Journalistic Bias

For years the Right Wing has vigorously tried to persuade Americans that journalists generally have a liberal bias. In the late '60s and early '70s, until the press ceased parroting the military's briefings and press releases about the Vietnam War and actually began to cover the war, liberals vigorously and correctly (in my not always humble opinion) tried to persuade us that the press had an "establishment" or even conservative bias.

I have no doubt that many journalists are anything from liberal-leaning centrists to liberal. But, as Robert Novak confessed yesterday (see blog title link), there are actually conservative jouralists. Shocking! He says of Robert W. Merry, whom he tells us is "a respected Washington journalist"

Merry over the years has been an objective journalist but considers himself a conservative and is said by friends to be a Republican who voted for Bush.

and then says

Merry, a former Wall Street Journal reporter, is president and publisher of Congressional Quarterly -- which seeks cool objectivity rather than passionate advocacy.

You will have to forgive me for thinking that Novak is engaged here in a bit of special pleading. And please forgive me for thinking that Novak doth protest too much in twice assuring us of the man's objectivity.

What is so astonishing about the Right Wing claim of a liberal journalistic bias is that conservatives, by and large, and even right wingers (think "Robert Murdoch") -- that is to say, rich people and corporations -- own most of the mainstream media outlets. In Columbus, Ohio, for instance, we are forced to endure two penurious Sinclair-owned television stations that on occasion unapologetically takes explictly right wing actions, especially during the recent election such as broadcasting an anti-Kerry documentary.

Now, I am supposed to restrict myself to language here so I must quit having fun at Novak's expense, who, after all, is a very easy target being autoparodic, as he is. Let me say, first, though that when one claims that this or that specific article is biased it is imperative that one factor in one's own bias in the assessment. Of course, we rarely, if ever, do that. But this is very important. Imagine that the MS Word Ruler represents the range of political positions of fairly normal people with, say, the range between "2" and "4" representing the political middle allowing for liberal-leaning and conservative-leaning centrists, with 0.5 representing a doctrinaire liberal position and 5.5 representing a doctrinaire conservative position. Naturally, if you are sitting at 0.5 and read a newspaper article on Iraq taking a 3.5 point of view, you will tend to see it as taking a conservative (as opposed to conservative-leaning centrist) or even right wing position because the position is so far from yours on the political scale.) The same, of course, is true of someone sitting at 5.5 reading an article that takes a 2.5 point of view. That article will likely be perceived as having a liberal or even leftist or downright Commie perspective. Putting aside my spatial metaphor, I think we could all agree (if we are honest and reasonably good at self-appraisal) that we tend to see our positions as "correct" and "objective" and those that deviate from them as incorrect and probably biased.

And now, finally, something about language. I replicated a study done years ago on how Time magazine reported on the speech of Presidents Truman, Eisenhower, and Kennedy in my book The Language of Politics", now available for the shocking price of $200 (don't buy it, of course). The study focused on the verbs of reported speech, the most neutral of which is "say." Time employed more than just verbs, often resorting to more colorful expressions: "said with a big grin/scowl" is the sort of thing I have in mind. The study found a negative bias toward Truman and a positive bias toward Eisenhower. In my study, I focused on how Time, Newsweek, and USN&WR reported on the speech of President Reagan and Senator Mondale over a ten week period following the political conventions. I took the various verbal characterizations of the speech of these two men, gave them neutral content, and asked subjects (my students) to evaluate each with respect to various scalar predicates. I asked whether the language used represented the speaker as excitable-calm, rash-cautions, inaccurate-accurate, unsuccessful-successful, and weak-strong. No matter how I bent the data, I could find no statistically significant, systematic bias for or against either candidate though there was a consistent tendency to portray Reagan more positively than Mondale. But, as I said, this result was not statistically significant.

Though it would never occur to Robert Novak to do so, claims of bias or objectivity need to be defended with facts for otherwise they will likely merely be products of one's own bias.