An analogy is useful when it transfers structure, not when it merely transfers resemblance. Two situations can look alike and behave differently. Two situations can look different yet share the same mechanism. Under examination pressure, students often reach for familiar examples because analogy is fast: “This is like that.” The danger is that the comparison can carry more argumentative weight than the similarities have earned.
This guide owns the timed-performance layer of analogy control. It does not replace general comparative reasoning, metaphor, literary analysis or compare-and-contrast vocabulary. SETC’s task here is narrower: how does a writer use one case to illuminate another, decide which similarities are structurally relevant, preserve differences that change the conclusion, and stop the analogy exactly where it breaks?
The central rule is: an analogy is a bridge for reasoning, not a certificate of equivalence. The writer must be able to name the common structure, identify the relevant difference, explain whether that difference weakens the transfer, and return to the actual case rather than letting the borrowed case replace it.
The quick route
Use six moves. First, state the target problem. Second, state the source case you want to use as an analogy. Third, identify the structural feature the two cases share. Fourth, identify the most important difference. Fifth, decide which conclusion transfers and which does not. Sixth, return to the target case with a narrower, tested claim.
A useful planning line is: A is like B because ___. But A differs from B in ___. Therefore the analogy supports ___, not ___.
Why analogies are attractive under pressure
Analogies compress thought. A writer who already understands one system can use it to organise an unfamiliar one. This is especially helpful when the essay prompt asks about a new technology, institution or social change. But compression creates risk. The writer may import properties that are vivid in the source case but absent in the target case.
Maren’s difficulty is narrative confidence. Once an analogy sounds elegant, she can build a paragraph around it and stop checking whether the relevant mechanism truly matches. Her training is to write the break point immediately after the similarity.
Iona’s difficulty is evidence hierarchy. She sometimes treats an analogy as if it were another piece of empirical evidence. Her repair is to label the analogy’s job: explanation, hypothesis, boundary test or structural comparison. It can clarify evidence; it cannot replace evidence about the target case.
Leonie’s difficulty is time. She can spend too long refining the source case. Her examination rule is that the analogy must become useful within two or three sentences. If it needs a page of background, it is probably too expensive for the job.
Analogy versus comparison
Comparison can place two cases side by side using common criteria. Analogy goes further: it suggests that because the cases share some structure, insight from one may transfer to the other. That inferential step is what needs control.
Analogy versus example
An example shows an instance of the claim. An analogy uses a different case to illuminate the claim. “A school timetable is like a transport network” is not an example of a timetable; it is a structural comparison.
Analogy versus metaphor
A metaphor can illuminate through imaginative language without requiring a full inferential mapping. In an argumentative essay, if the metaphor begins supporting a conclusion, it has become an analogy and needs testing.
Analogy versus precedent
A precedent is a past case used because its prior outcome or judgement may guide the present. An analogy can connect cases from any domain. Precedent reasoning therefore requires historical and institutional continuity; analogy control requires structural fit.
Analogy versus false equivalence
False equivalence occurs when important differences are ignored and two cases are treated as morally, causally or practically the same. Good analogy control prevents that slide.
The source and target
The source is the case you already know. The target is the case you are trying to understand. Strong analogical reasoning moves only the features justified by the mapping.
Surface similarity
Surface features are visible but may be irrelevant: both involve screens, both involve large crowds, both are government programmes, both happened during economic growth. Surface resemblance can help generate an analogy but should not decide its strength.
Structural similarity
Structural features describe relationships: both systems create network effects; both interventions change incentives; both tasks require repeated retrieval; both institutions depend on trust. Structural similarity is usually the stronger basis for transfer.
Mechanism similarity
The strongest analogies often share a mechanism. If both cases work by reducing search cost, increasing feedback speed or creating switching barriers, insight can travel even if the surface domains are very different.
Outcome similarity
Two cases can produce similar outcomes through different mechanisms. Similar outcomes alone do not justify assuming identical causes.
Context similarity
Institutional, cultural, legal, technological and economic context can strengthen or weaken transfer. A mechanism that works under high trust may behave differently under low trust.
Scale similarity
A small classroom and a national education system can share a mechanism but differ dramatically in coordination cost, heterogeneity and feedback speed. Scale is often the hidden break point.
Time-horizon similarity
An analogy between a short-term crisis and a long-term institution can fail because incentives and adaptation change over time.
Risk similarity
A low-stakes recommendation system and a high-stakes medical or legal decision system may share algorithmic architecture while requiring different tolerance for error. The analogy must preserve the risk difference.
Stakeholder similarity
Two policies can affect different groups even when total outcomes look similar. Distribution can break an analogy that works on average.
The mapping table
During practice, create two columns. Put source-case elements on the left and target-case elements on the right. Draw only the mappings that perform a real argumentative job. Leave unmatched features unmatched.
The one-mechanism rule
Under time pressure, one strong structural match is usually more useful than five weak similarities. Choose the relation that actually supports the conclusion.
The decisive-difference rule
Every analogy should survive one question: What difference between these cases could reverse the conclusion? If such a difference exists and is present, the analogy must be narrowed or abandoned.
The break-point sentence
A disciplined analogy often contains its own limit: “The comparison is useful because both systems rely on network effects; however, the platform can expand almost instantly whereas the transport network faces physical capacity constraints.” The break point prevents false equivalence.
Analogy strength is claim-specific
An analogy can be strong for one claim and weak for another. A library and the internet are both information systems, which may support a claim about retrieval. The analogy is weaker for speed, scale, ownership or physical preservation.
Analogy as explanation
An analogy can make an unfamiliar mechanism understandable. “A rubric is like a ruler” can explain that it provides a common standard. But the analogy should not imply that qualitative judgement becomes perfectly mechanical.
Analogy as hypothesis generator
If two systems share a mechanism, the source case can suggest what to look for in the target. This is a question-generating role, not proof.
Analogy as boundary test
A contrasting case can show where the mechanism stops. If the analogy only works at small scale, scale becomes a boundary condition.
Analogy as policy warning
Historical or institutional analogies can identify risks worth testing. They should not be used to announce that the same outcome is inevitable.
Analogy as design transfer
A design that solves one coordination problem can inspire another domain if the structural constraint matches. The writer must still account for domain-specific risks.
The category trap
Two cases in the same category may be less analogous than two cases in different categories. Two schools can use completely different incentive structures, while a school and a hospital can share a queue-management problem.
The label trap
Words such as “platform,” “market,” “network,” “community” and “ecosystem” can make cases sound structurally alike before the relationships have been tested.
The moral-equivalence trap
Similarity in one feature does not make two actions morally equivalent. Scale of harm, intention, alternatives, rights and context may differ. Analogy can illuminate a principle without equating total moral weight.
The historical-equivalence trap
Two periods can share economic or political stress while differing in institutions, technology and international context. Use history to transfer mechanisms cautiously.
The technology-equivalence trap
Calling two systems “AI” or “digital” does not establish comparable capability, autonomy or risk. Generalise from the relevant feature, not the umbrella term.
The biological-equivalence trap
Natural systems are frequently used as analogies for organisations or societies. These can clarify adaptation or feedback while becoming misleading if biological inevitability is imported into human institutions.
The machine-equivalence trap
Calling a school, government or mind a “machine” can highlight inputs and outputs but understate agency, values and emergence. Keep the analogy’s limits visible.
The market-equivalence trap
Not every system with choice behaves like a competitive market. Information, switching, pricing and entry must exist strongly enough for the market analogy to carry.
The war-equivalence trap
“War on drugs,” “war on disease,” “war on misinformation” can create urgency but may import assumptions about enemies, victory and acceptable collateral cost. Treat rhetorical analogy separately from empirical reasoning.
The disease-equivalence trap
Ideas and information are often described as “viral.” Network spread may justify part of the analogy, but people actively interpret information in ways viruses do not.
The ecosystem-equivalence trap
Markets and technologies can resemble ecosystems in interdependence and adaptation. However, deliberate regulation and human planning distinguish them from purely ecological processes.
The infrastructure-equivalence trap
Knowledge systems, roads and digital networks can share bottleneck logic while differing in marginal cost and replication speed. Transfer only the matching structure.
The classroom-as-laboratory trap
A classroom experiment can reveal learning mechanisms but lacks some controls of a laboratory. Do not let the analogy imply experimental certainty.
The family-as-company trap
Families and firms both coordinate resources, but their obligations and goals differ. Efficiency criteria cannot be imported without moral examination.
The country-as-company trap
A government and a company both allocate resources but have different obligations, time horizons, coercive powers and accountability. Corporate analogies can illuminate management while distorting citizenship.
The student-as-customer trap
Students receive services but are also participants in learning whose effort partly determines output. Consumer analogies can clarify experience but weaken responsibility analysis if transferred wholesale.
The examination-as-game trap
Games and exams both have rules, constraints and strategies. But an exam measures performance against learning goals rather than merely victory over opponents. Strategy matters without making learning irrelevant.
The analogy hierarchy
Classify analogies by strength:
- Decorative: mainly rhetorical resemblance.
- Illustrative: helps explain one relation.
- Structural: shares several relevant relationships.
- Mechanistic: shares the process driving the outcome.
- Predictive: supports a cautious expectation because structural conditions align strongly.
Most timed essays need only Levels 2–4. Do not claim prediction without evidence.
The source-case quality problem
An analogy built from a misunderstood source case transfers misunderstanding. Verify the source case before using it as a reasoning model.
The target-case knowledge problem
If little is known about the target, analogy can dominate merely because it is familiar. Keep uncertainty visible.
The asymmetric-detail problem
Students often know the source case in detail and the target case vaguely. This makes the analogy seem stronger than it is. Compare at the same level of abstraction.
The abstraction-level problem
“Both involve people” is too abstract to help. “Both reward visible metrics in ways that may encourage gaming” is more structurally precise.
The similarity-count fallacy
Ten superficial similarities do not outweigh one decisive structural difference. Relevance matters more than count.
The difference-count fallacy
Many irrelevant differences do not destroy a strong mechanism analogy. Ask whether the difference changes the relation you are transferring.
The directionality problem
If A resembles B, it does not follow that every inference can run from B to A. The source may have features the target lacks.
The analogy-chain problem
A is like B, B is like C, therefore A is like C can compound distortion. Keep chains short and test each mapping independently.
The mixed-analogy problem
Switching metaphors mid-argument can smuggle contradictions. A system cannot simultaneously be treated as a competitive market, a family and a machine unless the writer specifies which relation each analogy illuminates.
The reification problem
A useful analogy can become a literal model. Remember that calling knowledge a “network” does not mean it possesses every property of physical networks.
The emotional-transfer problem
Source cases carry emotional associations. Comparing a policy to war, disease, slavery or natural disaster can import moral weight without argument. Separate emotional resonance from structural similarity.
The inevitability-transfer problem
If the source case ended badly, the target is not therefore destined to do so. Analogy can identify a risk mechanism, not historical fate.
The success-transfer problem
If the source case succeeded, the target still needs enabling conditions. Analogy does not guarantee replication.
The scale-break problem
Mechanisms can change when moving from classroom to school, school to system or start-up to national infrastructure. Coordination and heterogeneity often become decisive.
The incentive-break problem
Two organisations may share structure but reward participants differently. Incentives can reverse expected behaviour.
The information-break problem
Markets, teams and institutions behave differently depending on who knows what. An analogy that ignores information asymmetry can fail even when other structures match.
The power-break problem
Relationships can look reciprocal while one side has coercive power. Power differences can invalidate simple exchange analogies.
The rights-break problem
Efficiency analogies may fail when one domain involves rights that cannot be traded away like ordinary costs.
The reversibility-break problem
A low-cost reversible experiment and an irreversible policy can share mechanics but require different evidence thresholds.
The uncertainty-break problem
One source case may have established outcomes while the target concerns a forecast. Confidence should decrease when moving from observed to future worlds.
Worked analogy 1 — School as hospital
Useful mapping: both diagnose problems before selecting interventions. Break point: education is not medicine; learning outcomes depend heavily on learner participation, and individual variation is not pathology. Safe transfer: diagnostic sequencing, not medical authority.
Worked analogy 2 — Essay as building
Useful mapping: structure supports the whole and parts have different jobs. Break point: essays can revise their thesis as evidence develops; buildings do not change foundations mid-construction so easily. Safe transfer: planned architecture, not rigid immobility.
Worked analogy 3 — Evidence as currency
Useful mapping: evidence has value only in relation to what it can support. Break point: evidential value is not fungible; one piece cannot always substitute for another. Safe transfer: scarcity and allocation, not exact exchange.
Worked analogy 4 — Attention as budget
Useful mapping: limited attention must be allocated. Break point: cognitive capacity can change with expertise and fatigue rather than behaving like fixed money. Safe transfer: opportunity cost.
Worked analogy 5 — Feedback as navigation
Useful mapping: feedback identifies deviation and supports correction. Break point: there may be multiple legitimate destinations in writing. Safe transfer: course correction, not one predetermined route.
Worked analogy 6 — Algorithms as editors
Useful mapping: both filter and rank information. Break point: automated systems may optimise metrics at scale without human editorial accountability. Safe transfer: gatekeeping function, not identical responsibility.
Worked analogy 7 — Social media as public square
Useful mapping: public expression and social interaction. Break point: algorithmic ranking, private ownership and global scale differ from a physical square. Safe transfer: speech and gathering concerns, not full constitutional equivalence.
Worked analogy 8 — Data as oil
Useful mapping: raw inputs can gain value after processing. Break point: data can be copied, reused and generated without depletion. Safe transfer: processing value, not scarcity economics.
Worked analogy 9 — Education as investment
Useful mapping: present resources can create future returns. Break point: education also has civic, personal and cultural value not captured by financial return. Safe transfer: delayed benefit, not purely monetary valuation.
Worked analogy 10 — Democracy as marketplace of ideas
Useful mapping: competing ideas can be evaluated by citizens. Break point: information asymmetry, manipulation and unequal voice mean the market mechanism is imperfect. Safe transfer: competition among claims, not automatic truth selection.
Worked analogy 11 — Memory as storage
Useful mapping: information can be retained and retrieved. Break point: memory reconstructs, compresses and changes; it is not static disk space. Safe transfer: retrieval constraints, not literal storage.
Worked analogy 12 — Learning as muscle training
Useful mapping: adaptation requires repeated challenge and recovery. Break point: cognitive knowledge depends on representation, understanding and transfer, not only load. Safe transfer: progressive practice and recovery.
Worked analogy 13 — Language as toolkit
Useful mapping: different tools serve different functions. Break point: language choices interact and create meaning rather than operate independently. Safe transfer: functional selection.
Worked analogy 14 — City as organism
Useful mapping: interdependent systems circulate resources and respond to stress. Break point: cities are governed by intentional agents and political choices. Safe transfer: interdependence, not biological inevitability.
Worked analogy 15 — Institution as memory
Useful mapping: records and procedures preserve knowledge beyond individuals. Break point: institutions can choose, enforce and reinterpret rules. Safe transfer: continuity across generations.
Worked analogy 16 — Regulation as guardrail
Useful mapping: a boundary can reduce catastrophic deviation while allowing movement. Break point: regulation can be complex, costly and adaptive rather than physically passive. Safe transfer: bounded freedom.
Worked analogy 17 — Exam rubric as map
Useful mapping: it identifies important dimensions and helps orient effort. Break point: a rubric is an evaluation standard, not a complete route to success. Safe transfer: orientation, not recipe.
Worked analogy 18 — Market competition as tournament
Useful mapping: participants compete under rules. Break point: market outcomes depend on entry, information and repeated interaction, not a fixed bracket. Safe transfer: competitive incentives.
Worked analogy 19 — Scientific theory as model
Useful mapping: simplified representations explain and predict selected features. Break point: models are judged by evidence and can coexist at different scales. Safe transfer: selective representation.
Worked analogy 20 — Revision as debugging
Useful mapping: errors are located, classified and repaired. Break point: writing quality is not reducible to binary bugs; style and judgement can have multiple valid forms. Safe transfer: diagnostic repair.
The analogy ledger
During practice, use five columns: source, target, shared structure, decisive difference, safe inference. The last column forces the writer to state what the analogy is actually allowed to do.
The one-line analogy drill
Complete: “X is like Y because ____, but unlike Y because ____; therefore this analogy supports ____.” Keep the entire reasoning within one sentence.
The false-equivalence drill
Give two emotionally charged cases with one superficial similarity. Require the learner to identify the missing structural difference before making any judgement.
The abstraction drill
Rewrite a weak similarity at a more structural level. “Both are online” may become “both lower distribution cost while increasing dependence on ranking systems.”
The break-point drill
For every analogy, generate the nearest condition under which the transfer would fail.
The source-removal drill
Remove the source case after drafting. Does the target argument still stand on evidence? If not, the analogy may be replacing proof.
The source-switch drill
Use a different source case with the same structural mechanism. If the inference survives, you probably identified the relevant structure rather than relying on prestige.
The emotional-detox drill
Replace loaded source cases with neutral labels. Reassess whether the structural reasoning still holds.
The mechanism-match drill
Write the mechanism in both cases using the same verbs: increases, reduces, concentrates, disperses, rewards, penalises, delays, accelerates. Mismatched verbs often reveal a weak analogy.
The scale-swap drill
Move the target from a small to a large system. Decide whether the mechanism still operates.
The risk-swap drill
Move from low stakes to high stakes while keeping the mechanism. Decide whether the inference changes because acceptable error changes.
The rights-swap drill
Add a rights constraint to a previously efficiency-focused analogy. Decide whether the decision rule changes.
The time-swap drill
Move from short-term response to long-term adaptation. Identify whether the analogy remains stable.
The mixed-analogy drill
Find two analogies in one paragraph. Check whether their implied models contradict each other.
Week 1 — Source and target
Learn to name the two cases and the exact inference being transferred.
Week 2 — Structural similarity
Move from surface labels to mechanisms and relationships.
Week 3 — Decisive differences
Practise finding the difference that can reverse the inference.
Week 4 — Safe transfer
State the narrowest conclusion the analogy actually earns.
Week 5 — False equivalence
Use moral, historical and technology cases where emotional similarity can mislead.
Week 6 — Add the clock
Build and test one analogy in sixty seconds during planning.
Week 7 — Full essays
Use analogies sparingly and make break points visible.
Week 8 — Unseen transfer
Use unfamiliar source and target domains while preserving the same reasoning procedure.
Failure mode — resemblance becomes proof
The writer says “this is like that, therefore the same result follows.” Repair by naming mechanism and decisive difference.
Failure mode — source case replaces target evidence
The paragraph becomes detailed history about the analogy source. Repair by returning to the target after one or two mapping sentences.
Failure mode — analogy is only decorative
The comparison sounds clever but does no reasoning work. Remove it or state the transferable relation.
Failure mode — decisive difference is hidden
The writer notices a break point but omits it because it weakens the rhetoric. Repair by making the boundary explicit.
Failure mode — every difference kills the analogy
The learner overreacts to irrelevant differences. Ask whether the difference changes the mapped mechanism.
Failure mode — emotional weight transfers
A highly charged source case makes the target seem morally equivalent. Repair by separating structural similarity from severity and intention.
Failure mode — history becomes destiny
A past pattern is imported as inevitable. Repair by distinguishing analogy from precedent and by checking changed institutions.
Failure mode — scale disappears
A mechanism from a small case is transferred to a system without coordination analysis. Repair with a scale check.
Failure mode — umbrella category does the work
“Both are AI,” “both are markets,” or “both are democracies” substitutes for structural mapping. Repair by naming the shared mechanism.
Failure mode — analogy chains multiply
A paragraph stacks analogies and loses the target. Choose one source and one inferential job.
Teacher feedback that transfers
Useful comments include: “What exactly is being mapped?” “Is that similarity structural or superficial?” “Which difference could reverse the conclusion?” “What can this analogy support, and what can it not support?” “Where is the target evidence?”
Parent support
Everyday analogies are easy practice. If someone says “School is like work” or “The internet is like a library,” ask: in what way, and where does the comparison break? The goal is not to destroy analogies but to make them precise.
Exam-day protocol
Use an analogy only if it makes the argument easier to inspect. Name the shared structure, state the decisive difference and return quickly to the target. Never let an analogy carry factual claims that the target evidence does not support.
The thirty-second recovery
If an analogy is running away, write one line: “The comparison is useful only because…” Then delete or ignore every source detail unrelated to that mechanism.
The two-minute recovery
If the essay relies heavily on a weak analogy, replace the borrowed conclusion with a conditional claim and add target-specific evidence. Keep the analogy as explanation rather than proof.
The advanced writer’s challenge
Advanced writers often generate elegant analogies quickly. Their challenge is restraint: the more memorable the source case, the more carefully they must test whether rhetorical force is exceeding structural fit.
The developing writer’s challenge
Use the frame: “A resembles B because ___. The analogy weakens where ___. Therefore it helps explain ___ but cannot establish ___.” Fade the scaffold once the boundary becomes automatic.
The independence test
Give an unseen topic and a tempting analogy. The learner must identify source and target, map one mechanism, identify a decisive difference, state the safe inference and decide whether the analogy should remain. Strong performance is visible when the learner can reject a clever but misleading analogy.
What mastery looks like
A mastered writer can use similarity without surrendering difference. They know that an analogy can explain, generate hypotheses, reveal boundaries and test design while remaining weaker than direct evidence about the target. They can say exactly what transfers, exactly what does not, and exactly why.
The final rule
Transfer the relationship, not the entire case. Map the mechanism, preserve the decisive difference, state the safe inference and return to the target.
That is analogy control under examination pressure: enough similarity to illuminate, enough difference to prevent false equivalence.
Part VI — Deep practice: 60 analogy cases and what actually transfers
The fastest way to become reliable with analogy is to stop asking only, “Are these similar?” and start asking, “Which relation is shared, which difference is decisive, and what is the maximum safe inference?” The following cases deliberately span education, technology, institutions, public policy, language and everyday reasoning so that the transferable procedure becomes visible beneath changing content.
Case 21 — Teacher as coach
Shared structure: both observe performance, diagnose weaknesses and design practice. Decisive difference: academic learning often depends on conceptual explanation and knowledge building, not only performance rehearsal. Safe inference: targeted feedback and practice design transfer; the sporting model does not define the whole teaching relationship.
Case 22 — Teacher as doctor
Shared structure: diagnosis should precede intervention. Decisive difference: learners are active agents and differences in learning are not diseases. Safe inference: diagnostic sequencing, not medical authority or pathology.
Case 23 — Teacher as architect
Shared structure: foundations affect later stability. Decisive difference: learners can reconstruct and compensate dynamically; buildings cannot self-repair in the same way. Safe inference: prerequisite structure matters, but educational pathways can be revised.
Case 24 — Revision as strength training
Shared structure: repeated challenge with recovery produces adaptation. Decisive difference: cognitive gains depend on representation, retrieval and feedback, not simply load. Safe inference: progressive difficulty and recovery matter; “more resistance” is not automatically better learning.
Case 25 — Memory as library
Shared structure: information can be organised and retrieved. Decisive difference: human memory reconstructs and forgets according to use, interference and meaning. Safe inference: organisation aids retrieval, but memory is not passive shelving.
Case 26 — Memory as muscle
Shared structure: use strengthens performance. Decisive difference: knowledge can be transformed, connected and represented symbolically. Safe inference: retrieval practice matters, not literal muscular growth.
Case 27 — Vocabulary as toolbox
Shared structure: different items perform different jobs. Decisive difference: words interact semantically and socially rather than working independently like physical tools. Safe inference: selection by function.
Case 28 — Grammar as traffic rules
Shared structure: shared conventions reduce ambiguity and coordinate users. Decisive difference: grammar evolves and permits stylistic variation where traffic rules are more explicitly enforced. Safe inference: convention enables coordination, not that every deviation is equally dangerous.
Case 29 — Paragraph as room
Shared structure: one space can have one primary function while contributing to a larger building. Decisive difference: paragraphs can blend functions and transition dynamically. Safe inference: job clarity and internal organisation.
Case 30 — Essay as journey
Shared structure: a starting point, route and destination need coherence. Decisive difference: an essay can discover and refine its endpoint through reasoning. Safe inference: direction and transitions, not predetermined inevitability.
Case 31 — Essay as courtroom
Shared structure: claims require evidence and opposing arguments matter. Decisive difference: essays do not always have binary parties, formal admissibility rules or burdens identical to law. Safe inference: evidence and rebuttal, not courtroom procedure wholesale.
Case 32 — Essay as scientific model
Shared structure: a good argument simplifies reality to explain what matters. Decisive difference: normative essays include values that scientific models may not. Safe inference: selective explanatory structure.
Case 33 — Examination as performance
Shared structure: preparation must survive a constrained live event. Decisive difference: academic examinations measure underlying knowledge and reasoning, not only showmanship. Safe inference: rehearsal under realistic conditions.
Case 34 — Examination as game
Shared structure: rules, constraints and strategic allocation matter. Decisive difference: an examination is not merely about beating others and should reward genuine subject competence. Safe inference: procedural strategy without reducing learning to gaming.
Case 35 — Rubric as ruler
Shared structure: both provide a standard for judging. Decisive difference: qualitative criteria require interpretation and may be multidimensional. Safe inference: common standard, not perfect mechanical measurement.
Case 36 — Rubric as map
Shared structure: it indicates important dimensions and helps orient effort. Decisive difference: a map describes a route, while a rubric specifies evaluation criteria. Safe inference: orientation, not full process prescription.
Case 37 — Feedback as mirror
Shared structure: it reflects aspects of current performance. Decisive difference: useful feedback also interprets and directs attention rather than merely reflecting. Safe inference: visibility of current state.
Case 38 — Feedback as GPS
Shared structure: current position is compared with a destination and route corrections are suggested. Decisive difference: writing and learning can have multiple valid destinations. Safe inference: course correction.
Case 39 — Practice as simulation
Shared structure: realistic conditions reveal whether skill transfers. Decisive difference: practice should also isolate and repair weaknesses rather than only imitate final conditions. Safe inference: transfer testing, not exclusive full-task repetition.
Case 40 — School as factory
Shared structure: processes, resources and throughput can be measured. Decisive difference: students are not interchangeable raw materials and educational goals exceed standard output. Safe inference: process bottlenecks only, not industrial treatment of learners.
Case 41 — School as ecosystem
Shared structure: many actors are interdependent and changes can have indirect effects. Decisive difference: school systems are intentionally governed and can redesign rules. Safe inference: interdependence and feedback.
Case 42 — School as network
Shared structure: knowledge and support move through relationships. Decisive difference: not every relation has equal bandwidth or purpose. Safe inference: connectivity matters, not that more links are always better.
Case 43 — University as marketplace
Shared structure: students choose among programmes and institutions respond to demand. Decisive difference: accreditation, public purpose, information asymmetry and long-term educational value limit ordinary market assumptions. Safe inference: choice and incentive effects only.
Case 44 — Student as customer
Shared structure: students receive services and can evaluate experience. Decisive difference: students co-produce learning through effort and do not merely consume a finished product. Safe inference: service quality matters alongside learner responsibility.
Case 45 — Data as oil
Shared structure: raw material can gain value through extraction and processing. Decisive difference: data is replicable, non-rival in many uses and can be generated continuously. Safe inference: processing creates value; scarcity economics does not transfer neatly.
Case 46 — Data as infrastructure
Shared structure: shared data systems can support many downstream services. Decisive difference: data quality, privacy and governance change its usefulness. Safe inference: foundational enabling role.
Case 47 — Internet as library
Shared structure: both store and provide access to information. Decisive difference: the internet has dynamic ranking, variable quality, social interaction and replication at enormous scale. Safe inference: retrieval function, not curation equivalence.
Case 48 — Search engine as librarian
Shared structure: both help users find relevant information. Decisive difference: algorithmic ranking can be optimised commercially and lacks full human contextual judgement. Safe inference: information routing, not identical stewardship.
Case 49 — AI assistant as calculator
Shared structure: both extend human capability by automating operations. Decisive difference: generative systems can produce open-ended, uncertain output rather than deterministic arithmetic. Safe inference: cognitive offloading, not equivalent reliability.
Case 50 — AI assistant as intern
Shared structure: useful draft work may require supervision and verification. Decisive difference: AI lacks human accountability, lived context and stable understanding in the ordinary sense. Safe inference: review requirement, not personhood equivalence.
Case 51 — Algorithm as editor
Shared structure: both select, rank and suppress information. Decisive difference: algorithms operate at scale and may optimise metrics rather than editorial judgement. Safe inference: gatekeeping power.
Case 52 — Algorithm as traffic controller
Shared structure: both route flows under constraints. Decisive difference: information flow affects beliefs and speech, not merely physical movement. Safe inference: routing and congestion logic.
Case 53 — Social media as public square
Shared structure: people gather, speak and encounter others. Decisive difference: platforms are privately governed, algorithmically ranked and globally scalable. Safe inference: public discourse function, not institutional identity.
Case 54 — Social media as newspaper
Shared structure: both distribute information to audiences. Decisive difference: social media hosts user-generated content, personalised ranking and peer interaction. Safe inference: information distribution, not editorial accountability.
Case 55 — Misinformation as virus
Shared structure: spread can depend on networks and contact. Decisive difference: people interpret, reject and reshape information consciously; ideas are not biological pathogens. Safe inference: transmission dynamics, not disease treatment logic wholesale.
Case 56 — Content moderation as border control
Shared structure: both regulate entry under rules. Decisive difference: speech, identity and private platform governance raise different rights and accountability issues. Safe inference: screening and rule enforcement.
Case 57 — Privacy as property
Shared structure: individuals may claim control over something associated with them. Decisive difference: personal information can be copied and relational; others may also have legitimate interests. Safe inference: control and consent, not full property law.
Case 58 — Privacy as boundary
Shared structure: boundaries regulate access between self and others. Decisive difference: information can cross boundaries invisibly and persist. Safe inference: controlled access.
Case 59 — Cybersecurity as castle defence
Shared structure: layered barriers reduce attack opportunities. Decisive difference: digital systems are connected, updateable and attacked remotely at machine speed. Safe inference: defence in depth, not static walls.
Case 60 — Cybersecurity as immune system
Shared structure: detection, adaptation and response to threats. Decisive difference: software systems are deliberately engineered and threats can imitate legitimate activity. Safe inference: adaptive defence.
Case 61 — Market as ecosystem
Shared structure: interdependence, competition and adaptation. Decisive difference: markets are governed by laws and purposeful actors. Safe inference: feedback and adaptation, not natural inevitability.
Case 62 — Market as voting system
Shared structure: choices aggregate preferences. Decisive difference: purchasing power is unequal and many public values are not priced. Safe inference: preference signalling, not democratic legitimacy.
Case 63 — Competition as tournament
Shared structure: participants respond to rules and incentives. Decisive difference: markets allow entry, exit and repeated interaction rather than fixed brackets. Safe inference: incentive effects.
Case 64 — Monopoly as bottleneck
Shared structure: one narrow point can control broader flow. Decisive difference: monopoly power depends on alternatives, law and demand, not physical geometry. Safe inference: concentration creates leverage.
Case 65 — Inflation as tax
Shared structure: purchasing power can be reduced. Decisive difference: inflation is not a legislated transfer with identical distributional effects. Safe inference: erosion of real value, not literal taxation.
Case 66 — Government budget as household budget
Shared structure: resources are finite and choices have opportunity costs. Decisive difference: governments tax, borrow, issue currency under some systems and affect macroeconomic conditions. Safe inference: constraint and prioritisation, not identical solvency logic.
Case 67 — Government as company
Shared structure: both coordinate people and resources. Decisive difference: governments serve citizens, hold coercive authority and pursue public goods beyond profit. Safe inference: some management lessons, not corporate purpose.
Case 68 — Democracy as marketplace of ideas
Shared structure: competing claims can gain support. Decisive difference: truth and civic rights are not allocated simply by demand; manipulation and unequal reach matter. Safe inference: competition among claims, not automatic truth discovery.
Case 69 — Constitution as operating system
Shared structure: foundational rules constrain and enable later operations. Decisive difference: constitutional interpretation involves human judgement, history and legitimacy. Safe inference: rule hierarchy.
Case 70 — Law as code
Shared structure: formal rules guide permissible actions. Decisive difference: law contains ambiguity, discretion and contested interpretation; software executes more mechanically. Safe inference: rule structure, not deterministic execution.
Case 71 — Institution as memory
Shared structure: procedures and records preserve knowledge beyond individuals. Decisive difference: institutions also wield authority and adapt strategically. Safe inference: continuity across personnel changes.
Case 72 — Museum as archive
Shared structure: both preserve information and artefacts. Decisive difference: museums curate for public interpretation and embodied experience. Safe inference: preservation and retrieval.
Case 73 — City as organism
Shared structure: interdependent systems move resources and respond to stress. Decisive difference: cities contain deliberate governance, markets and individual agency. Safe inference: systems interdependence.
Case 74 — Transport network as circulatory system
Shared structure: movement through connected routes supports system function. Decisive difference: travellers choose routes and demand responds to price and land use. Safe inference: network flow and bottlenecks.
Case 75 — Road widening as adding lanes to a data network
Shared structure: capacity expansion can relieve congestion. Decisive difference: transport demand can be induced by capacity and land-use change. Safe inference: capacity matters, not that more capacity permanently solves congestion.
Case 76 — Housing market as queue
Shared structure: scarcity allocates access through waiting or price. Decisive difference: housing is a durable asset, investment and necessity. Safe inference: scarcity and allocation, not queue mechanics alone.
Case 77 — Climate policy as insurance
Shared structure: resources are spent now to reduce uncertain future loss. Decisive difference: climate action can also change current technology and distribution. Safe inference: risk management under uncertainty.
Case 78 — Regulation as guardrail
Shared structure: boundaries reduce catastrophic deviation while leaving room for movement. Decisive difference: regulation changes incentives and can have administrative cost. Safe inference: bounded freedom.
Case 79 — Regulation as friction
Shared structure: rules can slow transactions. Decisive difference: some friction prevents fraud, builds trust and improves market function. Safe inference: process cost, not that all friction is harmful.
Case 80 — Tax as price signal
Shared structure: both change incentives by changing cost. Decisive difference: taxes also raise revenue and operate through legal authority. Safe inference: behavioural incentives.
Part VII — Analogy jobs and their evidence limits
Job 1 — Clarify an unfamiliar system
The analogy can make a mechanism easier to visualise. It should be followed by target-specific explanation so the reader does not mistake the analogy for the thing itself.
Job 2 — Generate a hypothesis
If source and target share a mechanism, the source can suggest what to test. The hypothesis still needs evidence from the target.
Job 3 — Reveal a hidden variable
A good analogy can suggest that scale, incentives, trust or feedback might matter in both systems.
Job 4 — Test robustness
Move the mechanism into another domain. If the same reasoning survives, the principle may be more general.
Job 5 — Identify a boundary
A failed analogy can be valuable because it shows exactly which condition differs.
Job 6 — Transfer a design principle
A solution from one domain can inspire another when the constraint is structurally similar.
Job 7 — Compress explanation
One well-chosen analogy can replace several abstract sentences. Compression is valuable only if it does not hide the break point.
Job 8 — Frame a counterargument
An opponent may rely on an analogy. The writer can test the mapping instead of simply rejecting the conclusion.
Part VIII — The eight break-point families
Scale break
The mechanism changes when system size increases.
Risk break
Same mechanism, different cost of error.
Rights break
Efficiency logic fails because non-tradeable rights become relevant.
Power break
Actors who appear symmetric actually possess different authority.
Information break
One domain has much greater information asymmetry.
Incentive break
Participants respond to different rewards or penalties.
Time break
Short-run logic does not persist after adaptation.
Materiality break
Physical constraints in one system do not exist in the other.
Part IX — Analogy under exam pressure
The fifteen-second source check
Why did you choose this source case? Familiarity alone is not enough.
The fifteen-second mechanism check
State one relation both cases share.
The fifteen-second difference check
Name the difference most likely to reverse the conclusion.
The fifteen-second transfer check
State the narrow conclusion that remains.
Part X — Recovery scenarios
Recovery 1 — The analogy is only rhetorical
Either remove it or add the mechanism it clarifies.
Recovery 2 — The analogy is replacing evidence
Return to target-specific data, examples or reasoning. Keep the analogy as explanation only.
Recovery 3 — The source is emotionally loaded
Neutralise the source and reassess whether the structure still holds.
Recovery 4 — Too many differences appear
Ask which differences affect the mapped relation. Ignore irrelevant differences.
Recovery 5 — One decisive difference appears
Narrow or abandon the inference even if many other similarities remain.
Recovery 6 — The analogy chain is too long
Return directly to source and target. Remove intermediate metaphors.
Recovery 7 — The analogy is historically tempting
Switch into precedent control: check institutional continuity and changed conditions explicitly.
Recovery 8 — The conclusion has become morally equivalent
Separate shared principle from total moral weight.
Part XI — Diagnostic matrix
Pattern A — Student uses vivid analogies constantly
The learner may rely on rhetorical fluency instead of target evidence. Train source-removal drills.
Pattern B — Student rejects all analogies as imperfect
Train mechanism-level mapping and relevance of differences.
Pattern C — Student counts similarities
Train decisive-difference reasoning.
Pattern D — Student imports emotional weight
Train neutral-source substitution.
Pattern E — Student imports historical outcome
Train distinction between analogy and precedent.
Pattern F — Student relies on umbrella categories
Train mechanism verbs and structural relationships.
Pattern G — Student knows source better than target
Train target-evidence-first paragraph design.
Pattern H — Student cannot state safe inference
Train one-line analogy frames.
Pattern I — Student ignores scale
Use scale-swap drills.
Pattern J — Student uses multiple conflicting analogies
Choose one model per inferential job and make abstraction level explicit.
Part XII — Twenty advanced drills
Drill 1 — Surface or structure
Classify ten similarities as superficial or mechanism-relevant.
Drill 2 — One mechanism
Reduce a long analogy to one shared relation.
Drill 3 — Decisive difference
Find the single difference most likely to reverse the inference.
Drill 4 — Safe inference
State the maximum conclusion after the break point is acknowledged.
Drill 5 — Source removal
Delete the analogy and see whether target evidence still supports the paragraph.
Drill 6 — Source switch
Replace the source with a different domain sharing the same mechanism.
Drill 7 — Neutralise emotion
Replace charged labels with Case A and Case B.
Drill 8 — Scale swap
Move from individual to institution to national system.
Drill 9 — Risk swap
Move from low-stakes to high-stakes use.
Drill 10 — Rights constraint
Add a non-tradeable rights consideration and reassess transfer.
Drill 11 — Incentive swap
Keep structure but change rewards. Decide whether behaviour changes.
Drill 12 — Time swap
Move from immediate response to long-term adaptation.
Drill 13 — Information asymmetry
Add unequal information to a previously symmetric analogy.
Drill 14 — Power asymmetry
Add coercive authority and check whether exchange language still works.
Drill 15 — Mixed-model audit
Identify incompatible metaphors in one paragraph.
Drill 16 — Predictive restraint
Rewrite “will” as the strongest supported conditional expectation.
Drill 17 — Historical analogy
Identify one structural continuity and two changed conditions.
Drill 18 — Policy design transfer
Borrow a principle from another domain without copying the whole policy.
Drill 19 — Timed analogy
Build, break and bound an analogy in sixty seconds.
Drill 20 — Unseen transfer
Use a source domain the learner has never practised.
Part XIII — Teacher conference protocol
Ask: “What are your source and target?” “What relation is genuinely shared?” “Which difference could reverse your inference?” “What is the analogy allowed to prove?” “Where is the target evidence?” “Would you still accept this analogy if the emotional source case were replaced?”
Part XIV — Parent conversation protocol
Use common comparisons: “School is like work,” “A phone is like a tool,” “The internet is like a library.” Ask what transfers and what does not. Keep the exercise playful and brief.
Part XV — Analogy scorecard
Track decorative analogies, target-evidence gaps, missing break points, false equivalences, strong structural mappings and successful safe inferences. Improvement should reduce analogy quantity while increasing analogy quality.
Part XVI — Analogy and intellectual flexibility
Analogies let writers borrow understanding from one domain without pretending the domains are identical. This is one of the fastest ways to think through unfamiliar problems—provided the writer remains willing to abandon the analogy when the mapping fails.
Part XVII — Analogy and examination efficiency
A strong analogy can compress an explanation, but only if it is short. The exam writer should spend words on the structural mapping and break point, not on retelling the source case.
Part XVIII — Analogy and evidence hierarchy
Target-specific evidence normally outranks analogy. Analogy can organise, interpret and test that evidence. When the analogy conflicts with strong target evidence, the analogy should lose.
Part XIX — Analogy and uncertainty
The weaker the structural fit, the lower the confidence of transfer. Use calibration language such as “this analogy suggests,” “the comparison is useful only at the level of,” or “the analogy breaks where…” rather than pretending equivalence.
Part XX — Final transfer test
Give three unseen target problems and three tempting source cases. The learner must select the strongest analogy for each, state the mapping, identify the decisive break point, define the safe inference and reject at least one attractive but structurally weak analogy. Strong transfer is visible when the learner can explain why a memorable comparison should not be used.
Final synthesis
Analogy is disciplined borrowing. The source case lends a structure, mechanism or design principle to the target. It does not lend every property, every outcome or every moral association. The writer’s responsibility is to control the transfer.
The mature writer therefore maps only the relation that matters, names the difference that can break it, keeps source and target evidence distinct, and states the maximum safe inference. The result is neither sterile literalism nor loose metaphor. It is useful similarity under control.
The best analogy does not erase difference. It makes the important difference easier to see.
Part XXI — Analogy reserve: adversarial mapping, false-equivalence control and unseen transfer
This final layer turns analogy into a repeatable examination procedure. A writer should be able to test a tempting comparison quickly, preserve only the useful structural relation, reject imported emotional or moral weight, and state exactly how much inferential force the analogy deserves.
The adversarial analogy sequence
Test 1 — Remove the labels
Replace “school,” “company,” “government” or “AI” with Source A and Target B. If the analogy loses force, category labels may have been doing the reasoning.
Test 2 — Remove the emotional association
Replace a charged historical or moral source with a neutral case that shares the same structure. If the conclusion weakens sharply, emotion was being imported.
Test 3 — Reverse the direction
If B is like A for one relation, does reasoning from A back to B also work? Failure can reveal asymmetry in the mapping.
Test 4 — Change the scale
Move from person to institution, institution to nation, or pilot to system. If the mechanism changes, scale is a break point.
Test 5 — Change the stakes
Move from low-cost experimentation to high-risk irreversible decisions. The same analogy may no longer support the same recommendation.
Test 6 — Add power asymmetry
What looked like exchange may become coercion when one actor controls rules or exit. Market and contract analogies often fail here.
Test 7 — Add rights
If the source case treats all costs as tradeable but the target contains rights or duties, the analogy may need a hard boundary.
Test 8 — Add time
Short-run resemblance may disappear after adaptation, learning or institutional change.
Test 9 — Ask what target evidence says
If direct evidence contradicts the analogy, the analogy should lose. It is a reasoning aid, not an override.
Test 10 — State the break point aloud
If you cannot articulate where the comparison stops, you probably do not yet control it.
Thirty false-equivalence traps
Trap 1 — Same outcome, same cause
Similar outcomes can arise from different mechanisms. Do not infer shared cause from shared result alone.
Trap 2 — Same label, same system
Two democracies, markets, schools or AI systems can differ structurally.
Trap 3 — Same problem, same solution
Different constraints can make one solution transfer poorly even where the problem sounds similar.
Trap 4 — Same mechanism, same ethics
Mechanistic similarity does not imply equal moral weight.
Trap 5 — Same technology, same risk
Risk depends on task, stakes, users and safeguards.
Trap 6 — Same rule, same effect
Institutions, incentives and compliance can change outcomes.
Trap 7 — Same historical pattern, same destiny
Past resemblance identifies a question, not inevitability.
Trap 8 — Same economic incentive, same behaviour
Norms, information and constraints can alter response.
Trap 9 — Same network, same diffusion
Ideas, diseases, money and information move through networks differently.
Trap 10 — Same scarcity, same allocation
Price, queue, lottery, rights and need-based systems embody different rules.
Trap 11 — Same average, same distribution
Two systems can share average outcomes while affecting subgroups very differently.
Trap 12 — Same error rate, same acceptability
High-stakes errors matter differently from low-stakes errors.
Trap 13 — Same growth, same wellbeing
Growth in one metric does not map automatically to broad welfare.
Trap 14 — Same participation, same democracy
Participation volume does not guarantee representation or accountability.
Trap 15 — Same choice, same freedom
Choice quality depends on information, capacity and meaningful alternatives.
Trap 16 — Same price, same value
Price is not a universal proxy for quality, utility or social worth.
Trap 17 — Same speed, same efficiency
Faster completion can create error or external cost.
Trap 18 — Same transparency, same trust
Trust depends on usability, accountability and what is disclosed.
Trap 19 — Same data volume, same knowledge
Information quantity does not guarantee relevance or interpretation.
Trap 20 — Same rule complexity, same burden
Compliance cost varies by capacity and user sophistication.
Trap 21 — Same innovation, same disruption
Incremental and radical innovation operate differently.
Trap 22 — Same diversity, same inclusion
Presence and participation are not identical.
Trap 23 — Same access, same use
Availability does not guarantee affordability, capability or uptake.
Trap 24 — Same information, same decision
Values and incentives can produce different choices from the same facts.
Trap 25 — Same confidence, same competence
Self-assurance and skill can diverge.
Trap 26 — Same score, same learning
Assessment performance may not capture durable transfer.
Trap 27 — Same popularity, same legitimacy
Popular support and fair procedure are different dimensions.
Trap 28 — Same law, same implementation
Enforcement and administrative capacity change practical effect.
Trap 29 — Same policy name, same design
“Carbon tax,” “phone ban” or “AI regulation” can describe very different implementations.
Trap 30 — Same analogy, same conclusion
The same source-target pair can support one claim and fail another. Analogy strength is always claim-specific.
Examiner-facing analogy clarity
A reader should be able to see four things: the source case, the target case, the relation being transferred, and the limit of transfer. If the analogy appears as a decorative phrase with no mapping, it adds style rather than reasoning.
Visible mapping
Use language such as “The comparison is useful at the level of…,” “Both systems share…,” or “The analogy works because…”
Visible break point
Use language such as “The comparison breaks where…,” “Unlike the source case…,” or “This does not mean that…”
Visible safe inference
End with the narrow conclusion: “The analogy therefore suggests X, but it does not establish Y.”
Visible return to evidence
Follow the analogy with target-specific evidence or reasoning.
The five-level analogy confidence scale
Level 1 — Rhetorical resemblance
Useful for style or intuition only.
Level 2 — Illustrative similarity
Clarifies one relation but should not support a major conclusion alone.
Level 3 — Structural similarity
Several relevant relationships align; useful for explanation and hypothesis.
Level 4 — Mechanistic similarity
The same underlying process appears to operate; stronger transfer is possible if context also aligns.
Level 5 — Evidence-backed transfer
The analogy is supported by target evidence, comparable cases and aligned conditions. Even here, the cases are not identical.
When to reject an analogy entirely
Reject it when the decisive difference directly contradicts the transferred mechanism, when emotional association is doing more work than structure, when target evidence points the other way, or when the analogy would require ignoring scale, rights or power conditions central to the question.
When to keep a weak analogy
A weak analogy can remain as an opening intuition if it is explicitly labelled as limited and quickly replaced by direct analysis. It should not carry the thesis.
When to promote an analogy into a principle
If several different source cases reveal the same mechanism and target evidence supports it, the writer can state the underlying principle directly. The analogy then becomes an illustration rather than the argument’s foundation.
Analogy and precedent
When the source is a historical or institutional case and the target asks what should happen now, switch from simple analogy to precedent reasoning. Ask whether rule, institution, incentive, technology and stakeholder conditions remain comparable.
Analogy and counterfactual reasoning
A source case can suggest what might happen under an alternative condition, but the counterfactual still needs a realistic baseline and evidence anchor. Analogy cannot create an unobserved outcome.
Analogy and generalisation
One successful source case does not establish that a structural principle generalises widely. Broader transfer requires multiple contexts or mechanism evidence.
Analogy and uncertainty calibration
Use stronger confidence only as structural fit, context alignment and target evidence increase. “Suggests,” “offers a useful model for,” and “supports a limited expectation that” are often appropriate.
Analogy and proportionality
Analogies can clarify whether a response is excessive—for example, regulation as a guardrail—but proportionality must still be tested directly against problem severity, necessity and less restrictive alternatives.
The final-minute analogy audit
- Circle every “like,” “similar to,” “as if,” “just as” and metaphor carrying argumentative weight.
- Identify source and target.
- Write the mapped mechanism in three words.
- Name the decisive difference.
- Check that target evidence still supports the conclusion.
- Remove any analogy that imports emotion without structure.
Thirty analogy myths to retire
- More similarities make an analogy stronger.
- Any difference destroys an analogy.
- Cases in the same category are naturally analogous.
- Cases in different categories cannot be analogous.
- Analogies are only decorative language.
- A good analogy is proof.
- A historical analogy predicts the same future.
- A moral analogy establishes equal wrongdoing.
- A technological analogy transfers risk automatically.
- A biological analogy reveals inevitability.
- A market analogy makes every choice efficient.
- A machine analogy makes people predictable.
- A family analogy makes institutions benevolent.
- A company analogy makes governments profit-seeking organisations.
- A game analogy makes education merely strategic.
- A library analogy makes the internet curated knowledge.
- A tool analogy makes technology neutral.
- A coach analogy makes all teaching performance training.
- A doctor analogy makes learning problems illnesses.
- A network analogy means more connections are always better.
- A source case can substitute for target evidence.
- Emotional resonance strengthens structural reasoning.
- Analogies should be explained in great detail.
- One analogy should govern an entire essay.
- Different analogies can be mixed without conflict.
- If an analogy sounds elegant, it is probably useful.
- Breaking the analogy weakens the essay.
- Qualified analogies sound indecisive.
- The conclusion can ignore the analogy’s break point.
- Good writers never reject their first analogy.
Progression from developing to advanced analogy control
Floor 1 — Identify source and target
The learner can state which case is being used to illuminate which.
Floor 2 — Identify the shared relation
Similarity moves from labels to structure.
Floor 3 — Find the decisive difference
The learner can state what could reverse the inference.
Floor 4 — State the safe inference
The conclusion becomes narrower and explicit.
Floor 5 — Preserve target evidence
The analogy no longer replaces empirical or textual support.
Floor 6 — Control emotion
Charged source cases no longer import moral weight automatically.
Floor 7 — Control scale, risk and rights
Structural break points become systematic.
Floor 8 — Distinguish analogy from precedent
Historical transfer receives additional institutional checks.
Floor 9 — Calibrate confidence
Language reflects mapping strength and evidence quality.
Floor 10 — Timed rejection
The learner can abandon a seductive but weak analogy quickly under examination pressure.
Ten unseen transfer prompts
Prompt 1 — AI
“An AI assistant is like an intern.” Identify one useful mapping, two break points and the safe inference.
Prompt 2 — Government
“A country should be run like a company.” Identify what management insight transfers and what civic obligations break the analogy.
Prompt 3 — Education
“Learning is like muscle training.” Preserve practice and recovery while rejecting literal load assumptions.
Prompt 4 — Media
“Social media is the modern public square.” Map expression and gathering, then test ownership and algorithmic ranking.
Prompt 5 — Data
“Data is the new oil.” Test processing value against replicability and non-rival use.
Prompt 6 — Regulation
“Regulation is a guardrail.” Use the analogy to explain bounded freedom without assuming zero administrative cost.
Prompt 7 — Exams
“An examination is a game.” Keep strategy and rules while preserving learning and competence as the actual target.
Prompt 8 — Cities
“A city is an organism.” Map interdependence while preserving governance and intentional design.
Prompt 9 — Cybersecurity
“Cybersecurity is like an immune system.” Test adaptive defence while preserving the differences between engineered and biological systems.
Prompt 10 — Democracy
“Democracy is a marketplace of ideas.” Examine competition, unequal voice and whether popularity selects truth.
The five-minute analogy warm-up
Take two everyday analogies. For each, write source, target, mechanism, decisive difference and safe inference. Then reject one analogy entirely. Rejection is part of mastery.
Post-essay analogy audit
Underline every analogy. Beside it write S for surface, T for structural or M for mechanistic. Mark one break point and one evidence source from the target case. Delete any analogy with no target evidence or no definable inferential job.
The final analogy contract
The writer promises not to equate cases merely because they share a label, not to import moral emotion without structural justification, not to let a historical resemblance become destiny, not to use analogy as substitute evidence, and not to hide the break point for rhetorical effect.
An analogy earns its place when it makes the target clearer without pretending the target has become the source.
Continue learning
Explore reasoning in timed essays · Related guide: Use precedent carefully · The Library