You are viewing the historical archive of The Philosophy Forum.
For current discussions, visit the live forum.
Go to live forum

Calculus

Devans99 November 27, 2018 at 10:33 10800 views 42 comments
We write:

lim x->? 1/x = 0

But x tends to but never actually reaches infinity, so the right side never actually reaches zero, so surely its more correct to write:

lim x->? 1/x ~ 0

So whenever a limit is evaluated, it’s correct to use the approximately equals sign (~) rather than equals.

This could explain some of the rather peculiar results in calculus?

Comments (42)

Devans99 November 27, 2018 at 13:00 #231608
Actually might be better to write:

lim x->? 1/x ~> 0

That way, we preserve the information that the limit expression always evaluates to > 0.
Mentalusion November 27, 2018 at 14:48 #231644
I think the approximation is built into the concept of limit so you don't need the extra notation
ArguingWAristotleTiff November 27, 2018 at 14:49 #231645
Reply to Devans99 I so wish you and my youngest (19 year old) son could talk.
He is a Sophomore at an Aeronautical University, and he is on his third year, literally 36 straight months of Calculus.
Me? I gave up math in about 5th grade. When they started mixing in letters WITH numbers I knew I was in over my head.
Oh and Math and I are such good buddies that my idea of Purgatory? Is to wake up each day, be taught the fundamentals of Algebra all day long, begin to understand what is being taught only to go to sleep, wake up the next day with absolutely no retention of what was taught yesterday.
Actually that is what happened in High School so.... :rofl:
I admire and stand in awe of those who can work math wonders. :clap:
Devans99 November 27, 2018 at 14:51 #231648
Reply to Mentalusion The problem being if you take the result =0 and use it somewhere else, you have lost the information that it never actually =0, if you see what I mean. That could lead to an error.
SophistiCat November 27, 2018 at 15:53 #231666
Quoting Devans99
So whenever a limit is evaluated, it’s correct to use the approximately equals sign (~) rather than equals.


No. Look up the definition of the limit in any modern textbook or online reference. Do not assume that what a mathematical notation "looks like" is what it literally means.

Quoting Devans99
This could explain some of the rather peculiar results in calculus?


No.
Devans99 November 27, 2018 at 15:58 #231667
Quoting SophistiCat
No. Look up the definition of the limit in any modern textbook or online reference.


I know the textbook definition, the question the OP poses is: 'is the textbook definition correct?'.
Mentalusion November 27, 2018 at 16:13 #231671
Quoting Devans99
'is the textbook definition correct?'


It's a defined concept so the only way that it might be "incorrect" is if someone introduced a concept that fulfilled a similar role that had better computing or explanatory results w/re to describing how functions behave.
Mentalusion November 27, 2018 at 16:16 #231674
Quoting Devans99
The problem being if you take the result =0 and use it somewhere else, you have lost the information that it never actually =0, if you see what I mean. That could lead to an error.


If you could provide an example of such an error that would make discussion a little easier. I tend to think several centuries of people successfully using calculus for a variety of applications weighs against there being any errors short of theoretical technicalities.
SophistiCat November 27, 2018 at 16:33 #231685
Quoting Devans99
I know the textbook definition, the question the OP poses is: 'is the textbook definition correct?'.


The question in the OP indicates that you don't know or don't understand the textbook definition.
Devans99 November 27, 2018 at 16:35 #231687
Quoting Mentalusion
If you could provide an example of such an error that would make discussion a little easier


What might happen is someone evaluates a limit and then they take the result as precise when it's only approximate. If it then depends critically whether this value is >, < or equals 0, then anything that includes the evaluation of a limit is suspect.
Devans99 November 27, 2018 at 16:35 #231688
Quoting SophistiCat
The question in the OP indicates that you don't know or don't understand the textbook definition.


How so?
Mentalusion November 27, 2018 at 17:07 #231716
Quoting Devans99
What might happen is someone evaluates a limit and then they take the result as precise when it's only approximate


If someone did that, they wouldn't understand properly what a limit is and would be trying to get out of it something which it doesn't purport to be able to achieve.
Devans99 November 27, 2018 at 17:11 #231719
Quoting Mentalusion
If someone did that, they wouldn't understand properly what a limit is and would be trying to get out of it something which it doesn't purport to be able to achieve


So you agree we should write:
lim x->? 1/x ~ 0
rather than:
lim x->? 1/x = 0
?
Mentalusion November 27, 2018 at 17:12 #231721
Reply to Devans99

Quoting Mentalusion
I think the approximation is built into the concept of limit so you don't need the extra notation


What you're suggesting isn't necessarily wrong, it's just unnecessary and inefficient. Math already has enough goofy notation for people to keep track off. Why introduce symbols you don't need?

fdrake November 27, 2018 at 17:18 #231727
What's the epsilon which is too large for convergence as determined by the standard epsilon-N construction for the series (1/n)?
Devans99 November 27, 2018 at 17:20 #231730
Reply to fdrake No idea.
SophistiCat November 27, 2018 at 17:23 #231737
Reply to Devans99 Crack open any textbook on calculus. The concept of the limit is one of the first things that is covered in a typical calculus course, right after the basics of set theory. You need to understand mathematics before you can discuss philosophy of mathematics.

Quoting Devans99
No idea.


Yes, that's exactly the problem.
Devans99 November 27, 2018 at 17:27 #231743
Reply to SophistiCat I don't need to understand advanced calculus to point out an obvious notational error.
fdrake November 27, 2018 at 17:34 #231750
Reply to Devans99

Perhaps if you understood elementary calculus you would realise it is not an error.
Devans99 November 27, 2018 at 17:37 #231753
Reply to fdrake

lim x->? 1/x = 0
is an error.
For no value of x does 1/x take the value 0.
fdrake November 27, 2018 at 17:41 #231756
Reply to Devans99

Limit points don't have to be part of the set of convergent function evaluations towards a point, they just have to be in (metric) topological space underlying them.

Edit: in real analysis this roughly translates to 'sets don't have to contain their supremum or infimum', 0 is the infimum of any increasing sequence of x's which are plugged into f(x)=1/x - the sequence of function evaluations, that the limit exists and is 0 is ensured by the properties of monotonic decreasing sequences in a complete space.
Devans99 November 27, 2018 at 17:46 #231760
Reply to fdrake You are avoiding my argument.

1/x = 0
Is false for all x (undefined for 0).

So writing
lim x->? 1/x = 0
is definitely wrong
fdrake November 27, 2018 at 17:51 #231767
Reply to Devans99

I'm not actually avoiding your argument, wondering why the limit can be said to exist when it is in the closure of a set (a supremum or infimum) rather than simply in the set itself is one of the first concepts you have to understand or teach when you're teaching convergence of sequences. This is literally what the resources I linked for you explain.

Limit points deal with the more general case of convergence in topological spaces, the mathematics of complete metric spaces are what covers this specific example. Study the links and you'll learn something, continue to ignore them and you'll continue spouting rubbish.

Though yes, I agree that what you are talking about is a sticky point for understanding convergent sequences. This does not make it wrong, this means it requires care to grasp.
Devans99 November 27, 2018 at 17:57 #231771
I don't see anything in the links to justify writing equal when something is plainly not equal.
fdrake November 27, 2018 at 18:03 #231779
Reply to Devans99

This means you don't understand the distinction between a limit of a sequence and its elements. You would if you spent more time studying the links. Who knows, it might take more than 18 minutes of study to understand!

18 minutes was chosen because that's how long you took to survey the links. 18 minutes isn't even a complete introductory lecture on the topic, which usually has at least 2 university level math courses devoted to it, and even more involving it... That's hundreds of hours. You tried for 18 minutes.
Devans99 November 27, 2018 at 18:05 #231780
All I need is a grasp of the english language:

equal
adjective
the same in amount, number, or size

https://dictionary.cambridge.org/dictionary/english/equal

You are defending the indefensible. For something to be equal it has to be the same.
Moliere November 27, 2018 at 18:22 #231787
Reply to Devans99 By that reasoning I could criticize literally anything from a purported position of knowledge as long as the position were written in English. :D

Statistical mechanics? So long as it is in English I know what it means, and can criticize it from the almighty authority of the dictionary. The causes of World War II? I don't even need to read a book on the history, but just need to know English and any opinion put forward on the topic can be criticized from my well-rounded knowledge of the language.

Surely you don't want to double down on that principle.


A bit more seriously:

Sometimes disciplines use specialized language because it's more precise and easier to deal with the technical nature of a topic, and knowledge consists of more than knowing the language that it happens to be written in.
Devans99 November 27, 2018 at 18:24 #231790
Quoting Moliere
Sometimes disciplines use specialized language because it's more precise and easier to deal with the technical nature of a topic, and knowledge consists of more than knowing the language that it happens to be written in


But to write equals when something is not equals?

Mathematics should follow logic.
Moliere November 27, 2018 at 18:39 #231796
Reply to Devans99 Alright, I'll try one more time --

It's not the "equals" part that you're not understanding, it's the "limit" part.
Devans99 November 27, 2018 at 18:46 #231800
Reply to Moliere Maths has a responsibility to make logical sense.

1/x = 0
Is false for all x (undefined for 0).

So writing
lim x->? 1/x = 0
is definitely wrong

1/x is always greater than 0 so it could lead to an error downstream.

jorndoe November 27, 2018 at 20:25 #231843
Reply to Devans99,

[math]
\mathop {\lim} \limits_{x \to \infty} \frac{1}{x} = 0
[/math]

has a concise meaning, defined in terms of the universal (?) and existential (?) quantifiers for x ? 0:

[math]
\forall \epsilon \in \mathbb{R} > 0 \hspace{8pt} \exists \delta \in \mathbb{R} \hspace{8pt} \left| x \right| > \delta \Rightarrow \left| \frac{1}{x} \right| < \epsilon
[/math]

"We can always squeeze the fraction arbitrarily close to zero."

Check (?, ?)-definition of limit (Wikipedia)

The former is just a different, perhaps more intuitive way, of writing it.
Devans99 November 27, 2018 at 20:30 #231844
Quoting jorndoe
"We can always squeeze the fraction arbitrarily close to zero."


But arbitrarily close to zero is not zero and is never zero. The limit expression is always > zero.
Relativist November 27, 2018 at 21:05 #231859
The limit is synonymous to asymptote in the case in question, and an asymptote is never reached. That's why "=" is appropriate.
fdrake November 27, 2018 at 21:06 #231862
Reply to Devans99

You know this is accounted for in the definition right? This is how the definition works. 'For all epsilon greater than 0...'
jorndoe November 27, 2018 at 22:38 #231897
Quoting Devans99
But arbitrarily close to zero is not zero and is never zero.


So? That's not what it means. Check the link, or some of the other online resources.

As an aside, these sorts of things are used all the time in physics and other areas (derivatives, integration, etc).
I can only guess how much throughout the cool InSight project - congratz to the team.

How about fractals with an infinite circumference and a finite area? :)
andrewk November 28, 2018 at 01:08 #231921
I've moved this to the Lounge as there is no philosophy of mathematics in it. The thread is simply about a misunderstanding of the meaning of the limit operator, which the respondents have amply explained.
Devans99 November 28, 2018 at 08:51 #231994
Reply to jorndoe Are you saying the equal sign means 'arbitrary close' rather than 'equals'?
Devans99 November 28, 2018 at 10:49 #232016
Quoting jorndoe
How about fractals with an infinite circumference and a finite area?


Fractals have a potentially infinite circumference (potentially depending on how many calculations we do). Fractals never have an Actually Infinite circumference.
jorndoe November 28, 2018 at 17:28 #232114
Quoting Devans99
Are you saying the equal sign means 'arbitrary close' rather than 'equals'?


No. The lim, as defined, is zero.
Devans99 November 28, 2018 at 17:31 #232117
Quoting jorndoe
No. The lim, as defined, is zero.


Then it's defined wrong. There is no value of x for which 1/x = 0. Perhaps you are thinking of Actual Infinity?

Actual infinity, if it existed, would be a quantity greater than all other quantities, but:

There is no quantity X such that X > all other quantities because X +1 > X

Further, actual infinity does not follow common sense or mathematical rules:

oo + 1 = oo implies
1 = 0

Nothing in the real world can you add to whilst it remains unchanged. This logical absurdity implies infinity is not a mathematical quantity.
jorndoe November 28, 2018 at 17:36 #232121
Quoting Devans99
Then it's defined wrong. There is no value of x for which 1/x = 0.


Quoting jorndoe
So? That's not what it means. Check the link, or some of the other online resources.


@Devans99, it seems like you're not reading (or understanding) the mathematics and/or definitions. There are reasonably good online resources, though it may take a bit of reading if you're new to this stuff.
Devans99 November 28, 2018 at 17:39 #232122
Quoting jorndoe
it seems like you're not reading (or understanding) the mathematics and/or definitions.


The point is I do not agree with some of the definitions. How exactly is the axiom of infinity from set theory logical? I just disproved the existence of actual infinity twice above (I note you passed on both arguments).

Actual Infinity is just magic and plain impossible/illogical.