Scepticism: What Exactly Are We Declining to Trust?

Epigraph

“Nature is always too strong for principle.”

David Hume, An Enquiry concerning Human Understanding, XII, §23. The exact standalone sentence from the selected primary text. Here “nature” belongs to Hume’s argument about returning to ordinary life, rather than a measured law of behaviour.

What this article is about

The word “scepticism” can turn a particular doubt into a description of an entire person. Someone disputes a date, asks for evidence, or declines to act, and all three are called distrustful. Yet their responses concern different things. We can trust someone’s good faith while withholding assent from an explanation, find testimony useful while leaving a conclusion open, or accept a small provisional step without confidence in a whole plan. This article treats doubt as a question about a specific claim, its grounds, and the limits of acceptance. Descartes, Hume, and the Buddhist Kalama Sutta offer different ways of approaching that question within their own philosophical and religious settings. An invented photograph caption helps distinguish checking a date from judging a person, suspended judgement from denial, and caution from endless demands for guarantees. The voluntary exercise suggests naming a doubt clearly enough to see what could still be established and which decisions belong to the participants.

I. The object lost inside general distrust

Imagine a small local-history group. A member brings an old photograph of a bridge and suggests the caption “Opening after reconstruction, 1924.” She received it from a relative; the year was added later on the back, while the description of the event came through a family story. Another member says, “I don’t trust this.” What does he reject: the photograph’s existence, the year, its connection to reconstruction, the speaker’s good faith, or the group’s right to display it? Each question calls for different grounds. A later pencil mark might weaken a dating claim, but by itself says nothing about the intentions of the person bringing the photograph. An inaccurate caption does not make the visible scene imaginary either.

A more specific objection might be “I don’t yet accept this year as established, because we don’t know where the inscription came from.” That leaves a way forward. The speaker might know who wrote it, find another copy, or agree to a qualified caption. “You cannot be trusted” gives little indication of which clarification would matter. Specific wording need not soften the criticism of the claim. The participant can state plainly that the grounds are insufficient. Its value lies elsewhere: the insufficiency concerns what is actually under discussion, without acquiring a wider target before anyone has examined the connection. A strong objection can therefore remain sharply focused rather than becoming a general judgement.

Nor must doubt be confined in advance to comfortable questions. There may eventually be grounds to reconsider the photograph’s reported provenance too. But extending the objection needs a reason of its own: incompatible accounts, a concealed substitution, or available evidence of another history. It does not follow automatically from one mistaken number. The objection becomes informative when it shows its progression: dating was questioned first, then a discovered circumstance also affected provenance. Others can examine that movement. Calling someone a “sceptic” is insufficient; the label still leaves us to identify the claim, the proposed way of checking it, and the limits of what has actually been called into question.

II. Withholding assent does not establish a denial

“The bridge opened in 1924” differs from “It has not been established that the bridge opened in 1924.” The latter supplies no alternative year and does not show that the event never happened. Someone asking to remove certainty from a caption may be proposing an open date. Someone proposing “1924 is incorrect” needs grounds for that additional assertion. The distinction matters when evidence is sparse. Not finding a document in an examined folder reports the outcome of a particular search. Concluding that the document never existed requires more than an empty folder and a confident voice. The failure to establish one claim does not quietly establish its opposite.

Suspended judgement has a scope too. It may concern the year while leaving other details available for discussion. Railings, a group of people, and lettering on a sign may be visible; those features can be considered separately from uncertain dating. The reverse is possible: an independent record establishes the year while the event’s identity remains disputed. Uncertainty need not spread evenly through an account. Dismissing everything because one question remains open hides the better-supported parts. Spreading partial confirmation over the whole story makes the opposite mistake: a discovered date begins to certify the event and its meaning as well, although those conclusions require their own connections to the evidence.

A provisional account can be accepted with its status made explicit. “Probably 1924; the connection to the reopening after reconstruction remains uncertain” conveys a different degree of confidence from the original caption. Cautious words do not make that version true. They describe the present grounds more accurately. Participants may still consider it insufficient for the intended display, or keep the photograph without a date. Doubt does not decide for them. It helps distinguish three acts: rejecting the sufficiency of the current grounds, asserting that the content is false, and choosing how to handle material whose interpretation remains uncertain. Each act can be justified or questioned in its own terms.

III. Descartes and the search for assurance

In Part III of Discourse on the Method, Descartes describes provisional rules of conduct while revising his opinions. Ordinary life does not wait for completed philosophy. At the beginning of Part IV he explicitly distinguishes practice from the search for truth: for the latter, he resolves to reject as false anything admitting the slightest doubt. Errors of sense, possible mistakes in reasoning, and dreams lead into the assertion of the thinking “I.” This movement has a definite purpose: finding a secure foundation. It is different from a proposal to suspect every message in daily life, and it is not a historical investigation of dates written on photographs. Its object and ambition need to remain visible.

In Part III Descartes contrasts his intention with those he portrays as doubting for doubt’s own sake. That contrast must retain its authorship. It describes his understanding of his project and opponents; it cannot alone reconstruct every form of ancient scepticism. A narrower question is useful in our example: what does the doubter seek? A participant may want to correct the caption, establish who supplied the inscription, or prove that no caption is possible in principle. Those goals differ. The first two allow an identifiable result from investigation. The third requires a separate argument, and does not become sound merely because this particular date remains unconfirmed. An unfinished inquiry supplies no shortcut to a universal conclusion.

We should not transfer the Cartesian search for ultimate assurance into every shared task without considering the consequences. The group might find enough grounds for a limited caption without settling whether all knowledge has an indubitable foundation. Descartes’s construction remains something to read, rather than a seal of approval on their decision. His provisional rules do not become a universal demand for submission to existing customs either. Present-day participants can discuss action and decline involvement. The comparison draws attention to doubt’s relation to its aim: are we seeking grounds for a definite conclusion, or quietly requiring all philosophy to be completed before one modest question can be answered? The two demands carry very different burdens.

IV. Hume and the limits of radical doubt

In Section XII of An Enquiry concerning Human Understanding, Hume contests the project of universal doubt preceding inquiry. In the selected paragraphs, he argues that doubt carried through completely could not be escaped through the very faculties already questioned. Yet moderate caution has value for him: freedom from prejudice, slow progress, and reconsideration of conclusions. Descartes and Hume therefore cannot be presented as two authors of the same recipe. They disagree about the route to an initial foundation. Placing their texts together reveals a dispute about doubt; it does not produce a jointly endorsed modern procedure for checking every piece of information. A comparison must preserve the disagreement that gives it substance.

Hume also discusses the distance between immediate perception and customary beliefs about objects. The later selected paragraphs explore the tension between radical arguments and the return to ordinary activities. The epigraph about nature’s strength belongs to this philosophical account. It does not prove that everyone inevitably stops doubting, promise relief from distress, or make habitual confidence sufficient evidence. For the photograph, we can formulate our own narrower question: does this objection actually undermine this conclusion, or declare the whole capacity to know anything unreliable without explaining the extension? That question belongs to our application. Hume’s argument provides context for it, without becoming an experimentally demonstrated account of the participants’ minds.

Moderation is not measured by how agreeable the process feels. Careful checking may be inconvenient and overturn a cherished version; further investigation may be unavailable, leaving the question open. Hume connects mitigated doubt with caution, modesty, and restricting inquiry to matters within human experience. We can compare that with a responsibility to avoid exaggerating a finding, while distinguishing his philosophical position from our suggestion. “I checked two albums and found no date” is a bounded report. “All local recollections have now been proved useless” is a much stronger proposition, unsupported by those two albums. The difference concerns what the evidence establishes, regardless of whether the speaker sounds patient or forceful.

V. Trusting a person and accepting an explanation

The speaker may faithfully repeat a family recollection and be mistaken about the year. We can accept “That is what I was told” as testimony about the account she received without treating its date as an established historical fact. There are two different contents here. Collapsing them forces a choice between complete agreement and an accusation of lying. A more accurate response is possible: “I understand that your family associated the photograph with that year; firmer dating needs another basis.” This neither certifies nor disproves good faith automatically. It identifies the unsupported transition from a report to a conclusion. The report can remain relevant even when it cannot carry the full proposed claim.

The reverse mistake treats respect for a person as sufficient grounds for every explanation they offer. A member’s knowledge of bridge history may help, but we must ask whether it applies to this photograph and question. Construction knowledge may identify features without establishing when a relative added a pencil mark. Family closeness gives access to stories without turning memory into an independent event record. Naming those limits does not make experience worthless. It clarifies what the experience enables. Sometimes the answer becomes stronger through that clarification, because it draws on material genuinely available to the speaker. Expertise and proximity are most informative when connected to a definite claim, rather than allowed to certify everything nearby.

Relational trust reaches beyond assent to factual assertions. It may include agreements to report mistakes, preserve another person’s property, and avoid disclosing personal information without consent. Checking a date does not cancel those obligations or turn relatives into available subjects of investigation. Someone can decline to share family details; that may leave provenance uncertain, but refusal alone does not establish deception. The group must then decide whether the material suits its purpose. The objection retains its force without acquiring extra authority: a particular caption can be declined while respecting the person and their choice to restrict access to private history. Access limitations and factual falsity should not be treated as interchangeable findings.

VI. The Kalama Sutta and ethical discernment

In the Kalama Sutta, the Kalamas describe teachers who praise their own doctrines and disparage others. They struggle to identify who speaks truthfully. The response acknowledges grounds for doubt and lists bases that should not simply be relied upon, including reports, tradition, scripture, reasoning, probability, and regard for a teacher. But the discourse continues. It considers unskilful and skilful qualities, blame and praise by the wise, harm and welfare; greed, aversion, delusion, and their absence are examined. Extracting only the list of insufficient bases converts a religious and ethical argument into a contemporary slogan that the complete selected text does not support. Its continuation changes what the opening caution can reasonably mean.

The closing passage includes goodwill, compassion, appreciation, equanimity, and four assurances concerning an afterlife and the effects of actions. The sutta is therefore neither a historical-dating method nor a rule that whatever feels pleasant is true. Discernment is connected with qualities, consequences, and the judgement of the wise, within its Buddhist setting. A limited comparison is possible in our story: what are participants doing around the uncertain caption? Does the question clarify the claim, or become a way to humiliate the speaker? Humiliation does not correct the year. Courtesy does not establish it either. Ethical and factual questions can be held alongside one another while retaining their different objects and grounds.

An uncomfortable question cannot be declared harmful simply because it disrupts habitual agreement. Clarification may upset people while preventing an inaccurate caption from spreading. Nor does every harsh remark become beneficial because a doubter made it. The present distinction is the author’s proposal, rather than a guaranteed procedure extracted from the sutta. The comparison returns doubt to conduct: how a person treats another, and what consequences their manner of questioning creates. It does not authorise us to determine spiritual maturity, hidden greed, or freedom from delusion from a single remark at a meeting. The religious text can inform a comparison without becoming a licence to pronounce upon another person’s inner condition.

VII. Doubt is answerable for its grounds too

“This sounds too much like a pleasing legend” may prompt a question, but does not refute the caption. Some events really do fit convenient narrative forms. Turning a stylistic impression into a denial requires a connection: perhaps a feature visible in the photograph appeared later than the proposed year. The objection then concerns a circumstance others can discuss. Without that link, the sceptic has an impression rather than a completed proof of error. This applies to personal preferences too. A disliked story need not be false, and a well-phrased objection does not become true by sounding sober. Doubt must distinguish what it notices from the further conclusion it wants that observation to support.

We can ask what evidence might change the doubter’s assessment. That does not demand a promise of agreement in advance. A new document may be unsuitable or raise another substantial question. But rejecting every conceivable witness without distinction makes the inquiry’s object unclear. A record is dismissed because it is written, a recollection because it is spoken, a photograph because it is incomplete. A list of limitations does not itself explain why they defeat this particular conclusion. The limitation needs to be compared with the claim’s strength. Sometimes several limited pieces of testimony support a narrow account, even though none supplies unlimited assurance about every aspect of the story.

This responsibility does not require proving every possible alternative. Whoever proposes a categorical year should show its basis; an objector can identify a deficiency without establishing the correct date. However, asserting that the caption was knowingly invented or information deliberately hidden adds a new responsibility. The symmetry is limited: each assertion answers for its own strength. It does not require equal attention to every version, or prohibit ending discussion of an unsupported claim. Taking a sceptical position does not erase the difference between an identified defect and an explanation added by the objector. The latter remains a claim, even when introduced in the language of suspicion rather than certainty.

VIII. Waiting for evidence and making a decision

The group might postpone the caption while its owner looks for another copy. That is a specific decision with an identifiable question and occasion for returning to it. If no copy appears, waiting need not renew itself automatically. Participants can omit the year, explain the conjecture’s provenance, or abandon the display. Each option has consequences to discuss with the photograph’s owner. Doubt does not remove the need for choice; sometimes it changes the form of action. A working account allows less certainty while requiring an honest statement of its limits. The word “probably” does not give participants permission to display anything that happens to attract curiosity. The content’s status and permission to use it remain separate questions.

Radical philosophical uncertainty and an ordinary missing record need not produce identical practical responses. Descartes distinguished revising opinions from conduct, Hume discussed returning to ordinary life, and the sutta linked discernment with qualities and actions. These different contexts cannot be combined into an instruction to act immediately. For the contemporary example, we separately suggest considering reversibility, affected people, and their agreement to the proposed form. A trial caption inside the group and public display of a family photograph may have different conditions. Incomplete assurance does not erase those distinctions or establish the owner’s consent. The action still needs to be described precisely enough for those involved to decide whether they accept it.

Someone may decline participation even when others consider the evidence sufficient. The disagreement can be recorded without making the meeting an examination of trust. One person’s refusal neither disproves the date nor makes everyone who agrees credulous. Shared action has its own questions: who accepts an obligation, which information will be presented, and whose choice is required? Doubt helps when it clarifies those conditions. It loses focus when every practical disagreement becomes supposed proof of intellectual weakness, or philosophical uncertainty is used to speak for someone who never consented. A disagreement about taking part can stand as a decision, without being inflated into a verdict about the truth or the character of the participants.

IX. Ending a check without pretending finality

Suppose another copy appears with an earlier inscription, alongside an independent report of the bridge’s opening in the same year. The dating assessment can change. The speaker need not be declared right about everything, or the sceptic defeated. The initial objection concerned weak grounds; new evidence altered the position. Participants can describe what was confirmed while leaving the event shown in the photograph uncertain. A check ends for a selected task when its answer’s scope is understood and considered sufficient for a definite caption. That neither proves future correction impossible nor prevents reopening the question when significant new material appears. Ending a bounded inquiry does not require claiming an end to every uncertainty in the account.

Sometimes the outcome is to abandon a confident version. Keeping “we did not confirm it” distinct from “we refuted it” remains useful. A record of examined material lets a later participant see where the search went and what it failed to establish. It need not catalogue every technical movement. Meaningful limits are enough: which evidence was compared, which inference remained unsupported, and why a qualified caption was chosen. This can prevent the same dispute restarting because its circumstances were forgotten. Yet a record gains no authority merely by existing. Its content can also be corrected if an error emerges or the extent of examination was described inaccurately. The account of checking is itself answerable to what actually happened.

An ending criterion should not ban inconvenient questions. A new detail directly contradicting the accepted year can reopen the caption despite earlier agreement. Repeating an old objection, however, does not by itself create a new circumstance. Participants can distinguish those situations and agree what would warrant returning. This is an authorial suggestion for a limited conversation, not a measured law of productivity or a demand to suppress doubt. Its purpose is to keep correction available without promising endless attention to one question. Someone can find the current answer sufficient while understanding why it does not deserve to be called the final truth about everything in the story. Closure and corrigibility can therefore remain compatible.

X. Three texts and three different limits

Descartes seeks grounds for assurance and describes methodical treatment of the slightest doubt. Hume challenges escape from completely radical preliminary doubt and considers more moderate caution. The Kalama Sutta addresses doctrines and qualities within Buddhist ethical and religious reasoning. Gathering them under one word does not resolve their differences over purpose, foundation, and limits of inquiry. This article takes questions from them for comparison while preserving their origins. It does not claim that the authors jointly designed a way of captioning photographs, or that our conversational steps have been established as a general decision-making method. The comparison remains valuable precisely because the texts do not become interchangeable expressions of a single doctrine.

Our own conclusion is narrower: declining trust should concern definite content. The year may lack sufficient support; the photograph’s connection with an event may remain ungrounded; conditions for agreed display may be absent. Those findings cannot substitute for one another. A useful statement of doubt names its object, reason, and degree of acceptance: what is open, what has actually been rejected, and which additional assertion the objector is making. It permits a strong refusal where grounds are strong, without demanding equal uncertainty about everything. It also permits a modest answer when available reading or testimony establishes no more. Precision changes the extent of the conclusion, rather than simply decorating the same certainty with cautious words.

Scepticism in this sense is no badge of moral superiority. Credulity cannot be established from one agreement, or honesty from one cautious remark. Content can be discussed without dividing people into the awakened and the deluded. Checking may strengthen the original account, alter it, or leave a question unresolved. In each case we can say more precisely what we know and which action we are responsible for. General distrust gives way to distinctions: this conclusion still lacks grounds, this witness can address only a limited object, and this decision requires separate agreement. Such clarity promises no universal assurance. It makes the next conversation more intelligible by giving each question a recognisable subject and limit.

What this means in practice

If you wish, choose one modest claim you currently doubt. Avoid using the exercise for a question requiring urgent professional assistance or involving someone else’s private information without consent. Write down the content, such as “the photograph was taken in this year,” and separately record the grounds offered. Identify where the objection falls: the inscription’s provenance, its connection with the image, a move towards explanation, or the proposed action. If those cannot yet be separated, “I don’t understand how this conclusion was reached” may be enough.

Formulate your response without adding unsupported accusations. Distinguish “I don’t yet accept this as established,” “I consider it incorrect for this reason,” and “I don’t agree to participate.” Then consider which available clarification would actually matter. You need not list every possible guarantee or promise to accept any new document. One substantive question connecting the objection to the conclusion is sufficient. The other person can decline to supply private information; uncertainty may then remain. A missing answer does not by itself determine their intentions.

After the response, note what changed and what remains open. If there is no new information, choose an appropriate limit to action: leave the question open, adopt a qualified working statement, postpone a definite step, or decline involvement. This is a voluntary authorial exercise, rather than an assessment of your character. Stop if it no longer helps distinguish content. Its purpose is not to compel trust, but to make clear what exactly you are withholding acceptance from, and why.

Key quotation

“my design was singly to find ground of assurance”

René Descartes, Discourse on the Method, Part III; the exact nine-word fragment in John Veitch’s English translation. It describes Descartes’s own purpose and does not replace a history of sceptical schools.

Connections between traditions

Comparison	Shared question	Substantial limit	Application in this article
Descartes, III–IV, and Hume, XII, §3–6	Doubt is discussed with grounds of knowledge	Hume contests completely radical preliminary doubt; no agreed position is claimed	Distinguish ultimate assurance from checking one caption
Descartes, III, and Hume, XII, §21–25	Philosophical argument meets ordinary action	Cartesian provisional rules and Humean caution have different grounds	State working conditions without a general command to act
Kalama Sutta and offered grounds	Tradition and a teacher alone do not complete discernment	Buddhist qualities, consequences, the wise, and religious context remain	Avoid substituting respect for the speaker for evidence of a date
Kalama Sutta and conversational ethics	Discernment relates to conduct and consequences	Courtesy does not establish facts; our conversation exercise is authorial	Check a caption without assuming authority to humiliate or expose private information
Hume, XII, §24–25, and a bounded check	Caution limits the strength of a conclusion	A philosophical text supplies no measured contemporary procedure	Name examined evidence and the point at which the chosen question ends

Primary sources

René Descartes, Discourse on the Method, John Veitch translation, Project Gutenberg 59. Part III was read completely, followed by the opening of IV through the discussion of a clear-and-distinct rule; browser lines 120–157. Relevant themes are provisional conduct, the stated purpose of assurance, and the opening of radical doubt. No complete-work reading, French comparison, or print collation is claimed. “Doubting for doubt’s sake” remains Descartes’s characterisation. Selected text.
David Hume, An Enquiry concerning Human Understanding, XII, §3–6 and §21–25, primary English text at DavidHume.org. All specified paragraphs were read: radical preliminary doubt, caution, Pyrrhonism’s difficulty in ordinary life, and mitigated scepticism. The epigraph is the exact standalone sentence in §23. Neither the whole section nor the entire work is claimed as read. Philosophical claims are not presented as contemporary empirical laws. Section XII.
Kalama Sutta, AN 3.65, Thanissaro Bhikkhu’s 1994 English translation, Access to Insight. The entire primary passage was read, from arrival at Kesaputta through the Kalamas’ closing response, browser lines 47–155. The translator’s note was read separately and is not attributed to the discourse. Comparison retains qualities, harm and welfare, the wise, four assurances, and Buddhist religious context. No direct quotation from this translation is used. No original-language comparison or whole-canon reading is claimed. Translated discourse.

About the authors

Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, the unity of traditions, intercultural dialogue, and contemporary spiritual discipline. Instagram: @andrey.zarinsh.

Maria Evilkut is a co-author of United Wisdom. Alongside her research, she seeks to test these ideas in practice and observe how principles of unity and mutual understanding work in real life between people and cultures. Instagram: @ilkutki.