When the Media Do Not Merely Report Reality, but Shape It

How the endless flow of events becomes the world we see in the news

Every second, countless events take place around the world. People are born and die, governments make decisions, companies collapse, armies advance or retreat, scientists publish research, courts hand down judgments, people protest, doctors treat patients, criminals commit crimes, and millions of people live entirely ordinary days. Yet most of these events never become news.

When we open a news website or turn on the evening news, we are therefore not looking at the world itself. We are looking at a selection of the world. Someone had to decide what was important enough to publish, what would become the lead story, which information would be reduced to a brief item, who would be invited in front of the camera, which words would be used in the headline, and which details would remain outside the article.

This process is not necessarily malicious in itself. No media outlet can report everything. Selection is unavoidable. But it is precisely in this unavoidable act of selection that media power begins. Between an event and the audience there is always a series of decisions, filters, and interpretations. Once we understand these filters, it becomes clear that the question of media truth is not only whether an individual sentence is correct or incorrect. Just as important is why this particular sentence was selected, why this particular story was published, why it was given so much space, and what remained outside the frame.

View original archival source

Before an event becomes news

In journalism theory, the process by which some information passes through editorial gates while other information does not is known as gatekeeping. The concept is far older than the internet. Research into the history of journalism shows that questions about what is worth publishing, who may judge the importance of events, and what characteristics make something “news” were already taking shape during the professionalization of journalism in the late nineteenth and early twentieth centuries. Editorial judgment gradually became one of the central functions of the journalistic profession.

Every editor therefore faces the same problem: there is more information available than can be published. The death of one person may be a local news item, while the death of another becomes the front page of newspapers around the world. A protest by hundreds of people may be an important political story or an event a media outlet does not mention at all. A new study may be presented as a scientific breakthrough or disappear among thousands of other studies published that same week.

Many factors influence selection: proximity, the number of people affected, political importance, conflict, drama, novelty, the prominence of the people involved, the availability of images, time constraints, audience interest, and editorial judgment. But whenever space is given to one event, there is necessarily less space for others.

Every front page is therefore both a list of what was selected and an invisible list of everything that was not.

View original archival source

What will we think about today?

In 1972, Maxwell McCombs and Donald Shaw published a study that became one of the foundational works in research on the influence of mass media. During the 1968 U.S. presidential campaign, they compared the issues emphasized by the media with the issues that undecided voters in Chapel Hill considered important. The study laid the foundations for the theory of agenda-setting. Its central finding was not that the media can simply tell people what to think. The phenomenon is subtler: by choosing which issues receive attention and how much attention they receive, the media can significantly influence which issues the audience perceives as important.

Imagine a country in which the rate of a particular type of crime remains roughly unchanged. If the evening news begins showing graphic footage of crimes every night, interviewing victims, and displaying maps of dangerous areas, the public's sense of the importance of crime may change even though the statistics themselves have not changed significantly. No individual story necessarily has to be false. The robbery happened. The assault happened. The victim exists.

But a month of daily coverage can create a different sense of the world than a month in which the same events received only a few brief mentions. Similarly, a media outlet can emphasize inflation, immigration, climate change, war, an epidemic, corruption, or crime day after day. The more space a topic occupies, the more it becomes part of the public agenda. The question “Is the story true?” therefore captures only part of the problem. We also need to ask:

Why is this particular story receiving so much of our attention?

The same event, a different story

Once an event has been selected, the next decision follows: how should it be presented? Researchers call this framing. Robert Entman described framing as the process of selecting certain aspects of perceived reality and making them more salient in a message. A frame can influence what we understand to be the problem, what we see as its cause, how we evaluate the event, and which solutions appear reasonable. Imagine the same protest and three possible headlines:

“Thousands of citizens demand a change in policy.”

“Protesters once again paralyze the city center.”

“Police prevent riots at mass protest.”

The event may be the same. The first frame directs attention to the protesters' demands, the second to the disruption caused to others, and the third to public order and the police. All three articles may use entirely accurate facts. That is precisely why framing is harder to recognize than a direct lie. The frame does not have to reside in an incorrect fact. It may reside in the choice of a word, a photograph, the first paragraph, an interviewee, or the question with which the article begins.

The difference between the terms “protesters” and “rioters” is not merely stylistic. The second term already contains an additional interpretation of the event. Likewise, it matters whether a person “dies,” is “killed,” “loses their life,” or becomes a “victim of an attack.” Different formulations can emphasize the perpetrator, responsibility, and the act itself in different ways.

You do not have to lie to shape the world

One of the most common misconceptions about propaganda and media bias is the idea that manipulation must necessarily consist of false information. It does not. Imagine two media outlets reporting on the same country for a year. The first systematically publishes stories about economic growth, new companies, scientific achievements, successful infrastructure projects, and positive social trends. The second publishes stories about corruption, poverty, failed companies, crime, and problems in the health-care system. Assume that every individual story is completely true.

After a year, the readers of these two outlets may nevertheless form very different pictures of the same country. The first will be more frequently exposed to evidence of progress, the second to evidence of decline. Neither outlet had to tell a single lie. It was enough to select different parts of reality. This is selection. Its other side is omission.

When we analyze a media outlet, it is therefore not enough to ask what it published. Sometimes it is also important to ask what it did not report. If five media outlets publish a particular story on the same day and a sixth does not, that alone proves nothing. There may be many entirely legitimate editorial reasons. But if, over a longer period, a recurring pattern emerges in which one type of information systematically receives space while a comparable type almost never does, selection itself becomes an object of analysis.

The media picture of the world can therefore be changed profoundly without inventing events.

The whole can be misleading even when the individual selected parts are true.

Who gets the microphone?

The next filter is the choice of sources. A journalist cannot personally verify everything they report on. They need people and institutions that provide information: governments, police, armed forces, doctors, universities, companies, non-governmental organizations, eyewitnesses, researchers, and other journalists. But the choice of source already affects the story.

If an article about economic policy is based mainly on government statements, its perspective will be different from an article based mainly on trade unions or business owners. If reporting on a war relies almost exclusively on information from one belligerent, the audience will receive a different picture of the conflict than if the journalist has access to several mutually independent sources.

Nor does the label “expert” by itself guarantee that there is only one possible interpretation of a given field. In complex fields, differences in methods, data, and interpretation exist. This does not, of course, mean that every opinion deserves equal space. Evidence is not a democratic vote, and a well-supported expert finding is not equivalent to an unfounded claim merely because two opposing statements exist. But an important question remains:

Who was given the opportunity to explain the event, and who was left without a voice?

In wars, epidemics, and political crises this is especially important. Information is often incomplete, time pressure is intense, and governments and other large institutions have far greater capacity than individuals to quickly produce statistics, statements, experts, press conferences, and footage.

View original archival source

The power of the first line

A large part of the audience encounters a news story mainly through a headline, a phone notification, or a social-media post. A headline is therefore not merely a label for an article. It is often the most important part of the message a person sees at all. Research by Ullrich Ecker, Stephan Lewandowsky, and colleagues has shown that subtly misleading headlines can influence what people remember and how they draw conclusions about the content, even when the article itself is more accurate. The initial frame can therefore leave a trace even after the full text has been read.

The headline:

“Scientists prove that X causes Y”

creates a different impression from:

“Study finds an association between X and Y.”

If the study found only a correlation, the first headline changes both the level of certainty and the meaning of the finding. Similarly:

“Country preparing an attack”

is not the same as:

“Intelligence service claims the country may be preparing an attack.”

The second version tells the reader three important things: who the source is, what the level of certainty is, and that the information is not a directly verified fact. Research shows that even small changes in so-called epistemic language in headlines can alter how readers understand the status of a claim. Aaron Chuey, Yiwei Luo, and Ellen Markman showed that the difference between verbs such as “believe” and “know,” for example, affects whether readers perceive a claim more as an opinion or as an objective fact.

Words such as “claims,” “reportedly,” “probably,” “believes,” “knows,” or “confirmed” are therefore not insignificant linguistic decoration. They are part of how a media outlet tells its audience how much certainty to assign to a piece of information.

Fear and the economy of attention

A media outlet is not only an intermediary of information. Most media also operate within an economic system in which attention is a valuable commodity. For commercial online media, a click means a visit. A visit may mean advertising revenue. More time on a page, more views, and more shares therefore have direct or indirect economic value. This raises an important question: which kinds of headlines are people most likely to open?

A large study published in 2023 in Nature Human Behaviour analyzed more than 105,000 headline variants from the online outlet Upworthy, shown more than 370 million times. Because different versions of headlines were randomly displayed, the researchers could directly compare their effect on clicks. The result was clear: negative words in a headline increased the probability of a click on average. For a headline of average length, each additional negative word was associated with an approximately 2.3 percent increase in click-through rate, while positive words reduced the click-through rate.

This result does not mean that every media outlet deliberately creates fear or that every negative headline constitutes manipulation. The study was conducted on a specific online outlet known for headline optimization, so its numerical findings cannot automatically be generalized to every newspaper and every audience. But it demonstrates something important:

negativity can generate attention.

If the business model rewards attention, that is an incentive we cannot ignore when analyzing contemporary media.

When repetition starts to sound like proof

There is another mechanism that is not unique to media but is especially important in a mass information environment. Repetition. If we hear a particular claim again and again, its familiarity alone can increase the feeling that it is true. In psychology this is known as the illusory truth effect. A systematic review and meta-analysis published in 2026 in Nature Communications combined 182 studies, 366 measured effects, and 31,184 participants. The authors found a small but consistent average effect of repeated exposure on judged truth, although the strength of the effect varied considerably between studies.

Repetition is therefore not magic. It does not mean people will accept any claim merely because they hear it often enough. Prior knowledge, the plausibility of the claim, warnings about its truth status, and context all matter. But the effect exists. Another study found that even a single previous exposure to false information increased the likelihood that participants would later want to share it. Part of this association was explained by an increased sense that the claim was accurate.

This is why it is not enough to examine individual articles when analyzing media. One headline may be a coincidence. A long series of almost identical headlines is a pattern.

Does this require a conspiracy?

No. And this is where the problem becomes more interesting. For different media outlets to create a similar picture of the world, their editors do not necessarily have to meet secretly in the same room and receive identical instructions. They may follow the same news agencies. Use the same government sources. Invite the same experts. Watch one another. Work under similar time pressures. Compete for the same audience. Apply similar criteria for what counts as news. They may also share a similar cultural or professional view of a particular issue.

The result can be highly uniform reporting without any central command. But the opposite possibility is not merely theoretical either. History contains examples of direct state propaganda campaigns, shared corporate directives, political pressure, and planned efforts to influence media. We will examine such cases later in the series. So the question should not be:

“Is everything a conspiracy, or is nothing a conspiracy?”

A more useful question is:

“Which mechanism can be demonstrated in this particular case?”

Sometimes the answer will be commercial interest, another time political sympathy, a third time routine reliance on official sources, a fourth time direct editorial instruction, a fifth time haste, and a sixth time an entirely ordinary journalistic error. If we explain everything with a single theory, we are no longer analyzing media. We are merely replacing one simplified story with another.

When the editor is joined by the algorithm

The classical media model had relatively recognizable gatekeepers. Newspaper, radio, and television editors decided what would reach the public. Today the system is more complex. The editor still chooses the article, but its visibility is then also shaped by search engines, social networks, recommendation systems, influencers, and users themselves.

The Reuters Institute's Digital News Report 2025 identifies a continuing shift in audiences away from traditional news channels toward social networks and video platforms. The role of podcasters, online creators, and other personalities is growing, while the information environment is becoming increasingly fragmented. At the same time, traditional news brands remain an important part of the information ecosystem.

The way content is ranked can itself affect what we see. Research on systems that rank content according to user engagement has shown that, in certain environments, such ranking can give greater exposure to emotionally charged and divisive content than chronological presentation. The old gatekeeper has therefore not disappeared. It has gained company. Between an event and a person there may now be a journalist, an editor, a news agency, a platform, a recommendation algorithm, an online creator, and finally our own friends who share the information.

Every link in the chain can increase or reduce the visibility of a story.

View original archival source

The most misleading picture can be made entirely of true pieces

Imagine a jigsaw puzzle. We have a thousand pieces available, but select only one hundred. Every one of them is a genuine piece of the original image. But if we choose mainly the dark pieces, the assembled picture will look different than if we choose mainly the light ones. None of the selected pieces is fake. But the whole depends on the selection.

The information environment works in a similar way. Media can publish real photographs, real statements, real numbers, and real events, yet still create an incomplete or one-sided picture of the world. This does not mean journalism itself is propaganda. On the contrary. Good journalism tries to reduce precisely these limitations: it verifies sources, seeks broader context, separates fact from opinion, signals uncertainty, corrects mistakes, and looks for information that may challenge the initial explanation of an event.

But we can assess the quality of journalism better if we also understand the points at which information can change.

What will we look for in this series?

In the articles that follow, we will not begin from the assumption that media are always malicious or that everything they publish is propaganda. Nor will we assume that an alternative source is automatically more honest simply because it opposes mainstream media. For important stories, we will examine what was selected and what was omitted, who was given a voice, who the original source of a claim was, what level of certainty existed, how the headline changed the information, which words were used, how many times the same frame was repeated, and whether later evidence confirmed, supplemented, or called the original story into question.

In wars, we will examine how public consent for the use of force is created. In war reporting, what happens when a journalist depends on access provided by one side. In Israel and Gaza, how language, source selection, and the representation of victims affect the image of the conflict. In epidemics, how crisis communication works. In disinformation, who draws the boundary between incorrect information, legitimate disagreement, and content labeled dangerous. And in Slovenian media, we will apply the same questions to a concrete corpus of headlines and articles.

We will not collect only bad examples. We will look for good ones as well. If we want to investigate selection and propaganda, we must not make the same mistake ourselves by presenting only material that confirms our initial view. The goal is therefore not to determine which media outlet we should trust blindly. The goal is to understand something more difficult:

how, between an event and us, a story is created that we ultimately experience as reality.