STEP SIX: Evaluating Inferences
The most persuasive arguments are deductive, and contain logical proof of their conclusion: ʺIf 2 + 2 = 4, then 2 + 2 + 2 = 6.ʺ No other conclusion is logically possible and no other proof is needed.
But most arguments are not so certain: their premises donʹt prove their conclusion through inescapable logic but only lead up to, or imply, it. In these arguments we have to ʺinferʺ a conclusion by extrapolating from the premises. These ʺinductiveʺ arguments can be based on limited causal reasons or on analogies, but they are usually based on probability: ʺOnly one airplane in 100,000 crashes, so itʹs safe for me to fly today.ʺ There is no ironclad logical proof between the premise and the conclusion here. My plane could, in fact, be the very one in 100,000 that crashes next. But itʹs probably safe for me to fly today anyway.
Inferences can be strong (that is, very likely) or weak (not so likely). Strong inferences have good evidence of likelihood behind them, such as research, statistical testing, or wide experience. Weak inferences show less evidence and rely more on hope, faith, or trust than on demonstrated probability. When a jury is instructed to decide a defendantʹs fate ʺbeyond a reasonable doubt,ʺ the judge is asking them not to convict unless they have deductive proof or a strong inference of guilt.
Look at the arguments you identified in your document and answer these questions:
1. Does the author ever use a deductive argument (one in which the conclusion is logically inescapable if the premises are true)?
2. Does the author ever use an inductive argument (one in which the conclusion must be inferred from the premises)?
3. Is the authorʹs inference a strong one?
Reliance inferences more often than on proof, and relying on weak inferences rather than strong ones, are reasons to doubt an authorʹs conclusions.