Skip to the content.

The use of roads in the ancient Mediterranean based on inscriptions

ANALYSIS


Project_status

Purpose

The system of roads in the Roman State represented complex network of stone paved roads connecting Rome with all parts of the Empire, stretching all accross the Mediterranean with more of 80,000 km of known paved roads, or some 400,000 km of unpaved roads. The road network stretched as far as the Iberian Peninsula and the Near East, connecting the major administrative and population centres, with amenities serving the state as well as the local people. The roads represented crutial infrastructure enabling fast movement of Roman armies, circulation of information and state related traffic.

Inscriptions on a durable medium, suach as stone, were often placed next to the course of the road to convey important information to the traveller. The text of inscriptions contained details of distances to the administrative centres, informed about the road building and its maintenance, but it also contained physical description of the amenities, such as inns or horse-stations, bridges, gates, and the daily use of the road. The information extracted from an inscription can help not only to locate the course of the road in space, but also in time: the textual references to known historical events or people can provide a date for building or maintenance work of the section of the road, previously unknown form any other archaeological or literary source.

This project aims to to extract information related the road system and its use in the ancient Mediterranean on the basis of surviving Latin, and potentially also Greek, inscriptions. The goal is to fill in any gaps in the known course of the road network, add more details on its development in time and provide more details on the usage of the road system by the state, but also by the common people.

Project 1 - Milestones

Milestone represent a specific type of inscription, typically made of stone and placed next to the road so the travellers could orientate themselves in the space. Milestones usually contained information on who built or paid for the reconstruction of the road, who was responsible for the maintenance and the administrative region of which city the road belonged. The most important information was the distance from the location of the milestones to the administrative centre (cities in charge or the road section, called ‘caput viae’), usually stated in Roman miles with the typical formula ‘milia passuum’, literally: thousand steps.

The location (geocoordinates) of milestones can help determine the course of the road, if uknown, or can help correcting the current course of the road, or alternatively, help discover new smaller roads. The mention of the caput viae can help determining the extent of the administrative regions, that are often unknown from any other sources. The date of an inscription provides an important piece of information on when the particular section of road was built or rebuilt and who was responsible and in charge of financing.

However, not all milestones were categorised systematically and the set of rules ‘what defines a milestone’ is often vague and based on small samples from geographically limited regions.

The purpose of the Project 1 is to find a subset of all milestones in the EDH dataset with 80,000+ inscriptions and on the basis of the subset determine a set of rules for future epigraphers to help them classify an inscription as a milestone. The set or rules will be then applied to find all milestones from the larger EDCS dataset, containing 500,000+ inscriptions.

The set of rules will ideally contain:

  1. the most common words occuring in a milestone, repeating formulae, etc.
  2. most common material
  3. typical range of physical dimensions etc.

FUTURE STEPS: The example of a milestone inscription will be later expanded to all types of inscriptions in order to create a handbook for epigraphers, see subproject 1.

Project 2 - Road use on inscriptions

According to a widely accepted theory, inscriptions were clustered along the roads or in their proximity. Their contents were often communicating with the travellers, evidenced by the frequent use of invocations for travellers such as ‘viator’ or ‘parodita’ in the text of inscriptions. Other inscriptions mention the traffic on the roads and their use by the army or by the local communities as well.

I have created a set of Latin vocabularies related to the use of roads, their construction and physical ammenities connected with the road traffic. I will search for the occurences of road-related terms in the text of the inscription in order to create a subset of inscriptions related to a road use.

I will then assess the usefulness of the method by spotchecking the text of inscriptions and the relevance of their text to the use of the road system. I will reevaluate the list of the vocabularies based on the findings.

I will analyse the frequency of topics occuring on the road related inscriptions, e.g. commercial or military traffic on the roads, individual travellers, physical ammenities, building activities etc. and compare them over space and time (by Roman Provinces, by centuries).

As a next step, I will overlay the subset of isncriptions with the known datasets of the Roman roads (e.g. Barrington Atlas of the Greek and Roman world, Pleaides) in order to detect any spatial clusters of inscriptions away from the course of known roads. This was a new (undiscovered) Roman roads could be detected, but also the relation between roads and spatial dispositions of inscriptions can be verified.

Authors

License

License: CC BY-NC-SA 4.0 CC-BY-SA 4.0, see attached License

Data

Epigraphic dataset - Epigraphic Database Heidelberg, see a series of scripts in EDH_ETL Repository used to access and clean the tabular attributes of the dataset and the text of inscriptions.

Tabular metadadata

Description of individual attributes and their source

  1. EDH dataset metadata
  2. EDCS dataset metadata

Input (training) dataset (EDH)

Public link:

https://sciencedata.dk/public/b6b6afdb969d378b70929e86e58ad975/EDH_text_cleaned_2021-01-21.json

Access with R (using custom sdam package)

resp = request("EDH_text_cleaned_2021-01-21.json", path="public/b6b6afdb969d378b70929e86e58ad975/", method="GET", anonymous = TRUE, cred = NULL)

Access with Python (using custom SDDK package)

!pip install sddk
import sddk
auth = sddk.configure("SDAM_root", "648597@au.dk") # where "648597@au.dk is owner of the shared folder
EDH = sddk.read_file("public/b6b6afdb969d378b70929e86e58ad975/EDH_text_cleaned_2021-01-21.json", "df", auth)

Target dataset (EDCS)

Public link: https://sciencedata.dk/public/1f5f56d09903fe259c0906add8b3a55e/EDCS_text_cleaned_2021-03-01.json

Access with R (using custom sdam package)

resp = request("EDCS_text_cleaned_2021-03-01.json", path="/public/1f5f56d09903fe259c0906add8b3a55e/", method="GET", anonymous = TRUE, cred = NULL)

list_json <- jsonlite::fromJSON(resp)
EDCS = as_tibble(list_json)

Access with Python (using custom SDDK package)

!pip install sddk
import sddk
auth = sddk.configure("SDAM_root", "648597@au.dk") # where "648597@au.dk is owner of the shared folder, i.e. Vojtěch
EDCS = sddk.read_file("SDAM_data/EDCS/public/EDCS_text_cleaned_2021-03-01.json", "df", auth)

Data output (public folder on Sciencedata.dk)

https://sciencedata.dk/shared/66cbabddae0e02c6ae6c15be9746990c

Repository structure

Folder data on Github contains mostly spatial data, such as the extent of Roman Provinces (DARMC) and the course of the known Roman Roads (Barington Atlas of Greek and Roman World).

Folder output contains visualisations and charts created by the scripts.

Folder scripts contains scripts numbered according to their related project and in the sequence the should run. Scripts are both in R and Python.

Scripts

Project 1 - Milestones

R script 1_1_r_MILESTONES_EDH.Rmd searches EDH database for all milestones as categorised by the type of inscription, type of object, and commentary. The script also provides basic overview of milestones from the EDH database, including their text, physical description, location and date.

HTML version: EDH_milestones

R script 1_1_r_MILESTONES_EDCS.Rmd searches EDCS database for all milestones as categorised by the type of inscription, type of object, and commentary. The script also provides basic overview of milestones from the EDCS database, including their text, physical description, location and date.

HTML version: EDCS_milestones

Python script 1_2_py_MILESTONES_research.ipynb - exploration of milestones, unfinished.

Python script 2_1_py_DECLINE_TERMS.ipynb creates full declension paradigms for road-related terms.

Python script 2_2_py_EXTRACTING_TERMS.ipynb selects inscriptions from the EDH dataset containing road related terms (using the full declension paradigm from script 2_1).

Python script 2_5_py_TEMPORAL_TERMS.ipynb explores the inscriptions containing road related terms from script 2_2 in time.