3.5 Datasets

3.5.1 Handlists and corpora

Typically, a handlist provides a simple listing of basic information on all the stones of a specific period and/or monument type within a defined geographical area. They may or may not be illustrated (e.g. Henderson 1987; Fraser 2008). They are useful in establishing the extent of a body of material and the research already conducted on it. These can usually be produced rapidly but possess significantly less detail compared to a typical corpus and do not contain any analysis.

A corpus is a complete survey representing the state of knowledge at the time of its compilation. It groups and categorizes material, typically through an advanced study of design and style (as opposed to materiality or other values). They tend not to consider context and landscape in any detail. Examples include Steer and Bannerman 1977; Allen and Anderson 1903. The strength of most corpuses is that they proved high quality, detailed and richly illustrated information. Data is presented in a consistent format, usually through a degree of comparative analysis. However, once completed, published corpuses are not easily revised or updated with new information. Digital publication, as an online resource, provides opportunities for regular refreshing of the data aiding comprehensiveness, and increased accessibility.

Historic surveys are useful as records of missing stones but there may be a finite amount of surviving data available to (re)record and (re-)analyse existing corpuses. There may be the perception that a subject area is deemed to have been 'done', which can stifle subsequent scholarship for a generation, as claimed for ECMS (Stevenson 1981a, 175). It is not always clear as to what questions were being framed by the data, however, the degree of consistency contained within most corpuses means there are still valuable opportunities for reassessment with new perspectives. In some cases there is a ceiling on studies that is likely only to be broken as the result of new technology to study design and materiality.

To date, many existing corpuses and thematic datasets (see below) have been driven by a single disciplinary approach, usually art historical. There are strong opportunities for such studies to be multi- and inter-disciplinary in character to enrich analysis and understandings.

There has been a bias towards documenting early medieval, rather than later medieval and modern, stones (e.g. Allen and Anderson 1903; Fisher 2001; see Section 2). However, there are some notable early medieval omissions, for example Anglo-Saxon sculpture in southern Scotland is excluded from the British Academy's Corpus of Anglo-Saxon Sculpture (except as comparative material).

3.5.2 Thematic datasets and studies

Thematic studies of carved stones include several different types of record sets. They can be incorporated in studies of other types of heritage assets such as church buildings, e.g. Scottish Church Heritage Research Places of Worship in Scotland project and the Corpus of Scottish Medieval Parish Churches, or graveyards, e.g. Scottish Association of Family History Societies Graveyards Inventory. In these cases information of carved stones may be difficult to access and not particularly detailed.

Thematic studies also include records only of carved stones but the method of compilation means they are not intended to provide complete coverage or drawn-together analysis (e.g. Willsher and Hunter's 1978 study of 18th-century gravestone carvings). In other cases thematic studies may look at just one facet of a carved stone design but intend to reflect the state of current knowledge in this area (e.g. Davidson's 1977 inventory of 17th-century gravestones in Angus and Mearns). As with corpuses, thematic datasets may not be easily updated or refreshed if published in paper form.

3.5.3 Audits

The scale of an audit is significant, involving tracing surviving monuments, as known through existing records as well as new discoveries not included in sources such as Canmore. Audits are often used to establish condition and to identify what has been lost. So far, there has been no national audit (and only limited regional examples) to measure and characterise what exists, or has been lost, on the ground. The scale of architectural, later medieval (with the exception of funerary monuments) and early modern carved stones in particular are an unknown quantity.

Without any form of overview, it is impossible to identify gaps in current knowledge and begin to fill these in strategically. While we know anecdotally that carved stones are a resource under threat, such as loose and vulnerable stones, particularly at church sites, the lack of an overview means we cannot understand the nature of their vulnerability in terms of the relationship between different types of carved stones, stewardship issues and the rate of loss. With an audit it becomes possible to assess both condition and risk, and values/significance to help to target the finite resources available for maintenance and repairs more effectively. This is especially pertinent in the case of graveyards (see Graveyard recording: Case Study 3).

3.5.4 Canmore

Canmore is Scotland's principal heritage dataset, representing over a century of accumulated information. Its key strengths include regular update of content, the facility to crowd-source information, inclusion of images for many entries and its design for use by a broad user-base spanning researchers, heritage managers and the wider public.

As a database, Canmore does not present any analysis of information that it collates and indexes from different sources but is a valuable 'building block' resource for researchers. However, the level of detail within Canmore (in common with other Historic Environment Records (HERs) and statutory list compiled by Historic Environment Scotland: HES) is variable and there is no standard format or terminology to describe the different types of carved stones. This variability coupled with the limited number of search terms considerably limits the database's search capabilities, particularly for groups of records (see Canmore upgrade: Case Study 30; Canmore upgrade example: Case Study 4).

Canmore does not include information on monument condition, ownership or statutory designations. Nor is it directly linked to other datasets containing such information, although the portals of PastMap and Scotland's Places provide access to HES statutory lists, HERs, National Library of Scotland (NLS) maps and selected records from the National Archives of Scotland (NAS). There are records kept (e.g. of condition, such as HES monument warden reports) that are not publicly accessible. There are lists of monuments classified for designation purposes (HES Decision Portal etc.), but the schemes used for this are very simplistic and outdated. Some information is downloadable in GIS format and usable for those with the skills do to so. The fact that these databases do not 'talk' to each other, and, especially with Canmore (which lists all sites), is a hindrance to research. So, even if Canmore terminology is updated, it still cannot be readily interrogated to establish what of these categories is designated and how.