The army used different means to obtain supplies (Breeze 1984). They might make their own goods in central workshops, such as arms and armour. The provincial army, or an individual unit, might purchase supplies at fixed rates. In the East, orders are known to have been placed for items with suppliers in another province. Documents indicate that the army provided the soldier with certain items, including for example a share of the tent. But he might also purchase items of equipment or food for himself or ask his family to send them. It is possible, at least in the second century, that he purchased pottery vessels, as well as other goods, in the civil settlement outside the fort.
The Roman world was sophisticated and the pattern is complicated, and changed through time. In Scotland, there is a general trend from the army supplying itself in the first century, to a greater reliance on civilian suppliers in the second century; there is also a shift from reliance on more imported material to local production. The resulting pattern is complicated with soldiers, according to documents from elsewhere, cutting stones at quarries, travelling considerable distances to collect supplies and escorting supplies, and civilians delivering goods to forts. Some of these activities would have led to soldiers being away from their units for considerable periods, thereby reducing their operational efficiency.
Insights into economic activities can come from a range of sources. Coin supply and use is one which has received attention (Robertson 1978; Abdy 2002), and as datasets grow this will remain an important area for analysis.
A major research requirement is to improve knowledge of supply patterns and their development with chronology and geography: for example, does material from forts placed on readily-accessed waterways differ from those served by road? Evidence is plentiful; a review of the most ubiquitous artefact, pottery, would help, as it serves as a proxy for other materials.
For such analyses to be easier, pot assemblages should be analysed following the standards of the Study Group for Roman pottery (http://www.sgrp.org.uk/07/Doc/Contents.htm).
A research project for comparing patterns of imports between the two walls would be valuable in revealing similarities and differences.