Fallacies-malady or remedy?
As a novice myself I've read a handful of introductory books on logic. All these books always have a chapter devoted to fallacies, mistakes in reasoning. We're then repeatedly told to avoid making these mistakes if we are to be truly rational in our thinking.
Keeping the above in my mind let us now consider the theory of evolution whose basic message is we retain and pass to our progeny traits with survival advantage. If this is true am I wrong in inferring that our minds, its processes (including fallacious thinking) are life-critical traits we should be actually cultivating and reinforcing instead of avoiding and purging from or minds?
To give strength to my view let me illustrate with some examples:
1. Fallacy of affirming the consequent.
The above is a fallacy in deductive logic. Consider the following argument:
If a lion is in the bushes then the leaves will rustle
The leaves are rustling
Therefore there's a lion
While the fallacy is apparent it appears very very reasonable to use this form of inference if you were a deer or any other prey animal. It could make the difference between life and death.
2. Fallacy of hasty generalization. This is also a fallacy but think yourself as a deer. Seeing one tiger attack and devour another deer should be logically sufficient to realize tigers kill deer. In such cases NOT committing the fallacy could prove fatal for the deer.
These are just a few cases where fallacious thinking is life-saving. There could be others.
If you agree with all I've said until now don't you think we're making a mistake in so confidently blacklisting so-called fallacious reasoning?
Keeping the above in my mind let us now consider the theory of evolution whose basic message is we retain and pass to our progeny traits with survival advantage. If this is true am I wrong in inferring that our minds, its processes (including fallacious thinking) are life-critical traits we should be actually cultivating and reinforcing instead of avoiding and purging from or minds?
To give strength to my view let me illustrate with some examples:
1. Fallacy of affirming the consequent.
The above is a fallacy in deductive logic. Consider the following argument:
If a lion is in the bushes then the leaves will rustle
The leaves are rustling
Therefore there's a lion
While the fallacy is apparent it appears very very reasonable to use this form of inference if you were a deer or any other prey animal. It could make the difference between life and death.
2. Fallacy of hasty generalization. This is also a fallacy but think yourself as a deer. Seeing one tiger attack and devour another deer should be logically sufficient to realize tigers kill deer. In such cases NOT committing the fallacy could prove fatal for the deer.
These are just a few cases where fallacious thinking is life-saving. There could be others.
If you agree with all I've said until now don't you think we're making a mistake in so confidently blacklisting so-called fallacious reasoning?
Comments (56)
Think Zeno's pardox. Logically/mathematically one CANNOT travel any distance. However one can easily walk from one place to another. This is a perfect example that our world and this universe itself is not limited in any way by our logic and its rules. That leaves plenty of room for perfectly applicable fallacious thinking.
Paradoxes, like Zeno's, have potential answers.
The point is to resolve the paradox by rejecting the premises. Zeno's paradox is not really a paradox per say, but an argument meant to illustrate that changes we observe, such as motion, are illusions. Unless we can illustrate an error in Zeno's thinking in some way, it holds true.
The errors with all syllogisms is always the propositions.
Hence, take my above proposition with a grain of salt.
This is a deductive fallacy, but it is a textbook case of retroductive (or abductive) reasoning. "There's a lion" is a hypothesis, a plausible explanation for the rustling leaves. We have reason to suspect that it is true (and behave accordingly), but no warrant for claiming to know that it is true.
Quoting TheMadFool
It is logically sufficient to conclude that some tigers kill deer. Since a deer cannot know in advance whether any particular tiger that it encounters happens to be one that kills deer, it has reason to suspect that all tigers kill deer (and behave accordingly), but no warrant for claiming to know that this is the case.
The bottom line is that deductive logic is about explicating our premisses - figuring out what else we know, based on what we know that we know - not determining what actions we should take accordingly; especially in scenarios like these, where it is obviously prudent to err on the safe side rather than find out the truth of the matter.
Evolution developed the ability to think logically--at least sometimes, in instances where there was time to think.
"Logic" isn't a product of evolution; logic is the product of our ability to think and the capacity of our culture to develop certain kinds of thinking.
In any event, when it comes to lions, tigers, and bears in the bushes, deer and humans rely on flight or fight responses which have nothing to do with logic, or thinking either.
It depends on what you are after.
You can have a contented healthy life even if you go about it never knowing of logical fallacies or cognitive biases.
But when you are after a deeper understanding you will not make progress until you come to terms with these shortcomings.
I agree here.
Evolution got us to the stone age and we were content there for thousands of years.
Critical thinking is what has allowed us to go beyond that technological level.
Logic is concerned with validity.
An argument is valid when it is impossible for the premises to be true and/or the conclusion false.
The problem with fallacies is that they do not guarantee validity.
So yeah, it's wise for an animal to assume that there's a lion in the brush if the leaves are rustling in a particular way, but it's still a fallacy because that assumption does not guarantee validity. In other words, it's possible that it's not a lion in the brush after all. Therefore that's not a valid conclusion given the premises--because the premises are true, but the conclusion could be false, and validity only obtains when it is IMPOSSIBLE for the premises to be true and/or the conclusion false.
Conclusion: ban deer and other such prey from being taught logic. They must be kept dumb for the sake of their own survival.
Quoting Bitter Crank
No, no, no. They just didn't concentrate enough in logic class. Typical deers. The deers in my class were always slacking off. But it worked out well for them, because the clever ones ended up being devoured by lions.
I don't think you've read these books that carefully myself. The very small one by Graham Priest all on its own will put you right, and that's despite the fact that he believes in wacky logics that few other people believe in.
When positions and arguments are framed properly, we call them sound and valid; not fallacious. A fallacy is by definition an argument whose conclusion we choose not to accept because it's conclusion is either not necessitated (deduction) or not made adequately likely (induction) by it's premises.
The bushes rustling for instance is decidedly NOT fallacious when one takes into account overall survival strategy and the general context in which such a decision would be taken. If lions are known to leap from rustling bushes, even in only 1/1000 instances of bush rustling (it's rare for it to actually be a lion), it will still be quite rational to presume it is a lion for safety reasons and flee none the less. Depending on the prevalence of lions, bush rustling, and the combination of the two, it might be entirely rational for a person to assume that every bush contains a lion and for survival purposes burn them all down. The (proper) conclusion is actually that there is a chance that there is a lion in the bush, and based on the adequacy of that chance, a decision is made to flee or not flee based on probability (lacking a better term). Here the fallacy is not the argument or the conclusion, it's the very acceptance of the conclusion (in deciding to flee) when it is actually not strategically productive to do so (pertaining to the goals of the individual in question).
p1.1 Meteorites could kill someone if struck by one
p1.2 Meteorites hit the earth every day
p1.3 You are on the earth
c1.1 [p2.1] You might get hit by a meteorite
c2.1 DUCK!
The first argument is sound and valid, it's precisely c2.1 that commits the fallacy here by grossly overestimating the actual likelihood of getting struck by a meteorite. We could call this a fallacy, but we could also call it a weak inductive argument. Fallacies tend to be the tricky mis-steps in the logic game which wind up being convincing. That is to say, we take pains to labels specific types of errors as certain fallacies because we need to train ourselves to avoid them.
We train ourselves to avoid fallacies because they don't work. If they worked more often than not then we would call them strong inductive arguments and happily employ them on a regular basis.
If fallacies worked we would embrace them. The precise reason why we take pains to identify and reject them is because we know from experience that they lead to unreliable conclusions. At best you might describe evolutionary endowed predispositions leading to behavior which might resemble actions resulting from fallacious reasoning (like instinctual bush paranoia vs concluding every bush rustle is a lion, or being addicted to the thrill of fishing vs the gamblers fallacy), but when we understand the logical or rational benefit of being a dedicated fisherman and a careful woodsman, we can re-frame the argument in such a way that it makes perfect sense without embracing either instinct or fallacy. How often do lions kill people by leaping out of bushes? Might want to consider not taking chances if the frequency is high enough. Never giving up fishing, during either a time of feast or famine, is always beneficial to long term survival where food is not guaranteed to always be available.
The rub here is that if you manage to present a rational argument which makes fallacious reasoning somehow beneficial to survival or other such goals, then it becomes rational to adopt the actions prescribed by the fallacious reasoning in a strictly logical and strategic way; not fallacious.
As well they should be. Where would predators be without stupid prey? One doesn't always feel like cleverly outwitting supper.
Of course being wrong has value. It is how everyone learns, grows, and evolves. It is absolutely fundamental to human existence.
Quoting TheMadFool
Yes, I agree. Bergson solved the paradox by observing that true time is continuous (non-divisible) and heterogeneous. Logic can never solve these kind of problems.
Exactly. So shouldn't logicians be cautious about condemning a useful way of thinking? Fallacies are part of the repertoire of our survival skills.
Ad hominem:D...a survival skill
These are all just learning tools for people to explore. Every syllogism is fallacious ii one way or another. One takes away from it with what they wish.
Not fallacies per se, but rather types of reasoning other than deduction - e.g., retroduction (hypothesizing a plausible antecedent from an observed consequent) and induction (generalizing from individual cases).
In traditional logics, probability has nothing to do with it. Reasoning is fallacious if it doesn't guarantee validity, where validity is when it's impossible for the premises to be true and/or the conclusion false.
The point of this is that this traditional approach to logic isn't necessarily very useful when it comes to making decisions, acting, etc.There it's often wiser to act on educated guesses and so on. We need modified logics to handle such situations, if we want to treat them logically.
Only in deductive reasoning. Retroductive reasoning is valid when it produces an explanatory hypothesis that is capable of experiential testing. Inductive reasoning is valid when it proceeds in such a way that it will be self-correcting in the long run.
Inductive reasoning is still reasoning. We use it all the time. Instead of making its conclusion guaranteed and therefore sound as in deductive reasoning, inductive reasoning makes it's conclusion "strong" or more likely or probable without actually necessitating it.
What does that have to do with what I was saying though?
If inductive reasoning works, then why don't scientists use it? If you think that scientists use induction, perhaps you could give an example of a theory that was induced, and how it was induced?
The scientific method is inductive in its entirety.
With the understanding that retroductive reasoning is a type of inductive reasoning.
Not at all - retroduction (or abduction) is a distinct type of reasoning that provides explanatory conjectures for deductive explication and inductive examination. I prefer the term retroduction because it proceeds "backwards" relative to both deduction (consequent to antecedent) and induction (experience to hypothesis).
Maybe you'd care to give an example of it at work? How about a rough idea of how general relativity was induced?
What we need are the repeated observations that Einstein made, from which he induced an explanation. Then the repeated observations he made to make general relativity more likely.
Whatever you call it, it's still supposed to be a method of inference: theories from data. Science on the other hand is problem solving, and there's no method for that.
Einstein hypothesized it (retroduction), he and others worked out some of its experiential consequences (deduction), and then various scientists conducted further experiments and made observations to see whether those predictions were falsified or corroborated (induction).
Quoting tom
Not exactly; it is more like the formulation of a plausible explanation for an otherwise surprising observation on the basis of other background knowledge. It typically involves making connections that had not been recognized before.
Quoting tom
Engineering is problem solving, and there are all kinds of methods for that. The same basic pattern of retroduction (design), deduction (analysis), and induction (testing) is evident.
These were both retroductions (experience to hypothesis), not inductions (hypothesis to experience). Inductive experimentation requires a retroductive theory and its testable deductive predictions before it can even begin.
So, GR was not inferred from data. We know this to be historically true.
Einstein worked out some crucial tests - the classical tests of relativity. Putting the word (induction) is meaningless.
The crucial test is of central importance in the methodology of science. It pits two rival theories against each other in an attempt to render one of the theories non-problematic. The theories render each other problematic up to that point.
Quoting aletheist
As I said, theory from data. Science on the other hand works from problem to solution, without method. The method solely deals with how the solutions are treated.
Back to general relativity, what was the surprising observation, and how was the explanation inferred from it?
Quoting aletheist
So you don't think the unification of GR and QM is a problem? Each theory renders the other problematic due to certain mutual inconsistencies. There has never been an observation, surprising or otherwise, that calls either into question.
What observations?
Quoting Metaphysician Undercover
Are you claiming this is not true for Newton's Laws?
Just to let you know, I don't agree with this. Induction is not "hypothesis to experience". It is a generalization derived from experience, such as a law in the sense of a law of physics, like Newton's first law of motion for example. It does not need a prior hypothesis, with experimentation, it only requires observation. So I don't believe that an inductive conclusion requires a retroductive theory. Nor do I believe that one can make such a clear distinction between a retroductive and inductive principle.
Quoting tom
The observations I am talking about are the observations concerning the motions of bodies made by physicists prior to Einstein, such as Ptolemy, Copernicus, Kepler, etc.. Relativity was developed by Newton and Galileo. The fact that the motions of the solar system could be described by either the geocentric or the heliocentric models, is an indication that motion is relative.
Quoting tom
No I'm not claiming any such thing, that's why it's called Newtonian relativity.
What were the observations that led to that inductive conclusion?
You are the one who insists on defining retroduction as "inferring from data." I describe it as "formulating an explanatory hypothesis," which is exactly what Einstein did.
Quoting tom
That was deduction, and the actual testing was induction.
Quoting tom
Claiming this over and over again does not make it true. There is no deterministic route from a perceived problem to a single "right" solution - not in engineering, and not in science.
Quoting tom
I am not a historian of science, so you tell me - what motivated Einstein to develop his theory? What problem was he trying to solve? What reason did he have to doubt the theories that were already in place?
Quoting tom
What gave you that idea?
Quoting tom
How about the fact that each theory renders the other problematic due to certain inconsistencies? That seems rather surprising, hence the desire to find a way to unify them.
I obviously disagree, but I see no point in arguing about it. There is a reason why the scientific method is also called the hypothetico-deductive method. Newton's laws were just explanatory hypotheses (retroduction) until they produced testable predictions (deduction) that were subsequently corroborated by experiments and observations (induction).
Newton's laws were not written as hypotheses, to produce testable predictions in order to determine the reliability of the hypotheses. They were written as statements of fact, laws. These laws were intended to act as premises from which deductive reasoning could proceed. Since these laws were assumed to be true, the conclusions derived were also assumed to be true. To describe them as explanatory hypotheses is just a false representation.
Yes, a theoretical problem is discovered and solutions to the problem is sought. Just like general and special relativities. No observations were involved surprising or otherwise. The methods of science are employed once the theory is proposed.
CONSIDER YOUR ERRONEOUS EXAMPLE
You gave the 'lion in the rustling bush' as an example of a 'fallacy', which is a fallacy in itself - for you would not know if your guess was a fallacy or not until you were eaten or not, so you could not use 'fallacy' as a 'tool'. What your guess was in that situation was a PROBABILITY, which you COULD use as a tool.
FALLACY AS A TOOL
Yes, you can use fallacy as a tool - for causing action (or motivation). As such, even when false, it may lead to a desired outcome (and this is what you were questioning).
My concern is that so-called fallacies may not be real fallacies at all. Like in my example a deer may be saved by committing the fallacy of affirming the consequent thereby endorsing it as a valid type of reasoning.
And like I said, it is a valid type of reasoning - retroductive reasoning, rather than deductive reasoning. The conclusion is thus merely plausible at best, rather than certain.
Our natural modus operandi involves lots of "unsafe" reasoning (induction, abduction).
In fact, we wouldn't get far if we insisted on "safe" reasoning (deduction) only.
Yet another reason that justification (like evidence) is important in knowledge claims.
Where did I say that? I was addressing your specific examples, which are indeed fallacious as deductive reasonings, but valid as retroductive or inductive reasonings. It all boils down to the purpose of the reasoning. If you want to guarantee that you will only derive true conclusions from true premisses, then you go with deduction. If you merely want to formulate a plausible hypothesis, then you go with retroduction. If you want to test a hypothesis, then you go with induction, but only after deductively explicating it to derive experimental predictions.
There is. (See @aletheist's response, or look up examples, or I could post some I suppose, once time permits.)