5

After Cantor, mathematicians realised that infinities can be graded by size (cardinalities). But just as we view the naive infinite as bigger than any finite number, is there an 'infinite' greater than any infinite?

But then the same pattern can carry on, in which case we are actually no nearer the absolute infinite than we were to begin with, that is Cantors mathematics of the infinite cannot do real justice to the idea of the Infinite.

To be more explicit, the Large Cardinal Axioms extend ZFC by postulating ever higher 'Cardinalities'. "However the observation that large cardinal axioms are linearly ordered by consistency strength is just that, an observation, not a theorem. (Without an accepted definition of large cardinal property, it is not subject to proof in the ordinary sense)...Saharon Shelah has asked, "[i]s there some theorem explaining this, or is our vision just more uniform than we realize?" Woodin, however, deduces this from the Ω-conjecture, the main unsolved problem of his Ω-logic."

Mozibur Ullah
  • 1
  • 14
  • 88
  • 234
  • 3
    "is there an 'infinite' greater than any infinite?" Mathematically, no. "Cantors mathematics of the infinite cannot do real justice to the idea of the infinite" Alternatively, one could hold that set theory changed our notion of infinity, in that it corrected a flawed folk understanding of it. – DBK Dec 01 '12 at 12:22
  • Just in passing: I think Cantor was religious, and held that the absolutely infinite was God. – Joseph Weissman Dec 01 '12 at 14:23
  • @DBK: I'm not so sure of that. It showed that the infinite have structure, but the absolute infinite, what you might call the folk infinite is still out of reach. Look at the proposed list of large cardinal axioms which extend the range that ZFC can reach. There is so far a finite list and they can be linearly ordered. – Mozibur Ullah Dec 01 '12 at 15:54
  • @Weissman: I think you're right, one could also infer from that he also believed the absolute infinite was also out of reach from mathematics. One can always go higher. So the folk idea of the infinite is corrected to the idea of the absolute infinite. Is there any connection with Hegels notion of the absolute infinite here? – Mozibur Ullah Dec 01 '12 at 15:56
  • 1
    Actually, what Cantor specifically called the "[Absolute Infinity](http://en.wikipedia.org/wiki/Absolute_Infinite)" (and went mystical about, as @JosephWeissman noted) can be treated mathematically. See also [my answer to another question here](http://philosophy.stackexchange.com/a/4276/1582). – DBK Dec 01 '12 at 17:37
  • @DBK: Sure you can introduce stuff like classes, but then what is the class of all classes? is it also a class? Or do we apply the same formalism to get a larger CLASS? Doesn't this formalism then carry on? I think the idea of large cardinal axioms to ZFC address issues like this, but am not sure. I don't see how the absolute infinity can be reached, and I can sympathise with why he went all mystical about it: Blake said it better - to see the world in a grain of sand, heaven in a wild flower, to hold infinity in your hand, and eternity in an hour. – Mozibur Ullah Dec 01 '12 at 17:53
  • 1
    @MoziburUllah: Yes, you can always "go up". Specifically, the collection of all classes is an **hyperclass**, but as far as I know, nothing interesting can be gained going above *hyperclass theory*. (Remember that mathematical logicians study these systems only insofar as they tell us something interesting about current mathematical problems and not simply for the "philosophical" sake of it.) … – DBK Dec 01 '12 at 18:25
  • @MoziburUllah: … As to "absolute infinite": AFAIK what ***Cantor*** meant by this was a very specific "size" of collection of sets, which he called "[inconsistent multiplicities](http://en.wikipedia.org/wiki/Limitation_of_size)". Now, you can certainly take "absolute infinite" to mean something else then Cantor, but you specifically asked about Cantor's notion. And it turns out that the latter can be treated consistently. As with many so called "paradoxes" in mathematics, they do do not involve contradictions *strictu sensu*, but simply anti-intuitive results. – DBK Dec 01 '12 at 18:27
  • @DBK: Well, grothendieck used what are called grothendieck universes to avoid the usage of classes in algebraic geometry, they're equivalent to adding an axiom stating that a strongly inaccessible cardinal exists, and this is only one of the smaller large cardinal axioms, so it appears that there is nothing to be gained by going beyond the notion of hyperclass, because its the wrong tack to take, we're just extending the idea in an automatic fashion. Large cardinal axioms, I think, are how Set theorists appear to address newer notions of size, and they appear to be highly interested in them. – Mozibur Ullah Dec 01 '12 at 18:36
  • @DBK:As I've alluded to above, they can be linearly arranged in terms of 'proof stength', so one can ask the question as to why this is... – Mozibur Ullah Dec 01 '12 at 18:37
  • @DBK: thanks for the clarification. I'm more interested in the idea of the 'absolute infinite' representing a point where no further 'climbing' can be done. I agree with you: you can always go up. – Mozibur Ullah Dec 01 '12 at 18:42
  • @DBK: just a little backup for my position, I asked a similar question in math overflow, and [Andreas Blass](http://mathoverflow.net/questions/115091/what-notion-captures-the-class-of-all-classes) backed up what I had to say, its not really an area of mathematics that I'm very familiar with. – Mozibur Ullah Dec 01 '12 at 19:14
  • I know, Andreas Blass has [specific philosophical view](http://mathoverflow.net/questions/71765/are-proper-classes-objects/71773#71773) on class theory. I am more more sympathetic to [Joel Hamkins' take](http://mathoverflow.net/questions/71765/are-proper-classes-objects/72386#72386) on the matter. However, I propose we should not abuse the comments for discussion any further. – DBK Dec 01 '12 at 19:28

4 Answers4

8

There is no "absolute answer" as to whether there is an absolute infinity, because whether or not you can have an absolute infinity is a function of what mathematical formalism you use.

  1. If you use ZF set theory (with or without the Axiom of Choice), you can for any infinite cardinal construct one which is larger by diagonalization. So even though not all infinities would necessarily be comparable, there would always be one cardinality which was demonstrably larger.

  2. If you use NBG set theory — a conservative extension ZFC — then there is an absolute infinity: the cardinality of proper classes, which in the usual construction of these things would be identified with the proper class of all von Neumann ordinals. Of course, this cardinality wouldn't exist as an element in the cardinality function using any conventional set- or class-theoretic construction of functions, because this would require the infinite cardinality to be an element of a class, which is impossible for proper classes by definition. But one could certainly make a statement, in NBG theory, about whether or not a class has a bijective mapping to the universal class, which would certainly indicate that it had the largest cardinality.

One criticism that you might make is that you could always make a model for NBG in some other set theory in which there are no proper classes. (This is essentially what you point out with inaccessible cardinals). Because one can always extend beyond any one model in which there is a largest cardinality, one can say that in the subject of mathematics — not in any one model, but in the discipline as a whole — there is no absolute infinite. But of course, if you do not fix any theory whatsoever, nothing whatsoever can be absolute. Non-standard models of the Peano Axioms subvert our usual notions of the natural numbers, so that we could not even say with any confidence that there is an absolute notion of a natural number. There isn't even an absolute notion of what it means for a cardinality to be infinite, if you allow the changing of models as you like. You can remedy this by fixing some class of set theories you wish to discuss, but this is just fixing some particular theory — a theory of set-theories — in the same way that selecting NBG or ZFC would be fixing a theory; your answer will depend on what theory you fix.

In order to give any sort of answer, you must fix your domain of discourse sufficiently to give an answer. So depending on how you want to fix the terms of your question, the answer is either

  • "yes there is" (as in NBG set theory, or a meta-theory which is like it),
  • "no there isn't" (as in ZFC set theory, or a meta-theory which is like it), or
  • "we can't answer the question" (because the terms haven't been defined well enough).
Niel de Beaudrap
  • 9,960
  • 4
  • 31
  • 52
5

John D. Barrow's The Infinite Book contains the following passages:

Cantor's most dramatic discovery was that infinities are not only uncountable, they are insuperable. He discovered that a never-ending ascending hierarchy of infinities must exist. There is no biggest of all that can contain them all.

also

Mathematical 'existence' meant only logical self-consistency and this neither required nor needed physical existence to complete it.

and (this after a mental breakdown that Cantor suffered in 1884)

Indeed, Cantor said, it would have diminished God's power had God only created finite numbers.

The author John D. Barrow is a Cambridge mathematician and physicist who has also written a play called Infinities where Cantor makes an appearance.

I sense in your questions a certain discomfort with the mathematical concept(s) of infinities: For instance if so, you may also find the chapters on Leopold Kronecker's finitism and on the replication paradox (Universes Where Nothing is Original in The Book of Universes by the same author) rewarding.

My own current view is that it is helpful to recall from mathematics that infinite is the inverse of zero. I understand there were many cultures in human history where the concept of zero was unknown or uncertain (try writing down zero as a Roman numeral, for instance), just like there were cultures where words for certain colors were absent (the adjective blue in Mediterranean Ancient Greece, of all places).

Nowadays the possible paradox "How can nothing be something?" does not usually strike us as dangerous when we think of zero. So maybe some sort of cultural shift will bring a similarly new understanding of (the idea of) the infinite in a linguistic, philosophical, or even mathematical sense.

Drux
  • 1,644
  • 1
  • 12
  • 26
  • I don't have any problems with Cantors Infinities, nor with Large Cardinals, going further than Cantor did. I also know about the Ordinals which is a finer-graining of Cantors Infinities. But my own naive concept of infinity: is that which cannot be gone beyond. It doesn't look like the mathematicians have come even close. – Mozibur Ullah Dec 09 '12 at 04:53
  • @Mozibur Ullah sorry for my misassumption. I looked up your profile and now realize that you must be already quite well versed in this field. Could you please indicate a bit more about what you refer to as _the_ absolute infinite (perhaps in terms of published work or articles on the web. Or is it just the naive concept?) This may also further improve the question. – Drux Dec 09 '12 at 06:50
  • 1
    @MoziburUllah, your idea of infinite is indeed naive, but that's natural because the concept goes beyond our current intuition. It takes time to develop an intuition for a new concept. There will be a time when everyone will be comfortable with the idea of infinite. – Cantor Dec 09 '12 at 12:25
  • @Drux: Please see my answer to this question:[Cantor & Infinities](http://philosophy.stackexchange.com/questions/4175/cantor-and-infinities/4385#4385). In short the naive idea of infinity, is a reflection of a philosophical concept 'the Absolute Infinite', and not actualizable in mathematics. No matter how far we go in formalising the infinite, there is always further to go, which is why Cantor himself became mystical about it. When I first read about Cantors taming of the infinite in my teens, I got all excited about it; when I realised he'd got nowhere near, I was disappointed and annoyed – Mozibur Ullah Dec 09 '12 at 13:13
  • @Drux: and irritated at the *hype* surrounding his achievements (this is not to belittle them). Writers on the Infinite should be more careful... – Mozibur Ullah Dec 09 '12 at 13:15
  • @MoziburUllah, why should there be an absolute infinite? – Cantor Dec 09 '12 at 13:56
  • @MoziburUllah thx, I'll look at the linked material. Meanwhile please consider than finite is (I suppose) rooted in the Latin word _limes_ for boundary line. Hence a literal reading of absolute infinite would amount to sg. like bounded unboundedness. Prior to reading the material I also wonder why there should be an absolute infinite indeed. – Drux Dec 09 '12 at 16:36
  • @Drux: limes to limit, I suppose. Bounded unboundness is a bit of a paradox, but then philosophy abounds in paradox, if it didn't it would be easy ;). – Mozibur Ullah Dec 09 '12 at 16:50
  • @MoriburUllah I should have typed _finis_ for border (typo -- thx for catching this), point about bounded unboundedness remains the same. Yes, paradoxes are fun and may be instructive, but why should there be an absolute infinite in the light e.g. of such an uneconomic beast (and say Occam's razor :) – Drux Dec 09 '12 at 17:08
  • 1
    Mathematics does not at all say that infinite is the inverse of zero. Infinite is undefined in terms of operations, whereas zero is the neutral element of addition. – RParadox Dec 13 '12 at 09:21
  • Well, then "inverse" as in $1 / 0 = \infty$ :) – Drux Dec 13 '12 at 09:39
2

Yes, if you want there to be.

There is a simple, boring thing you can do; given any ordered collection of things (such as cardinal numbers), you can take a brand new object (which we might call ∞), and make a new, ordered collection by adding ∞ to the original collection, extending the original ordering so that x<∞ for all x in the original collection.

In short, if having a quantity greater than everything else is something that's important to you, it's a triviality to have it.

You might argue that this is just a technicality; it has no substance to it. You'd be half right; it is a purely formal thing. But it only lacks substance if you don't give it any; if you have an actually useful notion of a "greatest magnitude", then this technical device gives you the mathematical structure that allows you to formalize those notions.

As an example of this sort of thing in practice, see the extended real numbers, which adds two elements to the real numbers, &pm;∞, and then sets up and generalizes enough mathematical structure to let you do calculus in a way that treats &pm;∞ as ordinary points of the extended real number line.

In the context of your question, ∞ obviously can't literally be a cardinal number. However, if we adopt the axiom of global choice (or the axiom of limitation of size), then we can give ∞ substance in the sense of viewing it as the (extended) cardinal number of the class of all sets.

0

Though the question has been adequately answered, I happened by coincidence to be reading the very quote where Cantor distinguishes between the "three contexts" in which "infinity" arises.

First, in the absolute, second in the "created world," and third as a mathematical magnitude, which he strictly distinguishes by calling the Transfinite. Here is his description of the Absolute Infinite:

"...the infinite in its most complete form, in a fully independent, other-worldly being, in Deo, where I call it the Absolute Infinite or simply Absolute."

Nelson Alexander
  • 13,331
  • 3
  • 28
  • 52