arXiv subjects
Publications and source records attributed to F. A. Doria.
If we establish that the counterexample function for P=NP, if total, overtakes all total recursive functions when extended over all Turing machines, then what happens to the same counterexample function when defined over the so-called Baker-Gill-Solovay (BGS) set of poly machines? We state and prove here a lemma that tries to answer this query.
This paper discusses a function that is frequently presented as a simile or look-alike of the so-called ``counterexample function to P=NP,'' that is, the function that collects all first instances of a problem in NP where a poly machine incorrectly `guesses' about the instance. We state and give in full detail a crucial result on the computation of Goedel numbers for some families of poly machines.
We formulate the $P<NP$ hypothesis in the case of the satisfiability problem as a $Π^0_2$ sentence, out of which we can construct a partial recursive function $f_{\neg A}$ so that $f_{\neg A}$ is total if and only if $P < NP$. We then show that if $f_{\neg A}$ is total, then it isn't ${\cal T}$--provably total (where ${\cal T}$ is a fragment of ZFC that adequately extends PA and whose consistency is of ordinal order). Follows that the negation of $P < NP$, that is, $P = NP$, is consistent with those ${\cal T}$.
We investigate the existence of a class of ZFC-provably total recursive unary functions, given certain constraints, and apply some of those results to show that, for $Σ_1$-sound set theory, ZFC$\not\vdash P<NP$.