Gödel's first incompleteness theorem, perhaps the single most celebrated result in mathematical logic, states that:
For any consistent formal, recursively enumerable theory that proves basic arithmetical truths, an arithmetical statement that is true, but not provable in the theory, can be constructed.1 That is, any effectively generated theory capable of expressing elementary arithmetic cannot be both consistent and complete.
but godel does not tell us what makes a statement true
until he does we cant tell if robinson arithmetic is incomplete
we need to know what makes a statement true in order to see that it cant be proven
godel could have said there are gibbly statemets which cant be proven
but untill we know what makes a statement gibbly we have no way of knowing what his theorem is talking about
so untill he tells us what makes a statement true robinson arthmetic cant be said to be incomplete
On Apr 12, 3:41 am, "elsiemelsi" <cyprin...@nosam.yahoo.com> wrote:
> but godel does not tell us what makes a statement true
> until he does we cant tell if robinson arithmetic is incomplete
Of course we can.
> we need to know what makes a > statement true in order to see that it cant > be proven
No, we don't.
> godel could have said there are > gibbly statemets which cant be proven
Right.
> but untill we know what makes a statement gibbly > we have no way of knowing > what his theorem is talking about
Wrong. It DOES NOT MATTER what OTHER attributes the statement may or may not have. If the theory proves NEITHER the statement NOR its denial THEN the theory is incomplete. ONE OR THE OTHER of the statement and its denial MUST be true, but we DON'T NEED to know WHICH, or what "truth" "is", in order to know to know that theory is incomplete.
But since you have never studied this stuff, you didn't know that.
> so untill he tells us what makes a > statement true robinson arthmetic cant > be said to be incomplete
This whole endeavor DOES NOT EVEN CARE about what's TRUE: it CARES about what's PROVABLE. And we all HAVE told you that BEING AN AXIOM makes a statement provable, AS does being INFERRABLE from the axioms ACCORDING to the INFERENCE RULES of the LOGIC.
"Truth" in the arithmetical sense that the endeavor is dealing with IS something that you could be or possibly even have been told about, if not by Godel then by YOUR ELEMENTARY SCHOOL math teachers. The theories have a "standard" model, where 0 is zero, s(0) or (0+1) is 1, s(s(0)) or 1+1 is two, ad inf. All those numbers (and many different numeration systems for them, in addition to the two mentioned HERE) were ALREADY OUT THERE BEFORE Godel started proving things, and EVERYbody was legitimately expected to know what made 43*69=3426 true (or false), WITHOUT any help from Godel or 1st-order logic. So the point is, WE DO ALREADY KNOW what "makes" these statements true or false. The ones involving "Every" natural number or "Some" natural number might require an infinitary process for confirmation or refutation, but the point is, we KNOW it is THAT process and NOT some other, so we DO know what "makes" those statements true, even when we can't practically perform the investigation.
> but untill we know what makes a statement gibbly > we have no way of knowing > what his theorem is talking about
Wrong. It DOES NOT MATTER what OTHER attributes the statement may or may not have. If the theory proves NEITHER the statement NOR its denial THEN the theory is incomplete. ONE OR THE OTHER of the statement and its denial MUST be true, but we DON'T NEED to know WHICH, or what "truth" "is", in order to know to know that theory is incomplete.
i say rubbish unless i know what a gibbly i cant not know what his theorem is talking about
george say
we DON'TNEED to know WHICH, or what "truth" "is", in order to know to know that theory is incomplete.
rubbish until i know what makes a statement true i cant identify a true statement thus his theorem is meaningless as with out knowing what a true statement is i cant say there are true statements which are not provable
george says
This whole endeavor DOES NOT EVEN CARE about what's TRUE: it CARES about what's PROVABLE. And we all HAVE told you that BEING AN AXIOM makes a statement provable, AS does being INFERRABLE from the axioms ACCORDING to the INFERENCE RULES of the LOGIC.
i say rubbish until i know what makes a statement true i cant identify a true statement thus his theorem is meaningless as with out knowing what a true statement is i cant say there are true statements which are not provable
goerge says
This whole endeavor DOES NOT EVEN CARE about what's TRUE:
i say rubbish it depends on there being identifibale true statements if i cant tell what a true statement is the theorem collapses into meaninglessness as i cant never find a true statement which i will know is not provable
if there are no such thing as true statement the theorenm is meaningless and if there are true statement but i dont know what the hell they are then again the theorem is meaningless as i cant make any identifications about anything ie unprovable statments
On Apr 12, 8:41 am, "elsiemelsi" <cyprin...@nosam.yahoo.com> wrote:
> The australian philosopher colin leslie dean points out robinson arithmetic > is not incomplete
Then the australian philosopher colin leslie dean is a buffoon.
Perhaps colin leslie dean would like to prove either (Ax)(0 + x = x) or its negation in Robinson Arithmetic as defined e.g. in Boolos/ Burgess/Jeffrey edn 4, or in my book.
I'm sure that George and I will chip in for a $1000 prize if he does ... and Aatu will double it! ;-)
The theories have a "standard" model, where 0 is zero, s(0) or (0+1) is 1, s(s(0)) or 1+1 is two, ad inf. All those numbers (and many different numeration systems for them, in addition to the two mentioned HERE) were ALREADY OUT THERE BEFORE Godel started proving things, and EVERYbody was legitimately expected to know what made 43*69=3426 true (or false), WITHOUT any help from Godel or 1st-order logic. So the point is, WE DO ALREADY KNOW what "makes" these statements true or false.
i say rubbish
you say
43*69=3426 true it is only true via a proof but godel is said to have made a distinction between true and proven you are just giving the old discredired idea of truth ie the hilbert idea that true statements are proven from axioms you need to tell us why 43*69=3426 is true proof
quote http://en.wikipedia.org/wiki/Truth#Truth_in_mathematics In addition, from at least the time of Hilbert's program at the turn of the twentieth century to the proof of Gödel's theorem and the development of the Church-Turing thesis in the early part of that century, true statements in mathematics were generally assumed to be those statements which are provable in a formal axiomatic system.
The works of Kurt Gödel, Alan Turing, and others shook this assumption, with the development of statements that are true but cannot be proven within the system
> Gödel's first incompleteness theorem, perhaps the single most celebrated > result in mathematical logic, states that:
> For any consistent formal, recursively enumerable theory that proves > basic arithmetical truths, an arithmetical statement that is true, but not > provable in the theory, can be constructed.1 That is, any effectively > generated theory capable of expressing elementary arithmetic cannot be > both consistent and complete.
> but godel does not tell us what makes a statement true
> until he does we cant tell if robinson arithmetic is incomplete
There are two claims you might make:
(1) Robinson Arithmetic is complete (2) Robinson Arithmetic has not been proved to be incomplete.
You started by saying (1), but you end up only arguing for (2). Do you want to go with (1)? Do you think you can give some sort of completeness proof for Robinson Arithmetic? Or do you just want to stick with the weaker claim that it hasn't been proved to be incomplete?
> we need to know what makes a statement true in order to see that it cant > be proven
Two points:
(1) We do have a definition of truth, due to Tarski (2) In any case, Gödel's argument can be done without using the notion of truth, there is a purely syntactic version of the argument which only needs the assumption that Robinson Arithmetic is consistent
> godel could have said there are gibbly statemets which cant be proven
The theorem which says "There exists a statement which is true but unprovable in Robinson Arithmeic" is the semantic version of the theorem. There is also a syntactic version which just says "If Robinson Arithmetic is consistent, then there is a sentence which is undecided by Robinson Arithmetic".
Perhaps colin leslie dean would like to prove either (Ax)(0 + x = x) or its negation in Robinson Arithmetic as defined e.g. in Boolos/ Burgess/Jeffrey edn 4, or in my boo
i say rubbish godel made a distinction between true and proof true is independent of a proof so tell us why Ax)(0 + x = x) is true independent of a proof
In addition, from at least the time of Hilbert's program at the turn of the twentieth century to the proof of Gödel's theorem and the development of the Church-Turing thesis in the early part of that century, true statements in mathematics were generally assumed to be those statements which are provable in a formal axiomatic system.
The works of Kurt Gödel, Alan Turing, and others shook this assumption, with the development of statements that are true but cannot be proven within the system
On Apr 12, 11:36 pm, "elsiemelsi" <cyprin...@nosam.yahoo.com> wrote:
> peter smith says
> Perhaps colin leslie dean would like to prove either (Ax)(0 + x = x) > or its negation in Robinson Arithmetic as defined e.g. in Boolos/ > Burgess/Jeffrey edn 4, or in my boo
> i say rubbish > godel made a distinction between true and proof > true is independent of a proof > so tell us why Ax)(0 + x = x) is true independent of a proof
Why does he have to tell you that?
You said Robinson Arithmetic is not incomplete. So you must think that either this sentence or its negation can be proved in Robinson Arithmetic. He's challenging you to actually do it.
> In addition, from at least the time of Hilbert's program at the turn of > the twentieth century to the proof of Gödel's theorem and the development > of the Church-Turing thesis in the early part of that century, true > statements in mathematics were generally assumed to be those statements > which are provable in a formal axiomatic system.
> The works of Kurt Gödel, Alan Turing, and others shook this assumption, > with the development of statements that are true but cannot be proven > within the system
Who's that Colin Leslie anyway...? Why is he so frustrated by the incompleteness theorem? weird....
Why do his followers keep claiming that Godel's theorem is false?
They keep claiming that there is a vicious circle in the demonstration but there's none. There's no faulty circularity at all.
It just happens that we take a proposition that asserts that a well- formed formula is unprobable. And then it turns out this well-formed formula is indeed the one by which the proposition itself was expressed.(cf note 15, page 598, Godel's paper inVan Heijenoort source book).
george wrote: > On Apr 12, 3:41 am, "elsiemelsi" <cyprin...@nosam.yahoo.com> wrote: >> but godel does not tell us what makes a statement true
>> until he does we cant tell if robinson arithmetic is incomplete
> Of course we can.
>> we need to know what makes a >> statement true in order to see that it cant >> be proven
> No, we don't.
>> godel could have said there are >> gibbly statemets which cant be proven
> Right.
>> but untill we know what makes a statement gibbly >> we have no way of knowing >> what his theorem is talking about
> Wrong. > It DOES NOT MATTER what OTHER attributes the > statement may or may not have. If the theory > proves NEITHER the statement NOR its denial THEN > the theory is incomplete. ONE OR THE OTHER of the > statement and its denial MUST be true, but we DON'T > NEED to know WHICH, or what "truth" "is", in order to know > to know that theory is incomplete.
> But since you have never studied this stuff, > you didn't know that.
>> so untill he tells us what makes a >> statement true robinson arthmetic cant >> be said to be incomplete
> This whole endeavor DOES NOT EVEN CARE about > what's TRUE: it CARES about what's PROVABLE. > And we all HAVE told you that BEING AN AXIOM makes > a statement provable, AS does being INFERRABLE from > the axioms ACCORDING to the INFERENCE RULES of > the LOGIC.
> "Truth" in the arithmetical sense that the endeavor is dealing > with IS something that you could be or possibly even have > been told about, if not by Godel then by YOUR ELEMENTARY > SCHOOL math teachers. > The theories have a "standard" model, where 0 is zero, s(0) > or (0+1) is 1, s(s(0)) or 1+1 is two, ad inf.
I'd say the theories have an *assumed* "standard" model, which would make a difference as far as the validity of GIT, as a meta theorem, is concerned.
> All those numbers (and many different numeration > systems for them, in addition to the two mentioned HERE) > were ALREADY OUT THERE BEFORE Godel started > proving things, and EVERYbody was legitimately expected > to know what made 43*69=3426 true (or false), WITHOUT > any help from Godel or 1st-order logic. > So the point is, > WE DO ALREADY KNOW what "makes" these statements > true or false. The ones involving "Every" natural number or > "Some" natural number might require an infinitary process for > confirmation or refutation, but the point is, we KNOW it is > THAT process and NOT some other, so we DO know what > "makes" those statements true, even when we can't > practically perform the investigation.
Suppose ~GC is not provable in Q, what process would you think would make us "know" GC be true?
> Suppose ~GC is not provable in Q, what process would you think > would make us "know" GC be true?
Ordinary mathematical research. What's the point of expressing "suppose Goldbach's conjecture is true" as "suppose ~GC is not provable in Robinson arithmetic"?
You said Robinson Arithmetic is not incomplete. So you must think that either this sentence or its negation can be proved in Robinson Arithmetic. He's challenging you to actually do it.
i say
i could not care less robinson arithmetic has not been proved to be incomplete untill we are not told what make a statement true
the theorem depend on two things consistency and the presence of true statements but if you cant tell us what makes a statement true-independent of proof- then the theorem is meaningles regardless if (Ax)(0 + x = x) or its negation in Robinson Arithmetic can be prooved or not
and even if it was proved what makes it true anyway
Aatu Koskensilta wrote: > On 2008-04-12, in sci.logic, Nam D. Nguyen wrote: >> Suppose ~GC is not provable in Q, what process would you think >> would make us "know" GC be true?
> Ordinary mathematical research.
Care to explain that in details?
> What's the point of expressing > "suppose Goldbach's conjecture is true" as "suppose ~GC is not > provable in Robinson arithmetic"?
My question doesn't say anything about "suppose Goldbach's conjecture is true". And my question is a technical question. Care to directly answer that?
> Suppose ~GC is not provable in Q, what process would you think > would make us "know" GC be true?
Ordinary mathematical research. What's the point of expressing "suppose Goldbach's conjecture is true" as "suppose ~GC is not provable in Robinson arithmetic"?
i say not good enough tell s us what would makes us know GC is true even if it is unprovable
>> Suppose ~GC is not provable in Q, what process would you think >> would make us "know" GC be true?
> Ordinary mathematical research. What's the point of expressing > "suppose Goldbach's conjecture is true" as "suppose ~GC is not > provable in Robinson arithmetic"?
> i say > not good enough tell s us what would makes us know GC is true even if it > is unprovable
For what it's worth, "suppose Goldbach's conjecture is true" still requires an *unsaid assumption* that Q be consistent. In my "suppose ~GC is not provable in Robinson arithmetic", the consistency of Q would be a corollary, hence my assumption would be more compacted, among other benefits I've not stated.
> "Wovon man nicht sprechen kann, daruber muss man schweigen" > - Ludwig Wittgenstein, Tractatus Logico-Philosophicus
If someone could somehow convince me that that sentence or its negation was provable in Robinson Arithmetic, I'd gladly pay 5000 Australian dollars for that, at the very least.
On Apr 13, 12:15 am, "elsiemelsi" <cyprin...@nosam.yahoo.com> wrote:
> rupert said
> You said Robinson Arithmetic is not incomplete. So you must think that > either this sentence or its negation can be proved in Robinson > Arithmetic. He's challenging you to actually do it.
> i say
> i could not care less > robinson arithmetic has not been proved to be incomplete untill we are > not told what make a statement true
Well, this is what I was saying before. You can say it's not incomplete, as indeed you did, or else you can say it hasn't been proved to be incomplete, which is a weaker claim. You should make it clear which claim you are making.
And, with regard to your argument, as I was saying we do have a definition of truth, but we can give an argument that Robinson Arithmetic is incomplete without using the notion of truth at all.
> the theorem depend on two things consistency and the presence of true > statements > but if you cant tell us what makes a statement true-independent of proof- > then the theorem is meaningles > regardless if (Ax)(0 + x = x) > or its negation in Robinson Arithmetic can be prooved or not
If neither the sentence nor its negation can be proved in Robinson Arithmetic, then Robinson Arithmetic is certainly incomplete and so you must retract your statement that it's not. As I'm saying, you must be clear about what you are claiming.
> > Suppose ~GC is not provable in Q, what process would you think > > would make us "know" GC be true?
> Ordinary mathematical research. What's the point of expressing > "suppose Goldbach's conjecture is true" as "suppose ~GC is not > provable in Robinson arithmetic"?
> i say > not good enough tell s us what would makes us know GC is true even if it > is unprovable
Aatu's point was: we can prove in a very weak theory that ~GC is unprovable in Robinson Arithmetic if and only if GC. And he's saying, to find out whether GC is true, we just attack the problem in the way mathematicians ordinarily do. We might succeed in showing that it can be proved in ZFC, for example.
On Apr 13, 12:57 am, "Nam D. Nguyen" <namducngu...@shaw.ca> wrote:
> Aatu Koskensilta wrote: > > On 2008-04-12, in sci.logic, Nam D. Nguyen wrote: > >> Suppose ~GC is not provable in Q, what process would you think > >> would make us "know" GC be true?
> > Ordinary mathematical research.
> Care to explain that in details?
We might show that GC is provable in ZFC, for example. That's what we're trying to do at the moment.
> > What's the point of expressing > > "suppose Goldbach's conjecture is true" as "suppose ~GC is not > > provable in Robinson arithmetic"?
> My question doesn't say anything about "suppose Goldbach's conjecture > is true".
"Q does not prove ~GC" is equivalent to GC in PRA.
Rupert wrote: > On Apr 13, 12:57 am, "Nam D. Nguyen" <namducngu...@shaw.ca> wrote: >> Aatu Koskensilta wrote: >>> On 2008-04-12, in sci.logic, Nam D. Nguyen wrote: >>>> Suppose ~GC is not provable in Q, what process would you think >>>> would make us "know" GC be true? >>> Ordinary mathematical research. >> Care to explain that in details?
> We might show that GC is provable in ZFC, for example. That's what > we're trying to do at the moment.
Except that if ZFC turns out to be inconsistent, ~GC would also be provable in this case!
Something about a formal system - as an axiom set - that has seemed to escape our attention: it's supposed to reflect the entire underlying set of concepts; hence the assumption of its consistency is a must however *assumed* it might be. Consequently, any meta assertion that *must depend on the consistency* must clearly state so to be a valid meta theorem!
There's actually nothing "sacred" about using Q as the encoding theory, or about assuming N as the standard arithmetic model of Q, in GIT. One could equally *assumes* ZFC be consistent, and perform certain "set-ization", instead of "numerization", and still arrive at GIT, as a hypothetical meta theorem. "Hypothetical" because of the assumption ZFC be consistent in this case. It's only a matter of how explicit or implicit we'd make this assumption for the encoding theory!
>>> What's the point of expressing >>> "suppose Goldbach's conjecture is true" as "suppose ~GC is not >>> provable in Robinson arithmetic"? >> My question doesn't say anything about "suppose Goldbach's conjecture >> is true".
> "Q does not prove ~GC" is equivalent to GC in PRA.
On Apr 13, 2:41 am, "Nam D. Nguyen" <namducngu...@shaw.ca> wrote:
> Rupert wrote: > > On Apr 13, 12:57 am, "Nam D. Nguyen" <namducngu...@shaw.ca> wrote: > >> Aatu Koskensilta wrote: > >>> On 2008-04-12, in sci.logic, Nam D. Nguyen wrote: > >>>> Suppose ~GC is not provable in Q, what process would you think > >>>> would make us "know" GC be true? > >>> Ordinary mathematical research. > >> Care to explain that in details?
> > We might show that GC is provable in ZFC, for example. That's what > > we're trying to do at the moment.
> Except that if ZFC turns out to be inconsistent, ~GC would also be provable > in this case!
We might prove GC in ZFC, then find an inconsistency proof for ZFC, and so re-assess our belief in GC. Certainly that could happen. In that sense no mathematical knowledge we obtain is completely certain.
> Something about a formal system - as an axiom set - that has seemed to escape > our attention: it's supposed to reflect the entire underlying set of concepts; > hence the assumption of its consistency is a must however *assumed* it might > be. Consequently, any meta assertion that *must depend on the consistency* > must clearly state so to be a valid meta theorem!
> There's actually nothing "sacred" about using Q as the encoding theory, > or about assuming N as the standard arithmetic model of Q, in GIT. > One could equally *assumes* ZFC be consistent, and perform certain > "set-ization", instead of "numerization", and still arrive at GIT, > as a hypothetical meta theorem. "Hypothetical" because of the assumption > ZFC be consistent in this case. It's only a matter of how explicit or implicit > we'd make this assumption for the encoding theory!
> >>> What's the point of expressing > >>> "suppose Goldbach's conjecture is true" as "suppose ~GC is not > >>> provable in Robinson arithmetic"? > >> My question doesn't say anything about "suppose Goldbach's conjecture > >> is true".
> > "Q does not prove ~GC" is equivalent to GC in PRA.
Rupert wrote: > On Apr 13, 2:41 am, "Nam D. Nguyen" <namducngu...@shaw.ca> wrote: >> Rupert wrote: >>> On Apr 13, 12:57 am, "Nam D. Nguyen" <namducngu...@shaw.ca> wrote: >>>> Aatu Koskensilta wrote: >>>>> On 2008-04-12, in sci.logic, Nam D. Nguyen wrote: >>>>>> Suppose ~GC is not provable in Q, what process would you think >>>>>> would make us "know" GC be true? >>>>> Ordinary mathematical research. >>>> Care to explain that in details? >>> We might show that GC is provable in ZFC, for example. That's what >>> we're trying to do at the moment. >> Except that if ZFC turns out to be inconsistent, ~GC would also be provable >> in this case!
> We might prove GC in ZFC, then find an inconsistency proof for ZFC, > and so re-assess our belief in GC. Certainly that could happen. In > that sense no mathematical knowledge we obtain is completely certain.
I'd not say "no mathematical knowledge...", only "some mathematical knowledge...". But right, my main point here is the foundation of *current* mathematics is based on the "all-knowing" human mind even in dealing with infinity, which imho is quite wrong. Isn't it true that the essence Lindenbaum's Compactness would sort of encourage us to be a little conservative, as far as our knowledge is concerned?
> So what?
So that we could *improve* our reasoning, to cope with the ever increasing demand that mathematics is "the language" of natural sciences!
>> Something about a formal system - as an axiom set - that has seemed to escape >> our attention: it's supposed to reflect the entire underlying set of concepts; >> hence the assumption of its consistency is a must however *assumed* it might >> be. Consequently, any meta assertion that *must depend on the consistency* >> must clearly state so to be a valid meta theorem!
>> There's actually nothing "sacred" about using Q as the encoding theory, >> or about assuming N as the standard arithmetic model of Q, in GIT. >> One could equally *assumes* ZFC be consistent, and perform certain >> "set-ization", instead of "numerization", and still arrive at GIT, >> as a hypothetical meta theorem. "Hypothetical" because of the assumption >> ZFC be consistent in this case. It's only a matter of how explicit or implicit >> we'd make this assumption for the encoding theory!
>>>>> What's the point of expressing >>>>> "suppose Goldbach's conjecture is true" as "suppose ~GC is not >>>>> provable in Robinson arithmetic"? >>>> My question doesn't say anything about "suppose Goldbach's conjecture >>>> is true". >>> "Q does not prove ~GC" is equivalent to GC in PRA. >> Will PRA be consistent?
> Seems quite likely...
If mathematical reasoning is always based on "seems ... likely" there wouldn't be the words "formal logic". And in such case, we might as well become Zen masters, to know the "unutterable" truths, which would be the only truths!
On Apr 12, 11:05 am, "elsiemelsi" <cyprin...@nosam.yahoo.com> wrote:
> i say rubbish it depends on there > being identifibale true statements
OF COURSE there are identifiable true statements. But the point is that there is also a statement with the property that neither it nor its denial IS identified as a true statement.
> cant tell what a true statement is the theorem collapses into > meaninglessness
No, it doesn't, since the theorem basically PROVES you CAN'T always tell which statements are "true". The theorem CONSTRUCTS a statement that COULD be EITHER of true or false, as far as the axioms are concerned. There are models BOTH WAYS.