Linking, evaluating and concluding

Two students can write about the same topic, name the same examples, and earn very different marks. The difference is almost always linking, evaluating and concluding — the three skills that turn raw content into a top-band argument.

Look at where the marks sit in Paper 1: AO3 (communication) is the largest at 45%, then AO2 (analysis and evaluation) at 35%, and AO1 (selection of examples) just 20%. So how you connect, weigh and resolve your argument is worth four-fifths of the paper — far more than the examples themselves.

What each skill does:

• Linking (cohesion) — the connective tissue. Discourse markers ('however', 'consequently', 'by contrast') and signposting ('this essay will argue...', 'a stronger objection is...') make the argument flow so the examiner never gets lost. This is rewarded directly under AO3.
• Evaluating — weighing, not listing. You judge how strong each argument is, handle the counterargument fairly, and decide what the evidence really shows. This is AO2.
• Concluding — the verdict. You reach a clear, supported judgement that the essay has earned, not a summary and not a fence-sit. This is where AO2 and AO3 meet.

The two rules that decide the top band:

1. No counterargument, no Band 5. The band descriptors require counterarguments to be 'acknowledged AND addressed'. However fluent the writing, omitting the opposing view caps the essay below the top.
2. 'There are arguments on both sides' is not a conclusion. Cambridge calls it an evasion of AO2 and AO3 — it refuses to do the very thing the command word ('To what extent', 'How far') demands.

Cohesion means the parts of your essay are visibly connected — each sentence and paragraph follows logically from the one before. The examiner credits this under AO3, and it is the difference between an essay that 'flows' and one that reads like disconnected notes.

Two tools do most of the work:

1. Discourse markers (connectives) — the small words and phrases that show the relationship between ideas. Choose the marker that matches the logical move you are making:

• Adding / building: moreover, furthermore, in addition, equally.
• Contrasting / conceding: however, by contrast, nevertheless, admittedly, while it is true that.
• Cause and effect: consequently, therefore, as a result, this means that.
• Giving evidence / example: for instance, as illustrated by, this is evident in.
• Concluding: in conclusion, ultimately, on balance, the weight of evidence suggests.

2. Signposting — short phrases that tell the reader where the argument is going: 'This essay will argue...', 'A stronger objection is...', 'Having shown X, I now turn to Y.' Signposting is especially powerful at the start of the counterargument and conclusion.

The paragraph-link technique. The cheapest way to look cohesive is to open each new paragraph by referring back to the last. Instead of starting a paragraph with a fresh point out of nowhere, begin with a bridge: 'While accountability is democracy's chief strength, its critics point to a serious weakness...'. The reader feels the argument moving, not jumping.

Mini-judgements as links. End each body paragraph with a one-line judgement that points forward ('so elections are necessary but not sufficient...'). This both shows evaluation (AO2) and links the paragraph to the next stage of the argument (AO3).

Evaluation (AO2) is the skill of weighing. A weak essay lists points for and against; a strong essay decides how much each point is worth and why. To evaluate, keep asking: How strong is this argument? How good is the evidence behind it? Does it really answer the question?

The counterargument is the highest-value evaluative move. The band descriptors single it out: for Band 4-5, counterarguments must be 'acknowledged AND addressed'. Done well, the counterargument does not weaken your case — it strengthens it, because answering the strongest objection shows your thesis can survive challenge.

The three-step counterargument move — Concede, Refute, Judge:

1. CONCEDE — state the opposing view at its strongest, supported by a real example. Never build a 'straw man' (a feeble version that is easy to knock down). For example, conceding that 'censorship can protect the vulnerable, as with restrictions on material that incites violence' is fair and confident.
2. REFUTE or QUALIFY — explain why this does not overturn your main argument. Refute = show it is wrong or limited; qualify = accept it partly but bound it ('this justifies narrow, accountable limits, not broad state control of opinion').
3. JUDGE / LINK BACK — re-assert your line in light of the concession: 'so while some censorship is defensible to prevent direct harm, censoring ideas merely because they are unpopular cannot be justified.'

Weigh the evidence, not just the claims. Part of evaluation is judging quality: is an example typical or a rare exception? Is a single dramatic case enough to prove a sweeping claim? An A* writer notices when one well-chosen example undermines a big generalisation, and says so.

Here is the move that quietly separates A* essays from merely competent ones: they set a criterion. A criterion is the yardstick you judge by — the standard against which you weigh the arguments. Without one, a conclusion is just a preference; with one, it is a reasoned verdict.

Why questions demand a criterion. Most Paper 1 stems contain a slippery key word: 'is democracy the best government?', 'is direct democracy the fairest system?', 'is censorship ever justified?'. These words mean different things to different people. The A* writer defines the key word early — usually in the introduction — turning a vague debate into a measurable one.

How to set a criterion (in the introduction):

• 'By "best" I mean the system that most reliably protects citizens' rights and lets them remove their rulers peacefully over the long term.'
• 'By "fairest" I mean equal participation on each issue — though I will distinguish fairness of process from fairness of outcome.'
• 'Censorship is "justified" only where it prevents direct, serious harm and remains accountable.'

Then apply it in the conclusion. Because you defined the yardstick, your verdict can be precise: 'Measured by long-term accountability, democracy is therefore best, even though it is not always the most efficient.' The criterion does the deciding for you — you are not asserting a feeling, you are applying a standard you set out and defended.

Criterion + distinction = nuance without fence-sitting. Often the cleverest verdict separates two senses of the key word and answers each: 'fair in process but not in outcome'. That is nuanced, but it still decides — which is the opposite of sitting on the fence.

The conclusion is the single most decisive paragraph in Paper 1. Examiner reports note that the gap between the top two bands is often simply a clear, supported conclusion that is reached and sustained — not the volume of knowledge.

Three things a conclusion must NOT be:

• A summary. Re-listing your body paragraphs ('I have shown X, Y and Z') wastes the most valuable paragraph in the essay. The examiner read those points; they want your verdict.
• A fence-sit. 'There are arguments on both sides' or 'it depends on your point of view' is, in Cambridge's own words, an evasion of AO2 and AO3. The command word ('To what extent', 'How far') demands a decision of degree.
• A new argument. Introducing a fresh point or example in the conclusion is a structural fault — there is no room to develop it, and it suggests poor planning.

The A conclusion formula:*

1. Apply your criterion. Restate (briefly) the yardstick you set, so the verdict is grounded.
2. Answer the strongest counter-point in a clause. 'Although Singapore shows clean, effective non-democratic rule is possible, it depends on the calibre of particular leaders...'.
3. Commit, with a "because". 'Democracy is therefore the best form of government, not because it is the most efficient, but because it alone guarantees citizens the means to protect their rights and replace their rulers peacefully.' One position, one reason.

Nuance is allowed — fence-sitting is not. A verdict like 'justified to an extent, but only where the threat is genuine and the powers are accountable' is nuanced and committed. The test is simple: could the reader say, in one sentence, what you decided? If yes, you have judged. If no, you have fence-sat.