See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271932478

The Roman Transport Network: A Precedent for the Integration of the

European Mobility

Article  in  Historical Methods A Journal of Quantitative and Interdisciplinary History · July 2013

DOI: 10.1080/01615440.2013.803403

CITATIONS READS

18 872

2 authors:

Cesar Carreras Pau de Soto

Autonomous University of Barcelona Catalan Institute of Classical Archaeology
112 PUBLICATIONS   369 CITATIONS    31 PUBLICATIONS   139 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Archaeology of the Conquest (Hispania - II BC) View project

Revalorization of archaeological sites through the application of non-destructive methods" is a specific research program leaded by the Merida Institute of Archaeology
in the framework of the RITECA II Project (Cross-border Research Network of Extremadura, Centro and Alentejo) View project

All content following this page was uploaded by Pau de Soto on 01 April 2016.

The user has requested enhancement of the downloaded file.

 The Roman transport network: a precedent for the integration of the European
mobilityI,II

CÈSAR CARRERASa , PAU DE SOTOb

a Departament de CC. de l’Antiguitat i l’Edat Mitjana. Universitat Autònoma de Barcelona
b Centro Superior de Inverstigaciones Científica (CSIC) Instituto de Arqueología de Mérida

Abstract
In the last years, the use of geographical information systems has been common in many historical and geographical
studies. Perhaps one of the least known applications is network analysis, a series of functions related to hydrology and
mobility. Since 1994, researchers have used network analyses to understand costs and time expenditure in transportation
in Roman times, with the aim of modeling commercial routes. Archaeological objects were employed to confirm those
commercial distributions whose quantities were related basically to transport costs. The present articles go far beyond
and attempt to understand decisions in building a transport network based on time series. It seems that not all Roman
decisions in transport infrastructure were due to economic reasons; some parts of Roman roads were built basically for
political reasons and affected later developments.
Keywords:
costs, GIS, network analyses, roads, Roman, time series, transport

As happens with any other land transport infrastructu-

re, it was the economic success that developed the urban
People normally identify the Roman Empire as a huge foundation and condemned some others. Actually, it is the
political extension covering parts of three continents cen- closest precedent of how transport network affects urbani-
tred in the Mediterranean basin. It covered half Europe zation in the same way that were effects by other transport
from the Western Atlantic coasts up to Rhine and Danube infrastructures such as railways.
rivers and all the Balkans and Greece. Besides, its bor- The uniqueness of the Roman road network as well as
ders reached the Near East up to the river Euphrates and its layout continuity over the centuries made compulsory
the whole Northern African coastal strip and Egypt. Such referring and comparing to it in any serious study on the
extension of land could not be rule without a sophistica- history of land transport in Europe. Perhaps people still
ted transport infrastructure and developed courier public believe that Train a Grand Vitesse has an original layout,
system (Chevalier, 1976; Eck, 1979; Kolb, 2000). but their main railways still followed the basic Roman rou-
Therefore, the Roman Empire as a political project tes designed more than 2,000 years ago.
could not have carried on without a well planned mobility
system, which fostered economic exchange and informa-
tion transference. Notwithstanding the temporal distance, Roman transport infrastructures
the Roman Empire constitutes an excellent example for
the integration of European mobility. When anyone refers to Roman transport infrastructures,
Moreover, the Roman road network was the first inland normally people think about Roman roads. The Roman
transport network in Europe and the basis of most national Empire was the first political European entity that built
road systems in the European nations over centuries up to a complete road system linking different parts of its terri-
the present. The first Roman road layout was designed tory. The reason of such important investment in terms of
to favor contact between Roman citizens and colonies in money and labour was facilitating army movement around
the way that they first built up road infrastructures, and the Empire as well as an effective control of all conquered
later cities were founded. Perhaps people still believe that territories. It was not an original policy; in fact the Persian
Roman roads were built after cities foundation, but in fact Empire (Vth c. BC) and the Macedonian kingdom (IVth
it was the other way around. c. BC) had already created a sophisticated road network
for identical aims. As usual, Romans made an outstanding
I Published in HISTORICAL METHODS, July September 2013, imitation that overcame the original model.
Volume 46, Number 3, Pages 117-133 However, they paid even more attention to other means
II Pre-print version submitted by the authors of transport such as maritime and river navigation, since
1
 Figura 1: River boat with barrels from Nemausus (France) (Trier Museum)

those transports generated lower costs. The city of Rome transport was very slow; a wagon of a standard 1380 kg co-
was developed on a Tiber river ford, so they recognised the vered 16-20 Km a day as normal speed in a plain terrain.
importance of rivers as cheap mean of transport inland. On the contrary, mountainous conditions could increase
On one hand, Roman ruled all over the Mediterranean – costs and time invested. Cato (De Agricultura, 22,3) re-
the so-called Mare Nostrum – most of its territories were cords a transport of an olive-oil pressing-stone from Sues-
bathed by Mediterranean waters. On the other hand, they sa to Venafro (circa 60 Km) that took 6 days and 6 days
attempted to conquer the Atlantic Ocean – the so-called salary for 6 people.
Mare Exterior – without identical success. In general, people avoided land transport in commercial
Nevertheless, they built impressive port infrastructures journeys except if it was required or when traders carry
all over their territories (i.e. Alexandria, Carthage, Ep- luxuries.
hesus) in order to keep commercial routes towards Italy A different approach should be taken to the Roman mail
and the city of Rome (i.e. Portus). There was a vivid tra- system (cursus publicus). The state- run courier was set
de amongst different territories, which is archaeologically up by the Emperor Augustus (Eck, 1979; Kolb, 2000) to
documented by shipwrecks and volume of ceramics and convey news from the armies in the border to Rome and
amphorae recorded in great amounts in different points of the other way around. The State required a fast service
the Roman Empire (Parker 1992). The value of such ar- involving different riders and horses relays to cover large
chaeological data appears if ancient transport conditions distances. That is why Roman roads had mutationes every
are taking into account. 20 Kms (places to change horses) and mansios every 40
For instance, according to the historic tradition from Kms (stopping places – accommodation, meal, changing
Greek period (Hesíodus Op. V 663-665, 678-684), the Me- horses). Cursus publicus was a complex administration to
diterranean ports were closed from mid November up to provide fast news of any territory in the Empire, reaching
mid March (mare clausum) so as avoiding possible shipw- distances of 800 Km a day.
recks (Rougé 1952; Casson, 1985; Arnaud, 2005). Therefo- As said before, transport infrastructures were used al-
re, communications between different sides of Mediterra- so for economic reasons and results of such commercial
nean were hindered in winter and opened with the Na- transactions are distributions of archaeological ceramics
vigium Isidis celebration with a procession described by and amphorae. In the last years, Geographical Information
Apuleius (Metamorphosis 11) and Clement of Alexandria Systems (GIS) has become a quite useful tool to model pat-
(Stromateis 6.4.37.1). This seems to be the reason why terns of distribution and spatial networks. In the early 90s,
official documents in Egypt registered the name of an Em- we developed a network analysis model of Roman trans-
peror after his death (i.e. Nerva, Domitian), since they died port in order to explain archaeological distributions in the
in winter and news did not reach this distant province on Roman province of Britannia (Carreras, 1994).
time (Duncan-Jones, 1991). Britannia was our first case study for testing a metho-
Besides, ancient sources remind us that land commercial dology of cost and time consumption in ancient transport,
2
 Figura 2: Analyses of three Roman provincial transport networks in GIS

and later comparing the resulting maps to other quantified transport in terms of transport cost and speed. Connecting
maps of archaeological distributions. Later on, the same each layer with the digitalised network of specific mean of
model was applied to other Roman provinces from the Ibe- transport and linked at particular nodes where more than
rian Peninsula (Carreras and de Soto 2010), whose pattern network coincided, a complete model of the whole Roman
identified at least two temporal stages in the construction Britain could be created (Carreras, 1994).
of their transport network with a coherent economic aim.
The earliest digitalisation project was undertaken in
However, our final study of Roman Italy, still underway,
AUTOCAD, and then exported to different GIS program-
shows the complexity of development of transport infras-
mes such as SPANS and ARC/INFO. Later on the follo-
tructures in the Italian Peninsula as a cumulative result
wing projects, as well as the data review from Britain,
of a myriad of temporal political and economic decision,
were carried out directly in ArcGis. Those GIS program-
which may lack a final coherence.
mes are a multiscale system that allows detailed digitali-
As this paper will illustrate, GIS have been applied to
zation and wide-scale geographical analysis. They also in-
analyse individually different transport networks in Ro-
clude functions (inside the Network Analysis system) that
man provinces, which were built up in relative limited
calculates the shortest route between one selected point
spans of time. Apart from solving mobility problems in a
to another combining the different network segments. In
particular moment, they became fossilised networks that
fact, those types of analysis are kind similar to the fun-
determined mobility for later periods. Transport infras-
ctionality of ours GPS Car Navigation Systems. In those
tructures involved huge economic investments in the past
Network Analysis Systems, is also possible to calculate the
as it is nowadays, and they were developed in dynamic
shortest routes between one point to the remaining nodes
processes without a complete view of the final resulting
in the map. If one chooses this second application, it can
outcomes.
generate an interpolation map from best values from all
the nodes in the map (Carreras, 1994; Carreras and De
Transport model: A GIS application of network Soto, 2010).
analysis
The whole model relies on the constant values of speed
Our initial hypothesis was that cost and time consum- and cost for every singular mean of transport in Roman ti-
ption affected the trade volume at every particular market. mes. Speed values can easily obtained from ancient literary
In other words, markets with higher costs of transport con- sources and ethnographical information, gathered in well-
sequently should document less quantity of particular good known publications (Lawton, 2004; Carreras and De Soto,
(i.e. ceramics or amphorae). Therefore, we digitalised the 2010), though they may lightly vary in different geograp-
whole transport infrastructure of Roman Britain at scale hies. With regards to Roman transport costs, apart from
distinguishing between maritime routes, river navigation, a myriad of Egyptian papyri (Drexhage, 1991) our main
roads and pathways. Every distance unit (i.e. Km) was gi- source for Roman times is the Diocletian’s Edict (Lauffer,
ven a different constant value according to each mean of 1971; Rouché, 1989), which compares the cost of the diffe-
3
Figura 3: Resulting transport costs from the Roman port of Dover Figura 4: Resulting accessibility model for Roman Britain

rent types of transport in the year A.D. 301. The resulting ment. Sometimes those decisions affected the development
cost ratios are approximately: of the whole territory and they were not suitable all over
1 (sea): 3,4 (downstream river): 6,8 (upstream river): the time. Romans when conquered new lands established
43,4 (pack animal): 50,7 (wagon) Likewise, the available a kind of hierarchies in the new territories aiming to ob-
speed values applied to the time model for Roman times tain the maximum resources (i.e. cereals, minerals, taxes)
were: 3,7 Km/h (sea): 2,51 (downstream river): 0,62 (ups- with the minimum effort (i.e. cost, time) and transported
tream river): 2 (land transport) to the city of Rome. This is what can be called a colonial
Those theoretical models of transport efficiency in Ro- transport model.
man Britain were compared to some archaeological dis- The model can be implemented in the same digitalised
tribution of local ceramics (i.e. Oxfordshire ware, Black- transport network providing a value for each individual
Burnished 2) and imported amphorae providing an in- node according to the number of segment of network that
teresting result. Distributions were normally matched by ends at that point. For instance, every secondary road en-
cost maps from either the production area in local wares, ding at that point would add 1, a primary road 2, a river
or cross-channel points in imported amphorae (Carreras, port adds 3 and maritime port 4. The final sum would gi-
2000). In other words, archaeological distributions fulfilled ve a rate for each node and an interpolation map could
roughly the rule that a lesser transport higher archaeolo- represent the territory as a whole.
gical quantities, so more potential commercial benefits. The accessibility model applied to Roman Britain shows
A second transport model developed later was the ac- a clear hierarchy in a series of cities such as Londinium
cessibility one (De Soto and Carreras, 2007). It is anot- (London) as the provincial capital from the early conquest.
her application of the graphs theory, initially developed In fact, Dicks (1971) already identified the route between
by Dicks (1971) for Roman Britain. Considering the de- London and Chester as the most important communica-
sign of the whole transport network from the start, nodes tion axis in the province. Likewise, Laurence (2001) recog-
that concentrate more segments of network such as roads, nized the role of itinerary II from Richborough to Chester
maritime and river ports would have a more prominent crossing London, and afterwards to York and Carlisle as
role in the network, so higher accessibility. In other words, one of the main axis. The GIS accessibility model (figure
planned network infrastructures define always a kind of 4) shows an indirect route from Richborough up to Lon-
hierarchy amongst cities (i.e. nodes) from the beginning, don and following Thames valley up to Cirencester and
which responds to decision taken in a very particular mo- Glouchester and then, central Wales and Chester. This
4
road together with another road leading from London up rritories in order to reassure routes to their winter camps
to Leicester, Lincoln and York to Newcastle and South in the coastal towns. There are a series of milestones in
Shields, gave structure to the whole mobility in the pro- the late IInd century BC that draw parts of this original
vince, combined with the maritime transport. network layout, some of which was in disuse later on.
Transport infrastructure in Roman Britain appears to Roman experience in the Iberian Peninsula show diffe-
be a quite well-planned colonial network to supply the le- rent stages in the development of a transport infrastruc-
gions in the Northern border as well as obtained resources ture. Decisions were taken at different points to solve par-
from the Welsh and South-Eastern mines and high popu- ticular problems at each moment. Initially Romans had to
lated south. If you combine the accessibility map to the face hostile Iberian tribes and the main roads were built to
cost and time consumption maps, it looks like the Ro- secure any possible withdrawal to the winter headquarters
mans planned carefully the whole transport infrastructure (Emporiae and Tarraco).
to facilitate movements of troops and merchandises in the However, after the final conquest of Astures and Canta-
island. They even take care about the distance between brians in NW of the Iberian Peninsula, M.Vipsanius Agrip-
mansios (stopping places) that was shorter than other pro- pa, lieutenant and brother in law of the emperor Augus-
vinces such as Italy and Hispania. Laurence (2001, 82) ar- tus, envisaged a throughout transport infrastructure for
gues that such difference may have been due to the kind the whole Iberian Peninsula taking into the river valleys
of road surface – being gravel the typical British material and their river navigation, the main mineral resources in
-, but it is same case as Hispania. One possible reason is NW gold mines, silver in Sierra Morena and near Cartage-
the weather conditions of Roman Britain that favored a na, and the suitable ports (Rodà 1998). The final transport
shortest travel journey. infrastructure represents quite well a “colonial model”, in
Britannia was a later conquered territory by Claudius which part of the earlier military roads were left aside per-
(AD 43), so organization of transport infrastructures was haps because they did not have any economic value any-
already tested successfully in other previous territories. more. Again, it was a political decision taken at particular
Therefore, knowledge of the new territory as well as requi- moment that shaped the accessibility in the Iberian Penin-
rements of the army and economic exploitation was key in sula in Roman times, but also affected medieval periods.
the planned network layout.
The accessibility map generated in GIS (Carreras and
De Soto, 2010) shows clearly the desire of Roman politi-
First overseas experiences: the Iberian Peninsula cians to obtain the maxim resources (i.e. minerals, grain,
taxes) from the territory and uploading all those goods
The Iberian Peninsula was one of the first overseas terri- at the main provincial ports following maritime routes to-
tories conquered by the Romans as a result of the Second wards Puteoli or Rome (Arnaud, 2005).
Punic War (late IIIrd century BC). Probably, Romans did The river valleys such as Ebro, Guadalquivir, Guadia-
not know much about the inner parts of the Peninsula, na, Tajo and Douro were the main mobility gates to the
and at the beginning, they only had military headquarters Meseta (Parodi 2001), which had problems of communica-
and allied tribes in the coastal regions. tion due to its dependence on land transport. Provincial
It appears that they built the first inland roads for lo- capitals were also located in accessible places as is the ca-
gistic reasons in the way legions advanced in hostile te- se of Tarraco, Emerita Augusta and Corduba. Difficulties

Figura 5: Republican road in the NE of Iberian Peninsula – numbers indicate presence of milestones (adapted from Ariño et alii, 2004)

5
 Figura 6: Accessibility map from the Imperial transport infrastructure in the Iberian Peninsula

in inland transport can be easily recognized with the cost efforts to communicate the Meseta with periphery, buil-
model maps. They identify a well communicated periphery ding not only roads in a system but also developing failing
and river valleys and a difficult mobility in the central Me- inland waterways from the Douro (Canal of Castile) or
seta. Tejo to Madrid.
Actually, this is an endemic problem of the Iberian Pe- Individual political decisions did not overcome geograp-
ninsula that continued until the XIX century when intro- hical conditions in the Peninsula, and perhaps the Ro-
duction of railways was supposed to overcome this inner man solution was the most suitable one in economic terms
mobility problem (Ringrose, 1970). Roman transport so- according to the degree of technological development of
lution to these geographical conditions attempted to take transportation until the XIXth century (Leighton, 1972;
full advantage of water transport according to the diffe- Lawton, 2004). From this second provincial example, we
rence in the unit cost by Km, creating basic corridors for may conclude that geographical conditions determine pos-
mobility. sibilities in mobility in a territory and despite individual
Roman experience and its colonial vision illustrate a decisions over the time; economic factors could undermine
kind of territorial exploitation where transport was eco- any illogical policy. It becomes feasible to make errors in
nomically feasible, except special cases such as Imperial a new conquered territory with still undrawn limits, but
quarries. The remaining territories were left aside; they less acceptable in a long tradition states with invariable
did not have access to global markets and kept a more borders.
local economic profile. Such a colonial vision in the deve- In terms of methodology, GIS provide an interesting ap-
lopment of transport infrastructures produced also a side plication known as time series that could analyse changes
effect, which was an increase in population. The Iberian in a particular landscape all over the time (Wheatley and
Peninsula demography reveals an increase in towns’ size Gillins, 2002). It is cumulative information that may not
in the coastal areas and territories with good transport in- have a straight interpretation from the resulting map, for
frastructures. Therefore, transport was a key issue in the instance the final infrastructure as an addition of temporal
demographic development (Carreras 1996). constructions.
As said, changes in the road infrastructure in the Ibe-
rian Peninsula due to new political borders (i.e. Christians The first transport network: mastering geography
and Muslims, Portugal-Spain) and centralizing monarchies of the Italian Peninsula
(i.e. Austrians, Borbons) did not alter the main economic
problem between centre and periphery in the Iberian Pe- Rome was from the very beginning a hub of communi-
ninsula until the XIXth century. Actually, such centralized cation, since the early settlement was built on seven hills
policies from the XVIIth century onwards to improve the closed to a river ford on the Tiber valley. Therefore, it was
kingdom capital to the rest of the Peninsula show worse in the middle of an ancient land route communicating the
economic global results than the ancient Roman network Tuscan cities in the North of the Peninsula with the Greek
(Carreras and de Soto 2010). The Borbon kings made great colonies in the further South, as well as the middle of the
6
 Figura 7: Cost map generated from the city of Hispalis (Seville)

Figura 8: Transport roads in the Iberian Peninsula (up-down and left-right): Medieval Christian kingdoms, Muslim Califate, XVIth century
and XVIIIth century

7
 Figura 9: Itinerary of the via Appia

river Tiber and the Salaria route coming to the Apennines

to Tiber mouth.
During the IVth century BC, Rome started expanding
its influence towards the South, helping the city of Capua
against the indigenous tribes of Samnites. In the Second Figura 10: The main Roman roads in the Italian Peninsula
Samnite War, Roman legions were defeated due to lack
of water supply in the so-called Caudine Forks (321 BC).
That is the reason why the censor Appius Claudius (312 related to the space conquered by Rome at that time and
BC) order to build the first Roman road (via Appia) lin- linked only the Roman colonies (Spoleto, Nania, Rimini)
king the city of Rome with Capua (115 Km), to avoid any following a straight line. This road was a still a complex en-
lack of supply to the Roman armies (Laurence, 1999, 11- gineering work since it crossed the Apennines range, with
26). This first road was constructed by military reasons a second alternative road called the via Flaminia minor.
and connected only Roman colonies on its route towards Roman road construction followed similar pattern con-
Capua. It was not exactly the shortest and the cheapest necting the city of Rome with the new conquered terri-
route, but the safest one at the end of the IIIrd century tories and Roman colonies, with some exceptions such as
BC, crossing the Pontine Marshes for a first stop at the the via Aemilia that connected Roman colonies amongst
coastal Roman colony of Terracina. themselves (Rimini-Cremona-Piacenza). Therefore, the fi-
Crossing the Pontine Marshes involved land reclamation nal centralized road network was a simple result of a series
to the sea and wooden-floors to some parts of the itinerary. of temporal decisions in the expansion of Roman Republic.
Besides, according to Diodorus Siculus (20.36) it was a hu- With regards to the maritime ports, Rome was not a
ge economic investment that represented the State revenue coastal town, so the port at the Tiber estuary was not im-
of a whole year. The new Appian road required the foun- portant until the Imperial period. Roman Republic used
dation of two new Roman colonies on the way, Sinuessa the infrastructures of Greek colonies in the South and
and Minturnae, the second of which was on the river Liris founded Puteoli (Puzzoles) in the bay of Naples, beco-
that allowed direct inland access. The via Appia, as the ming the main maritime port of the whole Mediterranean.
first Roman road, showed a clear pattern of linking only Puteoli and Rome were connected by the via Appia, and
Roman settlements and avoiding foreign cities. Military this axis combined the main transport infrastructure of
and political reasons were behind those initial transport the early State, combined with some navigable rivers such
decisions, since Roman politicians at that time could not as Tiber and Po.
foresee the State future expansion. A detailed analysis of the accessibility model reveals
The second Roman road, the so-called Flaminian road such development of transport network in Roman Italy as
(220 BC), changed the transport geography of the who- result of individual decision according to determined tem-
le Italian Peninsula, connecting the Tyrrhenian with the poral constrains (Fig 12). Those political decision favored
Adriatic sea (Laurence, 1999, 22). Again, it was a decision contact between communities of Roman citizens but did
8
 Figura 11: Accessibility map in the Italian Peninsula Figura 12: Transport cost from the city of Arretium

not foresee a political union of the whole Peninsula. The- lack of roads in less populated regions inland, but this also
refore, the accessibility maps show difficulties in crossing affects accessibility of most ancient coastal Greek colonies,
by road the Peninsula from Tyrrhenian to the Adriatic which were still basically connected by ship.
Sea, in other points not covered by the Flaminian road. Some Italian industries or exporting areas were located
Accessibility was considered only from the capital, Ro- in those regions with facilities for transporting their goods
me, which was the centre of the whole transport system to potential consumers. Figure 13 represents the transport
not only by road, but from other means. At least, the cons- cost model from Arretium (Arezzo), which was the main
truction of Portus (port of Ostia) in Claudius and Trajan production area of Terra Sigillata Italica (TSI). This fine
period converted Rome also in the main maritime port ware was distributed all over the Western Roman Empire
in the Mediterranean. Therefore, Puteoli lost part of the since it was located in a place in the Tuscan region well-
maritime traffic of Republican period, and Rome was the connected with coastal ports in the Tyrrhenian. Despite its
main hub of communication of the Empire. As is said, "all excellent accessibility towards the Tyrrhenian coast, land
the roads lead to Rome". transport increased product cost so distribution in inland
However, most part of the contact within the Italian Pe- Italy and other provinces was limited to some extent. That
ninsula appears to have been by sea according to the acces- is why branches of TSI were developed in other places
sibility map. There is a good access between the Northern such as Lyon and the Western provinces created their own
ports of the Adriatic from Ancona to Ravenna and Pata- industries shortly such as Terra Sigillata Gallica (TSG) in
vium with the Po valley as inner river route. The port of Southern and Central Gaul.
Aterno in central Adriatic was another well-communicated Arretium (Arezzo) location inland, close to Arno valley,
area, and the same happened with Brundisium and Tarent, did not give any advantage to other location closer to ma-
though with a lack of accessibility inland. A similar picture ritime routes and the final consumer markets. Italian road
can be drawn in the Thyrrenian coast with an excellent ac- network was not built up, as we have seen, with an econo-
cessibility in the Bay of Naples and around Rome. In the mic perspective but a political one. Therefore, some inland
Tuscan region, there is an extensive area well-connected towns in Italy could not compete in economic terms with
road network that reaches areas such as Florentia with other provincial towns. As soon as those provinces such as
the coastal ports (Luni, Pisa, Populonia, Cosa). Hispania or Britannia were romanised, they were in better
Difficulties in transport were evident in the inland conditions to access to distant and even central markets
mountainous regions in the Appenines. There is a clear such as the city of Rome.
transport barrier crossing the centre of the whole Penin- Because Rome, and other coastal cities in Italy such as
sula from North to South, with only a few roads connec- Ostia, Pisa, Puteoli and Ravenna, were the real hubs of
ting the two seas. Besides, some southern regions from the communication in the Italian Peninsula; so they could take
Magna Grecia suffered a limited investment in roads, as advantage as economic centers in Italy. A series of time
is the case of the present Puglia, Basilicata and Calabria, decision constructed the transport network of the Italian
which hardly depended on maritime transport in coastal Peninsula in a way that mainly favored the capital, Rome,
areas. and other main ports all over the Peninsula: Puteoli in the
Again, the transport network development in the Italian South (port of Magna Grecia), Pisa (port of Etruria) and
Peninsula does not follow any economic criteria. There is a Ravenna (north Adriatic).
9
 Figura 13: Transport cost from the city of Rome

The figure 14 reveals transport cost from the city of Ro- a conquered territory and quick connection to new esta-
me, and the cheap access from all the coastal regions of blished Roman colonies. This limited vision of the Italian
Italy except the area of Apulia, close to the port of Brin- Peninsula explains its road network centralised from the
disi. Besides, it shows difficulties in terms of high costs Roman capital without caring about other Italian citizens.
to access diverse towns inland. The Apennines constitu- The population increase in the city of Rome meant that
te a barrier for land transport to connect Tyrrhenian and the hinterland could not feed the new Roman citizens
Adriatic seas, which was increased by the limited number (Frayn 1993), so they require external resources of food
of roads such as via Flaminia that crossed the mountain and other merchandises to simply survive. That is why
range. Compared to the other provinces studied with those Roman territorial expansion outside its boundaries in the
GIS applications, Roman Italy showed a transport network Italian Peninsula was part of its economic model. From
that did not take into account mobility and access to the the IInd century BC, Rome expands its territories to both
markets to the main towns, except Rome. Other provin- western and eastern Mediterranean, so Italy was the cen-
ces conquered later by Rome presented a well- organized tre of this new Empire. Now, Rome invested in maritime
network that notwithstanding the main aim of exploding infrastructure creating large ports such as Portus and Os-
local resources, it favored easy communication with the tia, which could download merchandises carried by ships
coastal ports, and thereby the city of Rome. coming from all the Mediterranean ports. The system was
complemented with a mail system called cursus publicum,
to convey official news from all over those imperial lands.
Conclusions Provincial infrastructures layout responded to a new
economic ideology. New territories were not only controlled
Roman Empire was developed over a long period of time. politically, but were vital economic resources for Roman
Sometimes, we record in archaeology such developments as Italy. Therefore, the provincial transport infrastructure ai-
the case of transport infrastructures, believing that they med to obtain the maximum resources (i.e. grain, metal,
were the result of a single decision. Transport networks in slaves, taxes...) with a minimum effort. Combination of
Rome and other historical periods are “long durée” human maritime ports, river waterways and roads facilitate mo-
processes, which may or not have a unique sense when vement of those goods from any location inland up to the
analysed as a whole. GIS applications allow us to take closest port, and from there, they were shipped to Italy.
into account not only a global vision of transport network Evaluation of those transport network functionality was
but also short processes as time series. thanks to GIS applications. Those new computer appli-
As can be observed in the transport network of the Ita- cations permit modelling spatial data and also transport
lian Peninsula, roads and ports were built to solve military, infrastructures in such a way that new questions can be
political and social problems every time that the young Ro- raised. Network analyses are a type of application that in-
man Republic had to face them. Those infrastructures did volved movement according to a series of conditions such
not have a main economic aim, but a political control over as means of transport, costs or speeds. This type of appli-
10
cation has been employed for us to explain the Roman R.LAURENCE (eds.) Travel and geography in the Roman
transport infrastructures in different territories and pe- Empire. London: 67-94.
riods and also comparing the same territory (i.e. Iberian LAWTON, B. (2004) Various and indigenous machines.
Peninsula) over the time (Carreras and De Soto 2010) with Power generation and transport. Volume I. Brill, Leiden.
quite remarkable results. LEIGHTON, A.C. (1972) Transport and communica-
Modelling influence of military, political or economic in- tion in early medieval Europe AD 500-1100. Oxford.
terests sheds some light in how societies change their lands- PARKER, A.J. (1992) Ancient Shipwrecks of the Medi-
cape and helps us to understand their historical evolution terranean and the Roman Provinces. British Archaeologi-
and incidence of mobility or communication in their deve- cal Reports. International Series 580. Oxford.
lopment PARODI, J.M. (2001) Ríos y lagunas de Hispania como
vías de comunicación. Écija.
RINGROSE, D.R. (1970) Transportation and Economic
Bibliography Stagnation in Spain, 1750-1850. Durham.
RODÀ, I. (1998) “El papel de Agripa en la trama urbana
ARIÑO, E; GURT, J.M.; PALET, J.M. (2004) El pa- de la Hispania Augustea”. Los orígenes de la ciudad en el
sado presente: Arqueología de los paisajes en la Hispania Noroeste hispánico (Lugo 1996). Vol I. 276-293
romana. Barcelona. ROUGÉ, J. (1952) “La navigation hivernale sous
ARNAUD, P. (2005) Les routes de la navigation antique. l’Empire Romain”. Revue des etudes anciennes. 316-325.
Itinéraires en Méditerranée. Paris. ROUCHÉ, C. (1989) Aphrodisias in the Late Antiquity.
CARRERAS, C. (1994) Una reconstrucción del comer- Monographs JRS. London.
cio en cerámicas: la red de transportes en Britannia. Cua- WHEATLEY, D.; GILLINGS, M. (2002) Spatial tech-
dernos de Arqueología, 7. Barcelona. nology and archaeology. The Archaeological Applications
CARRERAS, C. (1996) “A new perspective for the de- of GIS. London.
mographic study of Roman Spain”. Revista de Historia da
Arte e Arqueologia n.2, 1995-1996; pp 59-82.
CARRERAS, C. (2000) La economía de la Britannia
romana: la importación de alimentos.Barcelona.
CARRERAS, C.; DE SOTO, P. (2010) Historia de la
movilidad en la Península Ibérica. Barcelona.
CASSON, L. (1985) Ships and Seamanship in the An-
cient World. London.
CHEVALIER, R. (1976) Roman roads. London.
DE SOTO, P.; CARRERAS, C. (2007) “Anàlisi de la
xarxa de transport a la Catalunya romana: alguns apunts”.
Revista d’Arqueologia de Ponent 16-17: 177-192.
DE SOTO, P.; CARRERAS, C. (2009) “La movilidad en
época romana en Hispania: aplicaciones de redes de análi-
sis (SIG) para el estudio diacrónico de las infraestructuras
de transporte”. Habis 40,pp.303-324.
DICKS, T.R.B. (1972) “Network analysis and historical
geography”. Area 4: 4-9
DREXHAGE, H.J. (1991) Preise, Mieten/Pachten, Kos-
ten und Löhne im römischen Äegypten.St.Katherinen.
DUNCAN-JONES, R.P. (1991) Structure and scale of
the Roman Economy. Cambridge.
ECK, W. (1979) Die staatliche Organisation Italiens in
der hohen Kaiserzeit. Beck, München.
FRAYN, F. (1993) Markets and fairs in Roman Italy.
Oxford
KOLB, A. (2000) Transport und Nachrichtentransfer im
Romischen Reich. Berlin: Akademie Verlag, (Klio. Beitrae-
ge zur Alten Geschichte, Beihefte, Neue Folge, 2).
LAUFFER, S. (1971) Diokletian’s Preisedikt. Berlin.
LAURENCE, R. (1999) The roads of Roman Italy: Mo-
bility and Cultural change. London.
LAURENCE, R. (2001) “The creation of geography.
An interpretation of Roman Britain”. In C.ADAMS and
11

View publication stats