[FOM] a real number for the Ackermann function

Ron Graham graham at ucsd.edu
Fri Apr 14 22:57:48 EDT 2006

This real is well known to be e^(1/e) where e is the base of the natural 

Ron Graham

Robert M. Solovay wrote:

>Do you have some other, more transparent, characterization of this real?
> 	--Bob Solovay
>On Thu, 13 Apr 2006, Andreas Weiermann wrote:
>>There might be some interest in the
>>following sharp phase transition result
>>for the Ackermann function
>>in terms of a specific (possibly transcendental) real number.
>>Theorem: Let a>1 be a real number. Define a hierarchy of unary functions
>>a_k as follows. Let a_0(x) be a tower of a's of hight x where x is a natural
>>number argument. Let a_{k+1}(x) be the x-th iterate of a_k applied to x.
>>Finally define the diagonal function a(x)=a_x(x).
>>Then x\mapsto a(x) is Ackermannian iff a>1.44466781...
>>Best regards,
>>Andreas Weiermann
>>FOM mailing list
>>FOM at cs.nyu.edu
>FOM mailing list
>FOM at cs.nyu.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/fom/attachments/20060414/f87e0b93/attachment.html

More information about the FOM mailing list