Magnus Anderson wrote:
There might be no textbook that speaks of \(999\dotso\) and no mathematician who has ever worked with such a symbol but that does not mean that the meaning of that symbol cannot be logically deduced from existing definitions.
To be mathematically correct, I should mention that, completely off-topic from the present discussion, there is a mathematical system in which such expressions make sense; namely, the
p-adic numbers. The point being that you can define any mathematical notation you like, if you can make sense of it. Reading ahead, I would say that you have not succeeded in making sense of the notation as you're using it. But I'm a little confused, it was @Clearly real who was using that notation. Are you endorsing it as well?
Magnus Anderson wrote:In fact, I would say that it can.
\(999\dotso\) stands for \(9 \times 10^{\infty} + 9 \times 10^{\infty - 1} + 9 \times 10^{\infty - 2} + \dotso\).
You already have a problem here. \(9 \times 10^{\infty}\) is not defined. And \(\infty - 1\) is not defined. You'll have to nail those down first.
Magnus Anderson wrote:I read the entire thread
That is very impressive, my hat is off to you. Seriously. It's way more than I've done.
Magnus Anderson wrote:and I don't really think you ever had much of an argument.
Against infA and infB, you're probably right. As I recall, by the time I got to the thread James had already been using these terms for a long time in other threads. I made an effort but never tracked their definitions back to the beginning. James posted mostly on physics topics that were far from my interests.
I do recall that he tried to link them to the hyperreal numbers of nonstandard analysis; and that his understanding of those numbers was flawed and incorrect. I do remember correcting him on many aspects of that. But on the definitions of infA and infB, you're right, I never did get to the bottom of what he was talking about.
Magnus Anderson wrote:Perhaps you should present one now? Note that what you're doing here is merely asserting that you think that "infA" and "infB" are non-sensical, not well defined and not logically consistent.
You said what you believe but you did not say why you believe that.
Fair enough. Let's start fresh then, because the origin of those terms is not to be found in this thread, but is in other threads that I've never even seen. And James was nothing if not prolific. I'm willing to start anew, regretting that James isn't around to give his own interpretation. So I'll have to take your interpretation as official.
Observer wrote:If you don't show why they are not then you're right - your words will not mean anything.
Magnus Anderson wrote:I would say they are well defined.
Ok. Let's start here then.
Magnus Anderson wrote:"infA" stands for "the number of natural numbers". That's how James defined it and that's also how Observer defines in the post that you responded to.
Well, "number" is not defined in this context. All we really know about is the finite natural numbers 1, 2, 3, and so forth. Sometimes including 0 but that's not important here. So the "number" of fingers on your hand is 5. We all agree on that.
In order to talk about "how many" elements an infinite set has, we have to
define what we mean by that. The "number of natural numbers" is not defined till we say what we mean.
I will accept that you DEFINE infA as the number of natural numbers. This is the same as what mathematicians define as \(\aleph_0\), the cardinality of the set of natural numbers. Another name for this set is \(\omega\), the Greek lower-case letter omega. It's the same set as \(\aleph_0\), but regarded as an
ordered set.
The distinction is illustrated as follows. Suppose I reorder the natural number to put 5 at the very end, like this:
1, 2, 3, 4, 6, 7, 8, 9, ..., 5.
To do this we would just define a "funny <" relation so that a < b whenever neither are 5; but n < 5 for all n except for n = 5.
Then the set {1, 2, 3, 4, 6, 7, 8, 9, ..., 5} has the same cardinality as \(\aleph_0\); after all, it's exactly the same set of elements, just rearranged.
But as an ordinal, {1, 2, 3, 4, 6, 7, 8, 9, ..., 5} has both a smallest and largest element; whereas in the usual order (1, 2, 3, 4...}, \(\omega\) had a smallest but not a largest element. So we can change the ordinal number of a set without changing its cardinal.
In fact the ordered set {1, 2, 3, 4, 6, 7, 8, 9, ..., 5} has the ordinal number \(\omega + 1\).
I mention all this because math already has a well-developed theory of infinite quanties and orders. You can make up a different theory if you like, but you have work to do.
Magnus Anderson wrote:As for "infB", it means "the number of even natural numbers". It was also defined in the post that you responded to.
Ok well here you are going to have some trouble, and unlike your use of "number of natural numbers," which is a minor issue, here you have a major one.
The problem is that the cardinality of the set of evens is the same as the set of naturals. That's because:
* We DEFINE two sets to have the same cardinality if there exists a bijection between them.
* In this case there is a bijection between the naturals and events, given by \(f(n) = 2n\).
Note that I am not saying that the "number" is the same, because I haven't defined that. All I've said is that they have the same cardinality. This is important, because
all I've done is give a definition. It's not right, it's not wrong, it's not good, it's not bad. It's just a definition. We define two sets as having the same cardinality if there's a bijection between them. We note that there's a bijection between the evens and the naturals; hence by definition these two sets have the same cardinality.
Magnus Anderson wrote:But if these definitions aren't good enough for you, I'd suggest explaining what's unclear so that a better definition can be provided.
Ok glad to do so. The problem is this: Suppose you show me a set of size (may I use that word, or do you regard that as incorrect?) infB. I say to myself, Ok, that's a set having the same size as the even numbers.
But now suppose I had a set of children's blocks of size infB. So I have blocks labeled 2, 4, 6, 8, 10, ... and so forth. But then I grab some paint and from each block I erase the number written on it, and replace it with n/2. So now I have blocks labeled 1, 2, 3, 4, 5, 6, ...
Voilà! I see that I actually had a set of size infA after all! So in fact I just convinced myself that infA and infB denote sets of exactly the same cardinality.
This is exactly why I find the notion of infB superflous and not well thought out. Because it's no different than a set of size infA, which is to say a set of cardinality \(\aleph_0\), as it's called in standard math.
So I'm willing to say infA is coherent. It's just another name for \(\aleph_0\). But now infB is just another name for \(\aleph_0\) as well. I see no difference.
To get ahead of some common concerns and objections here, let me note a couple of things:
* In some sense we can say that the evens are "smaller" than the naturals, because the evens are a proper subset of the naturals. But this is a tricky business, because \(f(n) = 4n\) is a bijection between the naturals and the multiples of 4. And the multiples of 4 are a proper subset of the evens! So you have your "smaller" set of evens properly contaning an even smaller set of multiples of 4, yet the multiples of 4 provably have the same cardinality as the naturals. So again, if you like I won't use the word incoherent, which sounds pejorative. Rather, I'll just say that you still have some work to do to make your idea coherent.
* There is in fact a perfectly sensible way that you can say that the evens are exactly half the size of the naturals; and that is the idea of asymptotic density, or what Wiki calls
natural density.
To calculate natural density, we take the percentage of evens that occur in any initial segement of the naturals; and we note that as the number of natural increases without bound, the percentage approaches 1/2. This is very convenient in many applications. Likewise the natural density of the multiples of 3 is 1/3, and the natural density of the multiples of 4 is 1/4.
It's not perfect, for example the natural density of the primes is zero. But still, it's better than nothing.
So you see mathematicians CAN make such fine distinctions and use them as appropriate in a given context. But I don't think this is what you are referring to.
[quote='wtf"]There's an obvious bijection between the even whole numbers and the entire set of whole numbers[/quote]
Magnus Anderson wrote:I don't agree with that.
How can you not agree with the fact that THERE EXISTS a bijection between the naturals and the evens?
Of course it's true that there also exist functions between these sets that are not bijections; but so what?
You can't deny that there's at least one.
Magnus Anderson wrote:This suggests that the two sets are both equal and unequal (a logical contradiction.) But for some reason, you choose to ignore that, instead merely focusing on the fact there appears to be a bijective function between the two of them.
There IS a bijection between them. You can't deny that.
I perfectly well agree that there are functions between them that are not bijections. But remember we are just applying a DEFINITION. We say two sets have the "same cardinality" if there is a bijection between them. That's it.
This is a bit like a guy convicted of bank robbery, and identified in the newspaper as a bank robber. He sues the newspaper claiming that although it's true that he robbed a bank once last week, he DIDN'T rob a bank every OTHER day of the year.
It makes no difference. If you rob even a single bank, even if you lived all your life and never robbed any other banks, you are a bank robber. It's an existential ("there exists") and not a universal ("for all") quantification.
It's perfectly true that some functions between the evens and naturals are not bijections.
But at least one function is a bijection, and that's enough to satisfy the definition.
It's a definition. It's not right or wrong, good or bad. If a thing matches the definition, that's it. You don't have to like it, it's just a definition.
Magnus Anderson wrote:But the fact is that by using your very own logic, I can "show" that the two infinite sets are unqual in size.
How do you deal with that?
I hope I just did. The definition is that IF you ever robbed a bank, you are a bank robber. It matters not that on every other day of your life you did not rob a bank. If you ever robbed even a single bank once, you are defined as a bank robber.
If there exists a bijection between two sets, they are DEFINED to have the same cardinality. I said nothing else. You can't disagree that there is at least one bijection.
Of course I do agree with you that there are lots of other functions between the naturals and the evens that are not bijections. That's perfectly obvious. But so what? The definition says that if THERE EXISTS a bijection, the cardinalities are defined to be the same. The guy is a bank robber even though he's lived for 15,000 days and on 14,999 of them he did not rob a bank. The definition is that if THERE EXISTS a day on which he robbed a bank, he is defined as a bank robber. If THERE EXISTS a single lonely pathetic little bijection between two sets, those sets are DEFINED to have the same cardinality. You can't argue with a definition.