Then, we evaluated several ccs according to different weights of the parameters. The best result we obtained (bottom left) is a potential area for salt winning around the 12th c. AD which overlays literary mentions and aerial photograph's crop marks.

Criticizing the method, the dataset, and the evaluations

Estimation vs quantitative evaluation?

Criticizing the method

Not at all. Sometimes we wish to forecast the location of something (prediction), sometimes we wish to know why it is right there (postdiction). The same for resources' exploitation.

Is a ccs always necessary?

No. It is not. The land evaluation map is sufficient to postdict the potential production for crops, vineyards, and pasture.

When do we need a ccs?Anytime we wish to predict or postdict the human behaviour in relation to movement, resources' exploitation, location.

Prediction vs postdiction?

Maybe this is an old issue. Quantitative approaches need a dataset's accuracy we shall never have in the humanities. But we use maths to estimate.


Criticizing the dataset

Literary sources Reliable an abundant after AD 1200

Archaeological data

Palaeoenvironmental studies

Historical maps

Digital cartographyLIDAR and RADAR surveys extent is not

sufficient

No georeferenced data before AD 1823

Not much data. We need more efforts

Anthropological data Good dataset, but geographically and chronologically not well distributed

Often, we have bad datasets, old excavations or, simply, no interest in the

environment


Criticizing the evaluations

How much accurate is an estimation?

Who decides the weights and how?

An estimation is as accurate as the input's dataset. A LIDAR survey is more accurate than a 10x10m DTM. A land units map is better than a land use map.

The archaeologist has to choose the parameters to include in the evaluation and how much should each weigh. Thus, to declare the weights, to verify with a survey, and to compare with all available data are crucial.

What does an estimation produce?

It produces a raster map whose cells show the result of our estimation. We need to make several evaluations and to test them.

Conclusions: and then?

This method does not produce a quantitative estimation. It is a set of tools to help building a model to explain the fate of past communities. Where model means tool, not dogma.

The results must be tested on the “battle-field”. A raster map can suggest new directions. It cannot explain the complexity.

In addition, we think it allows to raise new questions about long lasting debates. For instance:

Did the manorial system introduced by the Carolingians in the 8th c. AD fail by the early 11th because new local end more efficient powers came to the fore? Or, rather, because it was environmentally, but not economically sustainable in the long term?

Citter, Carlo 2014. Evaluating the environmental sustainability of settlements through cumulative cost surfaces. Towns and castles in medieval Tuscany.

Documents