Web Databases

Without the use of databases, a project of this scope would be almost impossible. Databases are a means by which data can be stored in a structured format and retrieved, and it is they that do the real processing work of the project.

Several databases have been created throughout the life of the Project to store and process the information collected:

• Concrete Evidence
• Documentary Evidence
• War Diary
• Order of Battle
• Archive Documents

The diagram at right shows the development of the databases.

Concrete Evidence was created in 2006 to contain data regarding surviving structures; as the project developed it was clear that a new database was needed that could contain a wider array of data whilst incorporating some of the existing functions. In 2008 a second database, Documentary Evidence, was built as a development to contain data from documents.

At this point it was anticipated that the two databases would remain separate but potentially linked. However, the logic in not storing the data within one structure quickly became apparent despite one or two minor problems that merging would create.

Documentary Evidence therefore needed to be able to also list existing structures for which there are no documents and to also contain photographs like Concrete Evidence. This finally happened in late 2009, with limited online access of the new Concrete Evidence replacing the original format in 2010.

War Diary was another original 2006 database, containing extracts from unit documents describing the anti-invasion preparations of 1940. The regimental nature of these extracts required an Order of Battle database to support them. As the diary extracts needed to be related to their respective sources, Archive Documents was created as an index to all primary sources and the whole merged under the latter name at the same time as Concrete Evidence was reconstituted.

So as of 2010 there are now just two main databases:

• Concrete Evidence
• Archive Documents

As can be seen from the diagram, they are related; it is this relationship that generates the full dataset for each record in Concrete Evidence relating the evidence with the defence works.

Concrete Evidence database

The database is being compiled through the meticulous scanning of documents for mention of anti-invasion defences in order to identify their location and how they fitted into the overall scheme. The intention is to cross-reference each defence between documents as much as possible to build up a picture from construction through to use, creating a 3-dimensional picture of how East Sussex was to be defended in the event of invasion.

At this point in time, for a defence work to be included in the database, there must be at least one credible primary source that mentions it. This may be a document, a photograph, or the fact that the feature (or evidence of it) still survives in the landscape. At this moment in time, I'm not including features for which there is only German mapping evidence, as I'm wary of the accuracy of some plots I've seen; aerial photographs that clearly show a feature are, however accepted. German map data is currently only used when it is confirmed by Allied documents or other sources, or there is little doubt that the intelligence assessment was reliable. I'm also cautious about oral evidence and stories that have been passed down over the years, as experience so far shows that myths start to creep in with every new generation that retells the story. Secondary sources are not used on their own to establish the existence of a feature.

The evidence of the defence work (e.g. document extract) is quoted in the database record, and multiple sources can be associated with each defence. The reason this seemingly lengthy process is utilised is so that for each record, it is known how we know that defence work existed and what we know about it. In this way we trace the history and use of individual works.

Each record has a Historian's Comments field, in which discrepancies between sources or other points of note can be discussed.

Each record consists of the following fields:

• Record ID number
• Type of defence work
• Location (placename)
• Cassini Grid reference (can have multiple)
• National Grid reference (can have multiple)
• Historian's Comments
• Document references (can have multiple)
• Parent/child relationships (can have multiple)
• Serial Number (can have multiple)
• 'Status' fields to indicate:
  • Ownership/access (e.g., private property)
  • Whether feature is Ex-situ (e.g. roadblock cylinders relocated)
  • Whether a feature was proposed (may or may not have been built)

Problems in compiling Concrete Evidence

The largest problem is that of location; many defence works mentioned in the documents do not give a map reference but just a place name, making it hard to extablish exactly where the feature was. In some cases it has been possible to speculate the most likely grid square, in which case the Historian's Comments will indicate this.

There is also the potential for duplication of sites in the database. For example, a R.E. war diary may give a grid reference for a completed pillbox (say 123 456), but a later defence scheme may list a pillbox at a reference which is perhaps one digit different (perhaps 124 456). Now, are these documents describing the same pillbox (such discrepancies are common), or two different ones? This illustrates the need for careful and multiple references; another source (perhaps an aerial photograph) may be able to disentangle the confusion. In the meantime, two database entries are made to record the possibility that another defence may have existed.

Concrete Evidence as a processing machine

The database currently contains 2572 records and work is ongoing to input data; at least another 500 potential sites have been identified from documents awaiting the full extraction process.

Although the database cannot process the data and make historical judgements, it is an invaluable part of the analysis.

The sheer number of sites that it is anticipated will be contained within the database make it beyond normal human ability to store and retrieve the information in an efficient way.

For more about the database and statistics please see Concrete Evidence.