Domesday Explorer can best be described as an electronic edition of Great Domesday Book (GDB). Domesday Book is an abstract of the survey ordered by William the Conqueror at Gloucester in 1085. Little Domesday (LDB) is an expansive account of the three eastern counties of Norfolk,Suffolk, and Essex, GDB a more abbreviated version of similar returns for the remaining thirty-one counties of Anglo- Norman England. It was first edited in 1783 and since then there have been a further four editions and translations. The size of the source is daunting and hitherto historians have largely depended on indices of varying quality for access to the materials. That has now all changed with the publication of this, the first machine readable and searchable text of GDB. By and large Domesday Explorer is a triumph.
Like all projects that have been more than twenty years in the making, it comes with a history and that history sometimes weighs heavily. Domesday Explorer is based on the Phillimore translation of Domesday Book. Its founding editor, the late John Morris, set out to render the Latin of the original in modern English with results that have raised not a few academic hackles. There are, however, more serious problems with the edition for anyone who seeks to use it as a base text for electronic retrieval. In the early volumes repetitive phrases were omitted from the translation, the principles of place-name identification change in the course of the series, and personal name forms, especially of pre-Conquest holders of land, are inconsistent. Domesday Explorer has largely maintained the integrity of the edition (only the names of major figures like Merleswein/Merlesveinn have been regularized) and in consequence all references to individuals cannot always be identified unless the different forms are already known. If, for example, you are interested in Barthi in Lincolnshire, you have to know that the same name (and individual) appears as Bardi in Northamptonshire and Leicestershire.
The text presents further problems. The Phillimore edition chose to reproduce Farley's Record Type transcription. Domesday Explorer rightly opts for a facsimile, but copyright conditions have dictated that it is the Ordnance Survey version of 1861-3 which is reproduced. That edition was made with, for the time, a revolutionary process called photozincography, but the result is less a photograph than a silhouette, and rubrication had to be overprinted from a hand- prepared plate. The result cannot compare with the modern facsimile published by Alecto Historical Editions.
In a perfect world the authors of Domesday Explorer would not have chosen these limitations. They are, however, in no way fatal to the enterprise. Those who need to know the subtlety of technical terms will trouble to find out. After all, soca et saca, for example, is now as intrinsically obscure as the 'full jurisdiction' of Domesday Explorer. Different name forms are more of a nuisance but are confined to the later volumes of the Phillimore edition, notably Lincolnshire and Yorkshire, and due adjustments can be made. Again, the Ordnance Survey facsimile may not be up to the demands of the finer points of paleographical analysis, but it is perfectly adequate for most purposes. It reproduces in fine detail much of the stratigraphy of the text; indeed in some cases it is easier to see differences in the hand than in the more modern facsimile or even the MS. Personally, I am very pleased to have the edition along side the Alecto Facsimile and I use both. What is novel and important about Domesday Explorer is the way in which text and translation are linked to analytical, mapping, and tabulation engines.
By default the program opens with three windows. On the left there is a record of search terms and to the right the translation of the selected text mapped entry by entry to the Domesday facsimile. Whatever passage is displayed, the correct lines of the relevant folio are indicated by an outline blue box. As new search terms are entered, the user can either skip from result to result or view each within the context of the folio in which it is found; modifications can be made to the selection as appropriate. Once the desired information is found, views can then be changed. Data can be presented as a table or, perhaps most revealingly of all, on a dynamically linked map. Any combination of views can be chosen and each supports the export of the information in standard formats to other programs such as word processors and databases. Interlinking--even between dots on the map and the translation --ensure that you are never more than one click away from the text. A series of searches can be saved for later reference as a 'project.'
At the heart of the system is a vast series of invisible tags or codes. Early attempts to computerize Domesday filleted the text by extracting 'the facts' from their context and presenting them in arrays. The result was a highly interpretative and largely unrealizable database: none was ever completed. Domesday Explorer eschews this approach. Rather it retains the whole of the text in translation, but inserts codes so that entries can be easily identified and called up when needed. Text tagging in this way is, of course, equally interpretative: what is flagged up and what is not depends on the understanding, interests, and energy of the author. It is inevitable, then, that there are lacunae. But the most superficial of glances at the manual will show to what lengths the authors have gone to categorize each word of the text. There are, for example, twenty-eight different codes to characterize 'landowners' independently of the land they held-- whether TRE, TRW, overlord, undertenant, etc. There are likewise series of codes for entries, counties, circuits, boroughs, administrative units, editorial additions, manors, manorial statistics, and themes. It is indicative of the scale of the coding that where Robin Fleming identified 3217 entries with a legal import for the whole of Domesday Book in Domesday Book and the Law, Domesday Explorer codes almost twice as many for GDB alone.
Looking up words and phrases is simply a matter of typing them into the search box and specifying the range--county, circuit, or section of the text--with all the usual aid of wild cards, Boolean operators, Venn diagrams, and the like that one might expect. Coding allows more complex searching. Entries can be retrieved or excluded according to any of the codes, singly or in combination, contained within them. The sequence \T66=TW3, for example, will recover the 371 instances in GDB in which the TRE holder of land was a tenant (but not a tenant-in-chief) of the same land in 1086. Entry by entry what can be retrieved and in what combination only depends on one's mastery of the codes. What cannot be retrieved is arrays of statistics for further numerical analysis. Social and economic historians will have to extract figures themselves until such a time as a statistical function becomes available.
Anything more than simple searching is not for the faint- hearted. Putting together the codes with confidence is a complex matter. The on-board manual is not very helpful here: all the codes are set out, but there is little indication of how they can be combined to achieve the really clever searches. The excellent Domesday Explorer support website (http://www.domesdaybook.net/) is more useful, but functionality would have been greatly increased had more user- friendly search forms been implemented in the program itself. An interim measure might be a comprehensive on-line library of functions. I cannot be the only person who regularly wants, for example, to gather together all the estates in a single hundred. There are many similar common operations that users will regularly want to perform without the great brain-strain at present incurred.
Lack of such guidance is a serious gap in the documentation. Otherwise, the manual affords an excellent account of the background to the survey. The entry-by-entry notes of the Phillimore edition are not reproduced, but an unparalleled guide to the state of Domesday studies is provided. There is an overview of the Domesday process and current ideas on its purpose and execution. This is complemented by an extended glossary of terms which acts as a general commentary on the text. At each stage a comprehensive bibliography is furnished, with especial regard to recent publications, which is a valuable update to the David Bates' Domesday Bibliography of 1986. The Domesday ingenue could hardly find a better place to start on a study of a notoriously difficult text.
The program is written for the PC with a Pentium 166 MHz or greater processor, running Win 9x, 2000, or NT 4.0 (some quick experimentation suggests that it works on XP too), and requires 64 MB of RAM, 150 MB hard-disc space, a 24-bit video display card, and a SVGA (800 x 600) or higher resolution monitor. It installed cleanly to my hard drive and runs without any conflicts. It is generally stable, although some procedures, such as editing a search string, can produce system errors and occasional crashes. On my somewhat antiquated machine the response time for even complex searches was impressive, with waits to display information rarely exceeding one second. The user interface mimics the familiar Windows Explorer, but for one whose mouse is not his best friend, support for keyboard shortcuts would have been welcome. It seems natural to hit return to enter a search term, but nothing happens until the OK button is pressed. Most functions run smoothly, although on the version supplied for review not all the codes had been implemented or they had been blocked in some way. No doubt bugs like these will be fixed in due course. There has already been an update to the Help file. Work is on-going and a Version 2 incorporating LDB is promised in the future.
It is inevitable that Domesday Explorer will now be compared with Alecto's Digital Domesday, published in 2002, which incorporates both volumes of the Domesday text in a standard and standardized translation and the modern facsimile. Both share a common text, it is true, but otherwise they are very different products. Digital Domesday uses a largely untagged text and only supports simple word and string searches. It is basically a convenient electronic index to the Alecto edition of the Domesday text and its critical apparatus. Domesday Explorer, by contrast, is something more. It is, no less, a heavy duty analytical machine. Once its coding system is mastered, it is capable of facilitating analyses that are otherwise practically impossible. This is the great strength and beauty of the system. Without doubt it will prove itself an indispensable tool in all future Domesday studies.