The Internet, Distribution Lists, and Gatekeeping
School of Journalism
University of Missouri
Submitted to the Communication Technology and Policy Division of
The Association for Education in Journalism and Mass
Communication for the 1994 Convention, August 10-14, Atlanta, GA.
This paper begins to explore the gatekeeping function of
discussion list owners on the Internet. The study finds that
while some discussion lists are unmoderated, others operate under
guidelines that lead list owners to delete messages sent for
distribution. At other times list owners say they halt lines of
discussion. But in most cases, list owners say they are
reluctant to tamper with messages.
In addition, the paper finds many areas of Internet
discussion lists waiting for study, from methodological problems
to issues of information flow.
The Internet, Distribution Lists, and Gatekeeping
The advent of a new technology gives our, or any other,
field a chance to evaluate the technology against existing
theories, and to test the theories against the new technologies.
Computer technology as a mass medium through the Internet opens
up opportunities for us to apply the old print and broadcasting
paradigms to see whether they can simply be extended to this new
medium or whether we will have to make wholesale modifications,
or even search for new theories to explain the process and effect
of e-mail, bulletin boards and list groups.
Edmund Carpenter, in a book he co-edited with Marshall
McLuhan, said each mass medium is a language.\1\ The new media
then were radio, television and films, and Carpenter said their
grammars were still unknown, even though radio was 40 years old
and movies were a few years older. More than three decades
later, whether we have figured out the grammar of radio,
television and film or not, the newest medium is computer e-mail
technology, and we have another grammar with which to grapple.
This paper begins the process of matching gatekeeping theory
to the new technology of list groups on the Internet. The
research question to be asked first is whether there is in fact a
gatekeeping function on Internet lists. If such a function does
exist, what role do gatekeepers see themselves in? And what sort
of background do the gatekeepers have that might help them in
The Internet is, as is the Interstate highway system, based
on a defense department plan to survive an attack from enemies.
Originally called ARPAnet, it linked together defense and
academic computers in a network designed to operate even if part
of the network was destroyed and to allow instantaneous,
uninterrupted communication and computer capabilities throughout
the country.\2\ The National Science Foundation co-opted the
idea, and the system built from there.
There are actually three computer networks accessible to
most computer users. The Internet has become almost a generic
term, since the other two networks can be accessed through the
Internet. Usenet was formed in the late 1970s to serve
universities and later commercial organizations. Bitnet was born
in the 1980s to serve academic and research institutes. ARPAnet
was dismantled in March of 1990.\3\
Anyone who has "surfed" the Internet knows the variety of
information available to users. A number of governmental, quasi-
governmental and educational agencies store information for
retrieval by anyone who can find it. Literature and reference
books can be obtained via the Internet, as well as texts of
presidential speeches and announcements, weather forecasts,
pictures sent from various NASA vehicles, phone numbers and e-
mail addresses and even some publications. Some predict the
Internet, or a future evolution of the Internet, will be our
An estimated 300 gigabytes of data flow through just the
U.S. portion of the Internet each day, equivalent to 125,000 one
hundred-page books.\5\ A large part of that is in electronic
mail, or e-mail. Besides acting much like the post office, in
which one can send person-to-person communications, which would
conform more to interpersonal theories, three types of
distribution lists have characteristics closer to mass
Internet lists can be unmoderated as well, but they may also
be moderated, and may require prospective subscribers to meet
certain standards before signing them up. An Internet list on
quantum physics, for instance, may ask subscribers to answer
questions about their background and education, and if the
subscriber doesn't have a degree in physics or math, or does not
read scientific journals, a subscription will be denied. Editors
or moderators may review postings to the list before they are
distributed, and can decide not to distribute the message, or may
modify it to conform to the charter and purpose of the list.
Some lists have as their sole purpose the distribution of
technical information from an agency, institution or
manufacturer. No discussion is allowed on these.
Bitnet lists take any subscriber and are not moderated.
That is, messages that are sent for distribution are not
submitted first for editing. They are unmoderated lists. They
do, however, have charters that outline the purpose of the list,
and there is an "owner." Owners do not own the list in the sense
that they derive an income from it, but they are responsible (as
responsible as they want to be) for the conduct of the list. A
list owner may occasionally direct the discussion or, somewhat
more likely, halt a line of discussion that has degenerated into
name-calling ("flaming" in the lexicon of the Internet).
In both types of lists above, once a subscription is
accepted, the subscriber will get postings directly into his or
her electronic mailbox.
The final type of list is a Usenet group. Access to Usenet
lists is not by subscription. An Internet user can simply tap
into the Usenet group and read the postings. Messages are not
distributed into the user's electronic mailbox; the user must go
out and retrieve the message. It is analogous to strolling
downtown to look at a bulletin board, or going to look at the job
postings at the employment service. Usenet groups are not
moderated and are not owned.
Gatekeeping as a theory
Kurt Lewin was apparently the first one to use the term
"gatekeeping," which he used to describe a wife or mother as the
person who decides which foods end up on the family's dinner
table.\6\ The gatekeeper is the person who decides what shall
pass through each gate section, of which, in any process, there
are several. Although he applied it originally to the food
chain, he then added that the gating process can include a news
item winding through communication channels in a group. This is
the point from which most gatekeeper studies in communication are
launched. Significantly, however, Lewin also said gates may not
be governed by a keeper, but by "impartial rules."\7\ Some of the
distribution lists would fall more under the latter category.
Unfortunately, Lewin died before he could expand more on the
David Manning White was the person who seized upon Lewin's
comments and turned it solidly toward journalism in 1950.\8\ His
case study of a wire editor, "Mr. Gates," on a Midwestern,
medium-market newspaper established the term as our own. He
profiled Mr. Gates, then analyzed what wire copy Mr. Gates
accepted for the paper and what he rejected during one week. The
study has been duplicated several times. In the mid-1960s,
Snyder\9\ criticized White's study for being based solely on
rejected items and not examining why Mr. Gates accepted the items
he did, and for allowing the use of the general term "no space"
as a reason for rejection. He recast the study in terms of items
retained, and found gatekeepers' decisions are based largely on
audience interest rather than simply the pressures of the
publication deadline or the space available. Bleske replicated
White's study in 1992, and although the profile of the gatekeeper
has changed somewhat and the latest study shows "Ms. Gates" using
a computer to edit copy rather than reams of paper, the result
and conclusion was basically the same as White's.\10\
White was also criticized by Bass\11\ who said White saw
everyone in the newsroom who made news choices as a gatekeeper.
Bass maintained that the telegraph editor is not the key decision
maker. This is an important point. White merely selected the
wire editor to study, and subsequent studies have followed along.
Bass also said the study of news flow should be divided into two
segments: news gathering and news processing. Perhaps yet a
third should be added, namely news policy. Buckalew\12\ took a
step in that direction when he studied news editors at television
stations, using a Q-sort. He found news editors' judging
patterns are characterized by five news facets: normalcy;
significance; proximity; timeliness and availability of visual
McCombs and Shaw took a different direction when they looked
at the effects of gatekeepers' decisions. They found the
audience learns how much importance to attach to a news item from
the emphasis the media place on it.\13\ Fico, et al., also
explored this in the context of electronically-delivered
newspapers that are indexed by topic area and eschew headlines.
They found that audience members didn't like the indexed papers,
because salience cues, set by the gatekeeper, were missing when
headlines were removed. This power to set the agenda is on
editors' minds. Willis found that managing editors are concerned
about their ability and right to select their paper's news
agendas. They also feel that people want more entertaining
stories instead of newsworthy ones.\14\
Finally, Dimmick has entirely remolded the theory into a
series of set theory formulas and uncertainty theory propositions
to more expansively explain the gatekeeping process.\15\ Those
propositions are: 1) Gate-keepers are uncertain which events are
to be defined as news;\16\ 2)gate-keepers' potential universe
identification uncertainty is reduced by six factors;\17\ 3)gate-
keepers' decision spaces are multidimensional;\18\ and 4)gate-
keepers' actual universe selection uncertainty is reduced by the
composition model used by the gate-keeper.\19\
Of course, on the Internet, gatekeepers can choose whether
to be constrained or not, and may be thought of in some cases
only nominally as gatekeepers. Some may choose to channel the
discussion rather narrowly, others are willing to let it spread
over a wide area. They are not constrained by time or space
(yet), nor by the pressures of advertising or the profit motive.
Perhaps the only constraining factor is their ability to read all
the messages being submitted for distribution. In some cases,
subscribers themselves act as gatekeepers; if no one responds to
a message, quite often that discussion line dies out.
From previous experience on the Internet, the author had an
idea how well questionnaires are received. While there are many
helpful people who subscribe to lists, others can be abrupt when
messages stray from the point of the list. With this in mind,
the author consulted several long-time internet experts before
setting out on the survey.
One suggested the questionnaire be kept as brief and non-
intrusive as possible. With this in mind, a short, 14-question
survey instrument was drawn up, the answers to which could be
gathered without much thought and without having to consult
records or references, but with the goal of gathering as much new
and helpful data as possible.
The author obtained a list of lists, which included names of
list owners and in some cases editors or moderators. The list
contained nearly 1,000 names, and the original intent was to
randomly select 300 names from that list for the survey.
Although most subscribers use the Internet for free through
their school or place of business, those who must access the
Internet though a commercial gateway often have to pay per
message. They do not like to see mail arrive that they have not
solicited. Others get such a volume of mail daily that survey
instruments are a nuisance. For that reason, and out of a
concern for Internet etiquitte, the e-mail coordinator at the
author's university suggested mailing 300 unsolicited pieces of
mail would bring a rain of complaints to both the author and
coordinator. A check with an experienced list owner with whom
the author was acquainted confirmed that, and he suggested
putting the survey on a list that serves list owners, LSTOWN-L.
Before doing that, the author checked with the list's owner,
who said the survey was perfectly within the purpose of the list,
and gave his approval to conduct the survey. The letter
accompanying the survey and asking people to respond mentioned
the list owner's permission. In this case, the author made a
calculated decision to sacrifice a random sample for access to a
group of listowners.
The letter and survey were sent to the list on a Sunday
evening in February and distributed to the 223 subscribers
moments later. By the next morning, 14 surveys had been
returned, and there were no complaints. Although this was a
promising start, the numbers dropped off quickly, with only a
handful returned over the next few days. Many recipients act
immediately on mail rather than laying it aside for attention
later, and if they don't respond, they delete it. One week after
the original mailing, a second mailing was made, which elicited
only one response. Two weeks later, a final reminder was sent,
which resulted in 11 returns, for a total n of 29. In actual
fact, fewer than 29 people responded, but several people own or
moderate more than one list, and two of them answered for each of
In order to answer the survey questions, respondents were
asked questions about the list and about themselves. A cover
letter explained the gatekeeper concept, although the word
"gatekeeper" was not used in the questionnaire.
Respondents were asked which of four categories the purpose
of their list fits into. The four areas were discussion of
technical information, discussion of an [academic] area of study;
discussion of an area of interest (such as Christian living) or
strictly for distribution of technical information.
As Table 1 shows, the purpose of the lists is fairly divided
between the four categories, with heaviest emphasis given to
discussion of an area of interest.
Table 1. Number of lists according to purpose.
Discussion of technical information 5
Discussion of an academic area of study 7
Discussion of an area of interest 11
Distribution of information only - no discussion 2
When asked if they ever delete messages before they are
distributed to the list, many of the 28 who responded to the
question pointed out that their list is a Bitnet list and/or is
not moderated. In this sample, 21 said they do not, but seven,
or 25 percent, said they do. They gave a variety of reasons:
Delete is not my word, I have gotten posts which are either
personally insulting and/or absurd or just outside the scope
of the list.
Announcement messages are altered (edited) to conform to a
standard form...At least one of the last three (maybe more)
has been rejected and returned to the submitter because it
did not fit the topic of the list OR there was insufficient
Irrelevant material like subscription-related requests.
Deleted part of a message because I knew the answer and sent
it to the subscriber directly.
Articles which are not news or do not relate to news are not
posted; [name of list] is an edited list.
One was borderline flaming, so I asked the writer to
reconsider and re-post another day. He did.
Table 2 shows the breakdown by list type of how willing list
owners were to delete messages.
Table 2. Willingness to delete messages by list type
Purpose of List Yes No
technical info 5
academic area 2 6
interest area 3 10
Distribution only 2
Two more were willing to halt lines of discussion. Ten of
the respondents indicated they have stopped lines of discussion
on the list they maintain. They gave these reasons:
Coming too close to a flame war; uncivil and trivial msgs.
One topic had wandered too far afield. Readers were
complaining. One topic we agreed ahead of time would last
two weeks, as a tool for gathering data for a meeting.
Antagonists were pursuing the topic for the sole purpose of
halting all other discussion, and for aggrandizing
Inappropriate to the list, which gets a lot of traffic
Discussion turned into personal flaming or started returning
to previously discussed materials.
Not relevant to the list and annoying some (most?) of the
Table 3 shows the breakdown by list type of how willing list
owners are to halt lines of discussion.
Table 3. Willingness to halt discussion by list type.
Purpose of List Yes No
technical info 2 3
academic area 3 5
interest area 4 10
Distribution only 1 1
One of the functions of a list owner, and more so a list
moderator, is deciding how far to let messages and discussion
wander from the purpose of the list. Table 4 shows how far list
owners say the discussion strays according to the purpose of the
Table 4. Range of discussion from purpose by purpose
tech info Discuss
acad. area Discuss
far 2 3 5
3 5 4
only 1 2
In at least two of the studies of gatekeepers, the editors
had an education in English, rather than journalism. All, of
course, had worked in news gathering organizations before getting
the editor's position they held; most news organizations would
not put an inexperienced person in that job. but since almost
anyone can charter and own a list, background becomes more
interesting. Respondents were asked whether they had education
or experience in any of several traditional information areas:
journalism, information management systems, communications, and
Tables 5 and 6 show that background doesn't seem to play
much of a role in the decision whether to actively engage in
Table 5. Willingness to delete or alter messages by
Background Yes No
Info. Mgmt. Syst. 2 6
Communications 1 4
Libraries 1 2
All of the above 1
None of the above 3 8
Table 6. Willingness to halt discussion by background.
Background Yes No
Info. Mgmt. Syst. 2 6
Communications 2 3
Libraries 2 1
All of the above 1
None of the above 3 8
The self-selection, non-random method limits the
generalizability of the survey, but non-randomness became less of
an issue when only 29 responses were received. Rather, the
results can be considered as a case study, in which the purpose
is to explore a medium rather than to build new theory.
The first research question asks whether there is a
gatekeeping function on Internet lists, despite what the "rules"
about moderated and unmoderated lists are. The answer is that
there does seem to be some sort of gatekeeping function, but it
is quite different from that of the gatekeepers of White, Snyder
and Bleske. The gatekeeper here has the capacity to serve much
more as a censor, although those in this study seem disinclined
to do it. Because space or time is not a consideration as yet on
the Internet, the gatekeepers do not have to decide which
messages will be kept and which will be discarded. Seven of the
28 valid cases in the study say they are willing to alter or
delete messages, and 10 say they are willing to halt a line of
Deleting messages may be seen as more drastic than gently,
or even forcefully, halting a line of discussion. That's because
halting a line of discussion does not involve tampering with
someone else's copy, merely a note among all the other messages.
In fact these gatekeepers seem willing to let the discussion
wander from the main theme from time to time, guiding it back
reluctantly. Their open-ended responses to the reasons they halt
a line of discussion show avoiding hostilities on the list as a
The second question asks what role gatekeepers see
themselves in. It is clear from the responses of those who see
themselves as having any role that they do not wield a heavy
hand. Their role is not to restrict the flow of information, but
rather to channel it to the purpose of the list. Occasionally
they see themselves, it would seem, in the role of breaking up an
argument that has degenerated into name calling, or shooing pests
and party-crashers out the door.
The third question asks what background gatekeepers have
that could help them in their gatekeeping role. Most have no
experience or education in any of the traditional communication
fields. This is not inherently bad, but it would be interesting
to study the gatekeeping habits of those moderating lists who
haven't had a background in any of the communication principles.
Moreover, while even an inexperienced person on the desk of a
newspaper or television station will gain experience and guidance
from peers and superiors, list gatekeepers work in isolation.
Areas for further study
The possibilities for further study are practically endless.
First, there are methodology questions to be answered. How does
e-mail etiquette fit with the ability to conduct random-sample
surveys? What are the implications of inadvertent survey
distribution? This is rarely a problem in postal surveys.
People don't take the time to photocopy a questionnaire they've
received in the mail, address envelopes and redistribute the
survey. But in e-mail, forwarding a questionnaire to another
entire list is a simple matter of a few keystrokes.
How does one overcome the disinclination of so many people
to ignore the survey? In the journalism area, a fifty percent
response rate has become almost a standard of acceptance. This
survey got far fewer than that, forcing it into the realm of case
study research. It was a simple survey, and yet more than 80
percent of the recipients ignored it.
Audience perception of e-mail
Some lists average more than 100 messages a day (see
appendices). This makes simply reading one's mailbox an enormous
chore each day. Even those who scan the "subject" lines of each
message must take the time, and, like headlines, if the message
lines don't contain salience cues, readers may ignore whole
discussion areas. What is the role of the subject line in reader
response? What kinds of messages do readers respond to? Are
list subscribers more likely to respond to list messages and
postings than to the newspaper or TV station?
Free flow of information
Some people are subscribing to several discussion lists and
using the Internet to retrieve information they want to know
about. But are these people also still paying attention to the
"traditional" media? Is e-mail narrowing society from having a
marketplace of ideas to having boutique shops of ideas with
homogeneous customers? Is the discussion, often quite in depth,
on discussion lists escaping those lists to be distributed to a
wider audience? Are the experts and opinion leaders on the list
looking for a wider audience as well so some of these ideas are
mainstreamed? If not, what are the implications, especially as
use of discussion lists increases?
No one as yet is in the list ownership business for profit.
In fact, owning a list can be a lot of work, with no financial
gain. On the other hand, subscribing to a list costs nothing,
unless the subscriber must use a commercial service such as
Prodigy or Compuserve for access. No one really has a vested
interest in the operation of the lists. How does that affect
messages, censorship, ethics and the like?
If an Internet-like service is really to be our lifeline,
there is a lot to explore. But even if it never becomes vital
enough to become a lifeline, electronic exchange of information
outside the lines traditional media are likely to find has
already become important to many people. They are exchanging
information (accurate or not), generating ideas, building
associations. An understanding of the dynamics will be vital if
society is to use the Internet, in any of its manifestations, to
its fullest advantage.
Average Number of Messages a
*This was an open-ended question. When respondents answered with
a range, e.g. between 700 and 800, the difference was split.
Most who gave an exact number qualified it with "about."
Appendix B. Average number of messages per week according to
type of list.
List type Up to 10 11 to 50 More than 50
info 2 3
academic area 7 1
interest area 8 1 5
information 1 1
1. Carpenter, Edmund, "The New Languages," in Exploration
in Communication, eds. Edmund Carpenter and Marshall McLuhan
(Boston: Beacon Press, 1960).
2. Krol, Ed, The Whole Internet, (Sebastopol, CA: O'Reilly
and Associates, Inc., 1992).
3. LaQuey, Tracy and Jeanne C. Ryder, The Internet
Companion: A Beginner's Guide to Global Networking, (Reading
Massachusetts: Addison-Wesley Publishing Company, 1992).
4. Eng, Paul and Julie Tilsners, "Up all night with the
Internet," Business Week, no. 3357, p. 14, February 7, 1994.
5. "Inside Internet," MacLean's, January 7, 1994, p 45.
6. Lewin, Kurt, "Frontiers in Group Dynamics," Human
Relations, v. 1, no. 2, 1947, p. 145.
7. Ibid, p. 145.
8. White, David Manning, "The 'Gatekeeper': A Case Study In
the Selection of News," Journalism Quarterly, 27: 383-90 (Fall,
9. Snider, Paul, "'Mr. Gates' Revisited: A 1966 Version of
the 1949 Case Study," Journalism Quarterly, 48: 419-27 (Autumn,
10. Bleske, Glen L., "Ms. Gates Takes Over: an updated
version of a 1949 case study," Newspaper Research Journal, v. 12
no. 4 pp. 88-97.
11. Bass, Abraham A, "Redefining the 'gatekeeper' concept:
a U.N. Radio case study, Journalism Quarterly, 46: 59-72 (Spring,
12. Buckalew, James K., "A Q-Analysis of television news
editors' decision, Journalism Quarterly, 46: 135-37 (Spring
13. McCombs, Maxwell E. and Donald L. Shaw, "Structuring
the unseen environment," Journal of Communication, v. 26 no. 2,
pp. 18-22 (Winter, 1976).
14. Willis, Jim, "Editors, readers and news judgement,"
Editor and Publisher, v. 120, no. 6, pp. 14-15 (February 7,
15. Dimmick, John, "The gate-keeper: An uncertainty
theory," Journalism Monographs, no. 37, 1974.
16. Ibid, p. 8.
17. Ibid., p. 10.
18. Ibid., p. 18.
19. Ibid., p. 21.