Argument Analysis Lab
The preceding eight lessons have given you a complete logical toolkit. This lab puts every tool to work simultaneously. You will encounter arguments drawn from science, technology, ethics, and public policy — some rigorous, some deeply flawed, some subtle enough to require careful analysis. Your task in each case is the same: apply the full analytical method systematically and produce written verdicts. This is the kind of critical thinking practiced daily by scientists, judges, investigative journalists, and policy analysts.
Lab Method: The Six-Step Audit
For every argument you encounter in this lab, apply the same six-step procedure. Step 1 — Reconstruct in standard form. Identify the conclusion, all explicit premises, and all significant implicit premises. Write them numbered, one per line. Step 2 — Classify the inference type. Is the inference deductive (does the conclusion follow necessarily?), inductive (does it generalize from cases?), or abductive (does it select the best explanation)? Step 3 — Evaluate validity or strength. For deductive arguments: is the form valid? Use counterexample testing. For inductive arguments: is the sample representative and large enough? For abductive arguments: are there competing hypotheses not considered? Step 4 — Evaluate soundness. Examine each premise independently. Is it true? What evidence supports or challenges it? Step 5 — Identify any fallacies. Apply your knowledge of formal and informal fallacies. If a fallacy is present, name it and explain precisely how it operates in this specific argument. Step 6 — Issue a verdict and propose improvement. State whether the argument is sound (or strong), unsound (or weak), and why. Then propose the smallest change that would make it stronger.
In this lab, hold every argument — including ones you agree with — to the same standard. Agreement with a conclusion is not evidence that the argument for it is sound. Disagreement with a conclusion is not evidence that the argument is flawed. The audit must be conducted independently of whether you like where the argument goes.
Argument Analysis Lab — Full Session
- Work through the following five argument cases. Apply the complete six-step audit to each. Write your analysis in full sentences — not bullet fragments. Each analysis should be substantial enough that someone who has not read the argument can follow your reasoning.
- CASE 1 — Technology and Work:
- 'Automation has historically created more jobs than it has destroyed. The Industrial Revolution eliminated manual weaving but created factories, railways, and urban trade. Therefore, AI automation will similarly create more jobs than it eliminates.'
- Case 1 Task: Reconstruct in standard form. Classify as deductive, inductive, or abductive. Evaluate: does the historical pattern constitute strong inductive evidence for the AI case? What disanalogy between the Industrial Revolution and AI automation weakens the inference? Is there an implicit premise about the pace and breadth of AI displacement that should be surfaced?
- CASE 2 — Medical Ethics:
- 'Patient autonomy is the foundational principle of medical ethics. Any medical procedure performed without genuine informed consent violates patient autonomy. Administering a vaccine under employer mandate without offering an opt-out violates informed consent. Therefore, employer vaccine mandates are unethical.'
- Case 2 Task: Reconstruct in standard form. Is this argument deductively valid? Identify the most contestable premise. What implicit premises about the nature of employer mandates and the definition of informed consent must be true for this argument to work? Evaluate soundness.
- CASE 3 — AI and Bias:
- 'The facial recognition system is 95% accurate on the test benchmark. Therefore, it is ready for deployment in high-stakes law enforcement contexts.'
- Case 3 Task: Reconstruct in standard form. Identify the implicit premises. What must you assume about the test benchmark (e.g., its representativeness of the deployment population) for the conclusion to follow? Is this a valid deductive argument? What inductive reasoning flaw parallels the distribution shift problem from Lesson 3?
- CASE 4 — Bayesian Case Study:
- 'Ten people on our team tried the new productivity method and eight reported feeling more productive. That is an 80% success rate, which proves the method works.'
- Case 4 Task: Reconstruct in standard form. Identify the inductive inference. What sample size and selection bias concerns apply? What Bayesian question is not being asked — specifically, what is the prior probability that a self-selected group will report feeling more productive from any change? What control condition is missing? What would a rigorous version of this argument require?
- CASE 5 — Original Argument Construction:
- You have 15 minutes to construct an original argument on the following prompt: 'AI systems should (or should not) be required to disclose that they are AI when interacting with humans in a customer service context.'
- Build the argument using the full eight-lesson toolkit: precise conclusion, verified premises, classified inference type, pre-emptive objection handling, explicit epistemic status. Exchange your argument with a partner. Your partner will apply Steps 1-5 of the audit. Discuss their findings and revise if warranted.
When the lab is complete, spend five minutes reflecting in writing: Which step of the six-step audit did you find hardest? Was it surfacing implicit premises, identifying fallacies, or verifying soundness? The step you found hardest is the step most worth practicing. Difficulty signals the boundary of current skill.
In Case 1 of the lab, the argument moves from 'the Industrial Revolution created more jobs than it eliminated' to 'AI automation will do the same.' What is the most precise name for the inference type and its key weakness?
In Case 3, the facial recognition system has 95% accuracy on its benchmark. Which implicit premise, if false, most directly destroys the argument's soundness?
Match each lab case to the primary logical concept it tests.
Terms
Definitions
Drag terms onto their definitions, or click a term then click a definition to match.