Records Management for Electronic Mail

Eron Main
Faculty of Information Studies
University of Toronto


© 1998 Eron Main emain@ermine.ca, Faculty of Information Studies, University of Toronto
Katharine Sharp Review ISSN 1083-5261, No. 6, Winter 1998 [http://edfu.lis.uiuc.edu/review/6/main.html]

    Organizations are increasingly using electronic mail (email) for internal communications. Such communication may be replacing more formal written systems which supported filing and records management. While there is a body of opinion which holds that email does not produce records, cases such as Armstrong vs. Executive Office of the President (1993) have upheld the existence of email records. Organizations using email may not have systems in place to ensure the capture of vital records produced in this new medium. This article will discuss the status of email messages as records, examine the records management challenges and advantages provided by email, and offer recommendations for the effective management of email records.


    INTRODUCTION

    You know, everyone really uses this thing for electronic mail.
    --Bob Kahn on the ARPANET, 1972 (cited in Hafner and Lyon 1996, p. 186)

    By far, the greatest use is communication by electronic mail.
    --Lyn Robinson and David Bawden on the Internet, 1995 (p. 21)

  1. That these two almost identical statements could be made almost 25 years apart makes the importance of electronic mail (email) abundantly clear. Kahn was speaking to an associate at the first public showing of ARPANET--the network that was to grow into the Internet. The "everyone" he referred to was the group of computer scientists, programmers, and networking experts who built the Internet and first used it. Robinson and Bawden were addressing information managers on the subject of business use of the Internet. Two disparate user groups, a quarter century apart, found one feature of networked computing useful above all others: communication via email.

  2. Communication between individuals and groups, based on the exchange of information between people, is the great strength of computer networking. This raises several questions for information managers. Does this communication result in the creation of a record? Are all email messages records? If so, what is to be done with such records? Are current records management practices adequate to deal with email records? If not, what should be done?

  3. While these questions do not have simple answers, finding those answers is vital. In many organizations, email is becoming the most widely used medium for the exchange of information. If that exchange produces records which are not adequately dealt with, organizations will lose that information, with possible legal and operational repercussions.


    RECORDS

  4. For the purposes of records management, what is a record? The Association of Records Managers and Administrators (ARMA) defines a "record" as "recorded information, regardless of medium or characteristics" (Robek, Brown, & Stephens, 1996, p. 4). Other definitions (e.g., Menkus, 1996) refine this to indicate that a record must be related to an event or transaction, which the record then gives evidence of. "Transaction" refers to a business transaction of some sort--an agreement or a decision relating to the day-to-day business of an organization. Duranti, Eastwood, and MacNeil (1997) describe a record as "testimony, produced on a medium in the course of practical activity, of facts taken into consideration by the rules recognized as binding by a social group" (tem1.htm). Definitions are sometimes further refined to include the types of information which must be recorded for the record to be complete or reliable. Such information, while significant, may not always be recorded when a record is created, but is instead captured by the records management system.

  5. For this article, a record will be defined as information recorded in some lasting form, which provides evidence of an event or transaction. The term "lasting form" implies no set format or duration; it simply means that the information is not transient: some positive action must be taken to destroy a record once it has been created. This definition is general enough to include business-related correspondence conducted via email, without going so far as to capture every email message. This is necessary because of the widespread use of email systems to carry personal as well as business correspondence.

  6. This notion of lasting form implies the need for rules regarding the retention and disposal of records. In records management practice, retention requirements must be specified for all records. These detail which records must be kept in what forms and for how long. Once kept, records cannot be destroyed except in accordance with their retention schedules.


    ELECTRONIC MAIL

    Definitions

  7. There are a wide number of definitions of email (Du-Rea & Pemberton, 1994; Bearman, 1994). Some provide a narrow definition, encompassing only the transmission of text by certain electronic means. Others extend the concept to include all information which is transmitted electronically, including electronic data interchange, fax, and voice mail. For the purposes of this article, email will be defined as a computer-based system which permits the transmission of messages from an individual to one or more recipients. These messages may include text or other forms of data, either in the message body or as attached files. For an email message to be transmitted from one individual to another, all that is required is that they be using compatible systems which are connected by a network or group of networks.

    Characteristics

  8. While there are a variety of email systems in use, all share some common characteristics. The basic unit of email is the "message," a single transmission sent by a single individual. Messages may be sent to one or more recipients; each recipient receives a separate copy of the message. Most systems can be configured to retain a file copy for the originator. A message can be broken into three components: the header, the body, and the attachments.

  9. The header provides routing and identification information about the message itself. Most important is the network address of the recipient. This must specify a particular computer account at a specific network location. The account specified does not have to belong to a specific individual; accounts such as "root" or "postmaster" denote administrative positions. In general, though, most accounts are assigned to individuals. Depending on the addressing system and the account configuration, the identity of the account holder may not always be apparent; account names such as "ax762@torfree.net" are fairly common. To overcome any resulting confusion, headers will sometimes include the name of the intended recipient, although this is only possible if the originator knows the name.

  10. Other information in the header includes the date of transmission, the network path taken by the message, a reply-to address, and references to previous messages, if the message is in response to them. Even within email systems, headers are often not standardized. From the earliest days of email there have been debates over how much information should go into a header, with little resolution (Hafner & Lyon, 1996). Most email standards simply define a minimum set of header information. Figure 1 shows a typical email header from the Simple Mail Transfer Protocol (SMTP), a widely used standard for Internet email. SMTP requires only that the author, the creation time, and one addressee be identified. All other fields in the header are optional (Crocker, 1982).


    Return-Path: <cowan@ermine.com>
    Received: from localhost (cowan@localhost) by fur.ermine.com (8.8.5/8.8.4) with 
    SMTP id XAA31682 for <emain@ermine.com>; Mon, 7 Apr 1997 23:41:06 -0400
    Date: Mon, 7 Apr 1997 23:41:06 -0400 (EDT)
    From: Darin Cowan <cowan@ermine.com>
    To: Eron Main <emain@ermine.com>
    Subject: Your key, signed
    Message-ID: <Pine.LNX.3.96.970407234000.31152B-100000@fur.ermine.com>
    MIME-Version: 1.0
    Content-Type: TEXT/PLAIN; charset=US-ASCII
    Status: RO
    X-Mozilla-Status: 0001
    Content-Length: 3277
    

    Figure 1. Typical email header


  11. The message body contains the basic information content of the message. Because of the wide variety of computer platforms and email reader software used, and the vagaries of networked communications, email systems tend to rely on fairly simple formats for content. Message bodies normally consist of plain text only. Email standards often restrict the character set which may be used. SMTP, for example, requires that only the basic American Standard Code for Information Interchange (ASCII) character set (in effect, standard typewriter characters) be used (Crocker, 1982). Message bodies do not as a rule contain any information about the message itself; one- or two-word bodies, such as "I concur" are not uncommon. Some users append signature files to all outgoing messages; these are user-created text files which normally provide a name and email address, perhaps accompanied by contact information or other notes about the sender. Their use is never guaranteed and so they cannot be relied upon as a source for message information.

  12. Because of the content limitations of message bodies, most email systems support the use of file attachments to allow the transmission of documents formatted for specific computer applications, graphics, and other more complex information. These generally use the email message as an envelope which contains a binary data file. These files must usually be extracted and read with another application, such as a word processor or graphics viewer. Some newer email readers, such as Netscape, allow the viewing of certain attached file formats (e.g., hypertext documents, image files) within the reader itself.

    Uses

  13. As email is so widely used, it is difficult to describe a typical use. A common characteristic of email systems is the mix of business and personal messages. As early as 1973, one of the developers of ARPANET used email to ask a colleague in England to retrieve an electric razor he had forgotten at a hotel (Hafner & Lyon, 1996). The initial perception of email was that it would primarily replace the telephone as a medium for informal, transient communications. Evidence suggests that it has gone far beyond that.

  14. A 1986 UNESCO study found two major categories of email use: email messages and email documents. The former consisted of brief, informal communications which did more than just replace the telephone. "The short note written out on a note pad, or marginal comments on a letter or report fall under the rubric of electronic mail messages" (Dollar, 1986, p. 79). A distinction was made between these and what were referred to as email documents, which consisted of more formal communications.

  15. A later UN study, examining 18 organizations which used email, confirmed these findings. Email was "often used to send and receive information that would in any other medium be official-record material" (Advisory Committee for the Coordination of Information Systems [ACCIS], 1990, p. 9). Of the organizations studied, about half indicated that email was used to circulate documents for review, to send memos, and to transfer files and official documents. Only three maintained any distinction between official and unofficial use of email, and there were, in general, no policies regarding long-term storage in electronic form.

  16. Markus, Bikson, El-Shinnawy, and Soe (1992) found similar patterns of use at private sector organizations. In addition, email was found to be used for communication more often than either fax or voice mail. Users noted that email was better when they needed to be able to retrieve a copy of a message for later use and when they needed a record of the message.

  17. In implementing email, organizations normally give responsibility for the overall system to those who are responsible for the provision of computing resources in general. This is sensible, given that email systems are intimately linked with the total computer system in a given organization. However, this arrangement creates problems for records management, as will be discussed in more detail below.


    EMAIL RECORDS

    Do they exist?

  18. Given the nature of email messages and their uses, are they records? It seems clear that they can be. Duranti et al. (1997) state that the "necessary components of a record are medium, content, form, persons, acts" (tem1.htm). Email messages have all of these. The medium is electronic storage. The content is provided in the message body. The form is present in the display particular to the email system in use. There are a minimum of two persons associated with all email messages, a sender and a recipient. Emails are used to record or initiate acts, and the sending of an email is in itself an act of information transaction.

  19. As well as meeting the theoretical requirements of a record, email messages meet the practical requirements. They are clearly regarded by those who send and receive them as records. This is apparent in the desires of individuals to preserve some or all of their email messages, and in the use of email by those who wish to keep a record of a message (Markus et al., 1992).

  20. In the United States, the interpretation of email messages as records has been upheld in court. In the case Armstrong vs. Executive Office of the President (1993) (also known as the "Profs" case, after the email system in question), the American National Security Archive filed suit to prevent the destruction of computer backup tapes containing White House email messages from the Reagan administration. The defendants argued, in part, that the electronic versions of messages were simply convenience copies, and that staff were instructed to print out a paper copy of any message which was deemed by them to be a federal record. The court held that the email messages themselves were federal records. Printed copies were deemed to be insufficient to serve as the copies of record as too much key information can be lost in the conversion from complete email message to paper copy.

  21. While this decision does not mention private sector records, and has no bearing outside the United States, it indicates that there is a solid legal footing for treating email messages as records. As one commentator notes, the destruction of records, once they are created, should only take place under a records retention program. Even without laws explicitly covering their retention, if email messages can be deemed to be records, their destruction without clear guidelines "might cause legal problems" (Skupsky, 1994, p. 40).

  22. Also of note in the Profs case is the court's position that having managers choose which messages are records is inappropriate. A record does not become so based on a managerial decision; it is a record because of its origin in a transaction and its evidentiary nature. If a document, paper or electronic, provides evidence of an event or transaction, then it is a record whether or not anyone explicitly designates it as such. Managers may choose not to control or manage certain records, but that decision does not strip those particular documents of their status as records, nor does it protect the managers should those records not be available when they are required.

    Email as Evidence

  23. If email records are to be useful as such, they need to be able to act as evidence of an organization's activities. Bearman suggests an equation (Figure 2) which details the requirements for a record to become evidence (cited in Duff, Thomas, & Wallace, 1994, p. 305).


    Evidence = Information Content + Structure + Context of Transactions

    Figure 2. Records as evidence


  24. The information content alone, either the body of the email message, any attachments, or both, is insufficient to allow a message to act as evidence. It is this which led the court in the Profs case to rule that the electronic forms of the messages themselves were records. The structure and context of an email message may be completely lost if the format is changed, either from an electronic to a paper copy, or from an email message to a plain text electronic file. Much of the necessary information is contained in the header component of the message, not all of which is necessarily displayed to the recipient or printed out with the body text. Attachments to emails are often left out of printed copies as well, further limiting the usefulness of a paper copy.

  25. This is a related issue to the question of form. Different email systems will display the body text, the form, of the same message in different ways. As a result, information such as emphasis may be lost when a message is viewed on a system other than that on which it was created or for which it was intended. Even if all header information is visible, a paper copy printed out from a system other than the original may lose important aspects of form from the message body. While the problem of retrieving the original form persists even when the original email is preserved, such preservation at least provides the option of viewing the message on the original system. That opportunity is lost if the message is deleted after a paper copy is made.

  26. Duranti et al.'s (1997, tem6.htm) interpretation of electronic records supports this. Their requirements for a complete electronic record are shown in Figure 3. Most of these are normally contained in the minimum header information for email messages. All systems certainly provide for them, although it is technically possible to exclude the actual name of the author and recipients as network addresses alone will suffice for both. The time and date of transmission is recorded, though not necessarily that of receipt. The convention is that a subject is included, although it is not a requirement of the system. The disposition, the expression of the will of the author, is contained in the body of the message. If there is no body, there can be--by definition--no record.


    1. Chronological date (of both transmission and receipt)
    2. Topical date
    3. Entitling (originating address)
    4. Attestation (name of author/writer)
    5. Addressee(s)
    6. Receivers (name of copied persons)
    7. Title or subject
    8. Disposition

    Figure 3. Requirements for a complete electronic record.


  27. As email messages are records, and as they have the potential to act as evidence given the inclusion of some necessary information, it follows that they should be subject to the same types of records management, including retention and destruction policies, as all other records. At the moment, this is not always the case. The next section will examine why, and what may be done about it.


    MANAGEMENT OF EMAIL RECORDS

    Challenges

  28. Perhaps the greatest challenge facing the management of email records is the lack of awareness on all sides that they actually exist. It is somewhat understandable that the information systems staff responsible for email systems are unaware; they are not trained in records management and their primary concern for preservation is the maintenance of backups to allow their systems to be rebuilt. However, records managers have often failed to consider email. One records management textbook from 1989 makes no mention whatsoever of email, even though by that time email was already in fairly heavy use (Lundgren & Lundgren, 1989). Other texts lump email together with all other electronic records, based on its format, rather than including it as a form of correspondence based on its primary function.

  29. There is more to managing email records than indexing, maintaining the media, and ensuring systems exist to read archived messages. While these are important to ensuring preservation, many records are lost before they ever reach this stage. Often, email records are never captured in the first place.

  30. A major reason for this is that the users of the email system, those who send and receive the messages, are often the ones who determine whether a record copy is kept. When they recognize the need to do so, they act on it by keeping local copies of the messages they deem important. These are often not organized in any rational manner, are not kept according to any retention schedule, and are not accessible to the rest of the organization. The result is a collection of large personal email boxes, with some messages preserved in multiple locations and others preserved nowhere. Even regular backups of these email boxes will not be sufficient, as keeping a copy does no good if it cannot be found. As just one example of this, I once watched an individual attempting to retrieve a single email message from his own storage space. All of the messages were filed sequentially by date, in a single email folder which, at that point, contained over seven hundred messages covering more than six months. This was not an unskilled user, but a computer programmer who was the leader of a technical project team.

  31. This is not a question of users being unaware of the general need for records management. When dealing with paper records, most users know the requirements. As McDonald (1995) observed, in paper "we would have thought twice about having a memo prepared and forwarded without following some rules" (p. 84). The difference is that when a paper memo, report, or letter is prepared, it is generally distributed through channels which have been developed and are maintained by staff trained in records management. That is not the case with email. Originators distribute their email messages directly to the recipients, without passing through a central registry of any sort, and they do so through a system designed by a computer programmer, not a records manager. Email systems are not designed, nor their administrators trained, with record keeping in mind.

  32. Compounding the ignorance of users and computer system staff about records management techniques is the lack of awareness of email on the part of otherwise skilled records managers. In the UNESCO study noted above (Dollar, 1986), records managers were generally unaware that their organizations even used email. In the Profs case, no less an authority than the United States National Archivist appeared to be unaware of the records potential of email (Bearman, 1994). While this situation is improving, there are still many records managers who do not see how email should fit into their systems.

  33. The actual identity of authors and recipients is another potential problem area for email records management. The only information required by the system is an account identifier. It is possible for an account to be used by more than one individual, or for an individual to use an account belonging to another. For records management purposes, it may be necessary to assume that the holder of a given computer account, as designated by information systems personnel, is the originator of all email messages from that account. This assumes a certain level of systems security.

  34. Preserving the original form of email messages is also a challenge. While it is possible to require that all personnel within a given organization use the same email system, thus ensuring consistent message forms, such control is not possible over messages originating outside the organization. In addition, large organizations, organizations which are geographically dispersed, or organizations whose business requires the use of different computer platforms and operating systems by different personnel, may not be able to ensure such consistency internally.

    Advantages

  35. In addition to its challenges, the nature of email presents some advantages to those responsible for managing email records. One major advantage is that email is an automated system . It is possible to capture a record copy of an email message without placing any additional burden on the originator. If email systems are set up appropriately, these copies can be retained automatically, with no intervention required by the originator, the recipient, records manager, or computer systems staff.

  36. Another advantage of email is the potential it has to support the automatic capture of metadata from each record. Email systems are already designed to prompt the user for some header information and to provide other information automatically. By retaining the message header with any record copy, and designing it so that it includes all the required record information, records managers can ensure they have all the information necessary for a complete record. It is this characteristic which led to the court decision in the Profs case that email messages rather than paper copies were the copies of record. Only the electronic copies retained this metadata in the header fields, so only they were complete records.

    Some Proposals

  37. The question of how to manage email records has been given a fair amount of thought by some records management professionals. There is broad consensus that something must be done, that email records need to be managed better than they currently are.

  38. Foremost among the suggested solutions are awareness programs to inform users, computer systems staff, and records managers that email records are as valuable to the organization as every other record kept. By making users and computer system staff aware on the one hand that they must preserve email records, and making records managers aware on the other that there is such a thing as an email record, organizations will go a long way toward solving their problems. In particular, such awareness is vital if the necessary parties are to cooperate with the requirements of records management.

  39. Automatic backup copies of email messages are also a popular recommendation. The need to go beyond simple backups is also recognized; in the computer world, backups exist so that the system can be rebuilt after a crash. Normal backup systems cannot support requests for the retrieval of a particular item. One example of an enhanced automated backup is provided by Cobham (1991).

  40. In 1990, W. R. Grace and Company found that while much important information was being exchanged via email, it was not being made available to all interested parties. They modified their email system to automatically keep a copy of every message transmitted. Each day, the copied messages were compacted, indexed, and placed in a full-text searchable database. Messages were kept in the active database for six months and then moved to archival storage. While this is not a complete records management solution, lacking retention requirements, it does begin to solve the problem of providing access to stored records.

  41. This solution points to a major element in ensuring effective email management: that of building records management right into the existing system, what Bearman (1994) refers to as "[satisfying] functional requirement through design" (p. 183). While it would be helpful if users were proactive in supporting records management, it must be remembered that they use email because it is seen as a convenient way to communicate. If burdensome records management requirements are added to the system, users will bypass it in favor of a simpler mode of communication, such as telephone or voice mail.

  42. That is not to say that some additional user input cannot be implemented. Realistically, users will have to take some steps to classify their email messages if they are to be properly handled by the records management system. The normal volumes of email present in most organizations would overwhelm any records manager who attempted to assess each message individually. While some early writers on the subject proposed the use of artificial intelligence programs or expert systems to parse email messages and determine their categories, such systems are not yet workable. Although a subject heading is not essential if full-text searching of stored messages can be provided, originators must identify, by means of a file number or other notation, the record series or classification of their message, or else there will be no way to associate it with a particular set of retention requirements. As a bare minimum, users should specify whether the message is corporate or personal, so that invitations to dinner are not captured alongside quarterly budget proposals.

  43. A final proposal, often stated in the literature, details the requirement for standardized email systems to ensure that all parts of an organization may exchange messages. This was particularly noted by those examining UN organizations and other governmental bodies where there is a sufficient number of users and diversity of bodies that incompatible systems could be acquired. The need for standards is quite valid; however, several writers (e.g., ACCIS, 1990) have gone beyond the general principle to recommend the adoption of a specific email standard: the X.400 system proposed by the International Standards Organization. This is always a dangerous move. While X.400 has the backing of several government bodies, the world of networking moves fairly quickly and end-users often pay little heed to bureaucratic dictates. While the email standards debate is not over, the SMTP standard appears to be gaining the upper hand. What is important in the age of the Internet is not so much that a standard be adopted, but that an open system be used, which allows different email packages to share the same information and view the same messages.


    RECOMMENDATIONS

  44. Based on all the above, a number of recommendations can be made for the management of email records. After a lengthy description of how email differs from paper records, the first suggestion may seem incongruous: treat email records in exactly the same manner as paper. That is not to say that the same capture or storage methods should be used, only that the principles should be the same. In particular, it is doubtful that email records will represent an entirely different class, or classes, of records. Most email traffic deals with routine correspondence, projects, and planning, all the areas which are also covered by paper records. File numbers and classifications exist for these paper records; retention requirements have been determined. These do not change simply because someone submits an email comment on a project plan rather than a written memo.

  45. As much as possible, email records should fit into existing records management structures. The evidence points to the fact that email messages are simply another form of correspondence, that they have the same value and retention needs as paper. While it is necessary to implement different storage systems, there should be no need to reassess retention.

  46. The retention of the same records series in different formats--print and email, as a minimum--can create additional problems. Some means is necessary to ensure that a search of a given records series will encompass all forms. A solution may be to automate all records management functions, recording information about print and electronic records in the same database for search purposes.

  47. Records managers should take advantage of the automated nature of email to automatically create record copies. Some users may be uncomfortable with having all their email messages captured, preferring to designate those which are records and leave personal messages out. However, the existence of a record is a matter of fact based on the content and origin of the record; it is not based on the judgement of an individual. Users should simply be warned that any personal messages they send may be retained. If the email system has been procured to support the operational needs of an organization, this policy is entirely justifiable.

  48. Email systems should be modified with an additional required header field for a file number or classification. Users should be instructed to insert this, just as they are required to insert an address. The retention of an email record can then be based on the existing requirements for that class. Some messages may not be easy to classify; if a file number is not inserted, the system should route a copy to the records manager for classification, although such messages should be the exception.

  49. In addition to the file classification, the subject field and the names of the originator and recipients should be mandatory header components. Stored email messages should be indexed on the originator, recipient, subject, date, and file number. In addition, where possible, the full text of messages should also be searchable.


    CONCLUSION

  50. Email produces records. These records have the same legal requirements as all other records of an organization. While current records management systems do not always adequately handle email records, there is no reason that they cannot be extended to do so. What is required is that all parties be aware of the value of email records, and that modifications be made to email systems to meet the needs of records management at the same time as records managers adapt their procedures to incorporate email records. If this is done, a great deal of potentially valuable information which might have been lost will be retained, to the benefit of the organization.


REFERENCES

Advisory Committee for the Coordination of Information Systems. (1990). Management of electronic records: Issues and guidelines. New York: United Nations.

Armstrong v. Executive Office of the President, 810 F. Supp. 335 (D.D.C. 1993).

Bearman, D. (1994). Electronic evidence. Pittsburgh, PA: Archives and Museum Informatics.

Cobham, D. B. (1991). Small is beautiful: Electronic mail document management at W.R. Grace. In P. Gillman (Ed.), Text retrieval: Information first (pp. 46-50). London: Taylor Graham.

Crocker, D. H. (1982, August 13). Request for comments 822: Standard for the format of ARPA Internet text messages. Available from http://www.cis.ohio-state.edu/htbin/rfc/rfc822.html

Dollar, C. M. (1986). Electronic records management and archives in international organizations: A RAMP study with guidelines. Paris: UNESCO.

Duff, W., Thomas, D., & Wallace, D. (1994). Working meeting on electronic records management, 1994. Archives and Museum Informatics, 8(4), 301-352.

Duranti, L., Eastwood, T., & MacNeil, H. (1997). The preservation of the integrity of electronic records. Available from http://www.slais.ubc.ca/users/duranti/

Du-Rea, M. V., & Pemberton, J. M. (1994). Electronic mail and electronic data interchange: Challenges to records management. Records Management Quarterly, 28(4), 3-12.

Hafner, K., & Lyon, M. (1996). Where wizards stay up late: The origins of the Internet. New York: Simon and Schuster.

Lundgren, T. D., & Lundgren, C. A. (1989). Records management in the computer age. Boston: PWS-Kent.

Markus, M. L., Bikson, T. K., El-Shinnawy, M., & Soe, L. L. (1992). Fragments of your communication: Email, vmail, and fax. Information Society, 8, 207-226.

McDonald, J. (1995). Managing records in the modern office. In S. Yorke (Ed.), Playing for keeps: The proceedings of an electronic records management conference hosted by the Australian Archives, Canberra, Australia, 8-10 November 1994 (pp. 84-92). Canberra: Australian Archives.

Menkus, B. (1996). Defining electronic records management. Records Management Quarterly, 30(1), 38-42.

Robek, M. F., Brown, G. F., & Stephens, D. O. (1996). Information and records management (4th ed.). New York: Glencoe/McGraw-Hill.

Robinson, L., & Bawden, D. (1995). Internet: The way forward. Managing Information, 2(3), 20-22.

Skupsky, D. S. (1994). The law of electronic mail -- The impact of the White House case on you. Records Management Quarterly, 28(1), 32-40.


ADDITIONAL READINGS

Barry, R. E. (1993). Managing organizations with electronic records. Information Management and Technology, 26(3), 115-121.

Bearman, D. (1995). NARA issues new rules on electronic records. Archives and Museum Informatics, 9(3), 338-341.

Postel, J. B. (1982, August). Request for comments 821: Simple Mail Transfer Protocol. Available from http://www.cis.ohio-state.edu/htbin/rfc/rfc821.html

Thibodeau, K. (1996). Reengineering records management: The US Department of Defence records management task force. Archivi and Computer, 1, 71-78.

Saffady, W. (1992). Managing electronic records. Prairie Village, KS: ARMA International.

Wigand, R. T. (1985). Integrated communications and work efficiency: Impacts on organizational structure and power. Information Services and Use, 5(5), 241-258.