This page was automatically generated by NetLogo 3.0. Questions, problems? Contact

The applet requires Java 1.4.1 or higher. It will not run on Windows 95 or Mac OS 8 or 9. Mac users must have OS X 10.2.6 or higher and use a browser that supports Java 1.4. (Safari works, IE does not. Mac OS X comes with Safari. Open Safari and set it as your default web browser under Safari/Preferences/General.) On other operating systems, you may obtain the latest Java plugin from Sun's Java site.

Introduction    Space & Representation     How the model works     How to use it

created with NetLogo

view/download model file: itineraries.nlogo

Please adapt and use the code for your own research, but I'd appreciate a note if you do!


This model explores the effect of route network topology on the diffusion of information in the Roman world.

It was built as part of my research as the Postdoctoral Research Fellow in Roman Archaeology at the University of Manitoba, funded by the Canada Research Chair in Roman Archaeology. It was first presented at the 35th Annual Medieval Workshop, "Cartography in Antiquity and the Middle Ages: Fresh Perspectives, New Methods" at the University of British Columbia on October 28th, in my paper "The Antonine Itineraries and Agent-Based Modeling". The comments and critiques offered by the participants of that workshop are gratefully acknowledged. I also am in debt to the participants at the Netlogo workshop at Mesa State College, Grand Junction Co., for teaching me how to develop models in Netlogo.

The information presented here is intended as a preliminary description; an in-depth description of the model and the implications of this simulation will be discussed in print in due course. To cite this model and this information page, please use:

Graham, S. (2005) "An Agent-Based Exploration of the Antonine Itineraries"

Most people have a mental image of how places interrelate, how they connect. There is an excellent cartoon by Gary Larson, where a man wants to get from Point B to Point A. He asks for directions, but the local responds, ‘Beats me sonny. Most people want to go the other way’. The joke indicates a truth about how people visualize space – often, it is as an itinerary, a sequence of place-names through which one must travel.

For the Roman, large-scale geographical space was sometimes viewed as a list, of what-comes-next (there is debate about the extent 2D maps were known/used in the Roman world). Details about the space between places could always be supplied by a knowledgeable local, but the global perception was courtesy of a list, an itinerary. How successful one’s journey was, depended on the quality of the itinerary. For one such journey, take Galen’s journey to Lemnos in the second century AD (discussed by Kai Brodersen 2001). Galen lists the towns he has been travelling through, and tells us how he persuaded a captain to divert to Lemnos. Unfortunately, on this leg both Galen and the captain of the ship he is travelling on have particularly sketchy information and they miss their destination. As it turned out there were two cities on the island and they had landed at the wrong one. They did not have time to sail around to the other. Galen later returns to the island and records his itinerary so that others might not make the same mistake he made (Galen, de simplicium medicamentorum temperamentis 9, discussed in Brodersen 2001: 8-9).

An itinerary then is a very utilitarian tool for helping one plan a journey. But might a Roman’s itinerary also reflect a larger vision of how the world worked? What interpretations of space might be implied by this mode of representation?

With the Antonine Itineraries, we have a collection of routes, ostensibly to every corner of the Empire (but with some exceptions), collected and collated from earlier routes. Some of the routes come from earlier collections, others bear witness to the influence of Greek periploi. Scholars have deduced breaks in the sequence which imply that up to ten separate smaller itinerary lists were collated together to create the collection. Indications in the text suggest that some parts were compiled in the third century, other parts in the early fourth. Some itineraries listed seem to suggest that a real journey was recorded and incorporated, perhaps a journey of an Emperor Antoninus. (Salway 39-43).

Mapping the Itineraries is not the same as mapping the Roman road network, for there were far more roads and towns than are listed on the Itineraries. Because there was no formal system of designating the roads, a Roman would travel from town to town listed in the Itinerary, inquiring about which road to take along the way to get from point A to point B. These Itineraries present Roman space to us in a very social way. They present space in a way that reflects the agenda of only a few individuals at most.

Regardless of who compiled these itineraries, and for what purpose, they do record something of how towns and settlements interconnected in the Empire, and for that reason alone, are worth our consideration.

In the rather simple simulation presented here, I am testing the effect of different patterns of connectivity on the diffusion of information. I am moving from a static exploration of the itineraries to a dynamic one. In the simulation, I introduce an ‘idea’ to a single agent. This agent then travels through a region, following the pathways laid down in the itineraries. Other agents are travelling as well. When an agent who has not ‘heard’ the idea comes into contact with our ‘carrier’, the idea is passed along and we now have two carriers, spreading the idea. What is this idea? Perhaps it is a new religious practice, or a new style in pottery making. Whatever it is, it is secondary to the journey and is not the reason for the journey itself.

The various Antonine Itineraries have been 'stretched' across the map, so that the listed towns are pinned to their correct geographic location. Agents are distributed at random along the various routes. The initial 'informed' agent is coloured blue; a 'spotlight' shows you where that agent is located. The agents travel randomly along the routes. When they encounter another agent, there is a chance the message will be passed along.

The routes are 'thick', which is to take into account the difference between the global representation of geographic knowledge (the itinerary) and the local understanding of routes and byways.  A traveler might not know the exact best way to get from one town to another, and could get lost, take minor roads, or otherwise not travel optimally.  By representing the itineraries with thick lines, a certain degree of randomness in terms of exact route progression is allowed (watch carefully, and you will sometimes see agents pass each other but not interact).

The percentage of agents who have 'heard' the message is plotted, as well as the maximum number of times a single agent has heard the message.

Select the base-map you wish to explore, from the chooser 'map-choice'.

Select the number of 'travelers'. (over the net, the optimal number, at least to illustrate how the model works, is around 50).

Select the 'importance-of-the-news'. The higher the importance, the greater the probability that an encounter will result in information being exchanged.

Hit 'setup'. Then to go once, hit 'step'. To run without interruption, hit 'go'.

To add your own base-map to the model, create your map so that it will fit in the view window (roughly, 300 by 300 pixels). Mark the routes you want explored by drawing them in red. Save it as 'your-map.png' in the same folder as the model. Then choose 'your-map.png' from the map-choice chooser.

The simulation used here is not very sophisticated. Extensions to the model might want to include the chance that a message not be heard, or might want to include greater intentionality (for instance having agents actively seek out others who have not been encountered). From an archaeological point of view, one could extend the model to ask, were the itineraries used by long-distance traders, or were products traded hand to hand through local networks? In this case, the model could be extended by including pottery manufacture sites, and dividing the agents into producers and consumers. Then the model could be used to track the trade and distribution of pottery, using the itineraries as the geographical framework. An output of the model might be an artificial distribution map of pottery types, which could then be compared against actual distribution maps. The combination of parameters in the model which created the best fit against the actual distribution would therefore be of some import!

The set-up and navigation structure in this model comes from the 'Cruise' model developed by Guerin, McKenna, and Jung.


Brodersen, K. "The presentation of geographical knowledge for travel and transport in
the Roman world: itineraria non tantum adnotata sed etiam picta" in C. Adams
and R. Laurence, _Travel and Geography in the Roman Empire_ Routledge, London.

Graham, S. "Models"

Guerin, S., McKenna, S., and D. Jung. "Cruise"

Laurence, R. "The creation of geography: an interpretation of Roman Britain" in C.
Adams and R. Laurence, _Travel and Geography in the Roman Empire_ Routledge,
London. 2001

Salway, B. "Travel, itineraria, and tabellaria" in C. Adams and R. Laurence, _Travel
and Geography in the Roman Empire_ Routledge, London. 2001