EDUCATION SUBJECT ATLAS · ETHICS

What Is Ethics?

Ethics is the disciplined study of what people, organisations and societies ought to do, what kinds of character and institutions are worth cultivating, which harms and benefits matter, what rights and duties should be respected, and how decisions should be made when important values conflict.

Ethics is not simply personal preference, social custom, religious rule or legal compliance. Those can all influence moral life, but ethical reasoning asks a harder question: what reasons justify treating an action, rule, institution or character trait as right, wrong, fair, admirable, harmful or responsible?

Ethics begins when “I want” and “people usually do” are no longer sufficient reasons.

The first distinction: fact and value

Ethical decisions often contain both empirical and normative claims. “This policy will reduce accidents” is an empirical claim. “Reducing preventable accidents should count heavily in favour of the policy” is a normative claim.

Confusion arises when facts are treated as if they automatically determine values or values are disguised as factual inevitabilities. Ethical reasoning becomes clearer when the layers are separated and then recombined explicitly.

Morality and ethics

In ordinary language, morality and ethics are often used interchangeably. In some contexts, morality refers to the norms or principles people hold, while ethics refers to the systematic study and justification of those norms.

The distinction is useful but not mandatory. What matters is whether the argument identifies the relevant values, reasons and consequences clearly.

Normative ethics

Normative ethics develops principles for deciding what ought to be done. Major traditions include consequentialism, deontology, virtue ethics, care ethics and contractual approaches.

These traditions illuminate different dimensions of moral life. Real decisions often require more than one lens because consequences, rights, duties, relationships and character can all matter simultaneously.

Consequentialism

Consequentialist theories evaluate actions partly or entirely by their outcomes. The central question is what consequences an action creates and whose welfare counts.

The strength of consequentialism is attention to real effects. Its difficulty is that outcomes can be uncertain, long-term and difficult to compare. A policy can help many people slightly while harming a few severely; the ethical analysis must decide how such differences should be weighted.

Deontology

Deontological approaches emphasise duties, rights and constraints. Some actions may be wrong even when violating the rule appears to produce better aggregate consequences.

This protects persons from being treated merely as instruments, but difficult cases arise when duties conflict or when following a rule produces severe harm.

Virtue ethics

Virtue ethics asks what kind of person one should become and what practical wisdom good character requires. Courage, honesty, justice, temperance and generosity are examples of virtues in different traditions.

Virtue ethics is especially useful where rules cannot anticipate every context. It asks what a wise, responsible agent would perceive and do in the situation.

Care ethics

Care ethics emphasises relationships, dependency, vulnerability and responsibilities created through human connection. It challenges models of moral agency that imagine isolated individuals making decisions without social ties.

Care is particularly important in family life, education, healthcare and other domains where people depend directly on one another.

Contractual and social-contract approaches

Contractual approaches ask which principles people could reasonably accept for governing shared life. They are especially useful when designing institutions because they force decision-makers to consider rules from more than one social position.

Rights

Rights identify protected claims or freedoms that should not be overridden casually. Rights can concern liberty, privacy, bodily integrity, participation, property, expression or other interests.

Rights can conflict. Privacy may conflict with safety; expression may conflict with protection from certain harms. Ethical reasoning asks which rights are engaged, how strongly and under what conditions limitation is justified.

Duties

Duties are obligations to act or refrain from acting. Some duties arise from general moral principles; others arise from roles, promises, contracts or professional relationships.

A teacher, doctor, engineer or director may have responsibilities that go beyond those of a stranger because role-specific trust creates additional obligations.

Justice

Justice concerns fair treatment, distribution, procedure, recognition and correction. Different theories emphasise equality, need, desert, liberty, sufficiency or opportunity.

Ethical disagreement about justice often comes from prioritising different dimensions rather than simple indifference to fairness.

Equality and equity

Equality gives people the same treatment or share under a defined rule. Equity adjusts for relevant differences in need, barrier or circumstance. Neither is automatically correct in every setting.

The ethical task is to explain which differences are morally relevant and why.

Autonomy

Autonomy concerns a person’s ability to make meaningful choices about their own life. Respect for autonomy requires more than obtaining a formal yes. Information, understanding, freedom from coercion and genuine alternatives may matter.

Autonomy is central to healthcare, research, education, employment and digital design.

Consent

Consent can transform the moral status of an action, but valid consent has conditions. A person should understand what they are agreeing to, possess decision-making capacity and be sufficiently free from coercion or deception.

Clicking “accept” on a long unread interface does not settle every ethical question about data use.

Harm

Harm can be physical, psychological, financial, social, environmental or institutional. Some harms are immediate and visible; others accumulate slowly or are distributed across populations.

Ethical analysis should identify who bears harm, how severe it is, whether it is reversible and whether the people benefiting are the same people bearing the cost.

Benefit

Benefit is also multidimensional. More income, longer life, convenience, autonomy, dignity and social trust are not interchangeable goods.

Ethics becomes difficult when an intervention creates one kind of benefit by sacrificing another.

Intention and outcome

Intention matters morally, but good intentions do not erase foreseeable harm. Conversely, a bad outcome can occur despite careful and responsible decision-making under uncertainty.

Ethical judgment should distinguish intention, process, foreseeability and outcome rather than collapsing all responsibility into the final result.

Moral luck

Moral luck describes situations where factors beyond an agent’s control influence how we judge them. Two drivers may take the same reckless risk, but only one happens to injure someone.

The difference in outcome matters, yet the similarity in decision quality also matters. Moral luck exposes tension between judging choices and judging consequences.

Metaethics

Metaethics asks what moral claims mean and what, if anything, makes them true. Are moral truths objective, constructed, relative, expressive or grounded in reasons that any rational person should recognise?

Metaethical questions sit behind practical disputes even when people do not state them explicitly.

Moral realism and anti-realism

Moral realists hold that at least some moral claims are true independently of individual preference. Anti-realist positions challenge or reinterpret that idea in different ways.

These debates matter because they influence what people think moral disagreement is: discovery of truth, negotiation of norms, expression of attitudes or something else.

Relativism

Moral relativism connects moral truth or justification to cultures, frameworks or standpoints. Recognition of cultural difference does not automatically prove relativism. People can disagree widely about matters that still have better and worse answers.

Ethical education should distinguish empirical diversity in moral beliefs from philosophical claims about whether any cross-cultural judgment is justified.

Applied ethics

Applied ethics examines concrete domains such as medicine, business, law, environment, engineering, technology, research and public policy.

Applied ethics is not simply applying one theory mechanically. Real cases involve incomplete facts, role obligations, institutional constraints and uncertainty.

Professional ethics

Professions receive trust because they claim specialised competence. That trust creates ethical duties around honesty, confidentiality, competence, conflicts of interest and protection of the public.

A professional code is a useful floor, not a complete substitute for judgment. Novel situations can arise before rules are updated.

Research ethics

Research ethics protects participants, scientific integrity and public trust. Consent, risk, privacy, fair selection, data integrity and responsible publication are central.

Bad research can be unethical even when no participant is directly harmed if data are fabricated or conclusions knowingly overstated.

Medical ethics

Medical ethics addresses autonomy, benefit, harm, justice, confidentiality and professional responsibility in healthcare. Decisions often involve vulnerable people and uncertain outcomes.

Clinical ethics is strongest when evidence, patient values and professional duties are considered together rather than allowing one dimension to erase the others.

Business ethics

Business ethics examines obligations to customers, workers, investors, suppliers, communities and the environment. Legal compliance is necessary but not always sufficient.

Metrics can create ethical risk when organisations reward outcomes without considering how those outcomes are achieved.

Engineering ethics

Engineering ethics centres on safety, honesty, competence and public consequence. Engineers may face pressure from budgets and schedules, but known safety risks cannot be treated as ordinary trade-offs without ethical scrutiny.

This connects ethics with engineering: technical decisions become ethical decisions when people must live with their consequences.

Environmental ethics

Environmental ethics asks what obligations humans have toward ecosystems, animals, future generations and the non-human world. Climate and biodiversity problems make temporal and collective dimensions especially important.

Future people cannot consent to present choices, yet present choices can constrain their options substantially.

Technology ethics

Technology ethics examines privacy, autonomy, surveillance, safety, labour, manipulation, inequality and power created by technical systems.

“Can we build it?” is an engineering question. “Should we build it, deploy it here, collect this data or automate this decision?” is an ethical question.

AI ethics

AI ethics studies automated decisions, bias, privacy, agency, transparency, accountability, labour effects, safety and the distribution of power created by intelligent systems.

AI systems complicate responsibility because outcomes may involve model developers, data providers, deployers, users and organisations. Ethical governance must therefore map the full chain rather than assign responsibility only to the final interface.

Fairness in algorithms

Algorithmic fairness is not one metric. Equal error rates, equal treatment, equal opportunity and calibration can conflict under some conditions.

The ethical question is therefore not “Is the model fair?” in the abstract, but “Which fairness principle fits this decision and why?”

Transparency and explainability

Transparency helps people understand how decisions are made, but complete technical disclosure is not always useful or possible. Explainability should be matched to the audience and decision.

A person denied a benefit may need a meaningful reason and appeal path, not thousands of model parameters.

Collective action

Some ethical problems arise because individually reasonable actions aggregate into harmful collective outcomes. Pollution, congestion, overfishing and misinformation can have this structure.

Ethics therefore needs institutions, not only individual virtue. Rules, incentives and shared systems can change what responsible action requires.

Institutional ethics

Institutions distribute power and opportunity through procedures. An organisation can contain well-intentioned individuals yet produce harmful outcomes because incentives, information or accountability are poorly designed.

Ethical analysis should therefore inspect systems as well as persons.

Conflict of interest

A conflict of interest exists when secondary interests could improperly influence a primary duty. The ethical problem can exist even when no misconduct has yet occurred.

Disclosure helps, but some conflicts require recusal, independent review or structural separation.

Ethical decision-making under uncertainty

Many introductory ethics pages discuss ideal cases where outcomes are known. Real decisions rarely offer that luxury. Benefits and harms may be probabilistic, long-term or deeply uncertain.

Ethical reasoning under uncertainty must distinguish uncertainty about facts from uncertainty about moral principles. A decision-maker may know what values matter but not what will happen, or may be unsure which value should dominate even if outcomes are known.

Expected consequences and moral constraints

Expected-value reasoning can help compare uncertain outcomes, but some ethical principles impose constraints that are not easily represented as simple utilities. Consent, rights and prohibitions can matter even when expected aggregate benefit is large.

A mature ethical process therefore does not reduce every value to one number unless there is a justified reason to do so.

Precaution

Precaution gives additional weight to preventing severe or irreversible harm when evidence is uncertain. It is useful in safety, environment and public health, but extreme precaution can also block beneficial action.

The practical challenge is proportionality: how much caution is justified by severity, uncertainty and reversibility?

Reversibility

Reversible decisions can often tolerate more experimentation. Irreversible decisions deserve stronger evidence and review because mistakes cannot easily be repaired.

This is an important bridge between ethics and management: decision process should scale with consequence.

Moral disagreement

People can disagree because they have different facts, different predictions, different values or different theories about how values should be combined.

Good ethical discussion identifies which layer contains the disagreement. Arguing about consequences will not resolve a conflict that is actually about rights.

Ethical pluralism

Ethical pluralism recognises that several moral considerations can be genuinely important without reducing all of them to one master principle. Harm, fairness, loyalty, liberty and care may all matter in a case.

Pluralism creates complexity but can better match real moral life than a framework that ignores inconvenient values.

The ethical decision pipeline

Define the decision. What exactly must be chosen?
Establish the facts. Separate known, estimated and unknown information.
Map stakeholders. Who is affected, including those with weak voice?
Identify values and duties. Rights, harms, benefits, fairness, promises, care and role responsibilities.
Generate alternatives. Avoid false dilemmas.
Trace consequences. Short-term, long-term, direct, indirect and distributional.
Check constraints. Are some actions impermissible regardless of benefit?
Test reversibility and uncertainty. How costly is being wrong?
Seek challenge. What would a reasonable critic say?
Decide, document and review. Preserve the reasoning so future evidence can improve it.

This operational layer is where ethics becomes usable rather than remaining a catalogue of theories.

Ethics and law

Law and ethics overlap but are not identical. Some unethical conduct is legal; some legal obligations are morally controversial. Law creates enforceable institutional rules, while ethics evaluates what rules and actions ought to be justified.

This makes law and ethics complementary rather than interchangeable.

Ethics and psychology

Psychology studies how people actually make moral judgments; ethics asks how they ought to reason. Bias, conformity and emotion can influence judgment, but describing those influences does not settle what is right.

Understanding human behaviour can nevertheless improve ethical systems by revealing predictable failure modes.

Ethics and politics

Political institutions make decisions affecting people who disagree about values. Ethical reasoning helps evaluate legitimacy, rights, distribution and public responsibility, while political science studies how institutions actually behave.

Ethics and business

Business decisions allocate risk, opportunity and information among customers, workers, investors and communities. Profitability does not automatically justify every method used to produce profit.

Ethics asks whether the value-creation process respects the people who make and receive that value.

Ethics and literature

Literature gives ethical problems narrative depth. Characters act with partial knowledge, mixed motives and relationships that abstract examples often remove.

This makes literature a useful laboratory of moral imagination, while philosophy supplies the tools for evaluating the reasons exposed by the story.

Ethics and logic

Ethical arguments still require valid or strong reasoning. A moral conclusion may fail because an empirical premise is false, because the inference is weak or because a hidden value assumption was never defended.

Logic does not supply moral values, but it makes ethical commitments and consequences more visible.

Common misconceptions

“Ethics is just personal opinion.” Ethical claims can be supported or weakened by reasons, consistency and evidence.
“If something is legal, it is ethical.” Law and ethics overlap imperfectly.
“Good intentions make an action ethical.” Foreseeable consequences and duties still matter.
“Ethics means following one rule.” Real cases often involve competing values and uncertain consequences.
“A code of ethics solves professional dilemmas.” Codes guide practice but cannot anticipate every case.
“AI ethics is only about bias.” Privacy, autonomy, safety, labour, accountability and power also matter.

Mini case: the useful shortcut

An organisation can save time by collecting more personal data than it currently needs. The technical team says the data may become useful later. Ethical analysis asks whether future convenience justifies present privacy risk, whether consent is meaningful, who can access the data, how long it will be kept and what happens if it is breached.

The ethical issue is not anti-technology. It is proportionality between benefit, power and risk.

Mini case: one risky school experiment

A new teaching method might improve learning but evidence is limited. The ethical decision depends on expected benefit, possible harm, reversibility, informed participation, monitoring and whether students who do not benefit can return quickly to a proven approach.

Uncertainty does not prohibit experimentation. It changes the safeguards required.

A systems model of ethics

ENTITY: persons, groups, institutions, technologies, environments and future stakeholders.
STATE: vulnerability, rights, consent, knowledge, power and distribution of benefits or harms.
OCCURRENCE: decisions, interventions, omissions, disclosures, harms and repairs.
RELATIONSHIP: duty, care, authority, dependence, trust, contract and accountability.
INTENT: aims, motives, role obligations and institutional objectives.
OBSERVATION: facts, forecasts, testimony, impact evidence and lived experience.
ARTIFACT: policies, codes, consent forms, models, systems and institutional procedures.
CLAIM: judgments about what is justified, permissible, obligatory or wrong.
VOID: unknown consequences, missing stakeholders, moral uncertainty and unmeasured harm.

Broad-context analysis searches for the stakeholder and consequence that ordinary framing leaves outside the picture. A systems pass then rotates each option through rights, harms, duties, uncertainty, reversibility and institutional incentives before the decision is released.

How to think ethically

Separate facts from values.
Identify everyone affected.
Make rights and duties explicit.
Trace consequences across time.
Ask whether consent is meaningful.
Test fairness from more than one position.
Identify uncertainty and irreversibility.
Check conflicts of interest.
Invite a strong opposing argument.
Document why the chosen action remains defensible.

Ethics across the learning journey

Young learners can begin with fairness, promises, harm, honesty and responsibility. Secondary learners can compare moral principles, analyse dilemmas and separate factual from value claims. Advanced study adds metaethics, political philosophy, professional ethics, technology ethics and formal decision-making under uncertainty.

The progression is from “what feels fair?” to “which principles, facts and uncertainties justify this decision?”

Why ethics belongs inside education

Education creates capability, and capability creates power. Learners eventually make decisions affecting other people, organisations and environments. Ethical literacy helps them recognise that competence without responsibility can produce efficient harm.

Ethics belongs across the curriculum because every domain eventually reaches a point where facts tell us what can happen but not what ought to happen.

What Is Ethics? | Values, Duties, Consequences, Character, Uncertainty and Responsible Action