Abu Ghraib aside, these folks have really pushed the envelope on legal support. This is the approach the Justice Department used to take down big tobacco - managing millions of records, spanning decades, including faded photocopies, the affidavits of thousands of parties, and hundreds of lawyers spread across the country.
Most social scientists aren't likely to have that kind of capital anytime soon, but we might still learn from some of these techniques.
Check out "document conversion" and the ADIIS package. Very interesting stuff - high speed OCR, backed up with fuzzy search tools and a code-and-retrieve management system. Again, the style of argumentation may be different (rhetoric v logic), but the rules of evidence are similar.
What can we learn here?