Should Journalism Take Sides?

In recent weeks, in Italy there has been quite a bit of debate about neutrality in journalism.

This debate stems from the Gaza war — and I'm deliberately using the definition from Wikipedia, because I don't want this newsletter to become yet another place where people divide over which side to condemn and which to absolve or defend (especially since there aren't just "two sides": beyond Israel, meaning Netanyahu's Israeli government, and Hamas, there are Palestinian and Israeli civilians, settlers, opposition movements, and so on).

This debate emerged in Italy through a public exchange between two prominent figures in digital journalism. Francesco Costa, editor-in-chief of Il Post (Italy's leading digital-first newspaper that insists on clear explanation, fact‐checking and providing context), dedicated an episode of his podcast Wilson to explaining why, if news organizations want to do their job properly, they shouldn't take sides. Costa argues that a newspaper shouldn't take positions — it should report facts as accurately as possible, give voice to diverse viewpoints, and let readers form their own opinions. The goal is objectivity, and taking sides means abandoning that ambition, turning journalism into propaganda (even when it's propaganda for a just cause), and above all losing credibility with anyone who doesn't share that position.

Alberto Puliafito, a journalist and media consultant, responded in his newsletter The Slow Journalist that this is an illusion. Creating a newspaper is inherently a political act: you choose what goes on the front page and what doesn't, which sources to trust, how to write headlines, how much space to give a story. The idea of being neutral risks becoming a convenient alibi for not having to explicitly justify your editorial choices. Better to be honest: declare where you stand, make your criteria transparent, and allow readers to evaluate information knowing from which perspective it's being presented.

Of course this debate is not limited to Italy or Europe. The question of journalistic neutrality has become particularly fraught in recent years across all the media landscape. Perhaps the most striking recent example came in October 2024, when The Washington Post declined to endorse either candidate in the US presidential race, breaking with decades of tradition. The decision, reportedly made by owner Jeff Bezos over the objections of the editorial board, sparked massive subscriber cancellations and staff resignations. The stated reason was a return to "neutrality" — and that brings back to our topics.

Now, I'm a journalist myself. I cover culture, which doesn't completely shield me from dealing with delicate topics, though not as systematically as political reporters. But this isn't just professional interest: I studied philosophy, and the question of objective knowledge is central to epistemology (the branch of philosophy that investigates knowledge itself). My impression is that the debate about neutrality and objectivity in journalism is often poorly framed — it uses categories that don't work well for understanding what actually happens when we produce or consume news.

Let's Talk About "Subjective"

Doing philosophy means, first and foremost, clarifying the terms we use. So this text starts with some terminological premises — hopefully not too boring but (I think) necessary. The fact is, there's ambiguity around the terms "objective" and "subjective." I discovered this reading John Searle's The Construction of Social Reality, where in one of the first chapters he explains how, when talking about subjectivity and objectivity, we need to distinguish between ontology and epistemology — between things as they are and how we know them.

Something is ontologically subjective when it depends on a subject's experience to exist. The pain I feel when I stub my toe on a coffee table is ontologically subjective: it exists only insofar as I experience it. If I weren't there, that pain wouldn't exist anywhere. Conversely, something is ontologically objective when it exists independently of any subjective experience: the coffee table and my pinky toe exist even if no one perceives them.

That's about things as they are. Then there's the problem of how we know them. A statement is epistemically objective when its truth or falsity depends on facts, not on the opinions or preferences of whoever formulates it. "The temperature in this room is 22 degrees Celsius" is epistemically objective: it's either true or false, regardless of what I think (even if stated in Fahrenheit: 72 degrees). A statement is epistemically subjective when its truth depends on the attitudes, preferences, or viewpoints of the subject. "This room is pleasantly warm" is epistemically subjective: it depends on how I feel, and someone else might find it cold. The interesting case of apparent temperature is ontologically subjective — it depends on how humans perceive temperature, accounting for wind and humidity — but epistemically objective, since we have a precise formula for calculating thermal flux.

Is this distinction relevant to journalism? I think so.

Consider a reporter covering a location devastated by an earthquake: destroyed buildings, people describing their fear, rescue operations. Their account — stretching the terms a bit — is ontologically subjective: they're recounting their personal experience, what they see from their particular viewpoint. But epistemically, their account is objective: they're reporting verifiable facts about what they saw, heard, and listened to, and their statements are independently verifiable.

The analyst's case is different — someone trying to be ontologically objective and reconstruct "how things went" beyond personal perspectives. They read reports, cross-reference sources, attempt a comprehensive view. But there's the risk of being epistemically subjective in choosing sources, evaluating information, and deciding which interpretations to follow. Puliafito gives the example of Israeli hostages and Palestinian detainees; it seems difficult to consider people captured by Hamas on October 7 as "prisoners of war," but calling Palestinians "detainees" means ignoring all the investigations into the illegality of those arrests.

The Viewpoint

The distinction between objective and subjective in an ontological sense is relatively clear and simple (even when Searle introduces institutional facts that constitute social reality, but that's another story). Objective is physical reality, what another philosopher, Karl Popper, calls "World 1"; subjective is the set of mental and psychological states, which Popper calls "World 2" (then there's "World 3," which is basically Platonic ideas, but again that's another story).

But when we talk about knowledge, things are more nuanced. There's a continuum, and ultimately what we call "pure objectivity" is more of a regulative ideal, a goal we aim for knowing we can never fully reach it.

Much of contemporary epistemology reminds us that knowledge is always situated. There's no "view from nowhere," as philosopher Thomas Nagel calls it to describe what, in theological terms, we might call God's perspective — seeing everything from above and outside. We're always within some context, with a body, a history, a social position, a set of assumptions we carry with us. And this applies to science too, which should be the realm of objectivity but always produces "situated knowledges." This doesn't mean those knowledges aren't true: the distinction between ontology and epistemology serves precisely to keep the two discourses separate. Vaccines are effective in preventing disease from any perspective, conceptual scheme, or cultural frame from which you view it. It's just that these vaccines were developed by people born and raised in a particular culture and social environment, responding to particular ethical, social, and economic incentives. And we need to acknowledge this context — also because it might not coincide with the context in which those same vaccines will be distributed.

Recognizing that knowledge is always situated doesn't equate to "all opinions are equal"; rather, it means recognizing that pursuing objectivity doesn't mean trying to see everything from above, but comparing views from the ground. It's an intersubjective construction that emerges from comparing different perspectives, from the continuous attempt to correct our biases through dialogue.

Again, how does this apply to journalism? One level is fairly mundane: a piece of news should always have two independent sources before being considered true. And then there's a second level: taking into account different, even contrasting viewpoints, putting them in relation and context. Not the pursuit of neutrality — which usually means self-certification — but providing a picture of complexity. Keeping in mind that some positions are more partial than others, in the sense that they're less capable of "opening up" to other viewpoints. Not (necessarily) out of bad faith: often it's simply a lack of practice in confronting minority or marginalized positions.

And here we arrive at a delicate point. Objectivity as a regulative ideal — as a horizon to strive toward while knowing we'll never reach it — can be useful. But objectivity as an accomplished claim, as an assertion of having already reached a neutral and impartial perspective, can become an instrument of power, because "objectivity" simply becomes the way to universalize one's own perspective and exclude other viewpoints, perhaps confusing ontological objectivity with epistemic objectivity.

So, Should Journalism Take Sides or Not?

Having exhausted this initial premise, let's return to the opening question. Should (good) journalism take sides? And here's the disappointing answer: the question is poorly posed.

Good journalism is what aspires to objectivity — it wants to tell things as accurately as possible — but realizes it's always situated, that it has a viewpoint that might be better than others (it has access to information, resources, or expertise that aren't available elsewhere) but is still partial.

Good journalism is what's aware of these limits and takes them into account. From this perspective, I identify more with Puliafito's observations. Saying that "journalism shouldn't take sides" risks hiding all this complexity, risks making people believe that a truly neutral position exists, when in fact every editorial choice — what to put on the front page, how much space to give a story, which experts to interview — is already taking a side.

However, there's another aspect we haven't considered. And therefore another philosophical digression (sorry).

Not Everything Is Description

Here we've taken for granted that we're talking about knowledge, about describing reality, about what the philosophy of language calls "apophantic discourse" — statements that can be true or false, that say how things are.

But language isn't only this. There are commands ("close that door"), promises ("I swear I'll come"), requests ("could you pass me the salt?"), and many other things that don't describe the world but try to change it, or establish relationships, or simply express something.

In journalism, commands or promises are rare, but many opinion pieces contain recommendations ("the government should..."), value judgments ("this is an unjust law"), exhortations ("we must react"). These are all prescriptive statements, not descriptive ones.

To understand the difference between a descriptive and a prescriptive text, John Searle (him again) spoke of "direction of fit" between words and world. There are statements that must fit the world: if I say "the cat is on the mat" and the cat isn't there, the problem is in my words. And there are statements where the world must fit the words: if I say "turn on the light" and you don't turn it on, the problem is in your failure to act.

The classic example involves grocery shopping (maybe we should update it to online shopping). The shopping list is prescriptive, the receipt is descriptive. The list says what I should buy and the direction of fit goes from words to world. If it says "jam" and I don't put it in the cart, I need to intervene by putting jam in the cart to make the world match the list. The receipt, instead, says what I bought and the direction goes from world to words. If it says "milk" but there's no milk in the bag, I intervene by correcting the receipt (along with appropriate apologies for the error).

Is journalism descriptive or prescriptive? At first glance, descriptive. Certainly news reporting is: it tells what happened; but even opinion pieces should start from facts before reaching judgments. However — and we're finally getting to the point that interests me — the philosophy of language loves to complicate things. There's a third dimension, beyond syntax (the structure of sentences) and semantics (their literal meaning): pragmatics.

Pragmatics studies how we use language in concrete contexts, and especially what we do with words beyond their explicit meaning. It's the realm of implicatures, implications, things that are communicated without being said directly. Paul Grice, the philosopher who founded modern pragmatics, formulated the "cooperative principle": when we speak, we assume our interlocutor is trying to be informative, relevant, clear. And we use this assumption to infer meanings that go beyond literal content. Classic example: to the question “Are you going to the party tonight?” I answer “I have an early meeting tomorrow.” At a literal level, I’ve said nothing about the party. Yet the reply is immediately understood: I’m implying that I won’t go, because staying out late would be inconvenient. The additional meaning—that I’m declining the invitation—is not in the words themselves but inferred through context and shared expectations. It’s what is called a conversational implicature.

Now, let's apply this to journalism. What makes something news? They teach this in all journalism courses: news is the account of an event that's new, relevant, and of public interest.

An event is new when it's not the norm, when it represents a break from the ordinary course of things. And it's relevant when it matters, when it has consequences, when — in some sense — it shouldn't be ignored. In other words: if something is news, it means it deviates from normality in an important way. And that it shouldn't simply be taken as a given: it requires attention, reaction, perhaps intervention.

In other words: journalism is indeed description, but it's description that pragmatically functions as prescription. When a newspaper writes "X happened," it pragmatically also communicates "X shouldn't have happened" or at least "X is something you should reflect on, form an opinion about, possibly act upon."

Take a concrete example. "Robbery downtown." It's a factual description, certainly. But pragmatically it communicates: robberies are a problem, they shouldn't happen (or at least not downtown, where respectable people and tourists are), it's important that you know they're happening, something must be done (surveillance cameras? armed citizen patrols? more funding for police? more resources to fight social exclusion?).

Moreover, in a city where ten robberies occur daily, a single robbery isn't news. It's normal. If in the same city robberies occur ten times a year, those ten will get some headlines (paradoxically leading to greater perceived insecurity). Sure, there are in-depth articles that partly escape this dynamic. But generally journalism lives on this tension between describing (telling the facts) and prescribing (suggesting that those facts require attention, judgment, action).

And here's the point: this pragmatic dimension is always present and is an integral part of the news selection and curation process. Thinking you're only describing what happens would be naive.

Balance makes the difference. If the prescriptive aspect completely overtakes the descriptive one, we have a problem. That's activist journalism that starts with a goal — usually political — and selects which facts to tell based on that goal, shaping the narrative to support a predetermined thesis. And here I find myself agreeing with Costa: journalism shouldn't take sides in this sense. It shouldn't start with a goal and then look for facts to support it. But — and this is crucial — it's not a matter of descriptive objectivity. It's a matter of prescriptive honesty.

Good journalism is what recognizes its prescriptive dimension but subordinates it to the descriptive one. It says: yes, I think this is something important to reflect on (prescription), but first I'll tell you as accurately as possible what happened (description), and then maybe I'll also explain why I think it's important (explicit prescription).

The Elephant in the Room

There's one last thing, and perhaps it's the most important (and perhaps also a bit embarrassing).

We've talked about descriptive discourse, which tells how we see reality. And about prescriptive discourse, which shows how we should change reality. But there's a third category, perhaps the most important in terms of quantity: discourse that doesn't describe or prescribe anything, because it talks about things that don't exist or simply have no practical consequence.

This is the case with fiction: films, novels, TV series that certainly can help us understand reality or imagine alternatives, but indirectly, in a mediated way. When I read a novel, I'm not acquiring information about the world (even though I may learn many things), I'm doing something else: I'm entertaining myself, I'm exploring possibilities. But it's also the case, more mundanely, with small talk that we use to kill time and, perhaps, build social bonds.

Journalism is descriptive, it's prescriptive. And it's also entertainment. How much information do we actually consume to know what happened or to understand what to do, and how much instead to pass the time, to confirm our ideas, to feel part of a community?

If you're thinking about gossip: yes, you're right. But actually it's a phenomenon that concerns all information. I read about the lawsuit that the "QAnon Shaman" (the guy with the horns who stormed the Capitol to prevent Biden's certification as president) allegedly filed against Trump. It's news that could have some relevance: it's the distancing of what was an important electoral base for Trump, which now feels betrayed. But actually I just read about the proposal to create a coin worth 40 trillion dollars, about how James Cameron allegedly stole the idea for Avatar from him, and other nonsense: I'm entertained and I feel superior to the many Trump supporters (including those who have nothing to do with the conspiracy fantasies embraced by the QAnon Shaman).

In this context, does it make sense to talk about objectivity, neutrality, description?

To Sum Up

I got carried away. I think a brief recap is useful.

Asking whether journalism should take sides doesn't help much, if you want to try to understand how we know and tell reality. And this for three reasons.

• Knowledge is always situated, and absolute objectivity doesn't exist. But this doesn't mean all perspectives are equal — some are more partial than others, and serious work consists in multiplying viewpoints, not entrenching yourself in your own.
• Journalism has a prescriptive component that's always present, even when it merely describes. Denying this dimension is dishonest; but if it becomes dominant, it turns into propaganda.
• Much of what we call "information" is actually identity entertainment.