On pages 19-20 of Logic: The Laws of Truth, Smith argues that "good reasoning" cannot be equated with the properties of necessary truth-preservation (NTP) or validity on the following grounds:
- On equating good reasoning with NTP:
[The view that good reasoning can be equated with NTP] seems too strong — if “good reasoning”is taken to have its ordinary, intuitive meaning. For example, suppose that someone believes that there is water in the glass, but does not go on to conclude that there is H2O in the glass. This does not necessarily mean that there is something wrong with her powers of reasoning: she may be fully rational, but simply not know that water is H2O...it would not be right to say that she had failed to reason well. So we cannot equate good reasoning with NTP.
- On equating good reasoning with validity:
Yet even the claim that reasoning is good if and only if it is valid (as opposed to simply NTP) would be too strong. As we shall see in §1.5, an argument can be valid without being a good argument (intuitively speaking). Conversely, many good pieces of reasoning (intuitively speaking) are not valid, for the truth of the premises does not guarantee the truth of the conclusion: it only makes the conclusion highly probable.
I think I understand why good reasoning cannot be equated with validity, for an argument can be valid without being "good" (Smith gives the example of a valid argument that is not sound), and an argument can be "good" without being valid (e.g. if it is inductively strong). But I'm not sure I understand the first quote above:
What does Smith mean by "if 'good reasoning' is taken to have its ordinary, intuitive meaning"? What is this meaning?
How exactly does the water/H2O example caution against equating good reasoning with NTP?
I understand that the argument "There is water in the glass, therefore, there is H2O in the glass," is NTP by virtue of empirical facts (as opposed to its structure, as in the case of a valid argument). But, what's confusing me is this bit of the quote: "if someone believes that there is water in the glass, but does not go on to conclude that there is H2O in the glass...". If the person does not draw a conclusion, then how can they be said to be "reasoning"? And if they cannot be said to be reasoning, then how can the question of whether their reasoning is good arise?
Or does Smith mean that the person is saying, "There is water in the glass, therefore, there is no H2O in the glass"? But this shows that non-NTP ≠ bad reasoning, and I'm not sure if that's the same thing as showing NTP ≠ good reasoning.
Smith concludes on page 20 by motivating the need for a general method of assessing validity:
The full story of the relationship between validity and good reasoning is evidently rather complex. It is not a story we shall try to tell here, for our topic is logic — and as we have noted, logic is the science of truth, not the science of reasoning. However, this much certainly seems true: if we are interested in reasoning — and in classifying it as good or bad — then one question of interest will always be “is the reasoning valid?” This is true regardless of whether we are considering deductive or nondeductive reasoning.
I don't quite understand what Smith means by the text in bold. How is the question "is the reasoning valid?" applicable in nondeductive reasoning? Isn't validity a property of deductive reasoning only?