Post No.: 0596
Furrywisepuppy says:
Logical validity is critical if we wish to make sound, reasoned arguments. An introduction to logic was written in Post No.: 0066.
Fluffystealthkitten says:
We now want to run through some logical fallacies that we think are crucial for everyone to learn about. Furrywisepuppy and I have probably fell for a few ourselves without realising it at the times, but it’s all too easy to do!
Furrywisepuppy says:
There are tons of them so we couldn’t think of a better way than to just zip through them briefly. If you want to learn more about a particular fallacy then you can search online – Wikipedia has helped us out here too. But the main goal here is for us to be at least aware of them so that we hopefully won’t make or fall for them.
Fluffystealthkitten says:
Let’s go!
Furrywisepuppy says:
We’ll start with some common fallacies that attack the person rather than the argument.
Ad hominem, poisoning the well or appeal to motive – this is about attacking/favouring the messenger rather than their message. Whether this appears apparently justified or not (like if they’re a layperson or expert), we should really critique the content of a person’s arguments rather than their credentials. This even includes if they have a special role to express a view or verdict (like a pundit, judge or referee, although the latter two may have the final words anyway!)
Conflicts of interest, motives or circumstances may affect someone’s impartiality but this alone shouldn’t be a reason to reject/accept a claim. An expert’s view may be weightier but this doesn’t count as any proof of a claim itself. So even if we discover information that damages another person’s reputation, or they make one or two mistakes or lies, this alone shouldn’t discredit every single thing they’ll ever say. Basically, something is right/wrong not because of who said it or their historical record. We should seek more specific and harder evidence concerning a specific claim in question itself.
Fluffystealthkitten says:
Bulverism or psychogenetic fallacy – here, rather than proving that an argument in favour of an opinion is wrong, you instead assume that the opinion is wrong and then go on to explain why the other person held it.
For example, a salesperson claims that a cat feeder is dope, but because they’re trying to sell you it, you assume that the claim must be false. However it’s wrong to assume that if the psychological origin of an idea comes from a biased mind then the idea itself must necessarily be false.
Furrywisepuppy says:
Appeal to spite – dismissing a claim simply because one has a personal bias or grrrudge against the claimant, or exploiting people’s bitterness towards an opposing party.
Fluffystealthkitten says:
Tu quoque or appeal to hypocrisy – attempting to justify an action merely because someone else is doing it, or assuming that someone’s words are incorrect merely because they’re not practising what they preach themselves.
Also, ‘the pot calling the kettle black’ may be considered hypocritical, but this itself doesn’t make ‘the kettle therefore not black’. A journalist who attacks the media is presenting a media piece too, or a documentary that talks about how governments spread fear is spreading fear itself. But should this diminish the points made?
Furrywisepuppy says:
Next are some fallacies regarding blindly trusting sources. Note that these categorisations are incredibly loose. You’ll also notice that some of these fallacies overlap with others.
Argument from authority, anonymous authority, false authority or false attribution – relying on trusting who said something rather than analysing the contents of what was said. This is similar to the ad hominem fallacy but also includes not disclosing one’s source to allow a claim to be verified, or appealing to an irrelevant, unqualified, biased and/or fabricated source, using an ‘expert’ with dubious credentials, or using just one (cherry-picked) opinion to sell a product or idea.
Fluffystealthkitten says:
Appeal to accomplishment – this is where an assertion is deemed true/false, or the right to comment, is based on the accomplishments, or lack of, of who proposed it.
Furrywisepuppy says:
Anecdotal evidence – favouring a few firsthand stories over systematically-gathered data. Anecdotes might be unreliable because of, for instance, self-reporting biases or mistaking correlation with causation, like when I recovered from my cold after drinking some rooibos.
Fluffystealthkitten says:
Ad populum or groupthink – appealing to popular support as the basis for something being deemed true/false or as the sole basis for something being deemed morally good/bad, or simply blindly following one’s ingroup.
Something should be right/wrong not because of how many people believe in it but because of the evidence or considered arguments for/against it.
Furrywisepuppy says:
Appeal to common practice – diminishing the sense of one’s wrongdoing by claiming that it’s okay because others (allegedly) do it all of the time too or are even worse than oneself, or claiming that something must be right just because it’s commonly practised.
Fluffystealthkitten says:
Genetic fallacy or argument from antiquity – believing that a belief, tradition or precedent must be true or good because it has (apparently) been around for ages or has ‘always been that way’, or alternatively attacking the history or origin of a claim rather than its substance, current meaning or context. It’s also about assuming something’s current utility must be constrained by its historical utility.
Furrywisepuppy says:
How about some fallacies regarding generalisations or correlations.
Faulty or hasty generalisations – there are insufficient, weak or ambiguous premises or evidence to merit a conclusion but one has already made a hasty and/or sweeping generalisation, or drawn a broad conclusion from a tiny sample of data.
Fluffystealthkitten says:
Biased generalisations – generalising based on an unrepresentative sample to increase the strength of one’s argument.
Furrywisepuppy says:
Ecological fallacy – making strong inferences about the nature of specific individuals based solely upon the aggregate statistics collected for the group to which those individuals (supposedly) belong. For instance, the average furriness of dogs doesn’t mean every single dog must be that furry.
Fluffystealthkitten says:
Overwhelming exception – making an accurate generalisation but one that comes with so many caveats that eliminate so many individual cases that what remains is much less impressive than what the initial broad statement might’ve led others to believe!
Furrywisepuppy says:
Composition or division – assuming that something that’s true/false for a part of it must also be true/false for its whole, or vice-versa. Or assuming that the characteristics or beliefs of some individuals within a group must apply to that entire group overall, or vice-versa.
Fluffystealthkitten says:
Jumping to conclusions – drawing a quick conclusion before one has fairly considered all relevant (and easily available) evidence, or failing to consider the effect of random chance.
Furrywisepuppy says:
Post hoc ergo propter hoc – attributing an effect to a cause solely on the basis of it occurring earlier.
Fluffystealthkitten says:
Cum hoc ergo propter hoc – two events happening concurrently doesn’t necessarily mean one caused the other.
Furrywisepuppy says:
Confusion of correlation and causation – ignoring a common cause or third factor, or even confusing effect with cause. So just because the Nazis supported socialism or liked uniforms, it doesn’t necessarily mean these were the causally dangerous beliefs (which were more likely fascism and racism).
Fluffystealthkitten says:
Argument from final consequences – starting with the desirability of an effect then assuming a cause to be true if the effect is desirable, or false if undesirable.
Furrywisepuppy says:
Suppressed correlative – redefining a correlative so that one option encompasses the other thus rendering the encompassed option impossible. For instance, redefining everything as being subjective hence believing that nothing can ever be objective.
Fluffystealthkitten says:
Affirming the consequent or denying the antecedent – this is when the antecedent in an indicative conditional is claimed as true because the consequent is true i.e. if A, then B; B, therefore A (affirming the consequent). Or when the consequent in an indicative conditional is claimed as false because the antecedent is false i.e. if A, then B; not A, therefore not B (denying the antecedent). For example, if I hid your keys, then I am a naughty cat; I am a naughty cat, therefore I hid your keys. But I obviously haven’t done such a thing… yet.
It’s about neglecting that there could be multiple possible explanations for an outcome even when one possible explanation has been established, or assuming that there can only be one cause for the observation one is making, or ignoring a possible overall factor when focusing on a very specific factor.
Furrywisepuppy says:
Association fallacy or guilt/honour by association – with guilt, it’s saying that something’s wrong or discredited solely because it’s proximal or apparently related to something that’s wrong or someone who’s undesirable. And vice-versa with honour. It could be believing that A is a B, A is also a C; therefore, all Bs are Cs. Or A is doing C, B is also doing C; therefore, A is working with/for/is B, as if you need to be a transgender person to support transgender rights for instance.
Fluffystealthkitten says:
Undistributed middle – assuming that just because two things share a property, it means they’re the same or the same thing.
Furrywisepuppy says:
Chronological snobbery – this is where a thesis is deemed to be incorrect because it was commonly held when something else, which proved to be clearly incorrect, was also commonly held.
Fluffystealthkitten says:
Short-term versus long-term fallacy – assuming that a current trend has remained constant throughout history so will eternally continue this way, even though no evidence suggests such an extrapolation.
Furrywisepuppy says:
Argument from design – saying that if something has all the properties required for it to work in a certain way then it must’ve been specifically designed that way.
Fluffystealthkitten says:
Now for some fallacies concerning statistics.
Observational selection or confirmation bias – cherry-picking only confirming evidence whilst ignoring disconfirming evidence. An extremely common bias.
Furrywisepuppy says:
Statistics of small numbers – reporting changes in absolute number versus relative percentage terms depending on whether one wants to make it seem like a small or large change. For example, if something has increased from 2 to 4, one could present the increase as just an extra 2 or as a doubling or 100% increase, depending on one’s agenda.
Fluffystealthkitten says:
Base rate fallacy – focusing on specific information or the conditional probabilities pertaining to a particular case whilst ignoring relevant base rate, general information or prior probabilities, as if that individual case no longer falls under the more general group too.
Furrywisepuppy says:
Conjunction fallacy – assuming that an outcome that simultaneously satisfies multiple conditions is more probable than an outcome that satisfies fewer of them. There logically cannot be a greater chance that a creature satisfies both being a floof and having a tail than just satisfies having tail.
Fluffystealthkitten says:
Misunderstanding the nature of statistics – the ignorance of central statistical assumptions such as mean or median, or the definition of metrics, for instance.
Furrywisepuppy says:
Appeal to probability – assuming that because something could happen then it will inevitably happen, or already taking something as given because it would most probably be the case.
Fluffystealthkitten says:
Regression fallacy – ascribing a cause when an event is actually due to natural fluctuations and the regression to the mean. So a relatively worse result than average is more likely to follow after a relatively better result than average, and vice-versa. And there’ll be little to read into if a long streak of above-expected results is followed by a long streak of below-expected results, and vice-versa.
Furrywisepuppy says:
Gambler’s fallacy – assuming that historical results will affect future results even though each trial is independent from each other. The results of one lottery machine draw won’t affect the results of another draw, thus drawing the number 37 last week won’t alter the odds of the number 37 appearing this week whatsoever.
Fluffystealthkitten says:
…This is going to take much longer than we thought.
Furrywisepuppy says:
Hmm, we’ll continue with more another time. Woof!
Fluffystealthkitten says:
Meow!
Comment on this post by replying to this tweet:
