Most regions and countries of the world are experiencing unprecedented demographic growth. Therefore, sustainable development of agglomerations and urban communities is one of the declared priorities of the United Nations in recent years and for the new few decades.

The starting point of the research was the application of traditional centrality measures (betweenness centrality, closeness centrality and degree centrality) to terrestrial and fluvial settlement networks systems in Latium vetus and southern Etruria from the Final Bronze Age to the Archaic Period to compare the behaviour and characteristics of the two regions and start detecting similarities and differences.

As we have seen in Section 1.2 of Chapter 1, between the Final Bronze Age and the beginning of the Early Iron Age, southern Etruria and Latium vetus underwent important processes of centralisation and nucleation of the settlement system that led to the formation of large proto-urban centres. These eventually evolved into cities during the end of the Early Iron Age, the Orientalising Age and the Archaic Period. A graph representing the trend of median settlement size in southern Etruria and in Latium vetus through time, shows how the two regions had a similar beginning and parallel development with different final outcomes (Fig. 5.1). Is it possible to explain the reason for this final result? Were the initial situations after all so similar? In Chapter 4, by analysing centrality measures calculated on the fluvial and terrestrial networks of the two regions we emphasised some similarities and differences. In this chapter we focus further on the infra-structural systems of the two regions (fluvial and terrestrial communication routes) and we analyse and compare their characteristics and functionality.

Urbanism in the past and present remains hotly debated in academia and the media (we could mention the Copenhagen Polis Centre project; the Reception of the City in Late Antiquity European Research Council project, Cambridge; the UrbNet project, Aarhus; the Social Reactors Project, Colorado; the OIKOS Dutch network; and the Cities series published by the Guardian in the UK media). What is an ancient city? When can we say that a nucleated settlement has become a city? Why does a city sometime prevail over others and why does it eventually decline? These questions are matters for lively debate that have not yet been answered definitively, especially with reference to central Italy and Rome in particular. The long-term trajectory of Rome is quite well known and established from the early supremacy within Latium vetus in pre-historic and early historic times, to the emerging power in Italy, during the Republican period, and finally the dominance over the Empire, in the first few centuries of the last millennium before the final collapse around the end of the fourth century ad.

In the first stages of this research, presented Chapter 5, we consciously decided to not undertake least-cost-path analyses because we were interested in exploratory and experimental applications of network science approaches, and we were aware of several issues raised in the application of least-cost-path analysis. Therefore the trade-off between costs and benefits of such an application did not seem remunerative enough or worthwhile in the first instance. However, it was also clear from the analyses that the variable of distance was relevant for the analyses and that an integration of network science approaches within GIS applications, now more and more common, is promising and profitable.1 Therefore, this chapter present a multi-scale analysis of transportation routes in Etruria and Latium vetus based on least-cost path analyses, although we are aware of the critique and problems of such applications.2

Our purpose, in this chapter, is to infer how settlements were organised at the regional level by analysing the structure formed by the roads that connected them. The basic idea is to compare different hypotheses and quantitatively assess which of them is (or are) more plausible and, we do this in three steps (see Section 3.2). Adopting a network science approach implies that the first step is to translate available information on pathways from the usual map format into networks, that is, mathematical structures made up of interconnected objects. Once the empirical system is mapped onto weighted geographical networks, one can apply the established analytic tools provided by network science for their characterisation.