How firmly should a belief be held? People confront this problem many times each week, usually without recognising it. Grip an idea too lightly and conviction disappears; grip it too fiercely and no argument can break through. Many people move between these extremes without words for the territory separating them. Under strains far beyond those most people will meet, one Roman advocate devised such a vocabulary. His answer remains among the strongest available ways to combine commitment with an honest willingness to reconsider.
In early 45 BCE, at a country house beside the Lucrine Lake northwest of Naples, Marcus Tullius Cicero was completing the work later called the Academica. Few people in Western history have attracted such fascination or dispute. Here, the Roman statesman set out what sceptical philosophy meant and why it mattered not only for extending human understanding, but also for leading a worthwhile life. The work connects advancing human knowledge with the problem of living well.
The timing of this scholarly withdrawal was remarkable. Cicero’s cherished Roman Republic was nearing destruction at the hands of would-be autocrats. Julius Caesar, once a close friend but now a political adversary, had made himself dictator for life, a decision that precipitated his assassination when the Ides of March arrived in 44 BCE. Cicero’s private life had also broken apart. He had recently divorced Terentia after 30 years of marriage, while his beloved daughter Tullia had died from complications of childbirth only months earlier. Such collapses might have left anyone unable to concentrate upon philosophy. For Cicero, however, writing became therapy. These works would outlast both his political defeat and the Republic he tried to save. Across the centuries that followed, they shaped medieval Christian thinkers, Renaissance authors, and eventually the founders of the young American Republic.
His adopted outlook, known as Academic Skepticism because it defined Plato’s Academy from 266 until 90 BCE, sustained Cicero through this work. Earlier scholars often dismissed him as an eloquent Roman populariser who merely transmitted Greek concepts he could not claim as his own. That ungenerous reading was discarded decades ago, once researchers recognised that he weighed competing sources and reached decisions instead of copying them. He also created a philosophical lexicon in Latin, supplying coinages from which much of Europe’s abstract language descends. On his school’s central problem—whether sceptics can maintain opinions, and under which conditions—Cicero developed a distinctive position unlike his teachers’. The distinction matters because selecting, translating and testing inherited arguments was itself a form of original philosophical labour in a new Roman setting. He certainly practised philosophy before the public, testing his conclusions in courtrooms, the Senate and, ultimately, while confronting death ordered by his political enemies. That was philosophy conducted amid civic life, not second-hand exposition.
That practical record makes Cicero relevant now. His story provides, above all, a usable discipline for belief, positioned between the staged modesty of a professional doubter and the fortified certainty of somebody who claims to have completed independent research. Cicero called the approach probabilia. It offers a way to judge whether an idea deserves acceptance whenever evidence must be assessed and a consequential choice must follow. Rather than inviting endless hesitation, it directs responsible action whenever the evidence cannot yield absolute certainty in advance.
An authentic sceptic revises a conclusion when forceful evidence appears; contrarians defend theirs instead.
The Greek skeptikos meant an enquirer, giving ‘sceptic’ a serious and ancient philosophical lineage. From the 1990s onward, however, science-denial movements captured this prestigious term and attached it to three familiar forms of rejection: climate research, vaccination and evolutionary theory.
Those movements saw the prestige attached to scepticism. The label sounds exacting, independent and resistant to authority, so denialists occupied it. Their clever but damaging manoeuvre recast denial as investigation. Conduct reveals the distinction. Faced with forceful new evidence, an authentic sceptic revises a conclusion; a contrarian shifts the standard of proof, invents another objection or disregards an inconvenient result. Such figures selectively cite research and invoke the few dissenting specialists who have filtered the evidence. Their uncertainty is not Socratic doubt that begins enquiry, but defensive rhetoric designed to close it. Despite the name, they are dogmatists who settle their conclusion first and then use each later argument solely to rationalise their position.
An opposing form of scientism has overreacted in another direction, insisting that only certain knowledge may properly be believed. Its slogan “follow the science” ends discussion rather than opening a process of enquiry. Real scientific enquiry does not require certainty before accepting a belief either.
Both distortions reduce the possibilities to doubt or certainty. They exclude the needed middle position: confidence calibrated closely enough for someone to act on the strongest account available while accepting that it may prove mistaken. The middle position says that the current account is best, warrants action and could nevertheless be wrong. What is required is committed yet provisional belief. Cicero had already developed precisely that philosophy, in Latin, two millennia ago.
The intellectual path began in 87 BCE, when Philo of Larissa, then leading the Platonic Academy, visited Rome after political upheaval drove him from Athens shortly before Roman forces sacked that city. Cicero, a 19-year-old prodigy of law and public speaking, heard Philo and became enthralled. Philo offered no remote Platonic metaphysics. Instead, he taught a practical method rooted in Socratic enquiry: wisdom starts by acknowledging the limits built into personal knowledge. For the young Roman, this approach connected intellectual modesty directly to choices that still had to be made in public life.
Academic Skeptics named their institution the New Academy to separate it from Plato’s Old Academy. Arcesilaus and Carneades, its commanding sceptical figures during the 3rd and 2nd centuries BCE, carried the dialogical techniques of Plato and Socrates towards a radical result. Certainty is probably unreachable because human knowledge arises through sensory experience and reasoning, and either source may mislead even the most attentive enquirer.
Yet they did not infer that every belief should be abandoned. Carneades proposed the Greek pithanon—what persuades or appears plausible—as the correct basis for action. Cicero rendered it in Latin as probabile, whose plural is probabilia and which gave English the term “probable”. That idea became the foundation of his entire theory of knowledge.
Cicero offered a disciplined, testable hierarchy for judgment, not merely a broad preference for open-mindedness.
Cicero states the contrast clearly in the Academica. Dogmatists pronounce some matters definite and settled; his school accepts many claims as probabilia, suitable to follow readily but almost impossible to affirm outright. The imbalance carries his complete approach: follow without difficulty, but affirm only sparingly, without pretending that their truth has been demonstrated beyond every remaining doubt.
Here probabilia is not “probable” in today’s quantitative, statistical meaning. It is nearer to “deserving approval”: a proposition that survives examination, fits the existing evidence and merits rational assent without becoming immune to correction. That standard provides enough justification to act while admitting the conclusion remains open to revision.
Importantly, probabilia is not one ungraded class. Carneades ranked impressions—initial judgments—by the examination each had endured. At level one, an impression merely appears likelier to be correct than incorrect. Level two grants it persuasive force. Reaching the third makes it “undiverted”: no other belief conflicts with it. At the highest, fourth level, exhaustive testing has made it comparable to a scientific finding. Everyday questions may require only the first or second grade. Medical choices and criminal verdicts call for the third or fourth. The scale prevents confidence from being all-or-nothing and ties its proper strength to both the stakes and the quality of inquiry in each case. Cicero therefore inherited and used a structured procedure that could be checked, not a vague posture of intellectual openness.
This ranking also separates Academic thought from antiquity’s other major sceptical tradition, Pyrrhonism. Pyrrhonists recommended withholding judgment on most questions while behaving according to appearances and surrounding customs. Three distinctions matter. First, Academic thinkers assent in a limited yet genuine manner: they endorse and defend a proposition, accepting personal responsibility should it fail. Pyrrhonists decline approval and simply proceed with what appears. That difference is not semantic: assent makes the Academic answerable for a judgment, whereas simply following an appearance avoids claiming that it deserves belief. Second, Academics rank appearances by the testing survived, allowing some to become better than others. For Pyrrhonists, appearances possess no comparative weight; each is no better or worse than another. Accordingly, investigation can strengthen one impression and weaken another, rather than leaving every appearance on an undifferentiated plane. Third, their ends differ. Here, suspending judgment serves a fundamentally different end. Pyrrhonists seek ataraxia, or tranquillity, using suspended judgment to reach it. Academics continue seeking truth while doubting that its arrival can ever be certified. One method treats anxiety; the other operates as a programme of research for fallible creatures still compelled to act in the world.
The Academica leaves Cicero’s position explicit. Wise people follow the strongest currently available judgment, not because its truth is guaranteed, but because life becomes impossible when every conclusion remains suspended. Academics did not say that nothing is true. They argued that false claims mingle with truths so convincingly that no infallible test separates them; all judgments must therefore remain provisional. This was not nihilism; it was an admission that close resemblance between error and truth prevents final reliable certification.
Once freed from its Latin terminology and Roman dress, the Academic method becomes a handful of practical habits that anyone can apply to ordinary decisions today:
- Express a confidence level, rather than only a conclusion. A bare statement of belief reveals almost nothing. Rating a belief at seven out of ten, while identifying what prevents the remaining three points, gives another person something specific to examine during a serious discussion and leaves a defined route for movement when stronger arguments or evidence arrive.
- Rank beliefs by the severity of their tests. Much that people accept occupies only Carneades’ second tier because it happened to seem convincing. High-confidence action should be reserved for conclusions that have survived careful scrutiny at the third or fourth levels.
- Keep values distinct from factual propositions. The principle that children must not experience hunger is a value that may be defended firmly. Whether one school-meal scheme actually reduces hunger is an empirical proposition and deserves only the confidence supplied by available evidence. Mixing these categories causes people to defend ineffective policies as if conceding failure would destroy their integrity.
- Identify beforehand what evidence would change your mind. If nothing could, the position is no longer a belief but an identity.
- Build the opposing argument seriously. The Academy’s characteristic practice was reasoning in utramque partem, meaning on both sides of an issue. In Rome in 155 BCE, Carneades famously supported justice on one day and dismantled its case the next. This was not cynicism. A person cannot properly understand a conviction until constructing the strongest available case against it themselves, on its own terms.
- Proceed despite uncertainty. Sceptics who use doubt to delay every decision misunderstand the method. Because uncertainty is a permanent and unavoidable condition of human life, deciding effectively within it remains a skill from which everyone can benefit.
- Regard an error as information, not humiliation. Following these sceptical practices should sometimes produce a changed mind. The real question is whether revising a mistaken belief feels like a humiliating defeat or instead provides evidence that the sceptical method is operating exactly as intended.
Cicero never treated these principles as theory alone. They directed every major choice he made, including his most bitterly disputed decision.
While serving as Consul, Republican Rome’s highest office, in 63 BCE, Cicero confronted a plot to topple the government. Its leader, the embittered senator Lucius Sergius Catilina, threatened the city from the field with an army. Cicero collected the intelligence available and proceeded upon probabilia. He submitted his evidence to the Senate, which adopted the senatus consultum ultimum, an emergency resolution empowering the consuls to take whatever action state security demanded. Cicero then asked the full Senate to decide what should happen to the captured conspirators. Caesar urged imprisonment; Cato the Younger demanded execution. Under the constitutional interpretation accepted by Cicero and the senatorial majority, the resulting deaths were lawful. Contemporaries and later observers contested that reading. Citizens possessed appeal rights, and years afterward Cicero was exiled under a retroactive statute designed expressly to claim he had breached them. The episode shows Romans never shared a single, uncontested interpretation of how far those emergency powers could lawfully reach. His procedure combined incomplete evidence with public deliberation and responsibility for an uncertain outcome. This history cannot establish that his decision was correct. What matters here is how he reached it: Cicero acted while knowing his conclusion might fail, asserted no unavailable certainty and defended the choice thereafter rather than renouncing it. Academic Skepticism under extreme pressure meant neither paralysis nor reckless assurance, but consequential allegiance to the best current judgment, including the direct personal price involved.
Cicero understood that returning could condemn him, yet judged the likely public benefit worth the likely price.
After Caesar’s dictatorship pushed him from politics, Cicero withdrew to his properties and produced philosophy with astonishing speed. His works included Academica and the Tusculan Disputations, alongside De natura deorum, De divinatione and De finibus bonorum et malorum. De divinatione brings Academic Skepticism to mysticism and superstition and is now regarded as the earliest treatise separating science from pseudoscience. Any practising scientist would recognise its method. Cicero does not immediately reject augurs and haruspices who inspect entrails or interpret birds in flight. He investigates their claims. Through what mechanism could an animal liver foretell a battle? Do predictions show genuine success, or are favourable cases selectively remembered afterward? At every stage, the evidence cannot force assent. Yet Cicero preserves the possibility that deities communicate through omens, finding only that the case fails standards he would apply elsewhere. This restraint is crucial: rejecting inadequate evidence did not require him to declare divine communication impossible, only unsupported by the evidence he had examined so far. Two millennia before Carl Sagan said extraordinary assertions demand extraordinary evidence, Cicero used this discipline against institutions Romans revered. Such scrutiny showed intellectual bravery rather than scholarly evasion.
Cicero’s correspondence with Atticus during the final months of his life supplies his most revealing surviving record. Aware that Mark Antony wanted him killed, he considered escape, a return to Rome and whether young Octavian, later the emperor Augustus, could be trusted. The letters show him deliberating under uncertainty, balancing probabilities and choosing among dangerous alternatives despite lacking any guarantee of safety. Cicero ultimately returned and renewed his attacks upon Antony in the speeches called the Philippics. Although this decision might amount to his own death warrant, he judged its probable public value greater than its probable cost. Murdered on 7 December 43 BCE, he reportedly stopped before the legionary dispatched to execute him, told the soldier that the act was wrong, and demanded at least that his head be severed properly.
Can Academic Skepticism still be recovered and used today? In three fields, the answer is an emphatic yes.
First comes science, which—despite common claims—does not operate through certainty. Scientific knowledge consists not of permanently established facts but a revisable consensus around the most thoroughly tested theories. Thomas Kuhn demonstrated that paradigms change across scientific history. Imre Lakatos explained why researchers reasonably continue programmes despite unresolved anomalies, adjusting surrounding hypotheses before surrendering a theory’s centre. This practice mirrors Cicero’s probabilia: the strongest current explanations merit confidence proportionate to their evidence, yet each conclusion must remain open to future revision. Modern Bayesian probability theory, which develops the theorem Thomas Bayes formulated during the 18th century, expresses that insight formally. Prior beliefs change with incoming evidence; assurance always stays proportional and never becomes certainty. Its mathematics may be modern, but its underlying wisdom is ancient.
This position is not relativism, which reduces theories to social constructions. It asserts almost the reverse. Contemporary philosophy of science describes convergence upon increasingly accurate pictures of reality through discarding weaker hypotheses and improving stronger ones, rather than progressing from one certainty to another. Like an Academic Skeptic, a scientist avoids claiming final truth without accepting that every account is equal, while continuing to discriminate carefully among rival explanations. In 1942, sociologist Robert K Merton aptly characterised science as “organised scepticism”.
The second domain is democracy, whose deliberations require citizens willing to be convinced. Political views must function as probabilia rather than dogmas, making the current political breakdown partly a crisis of knowledge. No past golden age was populated entirely by reasonable voters. Yet measurements reveal an unmistakable trajectory. Analyses of US congressional votes show the parties’ ideological overlap narrowing from the 1970s and virtually vanishing by the 2000s. Surveys from Pew and other researchers record a sharp increase since the 1990s in citizens with consistently ideological positions who regard their opponents as national threats. Much of this separation is affective rather than programmatic: actual policy preferences have diverged less than feelings toward the opposing party. The disparity matters because partisan hostility can become a commitment that resists contrary evidence, even when disagreements over concrete policy remain considerably narrower in reality. Evidence cannot easily alter that kind of attachment, which social media and the economy of fabricated news have intensified.
A self that never revises is not integrity at all, but simple rigidity presented as principled character.
How would a genuinely probabilist citizen behave? Such a person could maintain forceful political commitments because probabilia permits conviction and decisive action rather than indecision. Values, however, would remain distinct from factual beliefs about how society functions. Consider support for universal basic income. Someone may firmly value the principle that nobody in a wealthy society should endure destitution. Whether basic income offers the best means of achieving that aim is instead an empirical wager. Trials now underway in several countries could show it to be mistaken. The commitment to ending poverty remains intact even when the proposed instrument must be abandoned after disappointing evidence. A probabilist citizen could then acknowledge, without embarrassment, that the policy seemed promising but apparently failed and that the judgment must change. This is not opportunistic “flip-flopping”. It is the mental posture that permits democratic persuasion and represents consistency of a deeper kind: allowing evidence to determine where conclusions lead.
Finally, on a personal scale, Cicero’s epistemology applies as an answer to fixed identity’s attractions. Someone claiming to have discovered an unquestionable private truth has stopped behaving wisely. In The Prime of Life (1960), Simone de Beauvoir described the deliberate madness of a “philosophical system”, whose adherents obsessively elevate tentative patterns into universally applicable insights. Identity becomes dangerous when provisional interpretations are granted the authority of permanent truths and protected from later inquiry. An examined life instead continually recalibrates and revises, sustaining commitments strong enough for integrity and flexible enough to change when evidence does.
Such an outlook runs against a culture that equates authenticity with possessing a settled self and encourages everyone to claim an identity. Cicero would likely have considered that attitude self-congratulatory. His letters show him changing and worrying over assessments of allies, friendships and tactics. Yet a permanently fixed self is not evidence of integrity; it is rigidity performing the role of character. The Ciceronian person is willing to feel embarrassed by earlier certainties and treats that discomfort not as failure, but as evidence of a mind still revising itself through active enquiry when compelling new evidence demands it.
Cicero misjudged the Republic, which ultimately collapsed despite his efforts. Still, he behaved as though it might survive because, by his analysis, that remained the strongest available conclusion and merited action. His books preserved his legacy beyond the Republic’s defeat, even though he composed them beneath the threat of Antony’s vengeance.
A true sceptic neither questions everything indiscriminately nor demands proof before accepting any proposition. Scepticism means holding the strongest available conclusions with an accurately judged grip: sufficient firmness for action, and sufficient looseness for release when contrary evidence arrives. That is the meaning of probabilia. It demands more than either dogmatic certainty or nihilistic doubt. Cicero regarded it as the only intellectually honest position available to finite human beings: an exacting balance that demands conviction from fallible thinkers without pretending to certainty. He was probably right.
