4 ms·
As posted below: (1.0+0.99999...)/2
by RiderOfGiraffes 17y ago
As posted below: (1.0+0.99999...)/2
- ErrantX 17y agocan you solve that equation? (the answer is no, again it's proveable but it is beyond me to do so. Someone else will have to jump in)
- frig 17y agoSee, what's so awesome about that answer is it allows the idiocy to keep rolling. Here goes: Proof: assume .9... < 1 (b/c lol I know this); then: X = (1+.9...)/2 => 2X = 1 + .9... => - 2X < 1 + 1 = 2 x 1 (b/c 1 > .9...) - 2 x .9... = .9... + .9... < 2X (b/c .9... < 1) => 2 x .9... < 2X < 2 x 1 => .9... < X < 1 => proof there's a # between .9... and 1, assuming of course .9... < 1. What's cute about this is it's employing a similar conceptual error: the above proof is literally circular -- it's proving something equivalent to '.9... < 1' (that there exists a distinct # between .9... and 1) by assuming '.9... < 1'; it's just different-looking enough to hide the circularity / equivalence if you don't want to see it (the analogy to the 'debate' as to whether .9... == 1 should be apparent).
- scott_s 17y agoThat's the actual meaning of the phrase "begging the question."
- ErrantX 17y agoIndeed - or to put it more simply it's just rearranging the equation, not solving it ;P
- scott_s 17y agoThat's not an equation, it's a number. There's nothing to solve.
- frig 17y agoAnd if you did 'solve' X = (.9... + 1)/2 using the standard 'algorithms' for doing it by hand you will wind up with X = .9... + .0...1 (more on this in a sec); you get there by noticing 1.9... / 2 looks like it ought to be .9..., but "of course" 2 x .9... == .9...8, ergo you can deduce you have .9... + a remainder of .0...1. If you don't believe .0...1 exists (infinite 0s followed by a 1) then this ought to convince you that .9... == 1 (as (x+y)/2 == x => x == y) but since we're (hypothetically) arguing with someone who believes .9... != 1 it's quite plausible that the person is also going to believe in .0...1 existing, also. What I suspect is actually going on (I've also met otherwise-savvy people who will debate this issue also) is that the human numerical-cognition system uses something like a crude form of nonstandard analysis internally; and, under this theory, you see this internal system (and its intuitions) leak out in some people smart enough to get that .9... really does have 'infinitely many' 9s but who also for whatever reason are both undereducated on this point of mathematics and generally self-assured. Briefly: you can extend the real #s with a new entity (call it delta) such that: - delta > 0 - if r is a nonzero # then delta < |r| - (for now) delta^2 = 0 (in effect: delta is an infinitesimal, it's nonzero but smaller than any other positive #, and delta^2 is 0 is mainly an algebraic convenience). Once you've extended things by including delta you can write extended #s in the form r + s x delta (eg: 1 - delta). If we re-define ".9..." to be 1 - delta, then the following all drop out of that: - .9... != 1 (1 - .9... == delta != 0) - .9... < 1, (1 - .9... == delta > 0) - and we even get: (1 + .9...)/2 = 1 - (delta/2) ==> (1 + .9...)/2 > .9... And more generally in my experience most of the spinning-in-circles around people with beliefs about .9... != 1 is that the people who vociferously believe .9... != 1 don't even know about non-standard analysis (or infinitesimals, etc.) and thus can't articulate what their intuition is telling them, but what their intuition is using is basically (a nonrigorous form of) analysis with infinitesimals.
- slackenerny 17y ago(for now) delta^2 = 0 Oh, that's another wonderful story — of Clifford algebras: http://en.wikipedia.org/wiki/Dual_number http://en.wikipedia.org/wiki/Dual_number , useful in automatic differentiation. As to your general point of what constitutes "internal representation", someone below tries to convince people that definitions involved with Reals are arbitrary and can be made differently thus forming Hyperreals. But no. One first have to construct Reals anyways, then one arrives at infinity, which is not a number, hyperreal or otherwise, but a consequence of numbering per se. Nothing ad hoc here. Hyperreals are that or another notation, but the main thinking is still the same, just paradoxes pop in different places, the only gain is increase in confusion as hyperreals, tuns out, do not conform to that "internal representation" in many more places than the poor 0.(9) .
- grk 17y agoYou can only calculate the limit of that, which is: lim (x->1) (1+x)/2 = 1