What One Has to Know When Attacking \(\textsf {P}\) vs. \(\textsf {NP}\) (Extended Abstract)

  • Juraj HromkovičEmail author
  • Peter RossmanithEmail author
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10472)


Mathematics was developed as a strong research instrument with fully verifiable argumentations. We call any consistent and sufficiently powerful formal theory that enables to algorithmically verify for any given text whether it is a proof or not algorithmically verifiable mathematics (AV-mathematics for short). We say that a decision problem \(L \subseteq \varSigma ^*\) is almost everywhere solvable if for all but finitely many inputs \(x \in \varSigma ^*\) one can prove either “\(x \in L\)” or “\(x \not \in L\)” in AV-mathematics.

First, we formalize Rice’s theorem on unprovability, claiming that each nontrivial semantic problem about programs is not almost everywhere solvable in AV-mathematics. Using this, we show that there are infinitely many algorithms (programs that are provably algorithms) for which there do not exist proofs that they work in polynomial time or that they do not work in polynomial time. We can prove the same also for linear time or any time-constructible function.

Note that, if \(\textsf {P} \ne \textsf {NP} \) is provable in AV-mathematics, then for each algorithm A it is provable that “A does not solve \(\text {SATISFIABILITY}\) or A does not work in polynomial time”. Interestingly, there exist algorithms for which it is neither provable that they do not work in polynomial time, nor that they do not solve \(\text {SATISFIABILITY}\). Moreover, there is an algorithm solving \(\text {SATISFIABILITY}\) for which one cannot prove in AV-mathematics that it does not work in polynomial time.

Furthermore, we show that \(\textsf {P} =\textsf {NP} \) implies the existence of algorithms X for which the true claim “X solves \(\text {SATISFIABILITY}\) in polynomial time” is not provable in AV-mathematics. Analogously, if the multiplication of two decimal numbers is solvable in linear time, one cannot decide in AV-mathematics for infinitely many algorithms X whether “X solves multiplication in linear time”.

Finally, we prove that if \(\textsf {P}\) vs. \(\textsf {NP}\) is not solvable in AV-mathematics, then \(\textsf {P}\) is a proper subset of \(\textsf {NP}\) in the world of complexity classes based on algorithms whose behavior and complexity can be analyzed in AV-mathematics. On the other hand, if \(\textsf {P} =\textsf {NP} \) is provable, we can construct an algorithm that provably solves \(\text {SATISFIABILITY}\) almost everywhere in polynomial time.



We would like to thank Hans-Joachim Böckenhauer, Dennis Komm, Rastislav Královič, Richard Královič, and Georg Schnitger for interesting discussions related to the first verification of the proofs presented here. Essential progress was made during the 40th Mountain Workshop on Algorithms organized by Xavier Muñoz from UPC Barcelona that offered optimal conditions for research work.

Supplementary material


  1. 1.
    Aaronson, S.: Is P versus NP formally independent? Bull. EATCS 81, 109–136 (2003)MathSciNetzbMATHGoogle Scholar
  2. 2.
    Baker, T.P., Gill, J., Solovay, R.: Relativizations of the P =? NP question. SIAM J. Comput. 4(4), 431–442 (1975)MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Chaitin, G.: Information-theoretic limitations of formal systems. J. ACM 21(3), 403–424 (1974)MathSciNetCrossRefzbMATHGoogle Scholar
  4. 4.
    Gödel, K.: Über formal unentscheidbare Sätze der Principia Mathematica und verwandte Systeme. Monatshefte für Mathematik und Physik 28, 173–198 (1931)CrossRefzbMATHGoogle Scholar
  5. 5.
    Hilbert, D.: Die logischen Grundlagen der Mathematik. Math. Ann. 88, 151–165 (1923)MathSciNetCrossRefzbMATHGoogle Scholar
  6. 6.
    Kolmogorov, A.: Three approaches for defining the concept of information quantity. Probl. Inf. Transm. 1, 1–7 (1965)Google Scholar
  7. 7.
    Kolmogorov, A.: Logical basis for information theory and probability theory. IEEE Transit. Inf. Theory 14, 662–664 (1968)MathSciNetCrossRefzbMATHGoogle Scholar
  8. 8.
    Razborov, A.A., Rudich, S.: Natural proofs. J. Comput. Syst. Sci. 55(1), 24–35 (1997)MathSciNetCrossRefzbMATHGoogle Scholar
  9. 9.
    Rice, H.: Classes of recursively enumerable sets and their decision problems. Transact. ASM 89, 25–59 (1953)zbMATHGoogle Scholar
  10. 10.
    Aaronson, S.: P \({}\mathrel {\mathop =\limits ^?}{}\,\)NP. In: Electronic Colloquium on Computational Complexity (ECCC) (2017)Google Scholar
  11. 11.
    William, I.: Gasarch: guest column: the second P=?NP poll. SIGACT News 43(2), 53–77 (2012)CrossRefGoogle Scholar
  12. 12.
    Papadimitriou, C.H.: Computational Complexity. Academic Internet Publishers, Ventura (2007)zbMATHGoogle Scholar
  13. 13.
    Immerman, N.: Nondeterministic space is closed under complementation. SIAM J. Comput. 17(5), 935–938 (1988)MathSciNetCrossRefzbMATHGoogle Scholar
  14. 14.
    Szelepcsényi, R.: The method of forced enumeration for nondeterministic automata. Acta Inf. 26(3), 279–284 (1988)MathSciNetCrossRefzbMATHGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  1. 1.Department of Computer ScienceETH ZürichZürichSwitzerland
  2. 2.Department of Computer ScienceRWTH Aachen UniversityAachenGermany

Personalised recommendations