Texts and Close Reading
Why close reading matters: A Horse and Two Goats
R.K. Narayan's story (first published 1960, set in the South Indian village of Kirittam) shows how misreading happens without any lying. Muni, an elderly goatherd reduced to two goats, rests near an ancient clay horse statue. An American arrives in a yellow station wagon. Neither speaks the other's language (English vs Tamil).
- The American assumes Muni owns the statue (he is standing next to it) and starts bargaining.
- Muni, seeing khaki clothes, assumes a government official investigating a recent murder: "I know nothing about it."
- Shown a 100-rupee note (which Muni has never seen), Muni suggests the village lender for change. The American reads this as haggling and adds 20 rupees. Muni knows what 20 rupees looks like, so he thinks it is for his two goats.
- Both leave delighted: the American with the statue, Muni with 120 rupees. The goats follow Muni home, and he wonders why. Only the statue was sold.
Nobody lied. The wrong conclusions came from an unchecked assumption each man built the whole transaction on. Muni could only imagine a deal about goats (all he owned); the American saw a smiling man beside the thing he wanted and read agreement. Each read what he wanted and expected to hear and filled in the rest. That is exactly an implicit position read wrongly yet "obviously".
Close reading lets you catch others reading their assumptions onto your words, and catch yourself doing it to others. The story is humorous satire carrying serious layers: miscommunication, cultural difference, poverty and wealth, class, and the different values people attach to the same object.
Two layers: explicit claims and implicit positions
Every text (novel, business report, speech, email, article) makes claims on two levels at once.
| Explicit claim | Implicit position | |
|---|---|---|
| What it is | What the text directly states; you can quote the exact sentence | What the text assumes, values, or wants you to conclude without saying so |
| How you get it | Point to the sentence | Inference from the pattern of choices: what is chosen, omitted, emphasised |
| Fact-checking | Easy | Hard |
| Example | "Revenue grew 12% this quarter." | A report citing only metrics favourable to leadership's prior decision implies the decision was right |
Most persuasion happens at the implicit layer: it is hidden, harder to challenge, and so demands more careful reading.
Worked examples
- Explicit: "We grew active users 8% quarter over quarter." Implicit: opening with user growth, with falling revenue per user surfacing only later in an unprompted analyst question, implies growth matters more than monetisation, without arguing it.
- Explicit: "The new policy reduced permitting delays by 30%." Implicit: discussing only speed and never the policy's cost or who bore it implies clear success, argued by omission.
Key takeaways
- Explicit claims are quotable and checkable; implicit positions are inferred from patterns of choice.
- Misreading arises from unchecked assumptions, not only from lies.
- Persuasion lives mostly in the implicit layer, so close reading targets it.
- Close reading also audits your own reading of others.
Close Reading Techniques
Five techniques for finding a text's implicit position:
| # | Technique | Question to ask | Example |
|---|---|---|---|
| 1 | Repetition and emphasis | What word, phrase or idea keeps recurring, even if not argued for? | A memo mentioning speed 6 times and quality once implicitly ranks speed over quality |
| 2 | What is absent | Whose perspective would complicate the claim, and is it present? (links to selection bias: missing risk, counterargument, stakeholder) | A hiring announcement covering only the new role, not roles eliminated, implies the loss doesn't matter |
| 3 | Word choice | What attitude do the words carry beyond fact? | "Restructured" vs "gutted" for the same event; the choice itself takes a stance on whether it was justified |
| 4 | Structure and order | What comes first, and what is held to the final paragraph or an appendix? | A risk buried in paragraph 9 is told to matter less than an opportunity leading paragraph 1 |
| 5 | Assumed audience | Who does the text seem to assume is reading, and what do they already believe? | A text written as if its conclusion is obvious quietly says any reasonable reader already agrees |
Word choice draws on Patrick Hurley's distinction: cognitive meaning states a fact; emotive meaning carries an attitude toward it. Order is a framing choice, and framing carries implicit positions about what matters most.
Key takeaways
- Repetition ranks ideas without arguing for them.
- Absence connects to selection bias: ask whose view is missing.
- Same facts, different words, different implied judgment.
- Order sets a hierarchy of importance.
- "Obvious conclusion" tone is itself a position about the reader.
Annotating a Real Paragraph
Sample: "Our platform helped thousands of small businesses grow this year. As always, we remain focused on the merchants who make a community what it is."
- Explicit claim: the platform helped thousands of small businesses grow.
- Implicit position 1: "As always" implies a consistency the paragraph never evidences.
- Implicit position 2: no mention of businesses that left the platform. This is an absence, not a denial.
Annotating a Real Paragraph
Two traps
| Trap | What goes wrong | Safeguard |
|---|---|---|
| Mistaking a position for a fact | A position repeated often enough starts to feel like fact | Ask: could a reasonable person read the same text and hold the opposite position? If yes, it is a position, not a reported fact |
| Overreading (the opposite error) | Finding an implicit position supported only by your own assumptions | Every implicit reading needs a textual hook |
Discipline: state the implicit position, then name the specific word, omission or structural choice that supports it. Without that, you have only a hypothesis, not a close reading.
Key takeaways
- "As always" and unmentioned departures are typical hooks.
- Test positions with the opposite-reader check.
- No hook, no reading.
Two Layers, One Discipline: The Close Reading Checklist
Explicit claims tell you what a text says; implicit positions tell you what it wants you to believe. Close reading catches both and points to where each lives. It applies to novels, films, reports and datasets.
- 1Text
- 2What does it explicitly claim? A quotable sentence
- 3What repeats and gets emphasised?
- 4What is absent that a fuller account would include?
- 5What do specific words carry beyond fact?
- 6What comes first and what is buried?
- 7Who is assumed to already agree?
Two Layers, One Discipline: The Close Reading Checklist
Applied to A Horse and Two Goats
| Question | Finding |
|---|---|
| Explicit claim | The men never understood each other (different languages) yet each believes communication occurred. Muni is extremely poor, the American wealthy (yellow station wagon, crisp khaki clothes, English). Quotable: Muni cannot understand English and the American cannot understand Tamil, making every exchange a misunderstanding. |
| Repeated / emphasised | Muni's poverty (no food, debts, lost livestock); the communication gap; contrast of poor rural India vs affluent modern America; the clay horse as the centre of the misunderstanding. The misunderstanding is not just linguistic but cultural and economic. |
| Absent | What happens after the American leaves; villagers' reaction to the missing statue; the American never learns Muni didn't own it; whether Muni keeps the money or faces consequences; neither gains the other's perspective. Omissions keep focus on the irony, not on resolution. |
| Word choice | Two goats = reduced fortunes and fragile livelihood. The horse = cultural history and mythology for Muni, merely an art object for the American. Car, clothing, confidence = wealth, mobility, modernity; descriptions of Kirittam = simplicity, poverty, tradition. |
| First vs buried | Opens with Muni's daily life, poverty and setting before the American appears, so readers see through Muni first. The real misunderstanding emerges gradually; the final irony (goats following him home) comes after both leave satisfied. |
| Assumed agreement | Readers will recognise that language barriers cause real misunderstanding, that wealth and poverty shape interpretation, that cultural objects mean different things to different people, and will appreciate the irony of two people thinking they made the same agreement. It invites empathy for both men rather than judging either as foolish. |
Two people, two unchecked assumptions: both readings felt obvious, both were wrong. The techniques exist to catch this before it costs something real.
Key takeaways
- Use the six questions on any text, fiction or business.
- Ordinary descriptions encode class, culture and worldview.
- Order and omission steer whose perspective the reader adopts.
Applying Close Reading to Business Documents
Four techniques find a text's real position (structure, word choice, omission, emphasis); identifying assumptions finds what it quietly stands on.
| Document | Technique | What to check |
|---|---|---|
| Annual report | Structure | What leads the shareholder letter vs what appears only in the regulator-required, rarely read risk factors section |
| Word choice | "Challenging year" vs "difficult year with significant losses" can describe the same numbers but create different pictures | |
| Omission | Which segment, region or metric from last year was quietly dropped this year | |
| Press release | Emphasis | The headline is the number the company wants remembered, usually not the whole story |
| Structure | An executive quote near the top often carries the interpretation to attach to the facts that follow | |
| Assumption check | What would the release have to omit or reframe for the headline number to still sound impressive? | |
| Pitch deck | Omission | Which competitor risk or failed channel never appears |
| Word choice | "We are capital efficient" vs "Our margins are below what investors expect at this stage" for the same unit economics | |
| Assumption check | Growth projections almost always assume past growth conditions hold; is that named or only implied? |
Key takeaways
- The same numbers can be framed to lead to different conclusions.
- Each document type has predictable places where framing hides.
- Check whether key assumptions are stated or merely implied.
Close Reading in Practice
Method: read the excerpt once quickly (note the impression: "solid quarters, some challenges, confident management"), then slowly, one sentence at a time, like an analyst, journalist or auditor, and compare.
Close Reading in Practice
Annotating the excerpt (a retailer, "Northgate")
| Phrase | Diagnosis |
|---|---|
| "Broadly in line with expectations" | Whose expectations, set by whom and when? No baseline is given, so it cannot be verified. Confident language carrying no verifiable content is suspicious. |
| "Footfall in physical stores declined 8% year over year" vs "online engagement metrics trending positively" | Precise, checkable number beside a vague phrase with no figure. The pairing is rarely accidental; if the online result were strong, a figure would be given. |
| "Explore opportunities to optimize the store portfolio" | Common euphemism for closing stores. The report never says so and lets readers supply the meaning without committing in writing. |
| "Operating costs increased 6% and these investments position Northgate well for future growth" | A causal claim that is asserted, not demonstrated: no metric, comparison or timeline, yet stated with the same confidence as the hard revenue figure. |
| "The team has worked tirelessly" | Emotional and reputational work, not informational; says nothing about performance. Placed right after rising costs: a warm human sentence after the least comfortable numbers softens the transition. |
| "We anticipate Q4 will present similar dynamics... committed to delivering value for stakeholders" | Vaguest close. "Similar dynamics" commits to nothing; "delivering value" is a claim no company would contradict, so it carries almost no information. A strong close gives a number, target or timeline. |
The rewrite test
Rewrite so every vague claim is backed by a number or flagged as an unproven assumption. Two results: the rewrite is shorter (much of the original's length was persuasive, not informational) and less confident. The gap shows the original is not necessarily dishonest, but is written to feel more informative than it is.
Four questions for your own practice
- What is the comparison point for this number?
- What is an adjective doing?
- Who is the hidden actor in this sentence?
- What question is this sentence really answering?
Goal: not to catch a lie but to build the habit of telling confident language from real evidence.
Key takeaways
- Unanchored comparisons ("in line with expectations") cannot be checked.
- Precise vs vague pairing signals the vague side is weak.
- Euphemisms let readers draw conclusions the writer won't commit to.
- Assertion is not demonstration; look for metric, comparison, timeline.
- Rewriting with evidence exposes persuasion padding.
Narratives, Power, and Counter-Readings
The 5 Cs of a story
The fable of the tortoise and the hare: the hare mocks the tortoise's slowness; the tortoise proposes a race; a fox referees; the hare dashes ahead, is certain of victory and naps; the tortoise keeps steady, unbroken progress; the hare oversleeps, wakes too late, and the tortoise wins.
| C | In the fable |
|---|---|
| Characters | Hare and tortoise |
| Context | Public race with a referee |
| Conflict | Natural speed vs persistence and discipline |
| Choices | Hare naps; tortoise keeps going |
| Consequences | Tortoise wins, hare loses |
This is the underlying structure of almost every story told about businesses, leaders and organisations.
Same events, three readings
| Interpretation | Focus | Lesson |
|---|---|---|
| Conventional | Tortoise's persistence | Slow and steady wins the race |
| Hare-centred | Hare's choices | Exceptional ability can be undone by arrogance, overconfidence and complacency; the problem is assuming talent equals success |
| Marie von Ebner-Eschenbach | Aftermath | After winning, the tortoise thinks itself a fast runner: we credit our own ability when luck or the other's mistake may explain success |
No event changed; only the interpretation. A counter-reading does not invent or ignore events; it asks what happens when the same story is interpreted from a different angle.
Blockbuster
Late 1990s: over 9,000 US stores, dominant in video rental, hundreds of millions of dollars yearly from late fees alone. Bankruptcy in 2010.
- Dominant narrative: small DVD-by-mail Netflix saw where technology was heading; Blockbuster was complacent, attached to late fees, slow to change. Better technology won. Neat, common in case studies, but incomplete.
- Left out: people inside Blockbuster had proposed streaming and subscription years earlier. The ideas were ignored not because nobody saw the future but because senior leadership chose to protect short-term late-fee revenue and the physical store network.
- Better question: who inside Blockbuster saw what was coming, and what happened to their proposal?
A story is never a transcript of events; it is a version shaped by choices of who the characters are, what context matters, where the conflict lies, which decisions deserve attention, and what the ending/moral is.
Guiding questions: How do narratives persuade? How is storytelling tied to power? Whose perspectives are included or left out? How do we build a counter-reading? Which interpretation is stronger on logical grounds?
Narratives as persuasive structures
A narrative is events deliberately arranged to lead to a conclusion, often before we notice. The common pattern is SCR: Situation, Complication (conflict), Resolution.
- 1Situation: problem or opportunity
- 2Complication: conflict, something changes or goes wrong
- 3Resolution: company responds, adapts, succeeds
Persuasive power comes less from individual facts than from the order in which they are presented.
WeWork: its early pitch sold an idea (community, workplace revolution), presenting itself as a technology company rather than real estate: mission first, financial second. This helped justify its valuation for years. IPO prospectus scrutiny forced investors to look at numbers; story and balance sheet did not match, and the valuation collapsed.
When reading business stories ask: what was the setup, what was the disruption, what counts as resolution? Exercise: read a startup's About Us page, find setup/disruption/resolution, and ask what assumptions about its future its telling of the past asks you to accept.
Key takeaways
- Stories have five building blocks: characters, context, conflict, choices, consequences.
- A counter-reading keeps the events and changes the angle.
- The dominant version (Blockbuster) can omit the internal choices that mattered.
- SCR order does persuasive work; check the story against the numbers (WeWork).
Storytelling and Power
Whoever establishes the dominant narrative often decides what looks normal, what counts as innovation, and what is acceptable.
| Example | Narrative effect |
|---|---|
| Uber | Presented as disruption and consumer freedom (cheaper rides, choice, innovation against outdated regulation). Shaped public opinion and, for a long time, policy. Drivers classed as independent contractors; working conditions framed as the price of innovation. A labour issue was reframed as a technology issue. |
| IBM | Decades of a consistent story of professionalism, rigour and long-term thinking earned more benefit of the doubt in controversy than a younger company doing similar things. |
Identical facts can yield different public judgments depending on framing. This is the everyday process by which some ways of understanding seem reasonable and others do not, not simple propaganda.
Storytelling and Power
Narrative omission
Ask: who isn't in this story, and how might they tell it differently? This reveals whose perspective is privileged.
- Nike (1990s): marketing centred on elite athletes, achievement and performance. The campaigns were not false, but left out workers in supplier factories in countries such as Indonesia and Vietnam, whose wages and conditions later became the subject of major investigative reporting. The silence became part of the narrative.
- Facebook: mission described as connecting the world and empowering users; rarely mentioned third-party data buyers and the commercial ecosystem built on user data. Once known, people reinterpreted the same platform.
Exercise: take a company sustainability report, list three missing stakeholder groups, and ask whether their own account of the year would match the company's.
Building a counter-reading
A counter-reading is not mere disagreement; it is a different construction of the same story: change the narrator, the starting point, or the definition of success. Changing the central character often changes the conflict, which choices matter and the consequences.
- 1Identify dominant narrative's main character and implied measure of success
- 2Ask which stakeholder groups are missing
- 3Using evidence, rebuild conflict and key choices from their perspective
- 4Compare both versions using the 5 Cs
- 5Contradict each other, or highlight different aspects of same events?
Evaluating rival interpretations
A counter-reading is still an argument and must be evaluated on evidence and reasoning, not on which feels more satisfying or fits instinct.
| Test | Question |
|---|---|
| Evidentiary | Is it supported by verifiable evidence or mainly speculation? Noticing a missing group is only a start; you still need reliable evidence about their experiences and outcomes. |
| Consistency | Does it explain the evidence at least as well as the dominant narrative, including the facts that made that story persuasive, rather than ignoring misfits? |
| Predictive | Which interpretation better explains what happened next? A strong explanation fits the past and helps anticipate the future. |
Applied to the fable: "slow and steady wins" fits the single race but is weak predictively (silent on a rematch, or on a hare who doesn't stop). "Arrogance undoes talent" holds up better because it names a repeatable behaviour pattern. Ebner-Eschenbach's reading is strong on consistency: it extends past the finish line and explains how success leads people to overestimate their ability and misunderstand why they succeeded. This does not make it the one true moral; it survives closer scrutiny better.
Key takeaways
- Dominant narratives shape legitimacy (Uber reframed labour as technology; IBM gained benefit of the doubt).
- Narrative omission: ask who is missing (Nike factory workers, Facebook data buyers).
- Counter-readings change perspective, not events.
- Judge interpretations by evidentiary, consistency and predictive tests.
Case Study: Boeing 737 MAX
Between October 2018 and March 2019 two new 737 MAX aircraft crashed: Lion Air Flight 610 (Indonesia) and Ethiopian Airlines Flight 302 (Ethiopia), killing nearly 346 people in total.
| Dominant narrative (Boeing) | Counter-reading | |
|---|---|---|
| Central character | Pilots | The aircraft's design and certification process |
| Conflict | Pilot competence and training | Engineering decisions and regulatory oversight |
| Key choices | Pilots' actions in the final minutes | Management's design, testing and certification decisions made years earlier |
| Claim | Aircraft fundamentally safe; pilots responded incorrectly. Locations implied differences in pilot training, culture and aviation standards | New flight control system MCAS (Maneuvering Characteristics Augmentation System) pushes the nose down under certain conditions; faulty sensor data made it activate repeatedly though no stall was imminent; it kept forcing the nose down despite pilots fighting it |
| Supporting findings | Many pilots had little or no MCAS training because Boeing wanted the aircraft marketed as needing minimal additional pilot certification; internal communications showed employees' concerns about design and certification before the second crash | |
| Consequences | Better pilot training | Redesign, stronger regulatory oversight, scrutiny of Boeing's corporate decision-making |
Why the counter-reading wins (the three tests)
- Evidentiary: accident investigation reports, flight data recorders, internal Boeing communications and FAA documentation; not speculation or hindsight.
- Consistency: one mechanism explains both crashes, and why experienced pilots struggled to recover in near-identical circumstances.
- Predictive: if design was the problem, similar accidents would recur until the system changed. That shaped events: the fleet was grounded worldwide, MCAS redesigned, pilot training requirements revised, extra regulatory scrutiny before return to service.
The events never changed; evidence determined which interpretation was best.
Case Study: Boeing 737 MAX
Closing principles
- Recognising that narratives are constructed does not mean every corporate story is dishonest, nor that every counter-reading deserves equal credibility.
- Approach every narrative, including your own, with curiosity and discipline: who tells it, whose perspective is prioritised or missing, what definition of success is assumed, and does the counter-reading meet the same evidence and reasoning standards?
- Ebner-Eschenbach's lesson for companies: a single success can be mistaken for proof of brilliance, forgetting timing, circumstances or others' mistakes.
- Critical reading separates a good story from a good explanation.
Key takeaways
- Boeing shifted the central actors from pilots to management; facts stayed fixed.
- The better reading passed all three tests and predicted real-world consequences.
- Not every counter-reading is right; test it like any argument.
- Success may reflect luck or timing, not only ability.
Revise before your examAll 3 exam tips from this module in one list
- A valid close reading = implicit position + named textual evidence. Expect to be marked down for either unsupported speculation (overreading) or treating a stance as fact.
- Spot-the-weakness answers should name the phrase and the mechanism (no baseline, euphemism, asserted causation, emotional padding, vague close).
- Name the three tests (evidentiary, consistency, predictive) and apply each to both readings. A counter-reading is not right just because it is different.
Test yourself on this module
Questions from these notes, one at a time, with feedback. Your weak topics are saved.