Political science — the whole-degree cheat sheet

UTM undergraduate · basics through upper-year · built against your own POL243/244 foundations, POL320 and POL373 files · 24 Aug 2026

0 · The one thing the discipline is doing

Political science studies how people who disagree, and who cannot all get what they want, make binding collective decisions — and who wins. Every subfield is a specialisation of that sentence.

Two registers, and marks are lost by mixing them: normative (what ought to be — political theory) and positive/empirical (what is, and why — comparative, IR, domestic). A normative essay that smuggles in a causal claim, or an empirical essay that ends in moral assertion, is the single most common genre failure.

1 · Core concepts — fluent enough to use in a sentence

Power

Three faces (Lukes): 1 winning the open contest · 2 keeping issues off the agenda · 3 shaping what people want. The third face is where most upper-year critique lives.

Authority

Power accepted as rightful. Weber's three sources: traditional · charismatic · legal-rational.

Legitimacy

The belief that rule ought to be obeyed. It makes governing cheap; without it you pay in coercion, continuously. Regime durability arguments almost always run through this.

The state

Weber: the organisation claiming a monopoly on legitimate violence in a territory. Keep four apart — state (apparatus) · nation (identity group) · government (current occupants) · regime (the rules for occupying it). Also: infrastructural power (reach into society) vs despotic power (acting without negotiation) — Mann.

Institutions

Rules, formal and informal, that make outcomes predictable. North: institutions are the rules of the game, organisations are the players.

Collective action

Olson: shared interest does not produce shared action. Free-riding explains why parties, unions, revolutions and treaties need selective incentives, coercion, or small numbers. The default answer to "why didn't they just organise?"

Sovereignty · representation · citizenship

Sovereignty = final authority, internal and external. Representation = who may speak for whom, and on what mandate. Citizenship = the membership that converts a person into a bearer of claims. Each has a critical literature attacking it as historically particular rather than universal.

2 · Political theory — the canon as a chain of replies

ThinkerThe moveAnswering whom
HobbesWithout a common power, life is intolerable; authorise an absolute sovereign. Order is the first political good.civil war; classical/scholastic teleology
LockeNatural rights precede the state; government is a trust, revocable when it breaks. Property by labour.Hobbes's absolutism
RousseauObedience to a law you gave yourself is freedom; the general will, not the sum of private wills. Inequality is made, not natural.Hobbes and Locke's contract as a swindle for the rich
KantAutonomy and the moral law; right as the condition for freedoms to coexist; perpetual peace among republics.Hume, and Rousseau's radicalism
HegelFreedom realised through institutions (family, civil society, state), not against them; history as a rational process.Kant's formalism
MarxPolitics is a surface over material relations; the state secures class rule; critique of rights as bourgeois abstraction.Hegel, and the liberal contract tradition
MillLiberty, the harm principle, the argument from fallibility for free discussion; representative government.tyranny of the majority
RawlsJustice as fairness: original position + veil of ignorance → equal liberties, then inequality only if it helps the worst off.utilitarianism
NozickOnly a minimal state is justifiable; justice is historical (how you got it), not patterned.Rawls's difference principle
Communitarians (Sandel, Taylor, Walzer)The unencumbered self is a fiction; justice is internal to shared meanings.Rawls's abstraction
Feminist theory (Pateman, Okin, Young)The public/private line is itself political; the contract has a sexual contract underneath it.the whole contract tradition
FoucaultPower is productive and capillary, not just prohibitive; genealogy of the subject; discipline, biopolitics, governmentality.sovereignty-centred accounts of power
Postcolonial / decolonial (Fanon, Said, Mills, Dussel)The universal was written from somewhere; racial and colonial exclusion is constitutive, not an oversight.the Euro-canon's self-description

POL320 filing Place each text on two axes before anything else. Method (how it fights): genealogy · immanent critique · institutional design · reading the silence · materialism · defending the tacit · testimony · re-framing. Lens (what it thinks matters): reason · collective will · tradition · feeling & body · material life · power & knowledge. Then name the relation: real disagreement (same method, different lens) · talking past (different methods) · friendly radicalisation (an ally extended until it cuts).

3 · Comparative politics — the models that do the work

Regimes and the survival logic

Selectorate theory (Bueno de Mesquita): leaders do whatever keeps their winning coalition paid. Small coalition → private goods, patronage, repression. Large coalition → public goods. Acemoglu & Robinson: inclusive vs extractive institutions, and why elites block growth that would cost them power.

Democratisation

Modernisation theory (Lipset: wealth → democracy) vs Przeworski's amendment (wealth doesn't cause transitions, it prevents reversals). Transitions as elite pacts (O'Donnell & Schmitter). Class-coalition accounts (Moore: no bourgeoisie, no democracy).

Backsliding live agenda

Democracies now die legally, from inside — incumbents capturing courts, referees and media (Levitsky & Ziblatt; Bermeo on the change in form). Competitive authoritarianism: real elections on a tilted field.

Institutions and design

Duverger: single-member plurality → two-party; PR → multiparty. Lijphart: majoritarian vs consensus democracy. Veto players (Tsebelis): count the actors whose consent is needed and how far apart they are — that predicts policy stability better than ideology. Presidential vs parliamentary (Linz's perils: duelling mandates, fixed terms, no exit from deadlock). Federalism as both accommodation and veto-multiplier.

Political economy of the state

Rentier state: revenue from resources or aid rather than taxation breaks the tax-for-representation bargain (no taxation, no representation). Resource curse and its critics. Tilly: war made the state and the state made war — extraction capacity as the origin of European statehood.

Identity and cleavage

Ethnicity as constructed and instrumentalised, not primordial. Lipset & Rokkan's frozen cleavages. Clientelism as a contingent-exchange machine, not simply corruption.

4 · International relations — the three engines plus the mechanisms

Realism

Anarchy (no ruler above states) forces self-help. Classical realism roots it in human nature (Morgenthau); neorealism in structure (Waltz) — balancing, relative gains, the security dilemma (defensive arming reads as threat). Offensive realism (Mearsheimer): maximise power, seek hegemony.

Liberalism

Institutions lower transaction costs and lengthen the shadow of the future (Keohane); interdependence raises the cost of war; the democratic peace — democracies rarely fight each other (mechanism disputed: norms vs audience costs vs institutional constraint).

Constructivism

"Anarchy is what states make of it" (Wendt). Interests are produced by identity, norms and practice; norm cascades and norm entrepreneurs (Finnemore & Sikkink). Critical, feminist and postcolonial IR press further: whose order, secured by whose insecurity.

The portable mechanisms use these everywhere

Credible commitment — the deepest problem in politics. Bargains fail not because no deal exists but because no one can bind their future self. Explains war despite its cost (Fearon), central bank independence, constitutional courts, and collapsing peace settlements.
Principal–agent — voters→legislators→bureaucrats→international bodies. Every delegation buys expertise and sells control; the residue is slack, drift and monitoring cost.
Levels of analysis (Waltz's three images: individual · state · system). Most weak IR arguments are a level confusion.
Two-level games (Putnam) — negotiating abroad and at home at once; a tied hand at home is leverage abroad.
Path dependence — early choices lock in through increasing returns, so timing and sequence beat any equilibrium story.
Audience costs · private information and incentive to misrepresent · issue indivisibility — the three rationalist routes to war.

5 · The methods apparatus — where the marks actually are

From your own POL243/244 decks. Import the design logic, never the statistics vocabulary — regression talk in a qualitative essay reads as borrowed authority. Every "critically evaluate" objective is a validity + measurement + effect-size exercise in a substantive costume.

The nine moves that travel

  • Explanandum / explanans — force any reading into two lines. If you can't state both, you haven't understood it.
  • Name the puzzle — contra-expectations · divergence · convergence · variation over time · conceptual omission · measurement · epistemological.
  • Falsifiability — "what would prove this wrong?" A framework that accommodates both outcomes explains neither. Apply it to your own thesis in writing.
  • Most Similar / Most Different Systems — when a syllabus pairs two readings it is nearly always one of these. Name the design, then the variable doing the work.
  • Confounder vs mediator — controlling for a mediator is a mistake; most "not really causal" arguments are omitted-variable arguments in disguise.
  • Necessary / sufficient / scope condition — most strong claims are scope-conditional claims that haven't admitted it.
  • Selecting on the dependent variable — the most portable critique in the discipline: where is the negative case?
  • Latent variable → indicator — interrogate the indicator, not only the concept. Reliable-but-invalid is the dangerous case and the default of proxy measurement.
  • HARKing — write the prediction before opening the archive.

The evaluation checklist

  • Internal vs external validity — large-N buys external, small-N buys internal. Say which one the study has.
  • Random assignment vs random sampling — assignment buys internal validity, sampling buys generalisability. Almost nothing has both; saying which is a full-marks sentence.
  • Statistical vs substantive significance — size the effect against a base rate. And when harm sits in a tail, the mean is the wrong estimand.
  • Reverse causation — read the hypothesis backwards for ten seconds. No control variable repairs a two-way arrow.
  • Random vs systematic measurement error — bias doesn't average out; a bigger sample can't fix social desirability.
  • Ecological vs individualistic fallacy — an exam favourite.
  • Sampling frame and coverage — "what people are saying" is a non-probability sample of a hyperactive frame.
  • Levels of measurement — Freedom House / Polity are ordinal judgements rendered as numbers. Medians and ranks, not means. Catching this misuse is a free critique.
  • Correlation vs mechanism vs manipulation — the three routes to causal inference; say which one is taken and why the others are foreclosed.
  • Natural experiment · diff-in-diff · parallel trends — a before/after on treated units only has no counterfactual. Name the missing comparison group.
  • Concept stretching — a definition widened until it scores the wrong cases.

6 · Essay genres — the failure mode is genre, not ideas

GenreSpineHow it fails
Text-interpretiveA passage does something. State the move, then walk the text line by line to show it. Stay inside the text.Summarising the argument instead of reading the passage; importing secondary literature.
ComparativeOne shared question, then the split, then why the split matters. The comparison must be a claim, not a layout.Two mini-essays stapled together; "both thinkers discuss…".
Critical analysisTarget is the secondary literature or conventional wisdom, not a primary text. Reconstruct the strongest version, find the load-bearing premise, strike that.Attacking the conclusion rather than the premise; strawmanning.
Research proposalArgument · why the intervention is needed · state of the literature · structure. It sells a gap.Describing a topic instead of naming a gap and a payoff.
Research paperThesis that could be false · evidence that could disconfirm it · the strongest objection answered in the body, not the conclusion.Literature review with an opinion at the end.

Paragraph test — every body paragraph needs a claim in its first sentence that is contestable, and its last sentence must move the argument forward rather than restate. Introduction test — a stranger should be able to name your thesis, your opponent, and your evidence after reading one paragraph.

7 · Citations and honesty — the rules that protect the grade

Verify before you format

Publisher → Crossref/OpenAlex → author CV → library catalogue. A resolving DOI does not prove a citation is real — the resolved record's title must match. Evidence from your own audit: 9 of 25 citations on the POL373 syllabus were bibliographically wrong. Assume every reading list has errors until audited.

Never

Never invent a page number, quotation or attribution. Mark unknown pages ??. Never a pirate mirror — report the paywall instead. Distinguish the assigned edition/translation from whatever copy is on disk; pagination differs.

Honesty tags to use in your own notes

Per source: full text read · abstract + extracts · not read. Per claim: READ IN FULL · HIGH CONFIDENCE · VERIFY AT SOURCE. Never blur "I read this" into "I read about this" — that slip is where fabricated attributions come from.

Style

Ask the course which: Chicago author-date, Chicago notes-bibliography, or APA. Political theory courses often want notes; empirical courses usually want author-date.

8 · The study loop that actually works

  • Per reading (empirical): explanandum · explanans · puzzle type · design · the confounder the author most fears. Five lines, then stop.
  • Per reading (theory): method × lens · the one move · the common misreading · where to press. Four lines.
  • Weekly: one connection worth posting — which reading answers which, and whether that is real disagreement or talking past.
  • Pre-test: active recall on cards built from your own stamped notes, never from background knowledge. Spaced, not massed.
  • Pre-essay: name the genre first, write the thesis as a sentence that could be false, then write the falsifier line before you draft.
  • After a grade: what did the marker reward, which genre slip cost marks, which cards were missed. Write it down or it does not compound.

9 · Your machinery on this machine

WhatWhereFor
Contract + router~/.claude/polisci/KNOWLEDGE.md · /poliscievery poli-sci turn; resolves the course first
Foundations toolkit~/.claude/polisci/foundations/FOUNDATIONS.mdPOL243/244 apparatus, with deck locators
POL320 — theorycourses/POL320.md · ~/Desktop/modernity-helix · Six Lenses · Médaillier5 essays, 3 with passages assigned a week out — rehearse the form
POL373 — MENAcourses/POL373.md · ~/Desktop/pol373-assay (330 cards)two tests + essay; weights still ?? in the file
Cross-course search~/Desktop/armillary (8 courses, 730 cards)finding a concept across every course at once
Program map~/Desktop/pol-specialist-atlas · ~/Desktop/pol-atlasdegree requirements and course-by-course pages
Not included here: POSt admission thresholds, credit counts and deadlines. Those change by calendar year and the numbers on this machine were captured earlier — read them off the current UTM Academic Calendar and Degree Explorer before acting on any of them.