From my logic studies ten years ago I recall that an argument is considered 'Valid' if true premises must lead to a true conclusion (ie, no gaps in the logic); to be 'Sound', however, it must be Valid and have true premises.
This seems Valid to me so far. I guess the peer review (formal and crowd-sourced) will test for Soundness.
> Deductive reasoning is the process of drawing valid inferences. An inference is valid if its conclusion follows logically from its premises, meaning that it is impossible for the premises to be true and the conclusion to be false.
I have seen Rationalism mistaken for sound proof used in lot of arguments.Comfortably, ignoring that the validity of the conclusion is dependant on the validity of the premises. To be honest, logical reasoning is very useful in designing experiments, and multiple chains of logical reasoning is not my strongest point, but the refusal to define the conclusions beforehand is what makes me suspicious, when someone uses long chains of reasoning to convince me.
> There is no evidence that deduction has significant role in commonsense reasoning.
As a matter of fact the first logical system heavily used induction, not the duction, to extend regularity over unknown topics instead of exploring the unknowns systematically and deriving the regularities out of the experiments.
I think logic is more a proxy or a heuristic. It helps you make conjectures about the world even if you don't necessarily have direct evidence.
But if your observations contradict your logic? Something must be wrong. It could easily be the observations, but the more you have, the less likely it is. I think, at the very limit, deduction would have to give way to induction.
Exactly. There's good reason to think that most of human reasoning and the way we attain "knowledge" is not strict deductive logic from firm premises. Rather it's a mixture of Bayesian inference, abductive reasoning "to the best explanation", etc.
I think maybe he means something different from you here. I interpret deductive reasoning here as, "situations where you need to use logical arguments instead of being told what to do and repeating it with slightly different numbers."
You do not need objective truths to use deductive logic. That statement shows that you fundamentally misunderstand deductive logic.
You simply need premises (i.e. assumptions) to make use of deductive logic in a standard logical system. Those premises, when used with logic that has no fallacies, implies conclusions. Any real world conclusions are always based on premises, where those premises are based on experimental evidence. It's always the case that there could be problems with the experiments, and there always are limitations of those experiments. That doesn't mean you can't use logic to derive conclusions.
> I don’t have any sense for the various properties of inductive arguments
You can easily find a lot of resources on this topic, but I can't think of any to particularly recommend. To give my own high-level summary, there are three main types of logical arguments:
- Deductive arguments: from the premises, the conclusion necessarily follows. If the premises are true, and the argument is valid, the conclusion must be true. These are the kinds of arguments found in mathematical proofs, and classical syllogisms such as "All men are mortal, Socrates is a man, therefore Socrates is mortal"
- Inductive arguments: from many particulars, infer a generalisation. If all swans so far have been observed to be white, it is reasonable to conclude that "All swans are white" – and the more observations you make, the stronger the conclusion becomes – but that conclusion can never be certain, only probable, because maybe tomorrow someone will discover a black swan, proving the conclusion to be false after all–which is of course what actually happened, when Europeans came to Australia and found swans which were black. (Also, confusingly, "mathematical induction" is a type of deductive argument, not induction in the logical sense.)
- Abductive arguments: from a situation, infer the truth of its most plausible explanation: If I walk into a room and find toys spread all over the floor, I am going to conclude that our four year old did it. Now, maybe some crazy stranger broke into our house, threw toys everywhere, then left – not absolutely impossible, but far less likely than it being our four year old.
The big difference: in deductive arguments, if the argument is good, and the premises are true, the conclusion must be true; in non-deductive arguments, the argument can be good, and the premises can be completely true, and yet the conclusion can still be false. A deductive argument which produces a false conclusion from true premises is by definition a bad one; a non-deductive argument which produces a false conclusion from true premises can still be a perfectly good one, provided that unfortunate outcome is improbable.
The natural sciences use all three, but rely heavily on induction and abduction. Historians primarily use abduction – induction works best for repeatable events, whereas many of the events historians study are essentially unique and unrepeatable (World War II was in many ways a unique event in human history, nothing quite like it had ever happened before, even World War I was different in many ways, and if World War III ever happens it is going to be different again).
I think a lot of moral reasoning must be non-deductive, because deductive arguments generally deal with certainties, whereas it is quite natural for people to say about moral issues "I think it is probably wrong but I'm not 100% sure". I think moral reasoning often involves some degree of induction – if you repeatedly encounter particular situations which offend your conscience, you are likely to adopt a moral principle which condemns whatever those particular situations have in common – generalising a principle out of particular cases, which is essentially what inductive reasoning is. But, just like inductive reasoning, you can never be sure that there will not be some case which might convince you that your generalisation needs an exception – a moral black swan.
> This is the first time that I’ve heard of normative epistemology, which I gather is what you are talking about.
Yes, exactly. A lot of the concepts traditionally used in epistemology appear to be normative. For example, in the classical definition of knowledge as "justified true belief", justification appears to be a normative concept – a belief which is justified given some evidence is a belief which one ought to hold (or, at least, one is permitted to hold) given that evidence, a belief which is unjustified given some evidence is a belief which one ought not to hold given that evidence.
> One thing I will say though, thinking about this seems smoother than thinking about normative morality because (I think) I don’t have as many childhood memories of people telling me what I ought to think compared to those of people telling me what I ought to do.
I think a lot of people grow up exposed to conservative religious approaches to morality, which they often experienced as being pushed on them inflexibly and never with any convincing justification–and it is unsurprising many such people become attracted to the idea that "morality is subjective". However, I wonder, is that "throwing the baby out with the bathwater"? Most of these people would feel strongly that genocide, slavery, apartheid, torture, child abuse, theocratic totalitarianism, etc, are immoral – yet, having accepted "morality is subjective", those moral positions they endorse must be just as subjective as those which they've rejected. Subjective morality weakens their own moral positions. Indeed, once you believe that morality is subjective, you can no longer claim (in any objective sense) that anyone else's morality is wrong, including that of various religious conservatives – the most you can consistently say about their morality is that you don't like it – but why should they (or anyone else) care about what you like?
Yes; however deductive reasoning is stronger evidence if you have it. It seems that’s what the OP might be implying, although I haven’t seen much of this in model research (empirical evidence is much easier to gather).
Isn't that exactly the question that informal logical fallacies seek to address?
I don't think you can just throw away the word "logic" like that. We try to make arguments using deduction and inference, so there is a logical structure behind them. It's just that it isn't as well-defined and doesn't have the mathematical rigor of formal logic.
My university philosophy 101 critical reasoning course taught us to use this kind of abduction or reverse deduction to take the conclusion of somebody's argument and infer what premises might be needed to support it but are not being stated or are hidden or unexamined. A a very useful skill when in discussion with someone else who studied critical reasoning X'D
I am also trying to get better at this. I find that a lot of my own non-critical thinking hinges on assumptions I've treated almost as axioms. So perhaps a good place to start would be working on identifying the underlying assumptions in a statement or argument.
We might accept some assumptions for the sake of argument, but that doesn't take them off the table for unpicking once the main argument is made.
reply