Friday, 3 September 2010

Critical Thinking 1 - Argument Structures

Posted on behalf of Peter:

Critical Thinking provides a set of tools to assess the truth of arguments (argument = ‘attempt to persuade by reason’). Arguments that we need to analyse are everywhere: media commentators, financial advisors, business proposals – in fact anywhere where we need to decide if we should believe someone’s line of reasoning (including our own…).

Argument composition
An argument has one or more premises leading to a conclusion. Premise and conclusion are types of statement. Assessing an argument basically involves asking the following two questions. (1) Can I rely on the information used in the premises i.e. is it true, false, biased, complete etc. (2) Can I rely on the way the argument structure is used ?

Argument structures
There are two major types of argument: Deductive and Inductive. There is also a third type, Abductive, that tends to be used mostly in relation to scientific enquiry. Each major type has a set of associated thinking errors, or ‘fallacies’.

Deductive arguments
Deductive arguments deal in certainty. A ‘valid’ deductive argument is one where, if the premises are true, the conclusion must also be true. If the premises are in fact true then the argument is also ‘sound’. A deductive argument can be ‘invalid’ (even if the premises are true, the conclusion could be false) or ‘unsound’ (the premises are objectively false). An example of a valid, but unsound, argument would be: ‘All Greeks are green. Socrates was a Greek. Therefore Socrates was green’.

Inductive arguments
Inductive arguments are those seen most often in daily life, and are based more on probability / rational expectation than certainty. The equivalent of deductive validity is ‘inductive force’ (probability of conclusion being true greater than 0.5). The equivalent of deductive soundness is ‘inductive soundness’. An argument is inductively sound if its premise(s) are true and the structure is inductively forceful e.g.: ‘John rarely hands his homework in on time. Therefore it is likely to be late again tomorrow’.

Abductive arguments
Much scientific development is based on inductive reasoning i.e. if an experiment is testable and repeatable, then a general rule may be able to be built. However abductive reasoning is also used. Here, faced with an event or set of circumstances, a number of possible hypotheses are developed to explain the event. The most plausible explanation is then taken as a provisional explanation (subject to further testing).


  1. I have an impression that most politicians are prone to unsound Abductive Arguments, and solicitors/lawyers are most apt at turning a fallacy into a seemingly sound Deductive Argument.
    I suppose that's what they are getting paid for: to manipulate the arguments to suit their specific purpose.

  2. I agree that politicians, and lawyers, misbehave in the ways that you suggest. But what happens is also affected by the context, and that fact may affect our opinion of what such people do.

    Politicians have to decide on some policy or other, although doing nothing is nearly always one of the options. The evidence is rarely enough to establish, to a decent standard of proof, that any given option is the best one. So they do have to reach decisions that are supported by inadequate arguments.

    Lawyers work under specific terms of engagement. In a criminal case in England, the prosecution has to prove that the accused is guilty beyond reasonable doubt. That is such a high standard that a prosecutor will often be tempted to make out that the case is stronger than it actually is. A defence lawyer, on the other hand, only has to create enough doubt. One can do that without relying on bad arguments. In a civil case, the question must be decided on the balance of probabilities. In that context, both sides are likely to be tempted to use bad arguments.

    I do not excuse lawyers. The existence of a temptation does not excuse giving in to it. And I certainly do not excuse politicians who decide what they want to do and then look for arguments to support their decisions. But I think that contexts are worth bearing in mind.