Hacker News new | past | comments | ask | show | jobs | submit login

The only thing that shows is that reduction to absurds leads to absurdities.



Showing that a claim implies an absurdity is what the “reduction to the absurd” proof technique is, yes.

Generally, “implies a falsehood” is considered to be equivalent to “is false”.

What’s your point?


My point is - and it is sad that this needs elaborating - that what works well for physics and maths problems does not necessarily work well for societal problems.

Changing a company from the inside is something that can only be done by those that are on the inside, hence my qualification that you should only join those companies that you can still stomach and then push as hard as you can in the right direction, rather than to join the very worst organization that you can find in the world, and one that you - hopefully - can not stomach.


It is an error to think that reductio only applies to “physics and math problems”. Logic is logic, it applies to all truths. (Though one has to be careful with things stated informally, to ensure that one is working with the claims that were intended. There is a danger of parsing an informal statement as if it were a formal statement, and as a result interpreting it to mean something other than what was meant.

Also, there may be a risk of people being more likely to derive in invalid ways when attempting the reductio, like making some hyperbolic statements, exaggerating the implications of a statement, and treating those exaggerations as if they actually logically follow. But this doesn’t make reductio an invalid technique, just one that people can fail to use properly.)

Like I said in the other comment, the objection should be that the derivation of the absurd from the premise is invalid. For example, pointing to the “that you can stomach” part of your claim, and saying that the derivation was invalid in neglecting that part of the statement.


That appeal to absurdity is simply that, absurd.

Talking about changing ISIS from the inside as the IT administrator doesn't convince anybody that it's impossible to change an organization from within. It's a poor argument that might win Internet points for being a spicy hot take, especially on Twitter, but it does little to debunk the idea that, eg, an engineer could ethically join Facebook with the aims of elevating the company up, through working on the team to detect Russian trolling rings/Anti-vaxxers/MLM marketers.


If someone argues “X implies <you should do something stupid and bad>. It is absurd to conclude <you should do something stupid and bad>, so therefore, by reductio ad absurdum, X is false.” your refutation should not be “actually, reductio ad absurdum is invalid.”, but rather “It is not the case that <x> implies <you should do something stupid and bad>”.

Perhaps their derivation of the absurd conclusion from the claim <X> was based on a misunderstanding of the claim <X>, or perhaps their derivation of the absurd conclusion from it was based on some fallacy, or some false assumption.

But reductio is a valid logical step, so long as we have the validity of the argument that the absurd conclusion follows from the premise.


Hmm, thank you for the response.

In this specific case, X is:

"Personally, my take is the opposite: if you are an ethical person find the very worst company that you are still willing to work for and try to change them from the inside"

with the response, Y: "join ISIS' IT department"

What is the refutation of this particular "X implies Y" you would have used instead? Is there one that doesn't require refining X and instead attacks the Y? X is already verbose, and adding additional verbiage does not help the case (IMO)


Hmm, thank you for the response.

In this specific case, X is:

"Personally, my take is the opposite: if you are an ethical person find the very worst company that you are still willing to work for and try to change them from the inside"

with the response, Y: "join ISIS' IT department"

What is the refutation of this particular "X implies Y" you would have used instead? Is there one that doesn't require refining X and instead attacks the Y? X is already




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: