For journalists and political scientists, a principal task when reading and writing news stories is to shield their audiences from bias, especially those not self-proclaimed, lurking in the texts that people unconsciously take in.
I’ve been asked all the time what I do to keep an unbiased account of the news and the world at large, and usually, people asked me to name some unbiased newspapers or journals. By seeking an unbiased news medium, people are looking for singular neutrality, the neutrality brought by one single less biased newspaper.
While I did name some ostentatiously biased newspapers, I seldom gave out unbiased ones. The reason is simple: there are no truly unbiased information sources. Bias is more than straight lies. Partial truths can also distort people’s perception of the whole reality, and the manners that the truths are displayed can add to that distortion.
In fact, the reason that the major newspapers managed to grow is that they got the right bias that attracted specific groups of audiences. This in turn incentivizes them to be even more biased, triggering a positive cycle that leads to huge biased news media. As a result, there are seldom cosmopolitan, unbiased news sources that aim at all the people around the world, and even if there are, they rarely succeed in growing and thus failed to be exposed to us. Despite being biased, mainstream media are efficient at providing large amounts of information and professional insights specific to each piece of news. Totally discarding them would be wasteful.
Thus, our choices are all more or less biased newspapers, and looking for less biased newspapers is not the whole story. To be even more neutral than the most neutral available sources, we cannot depend only on one source.
I argue for a synthetic neutrality. Compared to singular neutrality, synthetic neutrality has to be acquired by accessing and “synthesizing” more than one source of information. We can use a mathematical approach to explain how this works.
News media have informal or formal mainstream ideologies that determine their stances on specific issues and news stories. In a simplified case, each stance can be attached a bias index. Its sign measures the direction of its bias, and its absolute value measures the severity of its bias. According to its bias index, a stance can be put at a corresponding location on a numeral axis called the bias axis.
In this sense, the bias index of your information and idea regarding specific news and issues is the sum of the bias indices of all the information that you accessed from different sources. Our goal of maximizing neutrality thus becomes minimizing the absolute values of the sum of all bias indices.
Based on our prior reasoning, no news media can have a bias index of zero. After brushing off the news media with intolerable bias values as high as, say, ±10,000, what we are going to do with the rest of the news media with bias values like +10, -8, +4, +6, and -11?
If we only choose one news medium, we would choose the one with a +4 bias index, optimizing our singular neutrality. However, if we choose all the news media, our synthetic bias index becomes +10-8+4+6-11 = +1, a value even smaller than the +4 by singular neutrality. By balancing off the respective bias of different media, we get less bias on our own.
In reality, synthetic neutrality requires some knowledge and reason from the audiences themselves. The estimation of the bias of an information source needs an understanding of news media and the issues being covered. Also, the synthesis of bias is not automatic like the mathematical summation above, which is a simplification. The summation process in our model here corresponds in the reality to the processing of the information we get after reading different sources of information. This can be active brainstorming in which we on purpose analyze the news stories we read, or it can be more passive as our intuitions and habits produce intuitive perceptions, feelings, and ideas based on the information we gathered behind the scene. The latter occurs in the subconscious part of our mind.
As we learn more about the world, our skills in active brainstorming would be instantly improved, while we have to polish the ways we see and think of the world inch by inch to gradually refine our inner intuitions and habits that help produce synthetic neutrality. To both, education and experiences are crucial. They equip us with the intellectual devices to synthesize the neutrality from news sources.
Thus, to maximize the neutrality of our understanding of reality, we should get information from many sources rather than just one, so that we can synthesize a more neutral point of view of our own based on the existing points of view that are more biased. To improve our performance in the synthesis process, education and experiences of different sources are vital.
The discussion of my model has not been exhaustive yet. Some clarifications and extensions of the concepts that I mentioned will help consummate my model.
Bias is an abstract concept and cannot be measured directly. Measuring bias is conceptually different from measuring the source of bias, which is usually more tangible. However, they are relevant to each other. Essentially, we can estimate one’s bias through measuring one’s source of biases. For example, political ideologies are a source of bias in news reporting. Although measuring one’s political ideology is not essentially the same as “measuring” one’s bias in news reporting, still, in our simplified case, we can consider the political ideology of a news firm as roughly proportional to its bias, thus estimating the news’ bias from its political ideology. In this case, the bias axis and the axis of the source of biases perfectly supersede each other. If a news firm has a value of, say, +7 on the axis of political ideology (a specific axis of the source of biases), then in a simplified sense, we can say its bias index is +7 as well.
Nonetheless, the world is more complicated, as one’s biases can be caused by more than one source of biases. Economic ties and social manners of a news firm, for instance, have a role in producing the firm’s bias as well as its political ideology. How can we take all the sources of biases into account in a conceptually precise way? This requires more intensive mathematics and is worth another article to elaborate on. In my next article in the series, I will talk about this more complicated form of my model as well as some of its real-life implications and some case applications.