In Manufacturing Consent, Edward Herman and Noam Chomsky0 propose a radical theory to explain the behavior of the US mass media. According to their model, the "raw material" of news is subjected to a series of filters: corporate ownership, dependence on advertising, reliance on official sources, influence of "flak" machines, and anti-communist bias. These filters serve the needs of the "elite," who are defined as the government, corporate owners, media executives, and a few powerful advocacy groups ("elite" is a clever choice of word, since any mention of elitism is sure to raise leftist hackles). The result, they claim, is a system in which stories that serve "elite" interests are emphasized, while those that undermine those interests are downplayed or suppressed.
Apparently, the existence of these filters has to be inferred from the number and kinds of articles that are published, since Herman and Chomsky do not present specific examples of reporters or editors who are willing to admit that this actually happened in the specific cases they describe.1 Moreover, anyone who actually pays attention to the media can come up with innumerable counterexamples. A few minutes' brainstorming yielded the following: Watergate, the Pentagon Papers, the Iran-Contra scandal, the Savings and Loan crash, Ivan Boesky, Michael Milken, the
Enron collapse, Arthur Andersen accounting scandals, antitrust litigation against Microsoft and Standard Oil and AT&T, the conviction of Imclone's CEO Samuel Waksal, Freddie Mac woes, and assorted scandals at major publications--Jayson Blair at the New York Times, Janet Cooke at the Washington Post, and Stephen Glass at New Republic. Then there are books published by major media corporations that critique the putative "elite": works by Ralph Nader and Michael Moore--and, indeed, Manufacturing Consent itself (the version I have was published by Pantheon Books, one tentacle of publishing giant Random House.2
To explain away such cases, all of which blatantly contradict their
thesis, Herman and Chomsky resort to a series of evasions and fudges:
1. Dependence on subjective standards. They claim that
such counterexamples are not really counterexamples at all--that the tone
and placement of articles matters, as well as their existence. This is a
rather clever standard, since it means that supporting examples can be
counted as evidence while counterexamples can all be dismissed out of hand
as exceptions. (They don't specify the number or type of counterexamples
that would be required to disprove their thesis. Nor do they explain how
the tone of Manufacturing Consent was changed to get it through
the filters.) Moreover, it means that any article can be counted as supporting evidence if in some way, no matter how minor, it doesn't quite say exactly what Herman and Chomsky think it should. For example, in contravention to the predictions of the "propaganda model," Time explicitly acknowledged that the Guatemalan government "obviously violates human rights" (p. 76). But this doesn't satisfy Herman and Chomsky, and it's not clear that anything would.
It's worth noting that this is the same kind of standard that conservatives use to argue for the existence of a liberal media bias. During the second Iraq war, it allowed conservatives to argue that the coverage was plagued with liberals gloating about quagmires, while leftists claimed that the coverage was nothing but jingoism. No doubt both sides could find something unsatisfactory about the coverage, but this just means that it fell somewhere in between their biases.
2. Shifting of standards from case to case.
When discussing "worthy and unworthy victims," they simply count the number of articles that each set of victims received. They show that the number of articles compatible with the US government's agenda was larger than the number of articles incompatible with it. In the section on "demonstration elections," however, this standard proves problematic; it works for El Salvador, but not for Nicaragua. According to Herman and Chomsky, the Nicaraguan elections were not favored, but the basic conditions for an election were better--yet the mass media paid "substantial attention to basic conditions" in Nicaragua (p. 137), and there were more articles incompatible with the government's agenda (p. 132-136). To avoid this problem, Herman and Chomsky simply change to a more complicated standard; in the rest of the book, they abandon this standard entirely. This is a classic social science error: you come up with a test that yields the results you want; then you apply it to a new situation; if it doesn't yield the desired results, you keep revising it until it does; then you stop and report that you've proven your thesis.
3. Uneven trust of sources. Herman and Chomsky
apply different standards to the same type of source, depending on what that source is saying. For example, sometimes the word of a single ordinary person is given credence (p. 119), and sometimes it's downplayed (pp. 124, 129). Refugee testimony is uncritically accepted in the case of Vietnam (p. 177) and in Cambodia in the early 1970s (p. 273); when they are evaluating reports of refugees from the Khmer Rouge, they suddenly cite the "truism" that refugee reports need to be evaluated carefully (see this for what Chomsky and Herman originally said). More generally, the potential biases of sources that support their thesis--i.e., those other than the US mass media--are never seriously considered.
4. Vagueness about the identity and powers of the
"elite." Herman and Chomsky seem to trust that their readers will understand that the "elite" refer to a muddle of icky corporations and politicians whom we all know
are in each others' pockets. This allows Herman and Chomsky to change the definition of the elite from example to example to fit the needs of their model. For example, in chapters 2 and 3, the Reagan administration is the elite, and is supposedly able to suppress an enormous amount of nasty "truths" about what's really going on in Nicaragua, Guatemala, and El Salvador. Suddenly, during Iran-Contra, a second elite appears that's able to bring that information to light. How did the elites in the Reagan administration suddenly lose their stranglehold over the mass
media? If this second elite, which is implied to be Congress, has the power and desire to make Reagan look bad, why didn't they do it before? If the information that Herman and Chomsky present is true, it would've been useful and easy for this second elite to paint Reagan as a deluded mass murderer, supporting the restriction of his power and the augmentation of their own. So why didn't they? Along identical lines, the Johnson and Nixon governments allegedly had incredible power to suppress putative truths about Indochina. But Watergate is explained away as a threat to the Democratic party, which part of the elite. But surely it would have been useful (and, once again, rather easy) for the Democrats to paint Richard Nixon as a murderer sending American boys to kill innocent people and die for no reason. Why didn't they? None of this is ever explained.
5. Refusal to consider alternative explanations. This
is another classic social science error. When testing an idea, it's never enough to claim that the results are consistent with your model; you have to rule out other competing models that could explain your results. Indeed, anyone who has actually written for a newspaper (as I have) will no doubt find this model to be rather incomplete. Incredibly, Herman and Chomsky pay almost no attention to the effects of reader choice3 ; everything is attributed to the whims of the "elite." Of course, there are many other factors that play a role. For example, a story about a gruesome murder might get top billing on a slow news day, but might be pushed to the back pages on a day when other headlines are judged more important or more likely to catch a reader's interest. Then again, maybe there have been a lot of gruesome-murder stories lately, and the editor wants something different this time around. Something like this probably explains Watergate. Herman and Chomsky state that Watergate actually supports their thesis, because the machinations against the Democratic party threatened elite interests, while those against the Socialist Workers' party did not. Alternatively, perhaps the news editors believed that their readers would be more interested in the crimes against the Democratic party than in those committed against a party to which almost none of them belong.
On a more general level, plenty of people (consciously or unconsciously)
seek out news sources that tell them what they want to hear.4 I suspect that most people don't want to hear that their government is engaged in murderous acts of terror, and wouldn't buy a paper that made such a claim (and thus they wouldn't see the ads,
etc.)5 (I'm not claiming by any means that any of this is a good thing. It isn't. I'm only claiming that it happens and is scarcely
accounted for by Herman and Chomsky's theories.) On the whole, it's easy to paint a
picture of an elite-dominated media if you systematically downplay every other force that plays a role.
I'm certainly not trying to claim that the US mass media are perfect or
even adequate. They're not. Most of what passes for TV news is
infotainment, and much of what passes for reporting is
heavily salted with opinion. As best as I can tell, the only
way to get anything like complete information is to hop back and forth
between left-wing, right-wing, and mainstream sites, keeping in mind the
various biases that are likely to plague each
one.6 While Herman and Chomsky provide some good examples of information that did not make it into the mainstream media, their model fails to account for the behavior of the media as a whole.
0Not "Chomsky," not "Chomsky and Herman." Herman and Chomsky. Herman is the first author, at least in the version in my library. Poor guy must get tired of everyone unconsciously relegating him to second place.
1 I am referring here to their examples in chapters 2-6. Early in the book, they do cite a source saying that the media receive substantial criticism when they point out Reagan's mistakes and are accordingly pressured into overlooking them. This is a good point, but here as elsewhere it's not clear whether this is attributable to the "flak" machines or to well-meaning (but misguided) ordinary citizens. There are also a few cases in which they assure us that unidentified journalists admit to this sort of thing, but no names are given.
2Corporate ownership is more
complex and intertwined than a redneck's family tree. But here's a brief
summary. My edition of this book was published in 1988 by Pantheon Books,
which has been an arm of publishing giant Random House since 1961.
(Random House also owns Bantam, Ballantine, and Knopf, among
others). At that time, Random House was itself owned by Advance
Publications (a newspaper publishing conglomerate that owned numerous
local newspapers as well as the the New Yorker and Conde Nast fashion
magazines). Pantheon is now owned by media conglomerate Bertelsmann
(which also owns Bantam, Doubleday, Dell Books, Arista, BMG, and
All of this means that--ironically enough--when you buy this book and put it on your bookshelf to show that you're, you know,
"with it" and "in the know" and stuff, you're not really rebelling; you're
giving money to a gigantic mass media corporation and two rich old white guys from the academic elite. I never
quite know what to make of this. Part of me thinks it's just a moral
compromise on their part; the more paranoid part thinks they're
tipping a wink to the nondeluded in the audience and
giggling all the way to the bank.
3It's barely and grudgingly
acknowledged in one footnote (p. 14). The section on the advertising
filter focuses on corporate attempts to squash embarrassing stories. There is scarcely any acknowledgement (a few sentences on pp. 17-18 and a mention in the conclusions that this might be a "nuance" or "secondary effect" (p. 304). It is not addressed in the main part of the book, and there is no acknowledgement and that this might work in the other direction--that advertisers want the TV executives to produce programs that the people want to see. Likewise, the "flak" filter might sound like an entry point for dissent, but it's attributed to--you guessed it--large and putatively conservative
organizations funded by corporations. Apparently leftists don't organize effective letter-writing campaigns.
4Of course, this would be
true of leftist publications as well, the readers of which seem to have a
strong emotional desire to hear that the great capitalist nation is
really a murderous failure. Publications or individuals who go against
this theme are likely to be attacked or "exiled"--see, for example, the
vicious reaction to Christopher Hitchens's post-9/11
5See also note 3 about the advertising filter. Incidentally, this seems to be a fairly
standard mode of thought among leftists. It goes something like this:
Corporations and governments are the root of every evil in civilization.
The poor innocent oppressed people (whom the leftists are claiming to
represent) have done nothing wrong at all, except by allowing themselves
to be duped--and of course they simply couldn't help it anyway.
Put another way, information about corporate malfeasances (which is useful
for leftist interests) is given great emphasis, while information about
the bad judgements of the populace (which would be contrary to leftist
interests) is downplayed or suppressed. Sound familiar?
6For example, during the second
Iraq war, there was some debate about the looting and destruction of
antiquities. I found the first reports of this on left-wing sites.
The right-wing sites only had a bit of grumbling about how it wasn't
really a problem, and why should we care anyway? Some time later, the
right-wing sites were the first to report that the original estimates were
wrong and "only a handful" of items were destroyed. The leftist sites
just grumbled skeptically. Then the leftist sites reported that there were far
more than a handful, the right-wing sites reported that all the major
items had been stored away and were safe, etc., etc., etc. The point is
that all these sites--both left and right, mainstream and not--are more
likely to report information that is in accordance with the biases of
their readers (which is not what Herman and Chomsky are asserting).
Stick to just one and you'll be deluded.