I tried to send in a fallacy for the Logical Argument FAQ for alt.atheism.
Only thing is, I can’t phrase it very well, nor do I know any good name for it.
Anyway, the fallacy (at least I _hope_ a fallacy) is as follows:
"You are doing X. Doing X causes lots of misery. Therefore you are causing
lots of misery."
"You want to do X. Doing X causes lots of misery. Therefore, you want to cause
lots of misery."
The first is a valid argument, and the second is not. I believe there might
even be some special word for how "want" acts in sentences like these. Can
someone give me some idea how to explain this well enough that I can submit it?
—
"On the first day after Christmas my truelove served to me… Leftover Turkey!
On the second day after Christmas my truelove served to me… Turkey Casserole
that she made from Leftover Turkey.
[days 3-4 deleted] … Flaming Turkey Wings! …
– Pizza Hut commercial (and M*tlu/A*gic bait)
Ken Arromdee (arrom…@jyusenkyou.cs.jhu.edu)
In article <C78p22….@blaze.cs.jhu.edu> arrom…@jyusenkyou.cs.jhu.edu (Ken Arromdee) writes:
>I tried to send in a fallacy for the Logical Argument FAQ for alt.atheism.
>Only thing is, I can’t phrase it very well, nor do I know any good name for it.
>Anyway, the fallacy (at least I _hope_ a fallacy) is as follows:
>"You are doing X. Doing X causes lots of misery. Therefore you are causing
>lots of misery."
>"You want to do X. Doing X causes lots of misery. Therefore, you want to cause
>lots of misery."
>The first is a valid argument, and the second is not. I believe there might
>even be some special word for how "want" acts in sentences like these. Can
>someone give me some idea how to explain this well enough that I can submit it?
The first idea you’re trying to express is that we think of cause as a
transitive relation on events. So if A causes B, and B causes C, we
say A causes C. Causal chains can extend very far — see the movie
"Back to the Future" for a dramatic illustration.
But we don’t think of volition acting transitively through causation.
So if (A wants B) and (B causes C), we cannot conclude (A wants C).
One can think of myriad examples of unintended consequences of
wanted events.
– Paul
—
———————————————————————–
| Paul Steckler | Desperately seeking clever |
| st…@ccs.neu.edu | new .signature slogan |
———————————————————————–
In article <C78p22….@blaze.cs.jhu.edu>
arrom…@jyusenkyou.cs.jhu.edu (Ken Arromdee) writes:
>I tried to send in a fallacy for the Logical Argument FAQ for alt.atheism.
>Only thing is, I can’t phrase it very well, nor do I know any good name for it.
>Anyway, the fallacy (at least I _hope_ a fallacy) is as follows:
>"You are doing X. Doing X causes lots of misery. Therefore you are causing
>lots of misery."
>"You want to do X. Doing X causes lots of misery. Therefore, you want to cause
>lots of misery."
>The first is a valid argument, and the second is not. I believe there might
>even be some special word for how "want" acts in sentences like these. Can
>someone give me some idea how to explain this well enough that I can submit it?
In the second sentence, the verb "want" is an intentional operator,
which introduces an oblique, or indirect contect. The situation is
complicated by the introduction of causality, which can be omitted
without changing the gist of your example. Thus if, unbeknownst to
you, X is Y, the following is not a valid inference:
"You want to do X. Doing X is the same as doing Y. Therefore, you
want to do Y."
In the work of Anscombe and Davidson, all actions are said to be
intentional only under a certain description; note that an exception
would have to be made for an omniscient agent, who presumably acts in
such a way as to be aware of all possible descriptions under which his
action will fall. See on this _Intention_ and _Essays on Actions and
Events_. Also see Leonard Linsky’s _Oblique Contexts_, for a good
introduction to the logical issues involved. The modern background of
this conundrum is to be found in Frege and Russell, Church and Quine.
But similar questions have been addressed by the Greeks.
cordially,
mikhail zel…@husc.harvard.edu
"Le cul des femmes est monotone comme l’esprit des hommes."
In article <C78p22….@blaze.cs.jhu.edu> arrom…@jyusenkyou.cs.jhu.edu (Ken Arromdee) writes:
>I tried to send in a fallacy for the Logical Argument FAQ for alt.atheism.
>Only thing is, I can’t phrase it very well, nor do I know any good name for it.
>Anyway, the fallacy (at least I _hope_ a fallacy) is as follows:
>"You are doing X. Doing X causes lots of misery. Therefore you are causing
>lots of misery."
>"You want to do X. Doing X causes lots of misery. Therefore, you want to cause
>lots of misery."
>The first is a valid argument, and the second is not. I believe there might
>even be some special word for how "want" acts in sentences like these. Can
>someone give me some idea how to explain this well enough that I can submit it?
Another example of the same type of argument that may be a little
easier to explain is:
John likes sprinting. Sprinting is painful. Therefore John like pain.
This argument is in the form of an inductive proof. Assuming that the
two premises are true, it may be true that John likes pain, but it
cannot be proven unless additional information is given.
The problem with the above argument is made more clear if you
rearrange it so that it is in the form, if A -> B, A, therefore B
(i.e. a deductive argument). Doing so gives you something like:
If you like sprinting, then you like pain. John likes sprinting.
Therefore John likes pain.
The obvious problem here is that the first premise may or may not be
true. If it is true, then the conclusion follows. If not, then the
conclusion is false.
In article <C78p22….@blaze.cs.jhu.edu> arrom…@jyusenkyou.cs.jhu.edu (Ken Arromdee) writes:
> I tried to send in a fallacy for the Logical Argument FAQ for alt.atheism.
> Only thing is, I can’t phrase it very well, nor do I know any good name
> for it.
> Anyway, the fallacy (at least I _hope_ a fallacy) is as follows:
> "You are doing X. Doing X causes lots of misery. Therefore you are causing
> lots of misery."
> "You want to do X. Doing X causes lots of misery. Therefore, you want to
> cause lots of misery."
> The first is a valid argument, and the second is not.
I don’t think the first is necessarily a valid argument; it depends whether
"Doing X causes lots of misery" means "Doing X *invariably* causes lots of
misery" or "Doing X is statistically correlated with a subsequent large
increase in misery".
The argument is only valid under the first interpretation, but I think that
"Doing X causes lots of misery" is more often used in the second sense, e.g.
"Adultery causes broken marriages". While it may be true on average, it isn’t
appropriate to generalise it to every instance. Thus "If you commit adultery,
your marriage will break down" is in general false.
I guess, on perusing the a.a logic FAQ, that this is the Fallacy of Accident.
Quoth the FAQ:
The Fallacy of Accident is committed when a general rule is applied to a
particular case whose "accidental" circumstances mean that the rule is
inapplicable. It is the error made when one goes from the general to the
specific.
However, it’s also pretty damn close to _dicto simpliciter_ or Sweeping
Generalisation. From the FAQ again:
A sweeping generalization occurs when a general rule is applied to a
particular situation in which the features of that particular situation
render the rule inapplicable.
In this case, it’s the bogus inference of an inviolate rule from a statistical
`rule’, i.e. a correlation. I’m not sure to which fallacy this example is
closest, since both appear to describe the example approximately. I wonder
if this is in fact a more general statistical fallacy, since it’s fallacious
to make such a generalisation *irrespective* of the
"particular case whose "accidental" circumstances…" (Accident)
or the
"particular situation in which the features…." (_dicto simpliciter_).
Input welcome. Maybe we need a Statistical Fallacies FAQ as well.
[second argument again]:
> "You want to do X. Doing X causes lots of misery. Therefore, you want to
> cause lots of misery."
(I’ll assume that "Doing X causes lots of misery" is inviolate rather than
a statistical observation, to factor out the possible objection above.)
I can think of two different ways of slicing this one.
The obvious problem is that doing X may cause any number of effects besides
lots of misery. We can assume, I think(?), that the realisation of some subset
of these effects is why the agent wants to do X (the effects can include
personal gratification alone — he may want to do X purely for its own sake,
in which case this is the only effect which he wishes to realise).
We can’t know *which* effects he wishes to realise unless he tells us, and
there lies the fallacy, I suspect.
This looks like _cum hoc ergo propter hoc_ per the FAQ:
It asserts that because two events occur together, they must be
causally related, and leaves no room for other factors that may
be the cause(s) of the events.
In this case, it’s `…other factors [effects of doing X, including pure
gratification] that may be the cause(s) of the events [the observation
that the agent wants to do X].’
The second way which occurs to me is to treat the argument as a bogus implied
identity; the argument would be valid if the middle sentence were
"Doing_X *is* causing_lots_of_misery".
But "Doing X causes lots of misery" is not an identity, and in this analysis,
the fallacy lies in the bogus implication that it is.
> Ken Arromdee (arrom…@jyusenkyou.cs.jhu.edu)
Cheers
Simon
—
Simon Clippingdale si…@dcs.warwick.ac.uk
Department of Computer Science Tel (+44) 203 523296
University of Warwick FAX (+44) 203 525714
Coventry CV4 7AL, U.K.
Oh well, I goofed. My example was a fallacy, but was compounded by other
problems, most of which can be removed by choosing a better example, one
where the connection between the two parts is more straightforward.
The best reply I got was the following (Carl Brock Sides):
>1) Mary believes that Plato was the greatest philosopher of antiquity.
>2) Plato was the teacher of Aristotle.
>3) Therefore, Mary believes that the teacher of Aristotle was the greatest
> philosopher of antiquity.
>This argument, when the conclusion is given what is called a _de dicto_
>reading, is not valid, since it could be the case that Mary is under a
>misapprehension about who the teacher of Aristotle was.
Anyway, is there some _name_ for this fallacy? ("De Dicto Fallacy"?) Please
respond by email this time (something which I left out last time)….
—
"On the first day after Christmas my truelove served to me… Leftover Turkey!
On the second day after Christmas my truelove served to me… Turkey Casserole
that she made from Leftover Turkey.
[days 3-4 deleted] … Flaming Turkey Wings! …
– Pizza Hut commercial (and M*tlu/A*gic bait)
Ken Arromdee (arrom…@jyusenkyou.cs.jhu.edu)
In article K…@blaze.cs.jhu.edu, arrom…@jyusenkyou.cs.jhu.edu (Ken Arromdee) writes:
> The best reply I got was the following (Carl Brock Sides):
> >1) Mary believes that Plato was the greatest philosopher of antiquity.
> >2) Plato was the teacher of Aristotle.
> >3) Therefore, Mary believes that the teacher of Aristotle was the greatest
> > philosopher of antiquity.
> >This argument, when the conclusion is given what is called a _de dicto_
> >reading, is not valid, since it could be the case that Mary is under a
> >misapprehension about who the teacher of Aristotle was.
> Anyway, is there some _name_ for this fallacy? ("De Dicto Fallacy"?) Please
> respond by email this time (something which I left out last time)….
It doesn’t look like any form of fallacy to me. It looks like a broken syllogism.
Part three does not derive from 1 and 2. The missing part would be:
2a. Mary believes teachers are better in their field of endeavour
than their students.
Anyone proposing the above as a syllogism has allowed assumptions to creep in.
—
"As far as the laws of mathematics refer to reality, they are not certain;
as far as they are certain, they do not refer to reality."
Albert Einstein Alan Pope <alan.p…@Eng.Sun.COM>