• Welcome to The Cave of Dragonflies forums, where the smallest bugs live alongside the strongest dragons.

    Guests are not able to post messages or even read certain areas of the forums. Now, that's boring, don't you think? Registration, on the other hand, is simple, completely free of charge, and does not require you to give out any personal information at all. As soon as you register, you can take part in some of the happy fun things at the forums such as posting messages, voting in polls, sending private messages to people and being told that this is where we drink tea and eat cod.

    Of course I'm not forcing you to do anything if you don't want to, but seriously, what have you got to lose? Five seconds of your life?

IMPORTANT: On Reasoning and Arguing


Still loves Joltik, though!
Staff member
In this forum there will be debates, and where there are debates there must be logic. While some people seem to think they can win a debate by shouting louder than the opponent or writing more indignant and incredulous responses, good debating technique is about logic, and to avoid coming off as silly you need to make sure your logic is sound.

Logical Structure

A logical argument is about three things: you start with some number of premises assumed to be facts about the world, and by applying logical reasoning, you derive a conclusion from these premises. If the premises are true and your argument is logically valid, you will have shown your conclusion to also be true. The conclusion can then be used as a premise in further reasoning; however, as the conclusion stands and falls with the premises and argument on which it was based, everything you base on it will topple if the original conclusion is proven false, so beware.

This means that there are two points of attack against an argument in a debate: you can question the premises, or you can question the logic used to derive the conclusion from the premises. Without providing a valid argument against one of them, you cannot knock down your opponent's stance, no matter how loudly you disagree with it.


If your premises aren't obviously apparent (or if somebody questions premises you initially thought were obviously apparent), either cite an accessible source for them or backtrack to make a well-reasoned argument to support the premise, as appropriate. Of course, if you're going to argue for the premise, that argument will in turn be based on other premises, so somewhere down the line you'll have to come down to base premises that everyone can accept, either because they're obvious or because there is a good source for them.

The more outlandish the claim, the more reputable a source do you need in order for your opponents to accept it. There can be no sensible arguing if the base premises aren't agreed upon by both sides; if your opponent believes one of your premises is incorrect, they'll dismiss your whole argument right off the bat. Sometimes a base premise can't be verified but is still accepted for the sake of the argument, but this only works if your opponents agree that it is interesting and meaningful to discuss further what would be the case if this premise were true. Whatever the case, everyone needs to agree on the premises used in the argument, or there can't be an argument.

If the base premises upon which your opinion is based are the kind of outlandish and unverifiable that makes your opponents unable to accept them based on the information available to them (personal experiences, say), and they're not willing to accept them for the sake of the argument, you unfortunately have no real place in a debate on the subject. Conversely, if you're going to waltz in here trying to be philosophical by questioning even obvious or well-supported premises on the grounds that "nothing is 100% certain" or something of the like and refusing to do any kind of accepting-for-the-sake-of-the-argument, you also have no place in a debate.

Most of the time, informal debates such as those conducted on the forums don't literally start with everyone outlining the precise premises their arguments are based on; instead the debate proceeds backwards, as contested claims are backed up by new arguments based on other claims. As this goes on, if the debaters pay attention and argue logically, one of them will either be convinced of the other's position or they will come down to premises that are sufficiently subjective or personal for them to agree to disagree or recuse themselves from the debate. It's hard to call an objective "winner" in most debates, which do not end with one party simply convincing the other; third-party observers can decide for themselves whose arguments they found more convincing, but ultimately debates aren't about winning so much as encouraging discussion and thought.


I won't go into the formal logic of inductive or deductive reasoning, because frankly, most of it is beautifully self-evident. Formal errors in logic are rare compared to informal fallacies, and most people kind of instinctively sense that there's something off about the argument "If it is raining, I am wet; I am wet; therefore, it is raining." However, I will tackle some of the most common logical fallacies appearing in Internet debates and why they're fallacious. If you show your opponent's argument to be fallacious, then you have knocked out their proposed path from the premises to the conclusion, meaning even if their premises are true, this does not imply the conclusion is true.

Ad hominem

Not synonymous with simply insulting the opponent's character, contrary to popular belief. This fallacy revolves around dismissing an argument not because of a flaw in the argument itself or its premises, but instead because of some property of the person stating the argument. It is fallacious because the properties of the person stating the argument in this particular case are (usually) not inherently attached to the argument at all; it could just as easily be stated by somebody else who does not have these properties.

Example: "Yes, Glenn Beck made this argument about taxes. However, Glenn Beck is a despicable human being, so obviously anything he has to say is not worthy of consideration."

Example: "You argued that if you flaunt expensive jewelry it is to some degree your own fault if you get mugged. However, you are playing devil's advocate and don't actually believe this. Therefore, your argument is meaningless."

It is also fallacious if the person attacked is somebody else who has made the same argument, such as the originator of the idea being presented (genetic fallacy). The identity of this person is still not relevant to the soundness of their argument.

Not an ad hominem:

Occasionally the properties of the person are in fact directly relevant to the argument, but this requires properties of the person to actually be premises in the argument:

Example: "Yes, you saw Jesus appear before you, Emily. However, you are a diagnosed paranoid schizophrenic with a history of vivid delusions, so your credibility is somewhat compromised."

Here the premise of the argument is that Emily saw Jesus appear before her and that her testimony is credible; even if somebody else stated the argument, it would still be about Emily's experience and credibility, and therefore the counterargument is merely attacking the premises of the argument normally. The fact Emily herself is stating the argument here is incidental.

Example: "That article you posted saying studies show men are inherently better at maths than women is from the Daily Mail, which is infamously inaccurate. Do you have a better source for this?"

Again, the argument presented by the opponent seems to be that the claim is true because it says so in the Daily Mail, which makes the Daily Mail's credibility into a premise. If the Daily Mail had presented an argument with credible, non-Daily Mail premises for why this must be true, however, the Daily Mail's credibility would be irrelevant and the argument would be an ad hominem.

Additionally, again, comments about a person's character that are not actually being presented as a basis on which to disregard their points are not ad hominems - they're just side comments. (Do remember that directing insulting comments at other members is never allowed, however.)

Example: "Glenn Beck is a despicable human being, and he's also completely wrong here, because [insert valid counter to Glenn Beck's argument]."

Appeal to consequences

An argument of the general form that because some premise would have desirable/undesirable consequences or implications, it is therefore true/false.

Example: "If Jesus is not truly the son of God, then he was evil or a lunatic. But the idea of such a huge portion of the world's population blindly following an evil man or a lunatic is abhorrent. Therefore, Jesus must truly be the son of God."

Example: "You think homosexuality is a dirty sin? Do you realize that when you say that you're accusing half of this forum of being dirty sinners?"

This comes in many forms and variations and closely related fallacies. Unfortunately, an argument just isn't invalid simply because it is offensive or draws a morally abhorrent conclusion. Feel free to report overtly offensive posts, however.

Not an appeal to consequences:

Of course, if the issue at hand is whether it would be good, wise or right to do something, rather than whether something is true or false, the consequences of doing it are usually highly relevant and the argument is not fallacious.

Example: "If we legalized gay marriage, the human race will stop having children and go extinct."

While the purported consequence in the example is absurd, it is perfectly legitimate for an argument about whether something should be legal to involve the consequences of legalizing it, and indeed, if the extinction of the human race were actually a direct consequence of gay marriage, that would be a fairly important concern to bring up.

Appeal to popularity

An argument saying that because some idea is popular or generally accepted, it must be true.

Example: "Most of the world believes in God. Isn't it unlikely that the tiny minority that are atheists are actually right?"

Related is the appeal to authority, which argues not from the opinion of the general population but from the opinion of some particular respected person or group.

Example: "You do realize you're arguing with the admin, right? Who do you think you are?"

In both cases, it is not relevant who believes the argument or how many; what matters is why. If their reasons for believing it are fallacious or based on false premises, they are just as likely to be wrong as a single random person making the same argument would be.

Not an appeal to popularity/authority:

Obviously, there is nothing wrong with appealing to authorities to establish premises to use in your argument - that's just citing sources.

Example: "The overwhelming scientific consensus is that global warming is happening and a serious problem, so we should invest more effort in developing clean energy sources."

Additionally, if you are disputing such a premise, it is legitimate to cite a source with more credible authority on the subject at hand:

Example: "You're not a physicist and neither am I, but this guy is and according to his article, you've completely misunderstood the second law of thermodynamics."

The fallacy happens when authority or popularity is used in itself to support or dismiss a conclusion, as opposed to a premise.

Argument from ignorance

The idea that because something has not been conclusively proven to be true, it must therefore be false, or vice versa.

Example: "Why do I believe in God? Well, there's no proof he doesn't exist, is there?"

This fallacy is closely related to the concept of burden of proof: when some nonobvious claim is made about the world, it is the one making the claim that has to support it with evidence; it is not the opposition's responsibility to prove the claim isn't correct. Meanwhile, if a claim has already been made and supported with plentiful evidence to the point of general acceptance, the burden of proof turns around and anyone challenging the claim has to produce even more convincing evidence for it not being correct.

If your argument is just that it is technically possible for your claim to be true, that's fine, but it leaves very little of interest to debate about, since generally the answer will simply be, "Okay, it's technically possible, but so unlikely we might as well ignore the possibility."

Not an argument from ignorance:

The fallacy of the argument from ignorance is strictly when truth or falsehood is inferred straight from a lack of conclusive evidence to the contrary. Absence of evidence can be evidence of absence if there is a legitimate expectation that evidence should appear if the claim is true, however.

Example: "If homeopathy did work, we would expect double-blind studies to show that patients given homeopathic remedies recover better than patients who receive only water, but given this hasn't happened, the claims of homeopaths don't appear to hold up."

Argument from fallacy

The idea that somebody making a fallacious argument supporting some claim is in itself evidence of its falsehood.

Example: "Pssh, your argument for why homosexuality isn't a dirty sin was fallacious. It only goes to show that I'm right."

This, by the way, is why even though the examples of fallacies used here frequently actually concern relevant real-world controversial subjects, the fact I have fallacious examples of arguments for certain positions has absolutely no bearing on whether these conclusions are true or false (or whether or not I personally agree with them).

Not an argument from fallacy:

Of course, an argument being fallacious definitely is a good reason to dismiss that particular argument.

Example: "Your argument for God's existence is just an appeal to ignorance, so I'm not convinced by it, I'm sorry to say."

Hasty generalization

When an obviously limited body of evidence is used to extrapolate massive conclusions that the evidence can't justify.

Example: "I hate Christians. All the Christians I've talked to are hypocrites."

Example: "I've known a lot of atheists, and the common denominator is that they were sullen, unhappy with life and socially isolated. Atheism is a cry for help from lonely individuals who have gotten distanced from God."

You just can't make generalizations about huge groups of people based on observations you've made about people you know who belong to that group. This is weak anecdotal evidence, subject to all manner of biases (sample bias, confirmation bias...), and comes from a tiny sample; you have no basis on which to consider it strong enough to extrapolate it to these populations as a whole, especially considering how different people are from one another in general.

Not a hasty generalization:

Simply naming examples while making an argument for why X intrinsically leads to Y is not a hasty generalization, since the conclusion isn't being extrapolated from the examples.

Example: "The idea of divine purpose, ultimate justice and eternal life is immensely comforting to the human psyche. How can you feel any real motivation to do anything, if life is just pointless movements of molecules designed to replicate genes? Why be good when ultimately we're all wormfood? The atheists I know tend to be pretty sullen; I think it's because humans need religion to function properly."

Here the argument attempts to establish a causal relationship between atheism and a lack of purpose and direction and only then brings up the poster's atheist friends. Since the atheist friend example isn't a premise and could be removed entirely without changing the argument, this isn't a hasty generalization, whatever its other problems.

Moving the goalposts

When any time your opponents counter your arguments, your response is to demand more evidence or otherwise an even stronger argument than that provided. If you honestly find your opponents are providing evidence too weak to convince you, you should establish immediately exactly what would convince you of their position, and they have a right to demand you do so so that you can't just back out of it again.

Example: "Okay, so the archaeopteryx is a transitional stage between reptiles and birds, but where's the transitional stage between reptiles and archaeopteryx, huh?"

Not moving the goalposts:

Simply clarifying what you meant if you've been misunderstood or mischaracterized isn't moving the goalposts - but do try to be extra precise about what you do mean if you find yourself needing to clarify, since repeatedly being deliberately vague is a way that people cover for goalpost-moving.

Example: "When I said taxes are stealing, I didn't mean in the literal legal sense; I meant in the sense that it's unjustly taking away something another person has earned, and I think that's wrong even if you're going to use it for a good cause."

Post hoc ergo propter hoc

Translates to "After this, therefore because of this." This fallacy is when you assume that because one thing follows another, the first thing caused the second thing, when the relation could be completely coincidental.

Example: "9/11 happened just a year after Bush was elected. I don't think that's a coincidence."

Related is the fallacy that correlation implies causation, namely that because two things are statistically correlated one must somehow cause the other.

Example: "The number of violent crimes is strongly correlated with ice cream sales. Obviously ice cream affects people's brains, turning them into violent thugs."

This is supposedly a true correlation, but there is no reason to believe that ice cream and violent crime are causally related in any way - rather, the standard explanation is that both of these statistics are incidentally correlated with the summertime, creating an illusory correlation between the two.

Not a post hoc ergo propter hoc:

Obviously, if you actually make an argument in favor of a causal relationship between the things you're talking about rather than just establishing a chronology or correlation and acting like that automatically implies causation, that's not this fallacy:

Example: "The Bush administration had intelligence about Al-Qaeda, but they failed to act on it, so they're partly responsible for those deaths."

Additionally, while correlation does not imply causation exactly, correlations can definitely hint at relationships between things and reveal avenues of research, so while they're not incontrovertible evidence and should be taken with a grain of salt, they shouldn't be automatically dismissed out of hand as irrelevant, either.

Red herring

When something is brought up as if it supports a conclusion, but in fact it is an irrelevant distraction. Red herrings can be obvious to the point of self-parody, but they can also be relatively subtle by attacking premises that seem relevant but actually aren't meaningful in the context of the argument.

Example: "Proof that the Christian God doesn't exist? That's easy. God is supposed to be omnipotent. Can he make a stone so heavy that he cannot lift it? If he can't, then he isn't omnipotent, whereas if he can, then he cannot lift the stone and is therefore not omnipotent either. Therefore, the Christian God is self-contradictory and cannot exist."

The obvious answer to this argument is simply "Okay, so he can't make a stone that's so heavy he cannot lift it. Your point being?" Nobody should care whether God's omnipotence extends to ridiculous logically contradictory thought experiments; focusing upon what "omnipotent" technically means in some abstract sense is irrelevant to the actual meaningful question at hand.

More on this kind of thing in "Debates and Words", below.

Not a red herring:

In order to be a red herring, the argument as a whole has to have no real bearing on the subject - valid arguments can bring up something apparently unrelated and relate it back to the topic at hand, such as in an argument by analogy. Many arguments by analogy are in fact red herrings, but then this is because they are false analogies, where the analogy ignores relevant, meaningful differences between the subjects - there is nothing inherently fallacious about arguing by analogy.

Example: "Suppose I told you there was a teapot orbiting the Sun, somewhere between the Earth and Mars. I can't prove to you that it exists, but neither can you prove that it doesn't. It would be ridiculous for me to expect you to believe in the existence of the teapot simply because you can't disprove it. It's about as ridiculous for you to expect me to believe in the existence of an invisible, undetectable god."

Here teapots may have nothing to do with God in themselves, but the form of the argument relates the teapot back to the question of whether it is reasonable to believe in God - if that's what the debate is about, then it's not a red herring.

Slippery slope

An argument unreasonably concluding that an action that can be seen as a step in some direction must lead to further steps in the same direction being taken until some outrageous conclusion is reached.

Example: "If we allow gays to marry, we'll end up allowing people to marry their dogs or their children! You don't want that, do you?"

Here no proper reasoning is supplied for the idea that endorsing gay marriage would lead to also endorsing bestiality or pedophilia; it is simply assumed that allowing one group that can't currently marry to do so will somehow lead to all groups that can't currently marry also being allowed to do so.

Not a slippery slope:

If a valid line of reasoning is provided for why accepting the original proposal must lead to accepting the absurd conclusion, that's a legitimate debating technique called reductio ad absurdum.

Example: "You say you want to ban homosexuality because the Bible says so, but the Bible bans all sorts of things - shellfish, pork, touching women who are menstruating. Do you want those banned too?"

Assuming the original argument truly is simply that homosexuality should be banned because the Bible says so, then it is reasonable to ask if accepting that reasoning shouldn't mean we ought to ban everything else that is banned in the Bible too - the one arguing for the banning would need to establish a distinction that makes it right to follow the Bible's instructions to ban homosexuality but not these other things.


Making an argument against a proposition that is not in fact the opponent's proposition at all, but a weaker, easily countered version.

Example: Person 1 says, "There is no proof that God exists, so there is no reason to believe in him." Person 2 replies, "There is no proof that God exists, no. But neither is there proof that he doesn't. That doesn't mean he does, but you simply can't claim that God cannot possibly exist."

Here the original claim wasn't that God cannot possibly exist; it was that when the evidence is lacking, there is no reason one ought to believe in him. Countering the former isn't a valid way to respond to the latter.

Not a strawman:

Much like with the slippery slope fallacy, the reductio ad absurdum is sometimes mistaken for a strawman. The key is that the legitimate form of the argument shows why accepting the originally stated argument must also lead to accepting an argument that is obviously false or abhorrent, while a strawman simply presents the poor argument as if it were the original.

Debates and Words

Here's the thing about words. Language is a tool that we humans use to communicate ideas from one brain to another. Words are labels that stand for certain ideas, nothing more.

Knowing this, when you are in a debate and you realize that the opponent took some different meaning away from the words you said than you intended, do you:

a) Apologize for the confusion, reword what you said to make it more obvious what you actually meant, and avoid using the contested word(s) in the future of the debate
b) Argue that your definition of the words you used is better, more useful or more widely used than your opponent's
c) Whip out a dictionary entry for the word you used and quote it with the definition you meant bolded, triumphantly thinking that ha! that will show them!

If you chose c), you win the obnoxiousness prize. If you chose b), congratulations! You have derailed the argument into a squabble about the definition of words instead of the presumably meaningful topic you were actually trying to discuss. If you chose a), you are one of the maddeningly few who can run into this sort of incident and actually stay on topic.

Unless the debate is about how a word ought to be defined, the words a debater uses to present their point are not relevant to the argument. If the words you've used are failing to communicate your point, you should look for other words to do it, not defensively try to establish that it's your opponent's fault they didn't get it - even if it actually is. You can point out if your usage is popular since that can be legitimately useful information to someone who wasn't aware of it, but focus your efforts on restating your actual argument in a manner that clarifies your meaning.

In a similar vein, a post explaining that some word is not applicable to a situation your opponent has described is almost never meaningful compared to a post explaining the actual interesting difference between the situation they described and a situation to which you might apply the word.

Example: "Minorities can't be racist. Racism is prejudice + power."

Instead of this, which will lead to a horrible meaningless back-and-forth about the definition of the word "racism", it would be better to for instance reply explaining that even if minorities can be prejudiced against majorities, they do not have the situational power necessary to actually oppress the majority, so majorities being prejudiced against minorities is far more harmful than the reverse. Then if you want you can go on to say that in fact the sociological definition of racism is prejudice + power precisely to make this distinction, the debate continues to be meaningful, and everybody goes home wiser.

In general it is a fairly good tactic in a debate to mentally break your opponent's words down into relevant definitions in order to get a better idea of exactly what they're saying. If you're not sure precisely what they mean by something they say, ask them - if there is more than air behind their argument, they will know the idea they were actually trying to communicate with that word and be able to clarify it. And if there isn't more than air behind the argument, this technique can be tantalizingly revealing. For example, take the basic form of the infamous ontological argument for the existence of God:

Premise: God is absolutely perfect.
Premise: Nonexistence is less perfect than existence, so something absolutely perfect necessarily must exist.
Conclusion: Therefore, God exists.

What does this argument actually say? Well, absolute perfection is a pretty muddy concept, so let's replace it with some placeholder list of properties that constitute absolute perfection:

Premise: God has some properties x1, x2, ... xn, which constitute absolute perfection.
Premise: Nonexistence is less perfect than existence, so one of the properties x1, x2, ... xn must be existence.
Conclusion: Therefore, God exists.

For the sake of the argument, let's say property xn is "exists", and then we can actually implement the second premise directly into the first:

Premise: God has some properties x1, x2, ... xn-1 and exists.
Conclusion: Therefore, God exists.

Well, isn't that illuminating! The ontological argument actually just boils down to "God exists; therefore, God exists." There is no need to bother with sophisticated analogies to explain why it's invalid (though those also work out fine). Really, if you look at the original argument, the way the debate should actually go is like so:

Person 1: "First, God is absolutely perfect. Agreed?"
Person 2: "Okay..."
Person 1: "Well, nonexistence is less perfect than existence, so something that's absolutely perfect must exist."
Person 2: "Oh, so God isn't absolutely perfect, then. Sorry."
Person 1: "Ha! Too late! Tricked you! You already agreed that God is absolutely perfect! I win!"
Person 2: "...Um, that's not how it works. May I direct you to the 'On Reasoning and Arguing' sticky?"

This also applies to vague questions, which usually only lead to meaningless squabbling parading as a debate. For instance, imagine a debate about the question "Can video games be art?" The word "art" is an extremely vague, muddy, subjective concept, and asking this question in a public capacity will lead to a horrible tangle of disdainful artists and offended gamers all going "Well, art is..." However, if the question were just to be redefined in some meaningful manner before people start responding to it, it would turn out that actually almost the only disagreement to be found in the whole shouting match is on the definition of the word "art" and not actually on the merits of video games; in fact there isn't any proper debate to be had about it because the answers to all the potential questions you could pose about the merits of video games tend to be ridiculously straightforward:

"Can video games be aesthetically pleasing?" Yes.
"Can video games, at least in theory, tell complex, engaging stories?" Yes.
"Are video games a pure, noninteractive expression of a single artist's intent?" No.
...and so on.

Similarly, such "timeless" questions as "Which came first, the chicken or the egg?" and "If a tree falls in a forest and no one is around to hear it, does it make a sound?" are ridiculous and meaningless and easily exposed for how utterly dull they are the moment you actually define properly what you're asking about:

"If a tree falls in a forest and no one is around to hear it, are sound waves produced?" Yes.
"If a tree falls in a forest and no one is around to hear it, was there a sensation of sound created in some brain as a result?" No.

(Seriously, this may be the least interesting question ever posed by anyone.)

If you come across a debate thread where the question being asked is vague, don't reply to it except to redefine the question and possibly answer the new, meaningful question. Otherwise it's a waste of everyone's time.

More logical fallacies I should tackle? Other stuff to add? Reply and make suggestions!
Last edited:


Still loves Joltik, though!
Staff member
That claim is really a form of ad hominem. "You say that X, therefore Y. However, you don't actually believe Y. Therefore, your argument is invalid."

Hm. I should probably add that. It's illuminating.