essay-structure

Files (3)

• SKILL.md
• essay-structure.skill
• routing.yaml · evals/

SKILL.md (first lines)

# essay-structure

## Core Philosophy
An argumentative essay is not an informational summary, an opinion diary, or a book report. It is a formal piece of intellectual persuasion. A strong essay does not merely announce that a topic is complex; it stakes out a **debatable, defensible, and non-obvious thesis statement**, marshals rigorous empirical and theoretical evidence, directly engages the strongest possible counterarguments, and synthesizes its findings into broader scholarly or societal implications. Structural integrity—ensuring every paragraph functions as an indispensable link in an unbroken chain of logic—is what separates compelling scholarship from aimless prose.

---

## 4-Step Essay Architecture

### Step 1: The Toulmin Argumentation Architecture
1. **The 6 Pillars of Toulmin's Model**:
   - **Claim**: The contentious, debatable assertion being argued.
   - **Data / Grounds**: Empirical evidence, textual citations, or historical facts supporting the claim.
   - **Warrant**: The underlying logical principle connecting the data to the claim.
   - **Backing**: Additional evidence establishing that the warrant itself is valid.
   - **Counter-Claim / Rebuttal**: The most formidable opposing perspective and why it fails or has bounded scope.
   - **Qualifier**: Words specifying the scope and limits of the claim (e.g. *under conditions of, predominantly, when constrained by*).

### Step 2: The Three-Part Thesis Statement Formulation
1. **The Anatomy of a High-Impact Thesis**:
   - A weak thesis states a fact: *"Artificial intelligence is transforming radiology."* (Obvious, non-debatable).
   - A strong thesis integrates **Concession + Claim + Rationale**:
     - *Formula*: *"Although [Concession / Opposing view], [Debatable Claim], because [Key Evidentiary Mechanism 1] and [Key Evidentiary Mechanism 2]."*
   - *Example*: *"Although diagnostic vision models achieve superior pattern matching in lung nodule detection, autonomous AI deployment in acute clinical triage remains premature because diagnostic neural networks fail on out-of-distribution comorbidities and obscure clinical accountability during lethal diagnostic misclassifications."*

### Step 3: Paragraph Architecture: The MEAL & PIE Frameworks
1. **The MEAL Paragraph Standard**:
   - **M (Main Idea / Topic Sentence)**: Direct assertion supporting the overarching thesis (no raw quotes here).
   - **E (Evidence / Concrete Illustration)**: Data, study findings, direct quotation with citation.
   - **A (Analysis / Warrant)**: Dissect the evidence. Explain *why* and *how* this data proves the paragraph's topic sentence.
   - **L (Link / Transition)**: Connect the insight back to the primary thesis and bridge into the subsequent paragraph.
2. **Reverse Outlining**:
   - Read only the first sentence of each paragraph in sequence. The thesis argument must flow logically as a coherent micro-essay.

### Step 4: The Strategic Counterargument & Rebuttal
1. **Steel-Manning over Straw-Manning**:
   - Articulate the opposing argument in its strongest, most sophisticated form before dismantling it.
2. **The Rebuttal Pivot**:
   - Acknowledge the counter-claim's validity in edge cases, then demonstrate why the primary claim holds superior explanatory power or lower societal cost.

---

## Deliverable Format: Comprehensive Essay Argument Outline Template

```markdown
# Working Title: The Delusion of Algorithmic Objectivity in Judicial Bail Decisions

## I. Introduction
- **Hook**: ProPublica's 2016 investigation into the COMPAS recidivism algorithm.
- **Scholarly Context**: The shift from subjective human judicial discretion to actuarial risk scoring.
- **Core Thesis**: Although algorithmic risk assessment tools promise to eliminate individual judicial racial prejudice, automated bail sentencing systems exacerbate systemic racial disparities because predictive models rely on historically biased arrest arrest metrics as proxies for recidivism and obscure constitutional due-process rights behind proprietary black-box intellectual property.

## II. Body Section 1: Data Poisoning and Proxy Variables (MEAL 1)
- **Main Idea**: Risk assessment algorithms do not measure criminal propensity; they measure likelihood of police contact.
- **Evidence**: Alexander (2012) arrest density data in historically redlined neighborhoods.
- **Analysis**: Because the training labels reflect discriminatory enforcement rather than true criminality, algorithmic scoring codifies historical injustice as objective mathematical fact.
- **Link**: This proxy distortion invalidates the core premise of algorithmic neutrality.

## III. Body Section 2: Trade-Offs in Algorithmic Fairness Metrics (MEAL 2)
- **Main Idea**: Mathematical impossibility of satisfying competing definitions of fairness.
- **Evidence**: Chouldechova (2017) proof: calibration across groups is mutually exclusive with error-rate parity when base rates differ.
- **Analysis**: Algorithms inevitably produce higher false-positive detention rates for marginalized groups unless explicitly constrained by demographic parity metrics.

## IV. Counterargument & Rebuttal Section
- **Opposing View**: Actuarial algorithms are statistically more reliable than human judges suffering from cognitive fatigue and implicit bias (Kleinberg et al., 2018).
- **Rebuttal**: While human judges are flawed, human discretion can be audited and appealed; proprietary algorithms legally shield scoring weights under trade-secret doctrine, denying defendants constitutional due process.

## V. Conclusion & Broader Synthesis
- **Restatement of Thesis**: Synthesized in light of empirical evidence.
- **Broader Implication**: Proposing hybrid audit models where algorithms serve only as non-binding advisories accompanied by open-source training data disclosures.
```

---

## Worked Example: Transforming an Undergraduate Essay from "B-" to "A"

- **Original Draft**: The student wrote 5 pages describing different renewable energy technologies without taking a stance.
- **Structural Overhaul**:
  1. Reframed the essay around a central controversy: Nuclear energy versus 100% renewables in achieving grid decarbonization.
  2. Crafted a 3-part thesis arguing for base-load small modular nuclear reactors due to rare-earth mineral extraction bottlenecks in battery storage.
  3. Reorganized 8 descriptive paragraphs into 4 MEAL paragraphs with dedicated counter-claim analysis on nuclear waste storage.
- **Outcome**: The revised paper received an "A", with the professor commending the rigorous logical progression and sophisticated counterargument handling.

---

## Verification Checklist

- [ ] Thesis statement is explicitly stated in the introduction and contains a debatable, non-obvious claim.
- [ ] Every body paragraph opens with an original analytical topic sentence (no quotes in sentence 1).
- [ ] Every piece of evidence is followed by original analysis explaining *how* it supports the thesis.
- [ ] Strongest counterargument is articulated and rebutted with evidence.
- [ ] Conclusion synthesizes insights rather than mechanically restating earlier sentences word-for-word.

---

## Anti-Patterns

- **The Quote-Dump Paragraph**: Dropping a 4-line block quote into a paragraph without explaining or analyzing what the quotation means.
- **The Descriptive "Book Report"**: Spending 4 pages summarizing the plot of a novel or the history of a war without presenting an argument.
- **The Strawman Counterargument**: Presenting an absurd, easily dismissed counter-argument (e.g. *"Some people think the earth is flat..."*) instead of engaging serious intellectual critics.

From poorvith-mp/skills-education, MIT.

More from poorvith-mp/skills-education