Gleanings from the 1253 Building Accounts of Westminster Abbey
By A. Richard Jones
AVISTA Forum, Vol.11:2 (1999)
Introduction: Between 1220 and 1266, Salisbury Cathedral was built at a cost of £28,000. My interest in recasting that figure in terms of labor and materials led me to the examination of a table, Men Employed at Westminster Abbey in 1253, in Jean Gimpel’s book The Cathedral Builders. That table contains columns for tallies of nine kinds of workers and a total column. Each row covers a time period in 1253, taking its start as 1 January: the first row covers eleven weeks; the remaining thirty-one rows, one week each. Some of the year is absent from the table, presumably because no work was done then: before 1 February; 19-27 April (Easter week); 9-15 June (Pentecost week: actually present, but with no entries); and after 6 December. The accounts from which this table is abstracted are among the earliest known complete records of a year in the construction of a cathedral or large abbey church.
A spreadsheet containing the numbers from Gimpel’s table immediately revealed that the totals are not equal to the sum of the worker tallies in the two weeks beginning 28 July and 17 November. This led to an investigation of Gimpel’s sources, which are H. M. Colvin’s two publications of a similar table. Unfortunately, Gimpel apparently chose the earlier one, not containing the corrections Colvin made in his 1971 Building Accounts of King Henry III. Better yet, Building Accounts contains the accounts on which the table is based. These are presented as the original account expanded into unabbreviated Latin, with an English translation on the facing page. It is instantly obvious that Gimpel’s table contains far less than half the account information: the actual accounts have money amounts that the table lacks, there is information about more workers, and each time period has a section for purchases that is absent from the table. The next step was therefore to create a computerized super table, or data base, incorporating as much of the account information as feasible.
The initial data base identified some numerical discrepancies, which could be either original or transcription artifacts. To avoid spurious conclusions based on the latter, it seemed a good idea to check the transcription against the original documents, which are held at the British Public Record Office (PRO). Six days were spent checking some 1,700 numbers against the originals at the PRO in April 1997. Robert Willis’s transcription of the 1253 account roll for 28 April to 6 December, published in G. G. Scott’s 1863 book, Gleanings from Westminster Abbey, was enormously helpful, because it was not expanded into unabbreviated Latin, and so was an interpretation of exactly what was on that roll. A second visit to the PRO in June 1998 served to reexamine some questionable numbers.
The original Salisbury Cathedral inquiry has not been further addressed, and the work on the Westminster Abbey accounts is not finished, but this paper describes some of what has been gleaned thus far by analysis of the computerized form of these accounts.