There have been several attempts to produce a research agenda for Roman Scotland from 1948 onwards (summarised in Richmond in Hawkes and Piggott 1948, 56-62, 100-4; Hanson and Breeze 1991; Barclay 1997, 31-2). A fundamental requirement of the earliest such agenda was refining the dating of the periods of occupation and understanding better the nature of the occupation and the internal arrangement of forts. This was the impetus for the Scottish Field School for Archaeology's work on the forts of Birrens, Cardean, Castledykes and Strageath intermittently from 1951-1986 (Robertson 1964; 1975; Woolliscroft and Hoffmann 2006, 158-164; Frere and Wilkes 1989). The agenda has subsequently developed, with a growing appreciation of the need to know more about civil settlements and the impact on the local population of their existence. As knowledge accrued about the details of the building history of the Antonine Wall, new models of its form, purpose, chronology and development have been proposed and tested.
The advent of new techniques and knowledge allowed the agenda to be expanded. One of the most important developments was the expansion of aerial survey (see below), but palynology, was also important in allowing the exploration of the impact of the army on the landscape. Geophysics has made a more variable contribution, with poor results along the Antonine Wall in locating activity around forts (even where this was known from other evidence), but considerable success in revealing plans of forts and their surroundings in Bradford University's work at Newstead, the Roman Gask Project's work on forts north of the Forth-Clyde line, and Time Team / work at Drumlanrig (R F J Jones, pers comm; Woolliscroft and Hoffmann 2006; Hunter 2005, 401-2). There is also ongoing survey of the fort at Dalswinton, Bankhead by the University of Glasgow as well as collaborative survey work between the University of Glasgow, RCAHMS and the Römisch-Germanisch Kommission at Dalswinton, Bankfoot and Ward Law camp.
The most influential, concerted and sustained campaign of research was that conducted by J K St Joseph of Cambridge University through his aerial survey programme, which was augmented by selective excavation to test the evidence of the photographs. The distinctive appearance on plan of Roman military sites meant that they could readily be identified by aerial survey, the pioneers of which were mostly Roman archaeologists. Early aerial survey was undertaken by O G S Crawford, archaeological officer with the Ordnance Survey in the 1930s, with the mantle passed to St Joseph in the 1940s. From 1976, RCAHMS commenced aerial survey under the direction of Gordon Maxwell. This resulted in the discovery of camps, forts, fortlets, towers and roads (Crawford 1930; 1939; St Joseph 1951; 1955; 1958; 1961; 1965; 1969; 1973; 1976; 1977; Maxwell 1983; 1984a, 1984b; Maxwell and Wilson 1987; see Jones 2005). St Joseph sought to relate individual camps to campaigns, while on the Antonine Wall he revolutionised knowledge with the discovery of camps and fortlets. Since Maxwell's retirement in the 1990s, aerial survey has been continued by RCAHMS, although rarely targeted specifically to Roman sites; the Roman Gask project have continued Roman-focused flights north of the Forth.
A more geographically-oriented research campaign by RCAHMS has sought to improve knowledge of Roman monuments within each county it surveyed; until the 1980s this included some excavation work to test site function and chronology. Scrutiny of vertical air photographs taken by the Royal Air Force and the Ordnance Survey for the production of county inventories led to the discovery of several more sites (eg. Oakwood: Steer and Feachem 1954; RCAHMS 1957). Operating more behind the scenes, Historic Scotland sought to maintain and develop the research agenda (e.g. Barclay 1997), and to use the money and influence at its disposal to ensure the survey and excavation of appropriate sites. This led to major excavations and subsequent research, including Duntocher, Barburgh Mill, Bar Hill, Croy Hill, Camelon, Bearsden, Elginhaugh, and Inveresk. Within local authorities, Falkirk is notably proactive in seeking to discover more about the Antonine Wall (eg. Bailey 1994; 1996).
Little active rescue work or research excavations are now undertaken directly by Historic Scotland. The protection afforded to Roman sites through scheduling has resulted in less rescue archaeology in the interiors of military sites. However, since many such sites lie under or near modern conurbations or development zones, continuing development leads to regular, generally small-scale interventions in the environs of forts, while large-scale road and housing schemes have led to the investigation of significant swathes of temporary camps at Monktonhall (Inveresk) and Kintore (Hanson 2002a; Cook and Dunbar 2008). A major advantage of developer-funded archaeology has always been the nature of the work leading to fortuitous discoveries (for instance in a greater understanding of the environs of Cramond and Inveresk forts; e.g. Bishop 2002a, 2002b; 2004; Cook 2004; Britannia 35 (2004), 269; Britannia 42 (2011), 333-4, 441-4; Leslie forthcoming; Masser 2006).
In recent decades, a few projects have sought to look at broader regional pictures. The Roman Gask project has re-evaluated the Roman and contemporary indigenous landscape north of the Antonine Wall through a combination of aerial and geophysical survey, field-walking and excavation (eg Woolliscroft 2002a; Woolliscroft and Hoffmann 2006). The Bradford University/Borders Region Newstead project had an ambitious research design to look at the interaction of Rome and the local population (Jones 1990); publication is awaited.
In the higher education sector, Glasgow University has played a significant role through publication, excavation and its support of research at doctoral and post-doctoral level; work at Edinburgh University has focused on sites relating to the indigenous population. Other Scottish universities have no recent involvement in the study of Roman Scotland. The National Museum has adopted a proactive role, supporting research at Newstead in order to contextualise its existing major collection from the site, and investigating the find spots of new discoveries such as the Birnie coin hoards (e.g. Manning 2005; 2006a; Hunter and Keppie (in press); Hunter 2007a, 2007c).
There have been some significant works of synthesis. In 1949, O G S Crawford published Topography of Roman Scotland North of the Antonine Wall, bringing together antiquarian references and archaeological research; further overviews and general works have collated and discussed the wider story of Roman Scotland for academic and wider audiences (these include Hanson 1991a; Maxwell 1989; 1998; Woolliscroft and Hoffmann 2006; Breeze 1982 ; 2006b; Fraser 2005; Robertson and Keppie 2001). South-western Scottish work was pulled together by Miller (1952). Inscriptions have been well-served by British-wide corpora (RIB I, II, III) and sculpture by the Scottish volume of an international series (Keppie and Arnold 1984); the Hunterian Museum's collection of sculpture and inscriptions has been subsequently updated (Keppie 1998). A major survey of temporary camps has very recently appeared (Jones 2011), while Roman finds from indigenous sites have seen fresh syntheses (Robertson 1970; Hunter 2001, 2007a, 2010; Wilson 1997, 2001, 2003, 2010).
Some items on previous agendas remain elusive, such as the location of civil settlements outside forts. There is now perhaps a greater appreciation of the intractable difficulties in answering some research questions owing to the nature of the archaeological evidence. The dating of temporary camps is notoriously difficult, and many remain without any so far retrieved dating material. The paucity of artefacts on the indigenous settlements has traditionally rendered it difficult to determine any effect of the Roman occupation on the local inhabitants, although substantial progress has been made, and the range of approaches and perspectives expanded. In summary, there has been no concerted attempt to produce and follow a single research agenda. Rather, the research questions have widened as more techniques and knowledge have become available and thus the scope of individual projects have also expanded. The role of the relevant central and local government bodies has been to comprehend these widening horizons and ensure that threatened sites are examined within the ever-growing research framework. Much research has also depended upon the personal interests of individuals working in Scotland. In the past, the study of Roman Scotland has been primarily focused on its military remains, but this has gradually widened to encompass investigation of the contemporary indigenous population, as well as seeking to use the evidence for Roman Scotland to help the study of the periods both before and after, providing a dated horizon which has aided and affected later prehistoric studies and offering a vital facility for testing models, as John Barrett has noted (1997b).
The study of Roman Scotland continues to develop, with changes in concepts such as 'Romanisation' and 'resistance', with the result that interpretations have become more subtle. In particular, the Roman army is now viewed as occupying a more proportionate place in a landscape which was principally peopled by indigenous farming societies. Understanding of the period is subject to all kinds of gaps and problems which with the richness of the data-set promise very considerable labour and, hopefully, rich rewards for the future researcher.