A Philosophical Inquiry into Knowledge, Evidence, Belief and the Limits of Certainty
Prabhash Chandra
Human beings have always lived between two conditions: the world as it is and the world as they understand it. The distance between these two constitutes one of the deepest problems of human existence.
We see, interpret, remember, infer, believe and act. We construct explanations from incomplete evidence and then live according to those explanations. Civilisations are built upon them; institutions are governed by them; sciences advance by challenging them; religions interpret existence through them; and individuals make the most consequential decisions of their lives because of them.
Yet beneath all these activities lies a deceptively simple question:
What is true?
The question appears elementary only until one attempts to answer it.
Is truth correspondence with an external reality? Is it coherence within a system of propositions? Is it something established through empirical verification? Is it revealed through consciousness? Is it mediated by language? Is it dependent upon the observer? Or does truth exist independently of our capacity to recognise it?
The history of philosophy may, in one sense, be understood as the history of humanity’s struggle with these questions.
But truth is not merely an abstract concern of philosophers. Every practical decision presupposes some conception of reality. A physician acts according to a diagnosis. A teacher acts according to an understanding of learning. A judge acts according to an interpretation of evidence. A scientist acts according to a model of nature. A leader acts according to an understanding of an organisation. An individual chooses a career, a relationship, an investment or a course of action according to assumptions about what is real, possible and desirable.
If the understanding is defective, action may be defective even when the intention is good.
The question of truth, therefore, is also a question of human action.
Truth Begins Where Certainty Ends
There is a fundamental distinction between believing something and knowing something.
Belief is psychologically natural. Knowledge is epistemically demanding.
A person may believe that an event will occur tomorrow. The event may indeed occur, but the correctness of the belief does not by itself establish knowledge. A person may strongly believe that a colleague is dishonest, but intensity of conviction does not convert suspicion into evidence. A person may be certain that a particular policy will succeed, yet certainty itself provides no guarantee of success.
Human beings frequently confuse psychological certainty with epistemic justification.
The distinction matters because confidence is a property of the believer, whereas truth is a property of the proposition.
I can be completely confident and completely wrong.
History offers countless examples of individuals and societies that were profoundly certain about propositions later shown to be mistaken. The geocentric model of the universe, racial theories presented as biological fact, medical practices once regarded as unquestionably sound, and innumerable political and social assumptions demonstrate the same epistemic lesson:
The sincerity of a belief does not determine its truth.
Nor does the antiquity of a belief.
Nor its popularity.
Nor the authority of the person who expresses it.
Truth requires a relationship between claim, evidence and justification.
This is where epistemology begins.
The Ancient Question of How We Know
The Indian philosophical tradition developed sophisticated discussions concerning the nature and sources of valid knowledge. The concept of pramāṇa occupies a central position in classical Indian epistemology: by what means does cognition become valid knowledge?
Different philosophical schools accepted different pramāṇas and disagreed profoundly about their scope, reliability and metaphysical implications. Nyāya, for instance, developed an elaborate logical and epistemological framework involving perception (pratyakṣa), inference (anumāna), comparison (upamāna) and testimony (śabda), while other traditions adopted different epistemic positions.
The significance of this intellectual heritage is not that it provides a ready-made modern scientific method. Such a claim would be historically and philosophically careless.
Its significance is deeper.
Indian thinkers recognised that knowledge has conditions.
We do not simply “have” knowledge. We acquire cognition through particular means, and those means can succeed or fail.
This is a remarkably modern question:
What makes a cognition trustworthy?
The question forces us to examine the route through which a conclusion has been reached rather than merely examining the conclusion itself.
That distinction remains fundamental today.
Perception Is Not Reality
Among the most immediate sources of human knowledge is perception.
We see the world, hear it, touch it, smell it and experience it.
Yet perception is not infallible.
An optical illusion can cause a perfectly functioning visual system to produce an inaccurate interpretation. Atmospheric conditions distort distant objects. Context changes perception. Attention determines what enters conscious awareness. Memory alters subsequent reconstruction of experience.
The philosophical implication is significant.
Perception is indispensable, but it is not identical with reality.
There is therefore a difference between:
“I perceived X”
and
“X is necessarily as I perceived it.”
The first is a report about experience.
The second is a claim about reality.
The distinction becomes even more important when the object of observation is another human being.
A person remains silent.
The silence is observable.
But the reason for the silence is not directly observable.
We may infer anger.
Perhaps it is sadness.
Perhaps embarrassment.
Perhaps fatigue.
Perhaps concentration.
Perhaps nothing significant at all.
The event and the interpretation are different epistemic objects.
Much human conflict begins when interpretations are unconsciously promoted to the status of facts.
Inference and the Architecture of Explanation
Human beings cannot function through perception alone. We constantly infer.
We see dark clouds and infer rain.
We see smoke and infer fire.
We observe a pattern in data and infer a relationship.
Inference allows us to move beyond what is immediately given.
But inference introduces another possibility of error.
A conclusion can be logically constructed from insufficient or misleading premises.
The ground may be wet because it rained.
It may also be wet because someone used a hose.
A student’s performance may be poor because of inadequate effort.
It may instead result from conceptual gaps, ineffective instruction, anxiety, illness, language barriers or a combination of factors.
Thus, the problem is not simply whether an explanation is possible.
The epistemically important question is:
Is this explanation better supported than its alternatives?
This is the foundation of rational diagnosis.
From Correlation to Causation
Few errors are more common in human reasoning than confusing correlation with causation.
Two phenomena occur together, and the mind immediately constructs a causal narrative.
Yet a statistical relationship does not, by itself, establish a causal mechanism.
If students who read more books perform better academically, several explanations may coexist. Reading may contribute directly to achievement. But students with stronger language skills may also be more likely to read. Family environment may influence both reading and academic performance. Socioeconomic variables may influence both.
The observed relationship may be genuine while the proposed explanation remains incomplete.
This is why rigorous inquiry requires more than finding patterns.
It requires identifying mechanisms.
The question changes from:
“What occurs with what?”
to:
“Through what process does one condition influence another?”
That movement—from association to mechanism—is one of the defining characteristics of mature reasoning.
The Problem of the Observer
Modern science often attempts to isolate the observer from the phenomenon being observed.
Human systems complicate this ambition.
In education, management, politics and social life, the observer is often part of the system.
A teacher’s expectations influence student behaviour.
A leader’s communication alters organisational culture.
A parent’s response influences a child’s future response.
A researcher’s framing can influence how participants respond.
An administrator’s policy changes the behaviour of the very people being evaluated.
Thus, in complex human systems, the observer cannot always be treated as an external spectator.
This creates an important philosophical problem:
How much of what we observe is produced by the system, and how much is produced by our way of observing it?
The question leads naturally toward metacognition.
Before examining the object, perhaps we must sometimes examine the instrument of examination.
And in human inquiry, that instrument is partly ourselves.
The Self as an Epistemic Variable
The Indian philosophical tradition repeatedly turns the question of knowledge back toward the knower.
The Upanishadic inquiry into ātman, consciousness and the nature of the self is not merely an exercise in metaphysical speculation. It raises a fundamental epistemological problem:
Who is the knower?
The question “Who am I?” therefore acquires a philosophical significance far beyond personal identification.
If the knower is affected by desire, fear, memory, attachment, social conditioning and prior belief, then the acquisition of knowledge cannot be entirely separated from the condition of the knower.
This does not mean that objective knowledge is impossible.
It means that epistemic discipline requires awareness of the conditions under which knowing occurs.
A frightened mind sees danger differently from a calm mind.
An ego-defensive mind interprets criticism differently from an open mind.
A person deeply attached to a conclusion evaluates evidence differently from someone genuinely willing to revise it.
The quality of reasoning is therefore partly dependent upon the quality of the reasoner.
“Neti, Neti” and Epistemic Restraint
The Upanishadic expression “neti, neti”—“not this, not this”—has traditionally been interpreted within a profound metaphysical framework. It should not be reduced to a modern problem-solving technique.
Yet as a philosophical metaphor, it offers an important lesson about epistemic restraint.
The human mind tends to identify reality too quickly with its first available description.
We name something and believe that we have understood it.
We classify something and assume that the classification exhausts its nature.
We construct an explanation and become attached to it.
“Neti, neti” can be read, cautiously, as a reminder that the map should not be confused with the territory.
Every conceptual description is partial.
Every model has a domain of validity.
Every theory has assumptions.
Every explanation leaves something unexplained.
This is not a weakness of knowledge.
It is a condition of finite knowledge.
Science and the Virtue of Revision
The strength of science lies not in possessing infallible conclusions but in possessing institutionalised mechanisms for correction.
A scientific hypothesis must be vulnerable to evidence.
A theory gains strength through explanatory power, predictive success, empirical support and resistance to serious attempts at falsification and refutation.
Scientific knowledge is therefore provisional—not because it is arbitrary, but because it remains open to revision when better evidence or better explanations emerge.
This is a crucial distinction.
Provisional does not mean unreliable.
It means revisable in principle.
A scientific proposition is stronger precisely because it is not protected from criticism.
This intellectual structure has an important implication for everyday reasoning.
We should not ask merely:
“How can I prove that I am right?”
We should also ask:
“What evidence could demonstrate that I am wrong?”
The second question is often more intellectually valuable.
The Psychology of Being Right
Human beings are not purely rational information-processing systems.
We are emotional, social and identity-bearing creatures.
We do not merely hold beliefs.
We often become attached to them.
A political belief can become part of identity.
A professional opinion can become a matter of reputation.
A religious interpretation can become intertwined with belonging.
A management decision can become associated with personal competence.
Once a belief becomes part of identity, evidence against it may feel like a personal attack.
This produces confirmation bias.
We search selectively.
We remember selectively.
We interpret selectively.
We notice supporting evidence more readily than contradictory evidence.
The danger is subtle because confirmation bias does not require dishonesty.
A person can sincerely believe that he is being objective while unconsciously selecting information that confirms what he already believes.
The remedy is not complete objectivity—which may be an unrealistic ideal—but procedural safeguards against our own biases.
Seek disconfirming evidence.
Invite disagreement.
Compare alternative explanations.
Separate evidence from interpretation.
Record predictions before outcomes are known.
Review decisions after results become available.
These practices do not eliminate human bias.
They make bias more visible and therefore more manageable.
The Question of Authority
Human civilisation depends upon trust.
No individual can independently verify every scientific claim, every historical fact or every technical proposition.
We therefore depend upon experts.
This dependence is not irrational.
It is a practical necessity.
But rational trust is different from intellectual submission.
The relevant question is not:
“Is this person an authority?”
It is:
“Is this person a reliable authority on this particular question, and what is the basis of that reliability?”
Expertise is domain-specific.
A distinguished physicist is not automatically an authority on constitutional law.
A brilliant mathematician is not necessarily an expert in clinical psychology.
A successful entrepreneur is not automatically an expert in educational science.
Authority deserves respect.
But claims deserve evaluation.
Tradition, Reason and the Discipline of Neither Blind Acceptance Nor Blind Rejection
Tradition presents another epistemic challenge.
The fact that an idea has survived for centuries does not prove its universal truth.
But the fact that an idea is ancient does not make it irrational either.
Traditions are repositories of accumulated human experience. Some contain enduring insights. Others are context-dependent. Some contain symbolic wisdom. Others may reflect historical limitations.
The rational response is neither uncritical acceptance nor fashionable dismissal.
It is examination.
What was the original context?
What problem did the idea address?
What assumptions underlie it?
What remains philosophically defensible?
What is empirical?
What is metaphysical?
What is symbolic?
What is historical?
Such discrimination is itself an expression of Viveka.
Viveka: The Discipline of Distinction
Few concepts are more valuable to rational thought than discernment.
Viveka can be understood broadly as the capacity to distinguish.
In practical epistemology, we need to distinguish:
fact from interpretation,
evidence from assertion,
cause from correlation,
possibility from probability,
probability from certainty,
authority from expertise,
tradition from truth,
emotion from evidence,
and confidence from competence.
Without such distinctions, the mind collapses categories that should remain separate.
With them, thought becomes more precise.
Perhaps this is why intellectual maturity is less about knowing more and more about distinguishing better.
The Bhagavad Gita: Knowledge as Transformation
The Bhagavad Gita presents knowledge not as mere accumulation of propositions but as something capable of transforming the orientation of the individual.
Arjuna’s crisis is not resolved simply by receiving additional information. The dialogue addresses his understanding of duty, action, attachment, selfhood and consequence.
This suggests an important distinction between information and wisdom.
Information can tell us what something is.
Knowledge can organise information.
Understanding can reveal relationships and causes.
Discernment can identify what matters.
Wisdom determines how understanding should shape action.
Thus:
Knowledge becomes wisdom when it changes the quality of judgement.
A person can possess extensive information and still make poor decisions.
The problem is not always informational deficiency.
It may be a deficiency of judgement.
Action Under Uncertainty
The Bhagavad Gita also offers a profound meditation on action under conditions where outcomes cannot be completely controlled.
The well-known verse,
“Karmaṇy-evādhikāras te mā phaleṣu kadācana”
is commonly interpreted as emphasising one’s responsibility for action without unhealthy attachment to the fruits of action.
A modern reading should not reduce this teaching to a simplistic productivity slogan.
Its philosophical depth lies partly in the distinction between agency and outcome.
Human beings can control some conditions.
They cannot control all consequences.
We can prepare for an examination, but not completely determine the questions.
We can make a responsible decision, but not completely control the environment in which it operates.
We can communicate carefully, but not entirely control another person’s interpretation.
The rational response to such uncertainty is neither fatalism nor illusion of total control.
It is disciplined agency.
Uncertainty Is Not Ignorance
Modern decision-making often struggles with uncertainty because human beings prefer binary categories.
True or false.
Success or failure.
Safe or dangerous.
Right or wrong.
But reality frequently exists in probabilistic states.
A scientific prediction can be highly probable without being certain.
A diagnosis can be strongly supported without being infallible.
A strategic decision can be rational without being guaranteed to succeed.
Therefore, intellectual maturity requires the ability to represent uncertainty honestly.
There is an important difference between:
“This is certain.”
and
“The evidence currently makes this the most probable explanation.”
The second statement may sound less decisive.
Epistemically, it is often stronger.
It acknowledges the possibility of revision.
Risk and the Consequences of Error
Not every uncertainty deserves the same degree of investigation.
If we are uncertain about which film to watch, the cost of error is trivial.
If we are uncertain about a major financial decision, the cost is greater.
If we are uncertain about a medical intervention, the consequences may be substantial.
Therefore, rational decision-making must consider not only probability, but also consequence.
This is the logic of risk.
A low-probability event with catastrophic consequences may deserve more attention than a high-probability event with negligible consequences.
Thus, epistemology eventually becomes decision theory.
The question of what is true becomes connected to the question:
What should I do given what I know, and given what I do not know?
The Architecture of Evidence
Evidence is not a single category.
There are observations, measurements, experiments, statistical associations, historical records, expert testimony, personal experiences and theoretical deductions.
Their reliability depends upon context.
An anecdote may provide a meaningful clue but cannot necessarily establish a general law.
A large dataset may reveal a strong pattern but still be vulnerable to sampling bias.
An experiment may establish causality under controlled conditions but have limited external validity.
An expert opinion may be valuable but must be distinguished from empirical evidence.
Therefore, evidence must itself be interrogated.
What is the source?
How was the evidence generated?
What assumptions were involved?
What limitations exist?
Could another explanation account for the same observation?
This is the architecture of rational inquiry.
Information in the Age of Algorithms
The epistemic problem has become more difficult in the digital age.
We no longer suffer primarily from scarcity of information.
We suffer from abundance.
Search engines, social media platforms and recommendation systems continuously select information for us.
The result can be intellectually dangerous.
We may encounter more of what we already believe.
The algorithm learns our preferences.
Our preferences shape what we consume.
What we consume reinforces our preferences.
The result is a feedback loop.
A person can therefore experience a highly personalised information environment and mistake it for reality itself.
The solution is not to reject technology.
It is to become more deliberate about epistemic diversity.
Read beyond ideological comfort.
Consult primary sources when possible.
Distinguish evidence from commentary.
Compare competing explanations.
Verify extraordinary claims.
And remain conscious of the fact that the information environment itself is not neutral.
Disagreement as a Test of Knowledge
A belief that has never encountered serious criticism has not necessarily been validated.
It may simply have been protected.
Intellectual disagreement can therefore serve a constructive purpose.
A rigorous critic identifies hidden assumptions.
An opposing model reveals neglected variables.
A contrary interpretation tests explanatory strength.
This is why serious intellectual traditions cultivate debate.
The purpose of disagreement should not be victory.
It should be epistemic stress-testing.
A proposition that survives serious criticism becomes more credible—not because criticism makes it true, but because plausible alternatives have been examined.
From Problem-Solving to Truth-Seeking
The connection with the previous article now becomes clear.
Problem-solving begins with the identification of a gap between reality and desired conditions.
But before solving the gap, we must establish what the current reality actually is.
This requires observation.
Then we require interpretation.
Then causal analysis.
Then evidence.
Then alternative hypotheses.
Then intervention.
Then measurement.
Then revision.
This is why methodologies such as PDCA and SDLC are more philosophically significant than they may initially appear.
PDCA institutionalises the principle that action must be followed by evaluation and correction.
SDLC institutionalises the principle that systems require requirements analysis, design, testing, deployment and maintenance.
Both embody an important epistemic humility:
Our first model may be incomplete. Therefore, reality must be allowed to correct the model.
That is the essence of rational practice.
The Difference Between Explanation and Understanding
An explanation tells us why we think something happens.
Understanding goes further.
It reveals the structure within which the phenomenon occurs.
Suppose an organisation experiences declining productivity.
An explanation may say:
“Employees are not sufficiently motivated.”
Understanding asks:
What are the incentives?
How are responsibilities distributed?
How is performance measured?
What information is available?
Where do bottlenecks occur?
What behaviours does the system reward?
What behaviours does it unintentionally punish?
What feedback loops exist?
What historical decisions created the present structure?
The second approach is slower.
But it is more likely to reveal the architecture of the problem.
When the Model Becomes the Prison
Every intellectual model simplifies reality.
That is necessary.
A map that contained every detail of the territory would be the territory itself.
Models are useful because they exclude information.
But the same feature that makes a model useful can make it dangerous.
If we forget that the model is a simplification, we begin forcing reality to conform to it.
This happens in management, economics, education, politics and even science.
A metric becomes the objective.
A category becomes an identity.
A theory becomes doctrine.
A procedure becomes more important than its purpose.
The model becomes the prison.
The mature thinker therefore asks periodically:
“What does my model fail to explain?”
That question is often more valuable than asking what the model explains.
The Limits of Certainty
Human knowledge is finite.
Reality is not obliged to fit within the limits of our current understanding.
This does not imply relativism.
The statement “we do not know everything” does not imply “anything could be true.”
There remains a crucial distinction between:
uncertainty
and
arbitrariness.
We can have incomplete knowledge while still possessing strong evidence.
We can recognise uncertainty while maintaining justified conclusions.
We can reject absolute certainty without surrendering rational standards.
This middle position is intellectually demanding.
It avoids both dogmatism and nihilism.
From Knowledge to Wisdom
Perhaps the progression can now be stated more precisely.
Information gives us data.
Knowledge organises data into justified propositions.
Understanding reveals relationships, mechanisms and context.
Discernment distinguishes what is significant from what is incidental.
Judgement determines what should be done.
Experience reveals consequences.
Wisdom integrates knowledge, judgement, consequence and ethical responsibility.
Thus, wisdom is not simply the possession of more information.
It is the ability to use knowledge appropriately within the complexity of life.
The Ethical Dimension of Truth
Truth-seeking without ethical responsibility can become dangerous.
Human beings have developed extraordinary capacities to discover what can be done.
The harder question is what ought to be done.
Technology can tell us what is technically possible.
It cannot, by itself, determine what is morally desirable.
Data can reveal patterns.
It cannot alone determine values.
Efficiency can reduce waste.
It cannot alone establish justice.
Profit can measure financial return.
It cannot alone determine social worth.
Thus, rationality requires a dialogue between epistemology and ethics.
The question of truth must eventually meet the question of value.
Truth and Dharma
This is where the concept of Dharma can enter the discussion—not as a simplistic equivalent of “morality,” but as a complex concept concerning order, duty, right conduct and the sustaining principles appropriate to a context.
The philosophical significance is that correct action cannot be determined solely by technical effectiveness.
A decision may be efficient and still be wrong.
A policy may produce the desired metric and still undermine human dignity.
A solution may solve one problem while creating a larger one.
Therefore, the final test of knowledge is not merely:
Does it work?
It is also:
What does it produce, for whom, and at what cost?
This is where responsible problem-solving becomes wisdom.
The Final Measure of Knowledge
Perhaps the deepest test of knowledge is not how confidently we can defend it.
It is how responsibly we can act upon it.
A person who knows the limits of his knowledge is often more reliable than one who claims certainty without justification.
A leader who says “I do not yet know” may be more trustworthy than one who provides an immediate but unsupported answer.
A scientist who publishes limitations demonstrates greater intellectual integrity than one who hides uncertainty.
A teacher who changes an approach after evidence shows that it is ineffective demonstrates greater professionalism than one who protects a familiar method merely because it has always been used.
Knowledge becomes credible when it remains answerable to reality.
Conclusion: Truth as a Discipline
The search for truth is not merely a search for conclusions.
It is a discipline of how we arrive at conclusions.
It requires us to distinguish perception from interpretation, inference from observation, correlation from causation, belief from knowledge, authority from evidence, possibility from probability, and confidence from certainty.
The Indian philosophical traditions remind us that knowledge has means and conditions. Pramāṇa asks how valid knowledge is established. Viveka asks us to distinguish. Svadhyaya turns inquiry toward the knower. The Upanishadic tradition challenges us to examine the foundations of our understanding. The Bhagavad Gita connects knowledge with action, responsibility and disciplined agency.
Modern science contributes another indispensable discipline: the willingness to be corrected by evidence.
Systems thinking reminds us that phenomena exist within relationships.
Critical thinking reminds us to interrogate assumptions.
PDCA reminds us to learn through feedback.
SDLC reminds us that every constructed system requires testing, maintenance and adaptation.
These are not identical philosophies, nor should they be artificially merged into one system.
But they converge upon a powerful intellectual principle:
Reality must have the final authority over our explanations of reality.
We may begin with a hypothesis.
We may construct a model.
We may hold a belief.
We may formulate a theory.
But if reality persistently contradicts it, intellectual integrity requires revision.
Perhaps, then, wisdom begins with a simple admission:
I may not know.
From that admission comes inquiry.
From inquiry comes evidence.
From evidence comes knowledge.
From knowledge comes understanding.
From understanding comes discernment.
From discernment comes responsible action.
And from action comes experience—the feedback through which knowledge is tested against life itself.
The journey is therefore not:
Belief → Certainty.
It is:
Question → Inquiry → Evidence → Understanding → Discernment → Action → Feedback → Revision.
And the cycle continues.
Truth, in this sense, is not merely something we possess.
It is something we approach through disciplined inquiry.
Perhaps the most intellectually honest person is not the one who has an answer for every question.
It is the one who knows:
what is known,
what is unknown,
why something is believed,
what evidence supports it,
what could disprove it,
and when the courage to revise is greater than the comfort of being right.
For ultimately, the search for truth is not a contest between certainty and doubt.
It is a movement from unexamined belief toward justified understanding.
And perhaps the architecture of truth is never completed.
It is built continuously—
by observation,
by reason,
by evidence,
by criticism,
by reflection,
by experience,
and, above all,
by the willingness to ask one question again and again:
“How do I know?”
Because sometimes the most profound discovery is not the truth we find.
It is the recognition of how much more carefully we must learn to seek it.
— Prabhash Chandra