Is There A Name For This Fallacy?

[Dr Adequate]

Arguing that a precaution was unneccessary on (in effect) the grounds that it worked.

"Wearing seatbelts is compulsory, yet the number one cause of injury in car-crashes is seatbelts."

Unexamined alternative : if wearing seatbelts wasn't compulsory, the number one cause of injury would be flying head first through the windscreen.

"We spent all that money on fixing the Y2K bug, and yet nothing went wrong on 1st January 2000."

Unexamined alternative : if we hadn't spent the money lots of things would have gone wrong.

"The government says we have to vaccinate our children against the measles, but when did you ever hear of anyone dying of measles?"

Unexamined alternative : in countries without vaccination, "Don't count your children before the measles" is a proverb used when we would speak of chickens and eggs.

You get the picture.

[athon]

I haven't come across a name, but have debated against people who have used this logic on several occasions. Man, it's infuriating.

I like to think of it as the 'alternative action' fallacy. An informed action was taken to prevent a consequence from occuring; that the consequence did not occur contributes no evidence to support the notion that the action was ineffective.

Athon

[Dr B]

Excellent thread!!!!

Yes, I have encountered this one lots as well in debates with woo's. Not sure of a specific name - but it is a kind of reasoning in a circle where a positive outcome is used to question the lack of a negative one. In other words - it kind of answers itself.....

[Anacoluthon64]

Is this not a special class of the irrelevant conclusion, a.k.a. ignoratio elenchi (http://philosophy.lander.edu/logic/ignoratio.html), mode of argumentation? In itself, ignoratio elenchi is a subclass of the non sequitur.

'Luthon64

[MRC_Hans]

Couldn't we just label it ...


..nonsense?


Actually, non sequiteur is quite fitting, IMO, because it works both ways:

"Wearing a white hat while walking the streets of New York keeps the tigers away."

"But there are no tigers in the streets of NY!"

"See? It works!"

You cannot, from the incidence of injury with seatbelts, infer anything at all about the incidence of injury without seatbelts.

.... You can infer form other things that seatbelts probably reduce various injuries but promote injuries from seatbelts, of course.

Hans

[Mojo]

If there wasn't one before, I think there is now:Unexamined alternative.All three examples rely on ignoring the potential consequences of not taking the precaution.

[Soapy Sam]

"Fallacy of the unexamined alternative".
Sounds good to me... but should we not examine some alternatives?

[UrsulaV]

Hmm, since it sounds like a total failure to appreciate the success of these programs, I'd call it "argument from ingratitude." Or perhaps "the ingrate fallacy."

[Stellafane]

I don't know whether there's an "official" name for this line of faulty reasoning. but I personally refer to this as the Challenger Syndrome. "Why do we have to be so careful when we launch space shuttles? None of them has ever blown up before."

[bob_kark]

Hmm, since it sounds like a total failure to appreciate the success of these programs, I'd call it "argument from ingratitude." Or perhaps "the ingrate fallacy."

I'd support argument from lunacy or argument from sheer stupidity.

[Rustle]

ummm how about "Denying the Consequent"?

Usually it is the other way around, post hoc propter hoc. In you examples, there is a causation, not just a correlation. If

x = kids are vaccinated against measles
y = kids don't get measles,

it isn't just that x and y occur at the same time, but x causes y

[Wudang]

I tend to think of it as cherry-picking data. Eg missing data - the cause, number and severity of injuries before seatbelts were made compuslory.
cf "missing the big picture"

[Hellbound]

Of course, the seat belt thing is not entirely an accurate comparison, either.

Seatbelts may be the number one cause of injury, because in any collision at more than walking speed their likely to leave bruising. However, they are FAR from the number one cause of death. And death rates clearly show that seat belts save lives. I think this one is a bit different, because seatbelts aren't meant to prevent any sort of injury, but to prevent death and reduce serious injury. I'd almost classify this as a strawman, because they are arguing that seatbelts don't prevent injury, when their primary purpose is to prevent death.

[Pardalis]

Of course, the seat belt thing is not entirely an accurate comparison, either.

Seatbelts may be the number one cause of injury, because in any collision at more than walking speed their likely to leave bruising. However, they are FAR from the number one cause of death. And death rates clearly show that seat belts save lives. I think this one is a bit different, because seatbelts aren't meant to prevent any sort of injury, but to prevent death and reduce serious injury. I'd almost classify this as a strawman, because they are arguing that seatbelts don't prevent injury, when their primary purpose is to prevent death.

Excellent.

[atari24]

Not really on topic, but as someone who would have died without a seatbelt, I had no problem with the severe bruising and rashing(it looked like someone slit my throat) that my seatbelt gave me. The alternative would have been a facefull of another car or an intimate relationship with my steering wheel.

[derail over, back on topic]

[Anacoluthon64]

Here (http://www.fallacyfiles.org/index.html) is a most excellent web resource concerning fallacies of argumentation. It can provide hours of informative recreation.

'Luthon64

[RSLancastr]

I believe it is a sub-class of Argumentum ad Rectum.

[CFLarsen]

"We spent all that money on fixing the Y2K bug, and yet nothing went wrong on 1st January 2000."

Beacuse we spent all that money on fixing it.

[Pardalis]

I think Huntsman said it pretty well. These arguments are fallacies because they are based on flawed assumptions:

"We spent all that money on fixing the Y2K bug, and yet nothing went wrong on 1st January 2000."

It assumes that there had to be something wrong to happen in 2000.

"The government says we have to vaccinate our children against the measles, but when did you ever hear of anyone dying of measles?"

It assumes that vaccination's purpose is to prevent death, when in fact it is to prevent desease and it's spread.

[SteveGrenard]

It assumes that vaccination's purpose is to prevent death, when in fact it is to prevent desease and it's spread.

except for one minor detail: measles does cause death and yes, vaccination also prevents measles and it's spread also.



Measles remains an important cause of childhood mortality, especially in developing countries. In the joint Strategic Plan for Measles Mortality Reduction, 2001--2005, the World Health Organization (WHO) and the United Nations Children's Fund (UNICEF) targeted 45 priority countries* with high measles burden for implementation of a comprehensive strategy for accelerated and sustained measles mortality reduction (1). Components of this strategy include achieving high routine vaccination coverage (>90%) in every district and ensuring that all children receive a second opportunity for measles immunization. In May 2003, the World Health Assembly endorsed a resolution urging member countries to reduce deaths attributed to measles by half (compared with 1999 estimates) by the end of 2005 (2). This report updates progress toward this goal and summarizes recent recommendations on methods to estimate global measles mortality.

http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5408a4.htm


edit: and somewhat closer to home than the so-called third world, there is this current news account:



The fact that England is in the grip of the biggest measles outbreak since 1988 has been brought into sharp focus by the death of a 13-year-old boy.


Experts have repeated their warning to parents in Lancashire that many children in the county currently have no protection against the potentially fatal disease.

http://www.lep.co.uk/ViewArticle2.aspx?SectionID=73&ArticleID=1574179

[Hellbound]

I think Huntsman said it pretty well. These arguments are fallacies because they are based on flawed assumptions:

"We spent all that money on fixing the Y2K bug, and yet nothing went wrong on 1st January 2000."

It assumes that there had to be something wrong to happen in 2000.

"The government says we have to vaccinate our children against the measles, but when did you ever hear of anyone dying of measles?"

It assumes that vaccination's purpose is to prevent death, when in fact it is to prevent desease and it's spread.

I have to disagree, my argument applies only to the seat belt issue. There are more injuries, sure, I can believe that. But there are fewer debilitating injuries and fewer deaths, and these are the things seatbelts were intended for (specifically to reduce death by keeping the occupant from being thrown outside the vehicle, other concerns were secondary).

The other two aren't strawman fallacies, they are of a different nature. Not sure what to call them, though, but Denying the Consequent seems pretty close:

Vaccination for measles should reduce measles deaths.
We vaccinate.
Measles deaths are reduced.

The view from the first post:
There are few measles deaths.
Vaccination is used to prevent measles deaths.
Vaccination prevents few deaths.

I dunno, though, may be a Correlation-Causation fallacy too. Hard to pin down. Seems more like a combination of other fallacies.

Disclaimer: I am not trained in formal logic, except by self-training. I f I am full of $hit, please feel free to tell me so :)

[Anacoluthon64]

I think Huntsman said it pretty well. These arguments are fallacies because they are based on flawed assumptions:Er, I think you missed Dr Adequate's central point. Each of the examples given in the OP are indeed fallacies stemming from flawed assumptions of causality. However, they have a common essence in that the effectiveness of a precaution is mistaken for evidence that the precaution was unnecessary.

Another example, if such helps, that might assist in further clarifying the matter goes as follows:
This example is typed in bold italics to make it noticeable. Since you noticed it easily from the rest of the text, I needn't have bothered about the bold italic typeface.

'Luthon64

[Pardalis]

I think SteveGrenard is right, the flawed assumption in this case is that the measles don't cause death, wich in fact it does in children.

I'm not trained in formal logic either Hunts. ;) But your right, it might be a combination of many fallacies.

[Hellbound]

Er, I think you missed Dr Adequate's central point. Each of the examples given in the OP are indeed fallacies stemming from flawed assumptions of causality. However, they have a common essence in that the effectiveness of a precaution is mistaken for evidence that the precaution was unnecessary.

Another example, if such helps, that might assist in further clarifying the matter goes as follows:
This example is typed in bold italics to make it noticeable. Since you noticed it easily from the rest of the text, I needn't have bothered about the bold italic typeface.

'Luthon64

Well, I think I had a valid point regarding seatbelts, as preventing all injuries was not the design goal...preventing death was (with reducing injury severity as the means). There's a bit of strawman in that one, even if it wasn't the good Doc's central point :)

But you're right about the other two, and I addressed it a bit in my post just before yours.

[Pardalis]

Er, I think you missed Dr Adequate's central point. Each of the examples given in the OP are indeed fallacies stemming from flawed assumptions of causality. However, they have a common essence in that the effectiveness of a precaution is mistaken for evidence that the precaution was unnecessary.

Another example, if such helps, that might assist in further clarifying the matter goes as follows:
This example is typed in bold italics to make it noticeable. Since you noticed it easily from the rest of the text, I needn't have bothered about the bold italic typeface.

'Luthon64

Ah, now I get it. That's a very good point, very enlightening. :)

I guess I was focusing on the initial false assumptions.

The one about what the end result of the precaution should be:

"Seat belts should prevent injuries" (they prevent death)

and the false assumptions about what the threat is:

"the W2K bug" (wich happened to never have been a threat)
"the measles doesn't cause death" (wich does)

ETA: btw, how do we conjugate "measles", is it plural?

[Arkan_Wolfshade]

Denying the antecedent maybe?
P1: If A then B;
P2: Not A;
C: Therefore, not B.

P1: If people die from measles then vaccination should be required.
P2: People don't die from measles.
C: Therefore, vaccination should not be required.

[Anacoluthon64]

... but Denying the Consequent seems pretty close ...

Disclaimer: I am not trained in formal logic, except by self-training. I f I am full of $hit, please feel free to tell me so :)By your leave, then, in inferential logic, denying the consequent and affirming the antecedent are both valid forms of argumentation, while affirming the consequent and denying the antecedent are not. Symbolically,

P => Q (if P then Q) is our basic rule of inference.


Affirming the antecedent: P is true, therefore Q is true - valid
Denying the consequent: Q is false, therefore P is false - valid
Affirming the consequent: Q is true, therefore P is true - invalid
Denying the antecedent: P is false, therefore Q is false - invalid


Try this with P = "you can read" and Q = "you know the alphabet":


You can read therefore you know the alphabet.
You do not know the alphabet therefore you cannot read.
You know the alphabet therefore you can read.
You cannot read therefore you don't know the alphabet.


The last two do not necessarily follow because you needn't have mastered the art of reading despite knowing the alphabet.

Also, I am not at all contesting the validity of your strawman assessment of the seatbelt case - it's an astutely reasonable one. I merely wish to point out that classifying it as a strawman can only be justified upon providing information not contained in its premises as given, and that even without such further information the argument is not a valid one.

'Luthon64

[Anacoluthon64]

ETA: btw, how do we conjugate "measles", is it plural?Hmm, "conjugate" is something one usually does to verbs, so I don't quite follow your meaning. As is the case with, say, hives or acne, we don't distinguish any numerical amount of measles, so it will always be a case of "measles."

I'll grant you though that someone contracting a solitary "measle" and then "conjugating" it raises several intriguing possibilities... :D

'Luthon64

[Pardalis]

Hmm, "conjugate" is something one usually does to verbs, so I don't quite follow you meaning. As is the case with, say, hives or acne, we don't distinguish any numerical amount of measles, so it will always be a case of "measles."

I'll grant you though that someone contracting a solitary "measle" and then "conjugating" it raises several intriguing possibilities... :D

'Luthon64

LOL

I meant is the measles a plural word wich we would have to conjugated plurally a verb to it? (sorry, I'm French) :D

Do I say: the measles do cause death
or: the measles does cause death

[Hellbound]

By your leave, then, in inferential logic, denying the consequent and affirming the antecedent are both valid forms of argumentation, while affirming the consequent and denying the antecedent are not. Symbolically,

P => Q (if P then Q) is our basic rule of inference.


Affirming the antecedent: P is true, therefore Q is true - valid
Denying the consequent: Q is false, therefore P is false - valid
Affirming the consequent: Q is true, therefore P is true - invalid
Denying the antecedent: P is false, therefore Q is false - invalid


Try this with P = "you can read" and Q = "you know the alphabet":


You can read therefore you know the alphabet.
You do not know the alphabet therefore you cannot read.
You know the alphabet therefore you can read.
You cannot read therefore you don't know the alphabet.


The last two do not necessarily follow because you needn't have mastered the art of reading despite knowing the alphabet.

Also, I am not at all contesting the validity of your strawman assessment of the seatbelt case - it's an astutely reasonable one. I merely wish to point out that classifying it as a strawman can only be justified upon providing information not contained in its premises as given, and that even without such further information the argument is not a valid one.

'Luthon64

I think Denying the Antecedent was what I mean. I can't remember all the names ;P I knew the idea I was thinking of, but I can never keep the names straight :) Thanks!

Hopefully my example made it clearer what I was getting at :)

And true about the seatbelt, it's a combination of factors, but even within the argument itself it falls into the same trap I was trying to point out but named incorrectly (I knew I should have double-checked that).

So, could this be some sort of version of Denying the antecedent?

[Anacoluthon64]

(sorry, I'm French) :DThat's hardly your fault... :D

Do I say: the measles do cause death
or: the measles does cause deathNormally one would say, "measles cause death," i.e. treat "measles" as a plural.

'Luthon64

[Anacoluthon64]

:) Thanks!Don't mention it.


So, could this be some sort of version of Denying the antecedent?Superficially it seems to follow a similar pattern, but in terms of the P => Q phraseology, it is more like saying "Q is true, therefore P has no effect on Q" simply by hiding the "if P then Q" part.

Therefore the label obscuring the antecedent may be more fitting.

'Luthon64

[brodski]

LOL

I meant is the measles a plural word wich we would have to conjugated plurally a verb to it? (sorry, I'm French) :D

Do I say: the measles do cause death
or: the measles does cause death
"Measeles does", we usualy don't bother with the "the" either.

[tsg]

LOL

I meant is the measles a plural word wich we would have to conjugated plurally a verb to it? (sorry, I'm French) :D

Do I say: the measles do cause death
or: the measles does cause death

The general rule most USians follow is, if it sounds plural, it's plural; if it sounds singular, it's singular; regardless of whether or not the thing it's describing is one or many. Especially with group names.

The Yankees do, but New York does. Measles do, rubella does. Etc.

The Brits have their own rules which appear to be largely based on "do the opposite of the Americans".

[Soapy Sam]

No we doesn't!

[Arkan_Wolfshade]

No we doesn't!

Just remember, "y'ins" is the plural of "y'all".

[Rasmus]

"The government says we have to vaccinate our children against the measles, but when did you ever hear of anyone dying of measles?"

It assumes that vaccination's purpose is to prevent death, when in fact it is to prevent desease and it's spread.

I almost died of the measles; I was told that another patient in the ICU at the same time wasn't quite as lucky, and I don't know what happened to the third.

[brodski]

The Brits have their own rules which appear to be largely based on "do the opposite of the Americans". Oh, we've got some brilliant ones, for instance in officialdom "the Government" is plural, but "the government" is singular.

[rdaneel]

As a variation on the Y2K question. I wonder how often IT departments find themselves having to fight for their budgets because "The current system runs so smoothly, why do we need to spend more money on it?"

[Rustle]

By your leave, then, in inferential logic, denying the consequent and affirming the antecedent are both valid forms of argumentation, while affirming the consequent and denying the antecedent are not. Symbolically,

P => Q (if P then Q) is our basic rule of inference.


One of the original statements was
"The government says we have to vaccinate our children against the measles, but when did you ever hear of anyone dying of measles?"

Which I would restate as something like
P = The gov. mandates measles vacc. (true)
Q = deaths from measles are rare (true)

So set up correctly, P=>Q, if the gov vaccinates, then deaths from measles will be rare

HOWEVER

From the way the question is posed, most people wouldn't draw the right causality. Instead of "if P, then Q" they would think just "P, and Q". So, would "Denial of Causality" work?

They are denying that P causes Q; denying that vaccines are the cause of measles deaths being rare.

I'm not sure its a real logical fallacy, so much as just being stupid.

[Anacoluthon64]

So set up correctly, P=>Q, if the gov vaccinates, then deaths from measles will be rareThat's a valid way of arguing provided that "P => Q" is itself, in fact, true (one must be careful to distinguish an argument's validity from its correspondence with observation, i.e. what one usually calls "truth" - an argument may be valid but untrue if, e.g., one or more of its premisses are faulty). The rule of inference "P => Q" (P is the antecedent, Q the consequent), if true, allows us to say that the truth of Q follows necessarily once the truth of P has been established, and this mode is called affirming the antecedent, or Modus Ponens. None of the labels given the four modes of argument say anything about the other half, i.e. P or Q, of the rule of inference. The reason for this is that the nomenclature assumes that "P => Q" is itself true.


From the way the question is posed, most people wouldn't draw the right causality. Instead of "if P, then Q" they would think just "P, and Q". So, would "Denial of Causality" work?

They are denying that P causes Q; denying that vaccines are the cause of measles deaths being rare.

I'm not sure its a real logical fallacy, so much as just being stupid.In an earlier post, I gave the valid and invalid modes of argumentation that are recognised in this context. The particular form that we are dealing with here seems to conform to the scheme "Q is true, therefore P is false." To clarify this, I'll revise your P and Q somewhat:
P = "Measles vaccinations are effective"
Q = "Measles deaths will become rare"

For the seatbelt case:
P = "Seatbelts are effective"
Q = "their (compulsory) use will bring about a change in the number one cause of motoring injury and death"

For the Y2K example:
P = "Y2K compliance is effective"
Q = "No Y2K problems will be encountered on 01/01/2000"

Phrased thus, it is fairly obvious that each example complies with the "Q is true, therefore P is false" formula - in the seatbelt case the truth of Q is tacitly asserted by failing to give details of the prior number one cause. In terms of the usual nomenclature, the formula affirms the consequent ("Q is true"), but goes counter to that label's usual deduction by taking this as necessitating P's falsity rather than it's truth. It must also be remembered that if "P => Q" is true, showing Q to be true permits no conclusion to be drawn about P's truth value, i.e. P could be true or false. From this view, it is equally invalid to conclude either P's truth or its falsity, so it seems that the fallacy is

indeed a logical fallacy, and
an unusual case of affirming the consequent.


On the other hand, each P in the paraphrased examples I gave above includes the word "effective." The OP suggests that the logical fallacy under debate concerns only such propositions as examine the effectiveness of a given methodology. All of these arguments seem to be adequately described as affirming the consequent so as to deny the antecedent. Too cumbersome and possibly too general a name, I know, but it's a start. Also, if the issue of effectiveness can be taken as a distinguishing feature then it could be woven into the naming convention somehow.

'Luthon64

[tsg]

As a variation on the Y2K question. I wonder how often IT departments find themselves having to fight for their budgets because "The current system runs so smoothly, why do we need to spend more money on it?"

It happens in the HVAC industry all the time. If we maintain the equipment too well, they eventually decide they don't need a service contract because the equipment never fails.

[Pardalis]

One of the original statements was
"The government says we have to vaccinate our children against the measles, but when did you ever hear of anyone dying of measles?"

Which I would restate as something like
P = The gov. mandates measles vacc. (true)
Q = deaths from measles are rare (true)

So set up correctly, P=>Q, if the gov vaccinates, then deaths from measles will be rare


But doesn't "when did you ever hear of anyone dying of measles?" wrongly suggests that measles don't cause death at all?

[Cyphermage]

I haven't come across a name, but have debated against people who have used this logic on several occasions. Man, it's infuriating.

I like to think of it as the 'alternative action' fallacy. An informed action was taken to prevent a consequence from occuring; that the consequence did not occur contributes no evidence to support the notion that the action was ineffective.

Athon

The problem with this kind of thinking, is that you are only considering aggregate statistics, not individuals. Suppose a vaccine kills 1 in a million people you administer it to, but protects against a disease which will wipe out 1 in every 10,000 in the population. In the public interest, you make it mandatory for every one in a country of 100 million.

Approximately 100 people die from the disease, and for each one of them, they had only a 1 in 10,000 chance of dying from the illness. As they lie dying, they can legitimately argue, that had they been allowed to not have the vaccine, they would have had almost no chance at all of harm.

Statistics are nice, but if it's you, it's 100%.

Since vaccinating almost everyone in a society will also wipe out a disease, the optimal strategy for survival, for any disease you are unlikely to get soon, is for everyone BUT you to be vaccinated. You then have no risk of harm from the vaccine, and also no risk of harm from the disease, since everyone you bump into is immune.

The people who argue against seat belts, or baby-decapitating air bags, argue that it should be their choice to make, even if in the aggregate, larger numbers of people in society as a whole perish.

It's an interesting argument, and I'm not sure where I stand, since I dislike both mass deaths and mandatory nanny governments.

[politas]

Not really on topic, but as someone who would have died without a seatbelt,

False statement. You cannot know that you would have died without the seatbelt. However likely death would have been in the circumstances, it is not certain.

This sort of statement gets particularly annoying when used in situations where the chance of death if (safety device) were not used are difficult to calculate with any degree of accuracy.

[tsg]

or baby-decapitating air bags, argue that it should be their choice to make,

The solution to that is to not put the baby near the airbag.