An interesting article by Mike McCabe from Stewarts Law looking at two recent cases - Castle Water Limited v Thames Water Utilities Limited and Bouygues (UK) Ltd v Sharpfibre Ltd - each that provides guidance about: known adverse documents and narrative documents.  The common theme of both of these cases is the need for parties to act reasonably and proportionately.

These are key phrases that need to be adopted across any decision making process in respect of disclosure and should not be limited to known adverse documents and narrative documents.  However, turning to these two cases, in  Castle Water Ltd v Thames Water Utilities Ltd [2020] EWHC 1374 (TCC) there was a distinction between having to search for these documents and having to check with relevant people whether they were likely to exist.  This is replacing an automated search with one that relies on human memory.

I do not want to propose some ongoing requirement to continually search across the entire population of documents infinitude.  But I would imagine on most cases, with documents loaded onto a document management system, that people are already searching for these documents as part of their ongoing process.  I would also imagine that this would extend to when there are fundamental changes to the case.  So I would expect a check to be a search in many cases!

In Bouygues (UK) Ltd v Sharpfibre Ltd [2020] EWHC 1309 (TCC) the topic in question was narrative documents and how whether it was proportionate to order a search for them.  Narrative documents are generally accepted as being relevant to the background or context of the case and are not directly linked to the issues whereby disclosure is required.  

From a technical perspective, I think that narrative documents need to be considered in two ways: firstly, from a data collection and processing perspective - they are likely to be already collected unless these documents are stored in a specific location unrelated to other potentially relevant documents - or a very targeted collection was performed, which has other risks.  Therefore there is no real issue there.  The second perspective relates to searching for and reviewing documents that may fall into this category.  That does create time and cost issues as I would imagine it would be hard to be very precise in search methodologies where the target contains background / contextual information that could be common across many documents.

In this case, the judge was not convinced that narrative documents were “necessary for the fair disposal” of the case.  As Mike sets out: "There were 20 broad Issues for Disclosure in an expansively pleaded case, so ordering disclosure of documents that were not directly relevant to one of those issues would not “assist [the court] in the determination of the issues."

As set out at the start of this post, the common theme of both of these cases is the need for parties to act reasonably and proportionately.  These should be considered at every stage of the process and there are lots of ways technology and the intelligent application of it can assist.  Please let me know if you'd like to know more!