What is the difference between logic and reasoning?
I can makes sense of what logic is about. But when it comes to reasoning what does it do more than logic does? Can you give an example which shows the difference between logic and reasoning?
For a start, it is worth noting that the word 'logic' is used in different ways. Up until about the middle of the 19th century philosophers treated logic as being concerned with the laws of thought. These laws were considered to be universal and normative and served to distinguish good reasoning from bad. With the development of boolean algebra, predicate logic and non-euclidean geometry in the latter part of the 19th century, our understanding of logic underwent a step change. Logicians stopped thinking of logic as being about stating things that are incontrovertibly true and then deducing other true things from them, but of being about relationships between thoughts or propositions, without reference to whether they are actually true or not. Boolean algebra, for example, treats a propositional symbol P as a variable that might have the value true or false. We do not assert P to be true, or to be false, we merely proceed to exhibit relationships between propositions on the basis that they might be one or the other. Euclid's approach to geometry is to state axioms and postulates that are supposedly incontrovertibly true and then deduce true theorems from them, but mathematicians discovered that we can assume different postulates (in particular, replacing the parallel postulate) and come up with other geometries, many of which have useful properties.
This development led to the idea that logic was concerned with the relationships of consequence between propositions, not with whether they are actually true. It is also not concerned with whether proceeding from the premises to the conclusion of some argument constitutes good reasoning, or whether it is justified. It is not about thinking at all, but about relationships between the propositions themselves. Logic progresses as we try to discover formal structures in these relationships that permit proofs and computations.
This is the modern sense of 'logic' that you will be taught if you take a logic class or pick up a logic textbook. The term is much narrower in scope than it used to be. Much of what used to be called logic we would now classify as epistemology. That said, 'logic' is sometimes still used in the broader, older sense, particularly in ordinary English usage, and by people who like to blog about logic on the Internet but who have never actually studied the subject. Worse still, Mr Spock must be blamed for promoting the use of the word 'logical' as a very broad term roughly on a par with 'rational'.
Reasoning, by contrast, is what goes on in the minds of rational agents. It is not a relationship between propositions but an activity of forming new beliefs on the basis of existing beliefs and suppositions. Some of the main differences are:
The simplest logical relationships are those in which truth is preserved from premises to conclusion. Reasoning is usually concerned more with grounds, reasons or justification. These can easily part company. For example, any question-begging argument is logically valid because if the premises are true then the conclusion must also be true, but a question-begging argument is not a good way to reason because there is no flow of justification from the premises to the conclusion.
Several logics, including classical logic, include the principle of explosion: any proposition is entailed by a contradiction. This is fine as a logical relation, but is hardly a good way to reason. If I discover I have inconsistent beliefs I will not take this as a license to infer anything I wish.
In logic we commonly define a theory to be a class of sentences closed under the relation of logical consequence, i.e. it includes a bunch of sentences and every other sentence they entail. Such a thing would be unthinkable when describing a reasoning agent. No human being is logically omniscient. We cannot know all the logical consequences of what we know, and as a corollary we cannot expect to have no inconsistent beliefs.
Most logics that are of interest to logicians are monotonic, which is to say that whenever a set of premises entails a given conclusion, adding another arbitrary premise entails the same conclusion. Real human reasoning is seldom like this. We reason all the time on the basis of rules of thumb that we take to work by default because they are usually correct, but which we know may allow for exceptions if some defeating condition arises. "If it looks like a duck, walks like a duck and quacks like a duck then it's a duck" is a pretty good way to reason, though not valid in classical logic. The conclusion does not follow if we add an additional premise that an ornithological expert assures us we are looking at a rare species of goose.
Logic typically lacks any kind of epistemic context or background. Logical proofs stand alone and can often be verified by a computer. Reasoning on the other hand takes place against a huge background of knowledge, beliefs and experience. This is why people can agree on the logic of a situation but disagree about the conclusion. If Alice and Bob both believe A is true and you prove to both of them that A entails B, Alice might judge that the truth of B is more plausible than the falsity of A and so infer B, while Bob might judge that the falsity of A is more plausible than the truth of B and infer that A is false. When it comes to reasoning, one person's proof is another person's reductio.
There is ongoing work in the field of epistemic logic, or more generally formal epistemology, to see if we can discover formal relationships that describe reasoning. This might be understood as an attempt to recover the broader meaning of logic that we abandoned in the 19th century.
Corollary: Gilbert Harman has been arguing since the 1980s that logic and reasoning are different things and that philosophers err in running the two together. His arguments can be found in Change in View: Principles of Reasoning (1986). One of his contentions is that the term 'inference' belongs properly in the realm of reasoning, not logic. Inferring is what people do and is not a logical relation. As such, he proposes that we stop using the expression 'rule of inference' to describe logical rules such as modus ponens but refer to these as rules of implication.
Let's push a fist through this problem with two examples - then a third.
Situation 1. You find you have locked yourself out of your flat. How can you get back in ? You look for the key you hid in the garden but can't find it. You think of phoning another flat holder but don't have your mobile and anyway can't remember any numbers. Your flat is on the second floor and the window is open. Is there a ladder nearby ? No. But there is a tree near the window and if you can climb that, you can probably make a short leap on to the widow sill and climb in that way. Up you go and in you get. This is an example of reasoning - of practical reasoning - reasoning about what to do.
Situation 2. You are presented with three sentences : 'All dogs have teeth', 'All dogs are carnivores', 'All carnivores have teeth'. What a jumble of trite statements. But the logician re-arranges them :
1 All carnivores have teeth
2 All dogs are carnivores
3.Therefore all dogs have teeth
It doesn't matter if any or all of these sentences are false. The sentences have been set out in an argument in which 3. (the conclusion) cannot be false if 1. and 2. (the premises) are true. The argument is valid regardless of the truth or falsity of 1. and 2. This is an example of logic.
In the reasoning example it mattered quite a lot whether your beliefs and assumptions were true. If you were wrong about the climbability of the tree, you'd break your neck. In logic the actual truth or falsity of the sentences is irrelevant - at least it involves no such catastrophic risks.
Situation 3. I've taken an example of practical reasoning but reasoning differs from logic even if we select theoretical reasoning - reasoning about the facts. A laptop has been stolen. You know you didn't steal it. Did your flatmate ? Possibly but s/he had no reason you can think of for doing so - not short of money, has own laptop, &c. There is a new tenant but it is unlikely they could have got access to the laptop. But there was a mystery caller to the house just before the laptop disappeared. Most likely they had a skeleton-key, nipped in and abstracted the laptop. This is sound and sensible reasoning but it is not logic. All your assumptions (premises) could be warranted and rational but the conclusion false. In the event it was the property-owner who stole the laptop.
Step back now and theorise the differences a little. Reasoning is psychological. It is about working out means towards ends, applying rules to cases, deciding which of two or more incompatible beliefs is better evidenced, puzzling out who stole the laptop, whether X will make a good partner. It centres on belief (formation and revision) and inference.
Logic by contrast is not psychological. At all. It centres on the correctness or validity of arguments. It is not concerned at all with whether the premises are true or false or whether the conclusion is true or false. Its sole focus is whether the conclusion must be true if (IF) the premises are true. If the conclusion must be true if the premises are true, then the argument is valid. The argument itself cannot be true or false; it can only be valid or invalid.
Logic and reason are not totally unconnected. One can reason logically. Suppose I reason as follows : I believe that p is the case; I also believe that if p is the case then q is, must be, the case; so I believe that q is the case. I might mentally chug through this. (I believe that x is red; I believe that if x is red then x is coloured; so I conclude that x is coloured.) Rather a simple example but it makes the point - that my reasoning is logical in the sense that it can be represented in logical form as modus ponens : p; p → q : q.
The catch comes - and reason and logic part company - when we adopt a certain type of logic. I don't know how much logic you know but certain forms of 'truth-table' logic can validate implications that make no sense for reasoning of the sorts we've looked at. In certain forms of logic, a conditional is true if the antecedent ('p') is false and the consequent ('q) is true. 'If the moon is made of green cheese (p) then 2 + 3 = 5 (q)'. Truth-table logic of this kind is capable of perfectly sound defence but that defence does not include its agreement with reasoning of the everyday kind considered here.
Like most questions about word meanings, this answer depends on the community of users.
By many users 'logic', being directly related to the word for 'words', is taken to limit you to exactly what the words involved imply, whereas 'reasoning' admits more context that sometimes cannot be captured in statements or has been injected without mention.
Many fallacies represent good reasoning that is bad logic. The slippery-slope fallacy is always bad logic. It assumes something that often happens, but does not necessarily happen in any given situations. But it is convincing reasoning if the problem has been observed before in a related context.
Reasoning is fundamentally the act of thinking. The connotation is that it's a higher form of thinking that leads to insights and knowledge that cannot be directly perceived. Reasoning can be inductive or deductive, meaning that you can think your way to understanding by either eliminating all other possibilities (deductive) or make a generalisation based on other arguably similarities (inductive).
Logic is a set of rules that helps someone reason in a structured and more effective way. Logical reasoning is going to yield far better results than illogical reasoning. Logic comes in various forms and contexts (programming, maths, etc). Some might argue it has the same underlying principles, that it is binary (ie. is not shades of grey) and rules-based. But in practice, there are different logical rules for different contexts.
Logical reasoning is based on rules that help the thinker avoid logical thinking errors. Examples of logical reasoning 'rules' are post hoc ergo propter hoc, confusing correction with causation, confirmation bias, occam's razor, etc. There are also other non-rules that serve as best practice to help a thinker avoid errors and reason more correctly. Avoiding group-think for example. A logical thinker should be less prone to group-think because they are focused on the logical validity of an argument irrespective of consensus.
But - in order to avoid falling into the psychological trap of self-delusion, sensitivity to consensus or contradiction by others is important to address. It's in fact logical, as humans are infallible, to open one's reasoning to the influence of others to consider and test alternative viewpoints.
How reasoning differs from logic is that general reasoning is what is used when we desire knowledge about x from human curiosity. That is, a human wants to know what x is and how does x relate to other things around that human. Why is x this way? and what is x?
Every academic discipline involves some kind of reasoning. However since the time of Aristotle logic has been about the study of correct reasoning in the field of Philosophy. Some people take logic to be more of an art than science. That is, the form of the argument and content matter. In this sense logic relies on pattern recognition and real world application. The pattern of argument should hold regardless of what the subject and predicates are in the argument. In this way finding exceptions to argument form prove a pattern is invalid. No math knowledge is needed. Recognizing patterns and building upon current and available knowledge is needed. In this way one must begin with only true premises. Beginning with true premises and applying correct argument forms guarantees the conclusion must also be true in reality. The term for logic that reflects reality and argument pattern is soundness. This does not hold for mathematical logic.
In Mathematics, one will frequently encounter that logic is the study of validity [and not soundness]. In math, one can encounter nonsensical premises and conclusions but the argument is still valid despite being false in reality. Validity is the relationship of the premises and the conclusion where if the premises are true then the conclusion must also be true. In both math and philosophy this holds true about validity. However, this does not hold true for all kinds of reasoning. For example, inductive reasoning does not guarantee it's conclusion from It's premises. So when you say logic I take it you mean "deductive logic" as opposed to other types of reasoning and other types of logic.
Deductive logic means certainty in the conclusion if the relationship between premises and are correct.
Reasoning is a general category, whereas deductive logic would be a sub category within the reasoning circle. So you can picture a large circle and smaller circles within the large circle. The large circle would be reasoning. The smaller circles in the larger circle of reasoning would be inductive reasoning, deductive reasoning, fuzzy logic, abductive reasoning, scientific method, etc.