In content, electronic documents are no different than paper documents. All sorts of documents are subject to discovery electronic or otherwise. Legally and technically, there are substantial differences between the discoveries of the two media.
“Some 93 percent of documents are now created electronically, according to industry reports. And 70 percent of documents never migrate to paper.1” No matter what the legal status of discovery of electronic documents will be, the prevalence of electronic documents makes them a major discovery issue.
The following is a list of discovery-related differences between electronic documents and paper ones. We assume that a paper document is a document that was created, maintained, and used on actual paper; it is not a hard copy of an electronic document.
This point is obvious to the majority of observers. Today’s typical disks are at several dozens gigabytes and these sizes grow constantly. A typical medium-size company will have PC’s on the desks of most white-collar workers, company-related data, accounting and order information, personnel information, a potential for several databases and company servers, an email server, backup tapes, etc.
Such a company will easily have several terabytes of information. Accordingly2, such a company has over 2 million documents. Just one personal hard drive can contain 1.5 million pages of data, and one corporate backup tape can contain 4 million pages of data. Thus the magnitude of electronic data that needs to be handled in discovery is staggering. In most corporate civil lawsuits, several backup tapes, hard drives, and removable media are involved.3
Paper documents can be ledgers, personnel files, notes, memos, letters, articles, papers, pictures, etc. This variety exists also in electronic form. But then spreadsheets are way more complex than ledger, for example. They contain formulas, may contain charts, they can serve as databases, etc. In addition to the additional information, e.g. charts, the electronic spreadsheet supports experimentation with what-if version the discoverer may want to investigate.
To demonstrate the variety possible in electronic documents it sufficient to consider the most ubiquitous of them: the text document. A Word4 document may contains:
An active spreadsheet
Links to Web address
Proliferation of new devices such as Personal digital assistants, pocket PCs, palm devices and BlackBerry devices adds more variants of electronic documents and increases the responsibility of discovery.
Computers maintain information about your documents, referred to as “metadata,” such as: author’s name, document creation date, date of it last access, etc. A hard copy of the document does not reveal metadata, although certain metadata items may be printed. Depending on what you do with the document after opening it on your computer screen, the actions taken may change the metadata collected about that document. Paper documents were never that complex.
Text documents allow you to pick fonts, use colors, use shade selectively, use watermark and change the background and text. Spreadsheets allow one to selectively display rows and columns, hide formulae and write complex macros. Many other document types have similar and additional attributes you may employ.
Attributes such as hiding parts of the document are significant to discovery that may tries to be informed about the hidden parts.
Document efficiency is not a standard term. Here, Document Efficiency means factors such as:
Use of less space
Easier to change
Cost of delivery
Faster to search
Personal file systems are physically smaller than a small cell phone
stored locally in filing cabinets5
Ease of change
can be edited, copied, modified and merged with almost complete ease
requires in most cases recreating document all over again
Delivery speed and cost
by networks, disks, flash memory and CD/DVD
by mail or manually
multiple users may access documents simultaneously
multiple users to access documents simultaneously one needs a set of documents per each accessing person
Document complexity is used quite widely in literature and industry. Recent work deals mainly with XML documents that do not directly pertain to this discussion. The presentation6 fits our needs. Documents complexity is the sum of item complexity and format complexity.
Item complexity is defined as the sum of items within a complex document. An item is discrete, discernable object associated with a document. For example, the abstract, the content table page, multiple content pages, and additional items like photographs, audio and video.
Format complexity is defined as the sum of all formats within a complex document. A unitary format document contains only one type of file encoding. A binary format document contains two types of file encoding.
There are other ways to define document complexity7, but the one above approach works well for us.
Using the document complexity make abandonedly clear that electronic documents have more items and more formats and, therefore, are more complex.
Paper documents are easy to destroy. They may be throwing away, shredding, burned, lost or stolen. Once such acts take place the documents disappear. Deleting an electronic document eliminates only the ubiquitous accessible copy. The document, i.e. its data, still exists and in systems such as Windows and Mac OS, an accessible reference to deleted documents may be in the trash bin. Restoring a document in the trash bin, i.e. a deleted document, revives the document to its original glory.
Even removing the document from the thrash bin does not erase the documents data off the disk. Once removed from the thrash bin, documents data areas on the disk go into a “fee list” that makes those areas available for future data creation needs. The free list contains all areas not currently allocated to active documents as well as to deleted documents still in the trash bin. How long will an area stay on the free list (thereby still containing the deleted documents data)? That is difficult to predict due the huge variability of factors such as: future demand for disk space, size of current and future files, the current availability of disk space, etc.
Even the complete deletion of a document, its trash bin instance and the allocation of the document’s data area on the disk does not typically extinguishes the document altogether. Certain habitual practices create copies of documents and are only marginally affected by document deletion:
Backups – most organizations and individuals regularly create back up copies of documents as precautionary actions. The backups are maintained independently of the document itself.
Documents may be exchanged by email, access through web pages and manually handed electronic copies. Thus copies continue to exist after the deletion of the original document.
Even work on a simple text document is quite frequently preceded by creating a copy of the document being edited. Once again, such copies persist beyond the deleted document unless specifically deleted.
Changes to an electronic document are fast and easy. The reason is obvious; all you need to do is make the change and save it. Changes to paper documents, however, require retyping the whole document.
There are many other reasons to the difference in speed and frequency. We already said that documents may be dynamic. Web pages are made dynamic in order to ease change.
For discovery, faster and frequent changes imply a need for a more meticulous and length monitoring of document discovery.
Paper deteriorates with time; paper documents can be destroyed by flood and fire. Although these factors have their parallels in electronic documents, e.g. a flooded computer loses its data; typical backups of the documents practices maintain copies away from the “office.” Paper documents may enjoy the same treatment, but the frequency, extent and usage of such backups is substantially lower.
Electronic document suffer from upgrades in technology. If one used a peculiar word processor, e.g. WordStar, to write a document 20 years ago, today it will be difficult to convert the document to current word processor, but a tool to convert the document can be located. Same holds for spreadsheets, databases, etc. Again, most companies have practices that avoid such problems by evolving documents with time.
There are several levels of redundancy to electronic documents.
Due to the type of recording used for electronic data, minor errors in a document can be corrected by existing tools. The tools rely on the redundancy of checksums and other devices. MS Word tries to recover defective documents.
Due to frequent changes in documents, individuals learn to save previous versions of the documents. Doing that generates redundancy of document versions.
Emails, flash memories, CDs all proliferate documents and result in high redundancy. One copies documents to flash memory, attaches a document to an email to a fellow worker or create a CD for distribution or archiving.
Most companies and many individuals backup documents regularly. Studies show that “about 70% of enterprises meet the criteria of verifying the integrity of their backup media at least weekly.8”
Tools to control versioning of files create built-in redundancy wherever they are applied. Versioning, i.e. version control9, widely used by the software industry has started to infiltrate word processor as well as other applications. Versioning, by its very definition maintains several versions.
MS Word supports “Document Collaboration.10” Where this term implies: “new objects, properties, and methods of the Word 10.0 Object Library shown in this article allow you to change the display of revisions and comments, accept and reject revisions, and start and end a collaborative review cycle.”
Another tool, Workshare 311, is an add-on to Microsoft Word that manages collaboration on Word documents and integrates this activity with email and the organization’s document repository tool.
Collaborations on databases (e.g. people using a bank’s ATMs update the bank’s database), spreadsheets (e.g. BadBlue12), and Web sites are commonly practiced.
This dwarfs collaborations on paper documents.
For discovery it implies that the author of a Word document may not be the only person involved in writing the document. One has to determine all the parties that collaborated on the document.
Paper documents are always written by human beings. That is not necessarily the case with electronic documents. We start with a simple, and rather common, example. The Quicken financial program can generate financial reports from a database of financial transactions.
This is an application generated document.
Using MS Word and its Autosummarize tool on a large document we got:
Patient Monitoring Techniques in Telemedicine
Through the leverage of these devices we can formulate distributed algorithms and create effective data structures to properly monitor patients. Every patient will have very specific needs and we need a real time system to properly monitor the status of every single patient.
Each individual patient will be uniquely identified with a combination of building, floor, room, and patient id. Senior Citizen Patients Monitoring Tree
Lastly, each room contains one patient.
The objects could be customized to contain all pertinent monitoring information of each respective patient. Our goal is to formulate a Medical Object Query Language (MOQL)
The medical devices can interface with each object api to continuously update each patient object (MP). Research Goals
The tool created the document within the box. In this case, discovery has to find the person that wrote the original document. That is not necessary with paper document.
The large volumes of data, its complexity, its variety of electronic documents have brought about many types of computer tools to help overcome the obvious difficulties.
Socha Consulting13 provides the following entries in its Tools section (we drop the commercial part and use just the generic description):
Electronic discovery software; allows users to evaluate and manage electronic documents
Automated litigation support software; allows users to organize, search, and retrieve e-mail with attachments
Open, view, print and convert various files types
Review, acquire and analyze digital information on individual machines or across a wide-area-network
View and access contents of various file types
Automated litigation support software; allows users to process electronic files
Electronic documents benefit from a large variety of search tools. Search goes through far more documents than human beings could review manually. Different techniques provide a rich set of options starting from keyword search, proximity search14 and semantic searches15. For discovery, this search potential end up producing results.
“Electronic data, unlike paper data, may be incomprehensible when separated from its environment.”16 The critical question is what is meant by environment. The report of the Sedona Conference takes environment to be the actual software structures used by the document. They say: “[i]f the raw data (without the underlying structure) in a database is produced, it will appear as merely a long list of undefined numbers. To make sense of the data, a viewer needs the context that includes labels, columns, report formats, and other information.” Actually, given just the numbers from a paper ledger without the labels and tags is quite meaningless as well.
Environment as in the folder in which a document resides can potentially influence the document content. Some documents are made Lego style. That is, the master document consists of independent sections, i.e. small identifiable documents that are brought together by linking. (Web pages tend to be thus constructed.) Once the master document moves to another folder, the links, or some links, may be severed resulting in a different document than intended.
Software serves as a good example for environmental dependency of documents. Paths, Include files and their location, location of executable files are involved in developing and testing programs. If any one of the elements is misplaced or wrong modified, the development process suffers.
Above, we mentioned text documents written with WordStar17. Although organizations undergo migrations of applications, platforms, methodologies and practices quite often, today’s technological mind set mandates keeping electronic resources up to date or ascertaining that tools to convert these resources from their old form to the new form are readily available.
The danger to discovery due to migration is limited and typically solvable. For instance, although WordStar documents may be 20 years old, the marketplace provides tools to convert the document into the latest MS Word version. After all, one can easily locate spare parts for a 60s Beetle.
Discovery does face difficulties due to old technology, but this stems mainly from legacy systems18. Large organizations or companies with huge investments in information technology found it too difficult to move on to newer technologies. Thirty year old computer systems, though clearly archaic in technological terms, are not uncommon. Discovery may have a handful with such systems. Expert may be difficult to find, discovery tools do not work on the legacy systems and, sometime almost unbelievable yet true, even the owning organization does not really know much about their system19 (all they know is input and output). At the very least, discovery will be expensive.
A claim is made that the ease and flexibility with which electronic documents are created, copied, moved and managed tends to result in too many copies of the document or pieces thereof. When one contrast that reality to paper documents, without that ease and almost costless space resources, it seems like moving from a disheveled office to one neatly organized. Obviously, the mess is “not good” for discovery.
Clearly, this is a potential problem; we do not have research results that help us know whether it is a problem or just an annoyance. Multiplicity and disorder in document management is not the only price an easy to use technology extracts. Following is a list of difficulties we tend to encounter:
Use of sophisticated document features backfires. For instance, word processors support use of macros. (A macro is a series of commands that is recorded so it can be executed later.) An uncontrolled use of macros may yield unhealthy, shaky and difficult to use documents.
Documents may be part of a set of document where the set has functional significance. Moving a file away, i.e. deleting the document from the set, may damage the set. For instance, a software product may come with a: installation guide, user guide, reference guide and a demo scenario. Removing one of these documents may make the product difficult to use.
Sets of documents may have their members spread over a network of servers in diverse geographical locations. A change in one of the locations may spell trouble.
Collaboration in document production and maintenance is typically encouraged. Yet, collaboration has obvious pitfalls. Coordination, agreement, accountability and scheduling are all supporting productivity and source of complication
The sky doesn’t get darker and electronic documents are not going to be replaced by paper documents in the foreseeable future. Cars kill more people than horses and buggy. We learned to enjoy the car and never compare it to old animal technology. In summary, it’s a problem but not a major one.
In the chapter dedicated to ED tools, we will discuss tools in a generic way and demonstrate their functionality.
1 Mary K. Pratt, Discovery channel: E-documents can mean headaches, Boston Business Journal - November 28, 2005.
2 High-Risk Insurance Company Reduces Risk Of Losing Documents, Business Solutions, March 1998, http://www.businesssolutionsmag.com/Articles/1998_03/980324.htm
3 Linda G. Sharp, The complexity of electronic discovery requires practitioners to master new litigation skills, Los Angeles Lawyer, October 2005, Vol. 28, No. 8.
4 Mary Millhollon and Katherine Murray, Microsoft Office Word 2003 Inside Out, Microsoft Press; Book & CD-Rom edition (November 5, 2003)
5 Content Management, Ryerson University’s Open College unit, Xerox Process Study, 5/30/2001.
6 Laura J. Smart, the Evidence Threshold of Complex Digital Documents, a Research Proposal, March 20, 2001.
7 Davis, B., & MacLean, M. (1998). Mapping the project, grasping the consequences. In M. MacLean & B. Davis (Eds.), Time & bits: managing digital continuity. Los Angeles, CA: J. Paul Getty Trust.
8 How many businesses are ignoring common sense security advice, TheInfoPro, Inc. February 2004.
9 Dave Thomas and Andy Hunt, Pragmatic Version Control Using CVS, The Pragmatic Programmers; 1 edition (September, 2003)
10 Lisa Wollin, Creating Custom Solutions for Document Collaboration, Microsoft Corporation, April 2001, Applies to: Microsoft® Word 2002.
11 Martin Langham, Closing the Collaboration Gap, IT-Director.com, September 2003. http://www.it-director.com/article.php?articleid=11205
14 Robert Krauthgamer _ James R. Lee, Navigating nets: Simple algorithms for proximity search, 11/3/2005.
15 R. Guha, Rob McCool and Eric Miller, Semantic Search, 12/2/2002.
16 THE SEDONA PRINCIPLES: Best Practices Recommendations & Principles for Addressing, The Sedona Conference, January 2004.
17 Mary Bellis, A Rising WordStar - The First Word Processor - Seymour Rubenstein and Rob Barnaby, Inventors of the Modern Computer, about.com
18 William M. Ulrich, Legacy Systems: Transformation Strategies, Prentice Hall PTR; 1st edition (June 15, 2002)
19 Giuseppe Visaggio, Comprehending the Knowledge Stored in Aged Legacy Systems to Improve their Qualities With a Renewal Process, 11/26/1997.