The idea of free speech sparked into life 2,500 years ago in Ancient Greece – in part because it served a politician’s interests. The ability to speak freely was seen as essential for the new Athenian democracy, which the politician Cleisthenes both introduced and benefited from.
Today, we debate the boundaries of free speech around kitchen tables and watercoolers, in the media and in our courtrooms. The right to freedom of thought, however, is more rarely discussed. But thanks to the growing influence of social media, big data and new technology, this “forgotten freedom” needs our urgent attention.
In democratic societies ruled by ballots not bullets, power is won through persuasion. Efforts at persuasion are ramping up: there will be more than 50 national elections involving half the world’s population in 2024, including in seven of the ten most populous countries. The results will shape our century, making it paramount that we protect people’s ability to think and vote freely.
But corporate and political actors know more about how our minds work than we do. They activate our biases rather than appeal to our reason, push us to share information without thinking, and control our attention to the point of addiction.
Advances in neuroscience may heighten this threat to free thought. Elon Musk and Mark Zuckerburg are among those in a race to read our minds with the help of artificial intelligence (AI). In 2021, the UN warned of the risks of neural technologies predicting, identifying and modifying our thoughts. Manhattan projects of the mind threaten to make lab rats of us all.
We could respond by calling on our right to freedom of thought. It’s there waiting for us, created in 1948 by the Universal Declaration of Human Rights (Article 18) and later becoming international law. But anyone reaching for this right may be horrified to find it hollow, bereft of any clear definition and unfit for purpose.
In recent years, the UN has sought to give this right more substance. One of its special rapporteurs, Ahmed Shaheed, has made a series of recommendations (which I will outline) that should, eventually, lead to a better defined, more muscular right to free thought. This process has promise – it could help shield our thoughts from prying eyes and protect our minds from manipulation.
Read more: To what extent are you truly free?
But it also has the potential for harm. In international law, freedom of thought is an absolute right. This means it could run roughshod over other important concerns. Activists could, for example, use this right to silence their political opponents by claiming their opponents’ speech is manipulating thoughts.
This right could also go wrong by failing to protect all forms of thought. We must ask where “thinking” ends and “speaking” begins in today’s world: should performing an online search, writing a personal diary, or asking a question in a WhatsApp group be regarded as forms of thought, or outright speech?
This article is part of Conversation Insights
The Insights team generates long-form journalism derived from interdisciplinary research. The team is working with academics from different backgrounds who have been engaged in projects aimed at tackling societal and scientific challenges.
History suggests the right to free thought will only become globally relevant if political factions or states use it to serve their purposes. And this looks increasingly likely as accusations of mental manipulation and “thoughtcrime” creation fly between the political right and left, and the US looks for new weapons in its developing cold war with China.
For both technological and (geo)political reasons, the right to freedom of thought’s time may have come. Whatever it is decided to mean, it will bind us all. As I argue in my new book Freethinking, this makes it crucial that we can all have input into its design.
Free thought past
The term “freethinker” came into common use during the Enlightenment in late 17th-century Europe, describing people who questioned religious authorities. Today, freethinking typically refers to being guided by evidence and reason, not authority – although this hasn’t stopped people who play fast and loose with evidence appropriating the term too.
Up against the freethinkers are those who seek to control thought to achieve and cement power. George Orwell’s classic vision of this threat, Nineteen Eighty-Four, actually came out 20 years after Japan’s Peace Preservation Law had already termed many people on the political left as “thought criminals”.
In Orwell’s novel, the ruling party aims to “extinguish once and for all the possibility of independent thought”. Beginning in childhood, people are taught to deny the evidence of their eyes and ears. No one is to be left alone to think – yet nor are they able to think with others. People are encouraged to stop themselves on the threshold of dangerous thoughts, as the terrifying Thought Police find and punish those who commit “thoughtcrime”.
The ruling party also uses extensive surveillance, parallels to which can be seen today. Consider the effects of Edward Snowden’s 2013 revelations that the US was heavily surveilling the internet. This led to a 10% drop in internet searches that could have got Americans in trouble with their government, such as “domestic security”, “nuclear” and “organized crime”. Americans’ suspicion that they were being watched harmed their freedom of thought.
New technologies drive new laws. Just as photography spurred a right to privacy in 1890, today scholars want to develop the right to freedom of thought in response to neurotechnology and the digital world. This means confronting the gulf between the extensive lauding of this right and the bewildering neglect of what it practically involves. Enter the UN.
Free thought present
In October 2021, special rapporteur Shaheed presented a report on the right to freedom of thought. To underpin this right, he proposed four pillars which I summarise as follows, including some questions I think we should consider about them:
1. Mental privacy. People cannot be forced to reveal their thoughts. This means we must scrutinise technological developments that open new windows into our minds. But should our minds always be private?
2. Mental immunity. People cannot be punished for their thoughts. This idea has existed since ancient Rome, though today we need to decide what exactly should count as a “thought”.
3. Mental integrity. People and organisations cannot alter others’ thoughts without permission. We know subliminal advertising is wrong because it bypasses our conscious mind – but beyond this, we enter a grey zone. When does persuasion become impermissible manipulation?
4. Mental fertility. This enshrines a government’s duty to create an enabling environment for free thought. But will governments really do this if free thought challenges their power? And even if they want to, how can they design a society that promotes free thought?
To build on these pillars, we need to answer basic questions about what thought is and what makes it free. Before we can protect thought, we must first define it.
Free thought future
Traditionally, the law views thoughts as happening inside our brains. Yet philosophers (and, increasingly, psychologists and technologists) have long claimed that thought “ain’t just in the head”, proposing that our minds extend into the world.
A notebook can be the functioning memory of someone with dementia. Writing in a diary, as Winston Smith did illegally in Nineteen Eighty-Four, can also represent thinking. Writing doesn’t only express thought; sometimes “the thinking is the writing”. Similarly, some internet searches can be a form of thinking as we use them to question, reason and reflect.
If the right to freedom of thought is deemed to cover our “extended minds”, this will have important consequences. Authorities often access the internet search histories of people accused of crimes, using this as evidence. In homicide trials, searches such as “how to get rid of someone annoying” or “chloroform” have been cited in court. But if such searches are deemed to constitute thinking, then accessing someone’s search history could become a violation of their mental privacy.
Speaking aloud can also be regarded as a form of thinking – we sometimes speak to find out what we think. As novelist E.M. Forster asked: “How do I know what I think until I see what I say?”
But we also speak aloud in order to think with other people – and we may think better with others than we do alone. Thought can be at its most powerful when it is social, rather than the solitary act depicted by Auguste Rodin. So, for thought to be truly free, we require public as well as private thinking spaces.
To facilitate this, we may need a new legal concept of “thoughtspeech”. This would represent the thinking aloud we do with others in the name of “good faith truth-seeking”. Thoughtspeech could be protected as absolutely as the thoughts inside our head: while one could (and should) still disagree with others, attempts to silence or punish thoughtspeech would be a human rights violation.
However, an obvious concern is that this concept could be misused to justify hate speech that Europe, but not the US, has deemed illegal. The protestation that “I’m just asking questions” can easily be employed as a cover to demonise people. At the same time, a creeping prohibition of asking difficult or challenging questions is also potentially dangerous, not least to society’s minorities who seek to challenge the status quo. Where necessary, courts would have to decide whether the claimed thoughtspeech was genuine truth-seeking in good faith, or had darker motives.
To see these thorny questions in practice, consider how, in Ireland, both the left and right have argued that the proposed criminal justice (incitement to violence or hatred and hate offences) bill 2022 will create “thought crimes”. Section 10(3) of this bill states that if you possess hateful material that you haven’t shown anyone, and it is reasonable to assume this material is not just intended for personal use, then you are presumed to be breaking the law. The police could then seek a warrant to enter your premises and access this information.
One politician claimed this bill would not create thought crimes because it involved “production of material”, and that “nobody is ever going to be prosecuted for what they’re thinking inside their heads”. This illustrates the restricted view of thought that some politicians hold – and suggests that legislators could leave much of our thinking naked and vulnerable.
Once we have settled on what thought is, we must work out what makes it free. The EU’s Digital Services Act forbids online platforms from deceiving or manipulating users, or impairing their ability to make “free and informed decisions”. But what counts as manipulation or impairing free decision-making?
Psychology suggests that free thinking requires us to control our attention, be able to reason and reflect, and to not need superhuman courage to think aloud with others. This makes platforms and products problematic that either capture our attention to the point of addiction, or employ “dark patterns” that undermine reflection and reasoning.
For instance, in the “false demand” dark pattern, a shopping website may falsely tell you that “Abby in London” has just bought a pillow. This could undermine your reasoning by triggering a panicky sense of scarcity in you, or setting a false norm of others buying the pillow, making you more likely to purchase.
The designers of systems need not even intentionally play on our mental biases for their products to be problematic. For instance, listing candidates in alphabetical order on a voting ballot paper may seem neutral, but the candidate named first gains a small but measurable advantage. This is partly because we have a mental bias to prefer the first item on a list.
Some US states don’t use alphabetical ordering on ballots for this reason. Instead they randomly rotate the order of candidates’ names on ballots across districts. Yet, elsewhere, including in the UK, this alphabetical practice continues. Alphabetical ordering arguably violates voters’ right to freedom of thought.
I believe the right to freedom of thought should protect thinking wherever it occurs – in our heads, our diaries, on the internet, or when we’re engaged in good faith truth-seeking when thinking aloud with others. And crucially, to keep thoughts free, our environment must be designed and regulated to let us control our attention, think logically and reflectively, and not fear punishment for our thoughts. Unfortunately, new technologies threaten this ideal.
The power to punish thought
New technologies have the potential to undermine three of the UN’s pillars of free thought – mental privacy, immunity, and integrity. It is well known, for example, that social media can use knowledge of how our minds work to hijack our attention, discourage reflection, and facilitate the punishment of wrongthink, thereby harming our autonomy. Less well known is how social media revives an old social pattern that threatens free thought.
Our hunter-gatherer ancestors lived in egalitarian communities with no dominant individuals. This was due to a “reverse dominance hierarchy” which meant that, if someone tried to rise above others, the group would work together to humble, exclude or even kill this would-be “tall poppy”.
My book Spite (2020) explains how anthropologists believe this was possible due to humans’ ability to moralise, wield weapons and use language. Language in particular, especially gossip, helped coordinate actions against tall poppies. When agriculture was invented, larger groups, private property and recognised authority figures came on to the scene, enabling a more hierarchical form of life.
Social media has bought back the reverse dominance hierarchy. People online can unite to moralise and gossip, sometimes with the effect of bringing down individuals. And while this can helpfully check people who abuse their power, it can also harm freethinkers who disturb the status quo and undermine what they see as society’s “noble” (or ignoble) lies.
And not only can new technologies facilitate the punishment of thought, they also have the potential to powerfully manipulate our thoughts. AI will soon know exactly what to say to us to maximise the chances of us performing a desired behaviour. As OpenAI’s CEO Sam Altman has warned in relation to the 2024 elections:
What if an AI reads everything you have ever written online – every tweet, every article, every everything – then right at the exact moment, sends you one message, customised just for you, that really changes how you think about the world?
The power disparity between us and AI means that courts could deem AI to have improper undue influence over our minds.
New technologies can also uncover our hidden thoughts. This goes beyond what Big Brother was capable of. As Orwell wrote in Nineteen Eighty-Four, even the ruling party “had never mastered the secret of finding out what another human being was thinking”.
Today, brain-reading technology that detects thoughts via brain scans or neural interfaces threatens to uncover this secret. Meta (Facebook’s owner) recently showed it could determine what people were seeing by examining their brainwaves using magnetoencephalography (MEG) technology. Behaviour-reading techniques that use our actions, such as musical preferences or what we “like” on Facebook, can also be used to infer our internal states.
Now, imagine if a government had someone in custody suspected of having planted a nuclear bomb in a city. There would be a strong temptation to use mind-reading technology to identify the location of the bomb from that individual’s brain – but this would violate the suspect’s right to free thought. Some people, perhaps many, would argue that this right should be violated in such circumstances.
Perhaps the future could include places where free thought is legally limited. While this is a challenging idea, it would be ironic if we failed to think critically about free thought itself. Legal scholar Jan Christoph Bublitz has speculated on the idea of “zones of restricted freedom of thought” in places vulnerable to terrorism, such as airports. In these zones, our thoughts could be permissibly accessed by the state to prevent calamities.
Likewise, the philosopher Sam Harris has suggested that once mind reading technology can detect lies, it could be used to create “zones of obligatory candour”. These would be locations, such as courtrooms, where lies would be automatically detected from your brainwaves.
Yet, concerns about mind-reading technologies are frequently blighted by hyperbole, alarmism and exaggeration. One does not simply fall into an MRI scanner; one must consent. Once inside, we must cooperate with researchers for brain-reading to work. The extent to which our mental content can be accurately identified is often over-hyped, as it requires extremely specific conditions.
Read more: Does AI have a right to free speech? Only if it supports our right to free thought
It’s also important to recognise that the same technologies that threaten free thought can also benefit thought. Brain-computer interfaces, where we interact with computers simply by thinking, could boost our mental bandwidth. AI systems such as ChatGPT can stimulate new ideas. So, over-regulating these technologies could be seen as harming free thought.
Clearly, we need to protect free thought in response to new technologies. But in my view, overreacting with unnecessary laws won’t lead to freer minds – it will simply enable other people’s anxieties to rule our lives.
Protecting employees’ and users’ thoughts
Traditionally, governments were seen as the main threat to our freedoms. Today, corporations, particularly those involved in controlling flows of information such as media and technology companies, vie for this crown.
Such companies influence what information we do and don’t see. They can also overwhelm us with too much content, creating “reverse censorship” that harms our ability to think. Corporations also threaten free thought through their ability to fire employees for thoughtcrime, potentially in response to a public outcry.
The British philosopher Bertrand Russell warned a century ago that “thought is not free if the profession of certain opinions makes it impossible to earn a living”. Russell said this problem would grow unless the public insisted that employers controlled nothing in their employees’ lives except their work. Today, employers can influence their employees’ morality, opinions, and even voting decisions. As a starting point, we need laws that protect employees from being fired for their thinking.
To take another example, while the anti-discriminatory aims of implicit bias training are laudable, corporations could require employees to reveal their thoughts when doing it. The designers of a common element of this training, the implicit association test, admit it is “a method that gives the clearest window now available into a region of the mind that is inaccessible to question-asking methods”. Forcing someone into this training could therefore be a breach of mental privacy.
Read more: Social networking sites may be controlling your mind – here's how to take charge
Turning from employees to users, perhaps big tech companies should be required to design their products to support, promote and protect their users’ free thoughts. For example, social media platforms could set their default options to those that minimise the risk of addiction.
Woodrow Wilson once noted that chewing tobacco “gave a man time to think between sentences”. Big tech could provide digital gum in the form of options that give users time for thought, like timeouts before responding to posts or making purchases. X (formerly Twitter) already asks users if they want to share articles they have not yet read.
Similarly, search engines could offer options to show information that challenges users’ existing views rather than confirming their opinions. Websites such as Ground News already highlight which stories are primarily featuring on left or right media, helping people see what is happening outside their own thought bubbles.
Big tech companies such as Google and Meta assess their human rights impacts, including through independent assessments. But freedom of thought is often overlooked. And while the UN has issued its Guiding Principles on Business and Human Rights, these have been described as “woefully inadequate” by Human Rights Watch as they lack any enforcement mechanism.
It’s not just ‘them’, it’s us
It is not just governments and corporations that threaten free thought – we the people do too. Free thinking has often been risky. “Tell the truth and run,” an old Yugoslavian proverb counsels.
Throughout history, though, some societies have tried to protect free thought and truth-telling. The ancient Athenians had the concept of a “parrhesiast”, someone who spoke truth despite the risks. An example of this came when the aged statesman Solon challenged politician Pisistratus’s quest for power in Athens. After arriving at the Greek Assembly dressed in armour to highlight Pisistratus’s aim to use force, Solon declared:
I am wiser than those who have failed to understand the designs of Pisistratus, and I am more courageous than those who have understood but remain silent out of fear.
To benefit from the parrhesiast, Athenians had to be willing to bear what philosopher Michel Foucault calls “the injuries of truth”. In this parrhesiastic contract, the truth-teller risked speaking out and the listeners promised not to punish them. There again, Solon was not thanked for his contribution, being labelled mad by his colleagues.
Creating a safe space for truth requires a “deep enlightenment” that goes beyond simply educating people to think critically. Designing a society that protects and promotes free thought among its population at all levels could even include city planning.
A Brazilian colleague once told me how the design of the country’s modern capital, Brasília, with its lack of street corners, was meant to prevent people assembling and thinking together – because such thinking could one day threaten the ruling powers. Indeed, the Portuguese for street corners can translate as “points of solidarity”. The design of Brasília is an offence against free thought.
Rather, we need to design physical and virtual spaces that protect, promote and support “people’s public use of their reason”. This function was partially performed by coffee houses during the Enlightenment. New spaces should allow a diverse range of voices to be brought together in debate – in order to help us best find truth. Yet all of this hinges on simultaneously building a culture of trust that makes people feel safe to think.
The oxygen of freethinking
The principle of free thought is in trouble. Today, public thinking is difficult unless you are rich, reckless or anonymous. Online public spaces, such as much social media, typically prioritise engagement and profit over truth-seeking, and can exclude challenging views. A corporate-controlled mainstream news media routinely excludes or distorts important perspectives such as labour issues. Some academics feel compelled to publish ideas anonymously in outlets such as the Journal of Controversial Ideas. These are all warning lights of flashing failure on the dashboard of democracy.
The first freethinkers challenged religious authorities and were associated with egalitarianism and the political left. Yet they had their own “faith of the Enlightenment” – the belief that developing one’s own reason could create a better life. Today, as well as sharing this faith in reason, many of us have faith that liberal democracy creates the best form of life.
However, some modern freethinkers are pushing back against these faiths. Such individuals tend to be pro-hierarchy and on the political right. The conservative psychologist Jordan Peterson argues that we’re at the start of a “counter enlightenment”, while legal scholar Adrian Vermeule maligns the “evidence-based freethinkers of the quiet car” who won’t speak out about liberalism’s problems. Alternatively, so-called “Dark Enlightenment” thinkers such as Curtis Yarvin and Nick Land question the benefits of democracy.
Whatever you think of these views, an important question is: will the descendants of the egalitarian left, who used freethinking to challenge societal norms, support the hierarchical right’s freedom to do the same? Or do they regard the political right as in need of silencing rather than debating?
Of course, freethinking on the left is silenced too – including those who oppose the “religion” of capitalism. Consider what happened when a declared socialist, Jeremy Corbyn, ran for prime minister in the 2017 UK parliamentary elections. An academic report on his coverage by the mainstream media concluded by asking whether it was:
acceptable that the majority of British newspapers uses its mediated power to attack and delegitimise the leader of the largest opposition party against a rightwing government to such an extent and with such vigour?
Whatever one’s views on democracy, liberalism, capitalism, or any other important topic, freethinking on these issues can prove profoundly valuable. If someone’s ideas have value, we may adopt them to live better lives. If we adjudge them mistaken, we will still emerge with a better understanding of precisely why our own ideas are valuable, having remade them as living truths rather than dead dogmas.
Free thought is not merely about gaining more perspectives. It is about duelling perspectives. The left and right could find common ground not in a commitment to mutual cancellation, but in a renewed dedication to debate. We must embrace the value of thinking.
Unfortunately, we often find thought a painful effort. Evolution has shaped us to make decisions using minimal energy, pressuring us to become cognitive misers who are “as stupid as we can get away with”, as psychologist Keith Stanovich argues. Many of us are not merely disinclined to think but actively prefer electrocution to being left alone with our thoughts, according to one 2014 study.
The rise of generative AI threatens to make this situation worse. One vision of the future imagines a singularity where we merge with machines by connecting our brains directly to AI. But what if we approach a bifurcation point rather than a singularity? Humans could become a mere source of animalistic appetites and desires, while machines do the thinking for us.
If we abandon free thought, homo sapiens will have been a brief candle between ape and AI. Humanity’s flame cannot continue to burn in an authoritarian vacuum – it requires the oxygen of freethinking. A right to free thought can give us this air, but we still have to breathe in.
For you: more from our Insights series:
To hear about new Insights articles, join the hundreds of thousands of people who value The Conversation’s evidence-based news. Subscribe to our newsletter.