Introduction
About me
Associate professor at the Department of Logic,
College of Arts ("Faculty of Arts"), Charles University in PragueFilozofická fakulta UK,
Palachovo nám. 2, 11638 Praha 1.
Courses Computability (an introduction to recursive functions),
Classical Logic II (with an emphasis on incompleteness and undecidability),
and some others related to metamathematics or to algorithms.
Research interests
Interpretability of axiomatic theories, arithmetization, related modal logics, algorithmic aspects
of non-classical logics.
Experience and education
During 1982-90 worked as a software developer for an industrial company ČKD Polovodiče.
Participation in writing a control system for PDP-11 computers.
The system was designed by Jan Pavelka*)Accidentally
the same Jan Pavelka who, a few years before turning into a respected software developer,
wrote innovative papers about fuzzy logics., and was successfully applied in
several branches of industry like steel rolling or international transport of natural gas.
Ph.D. in mathematical logic, Ph.D. study in the Institute of Mathematics of the Academy of Sciences
of the Czech Republic.
Dissertation Self-Reference and Modal Logic, 1982,
supervisor Petr Hájek. Graduated in mathematics from Charles University, 1978,
master's thesis Degrees of Interpretability.
Office info, events, miscellaneous updates
Office hours Thu 15:00, or by appointment.
For more info please see the image above, or the web of the
department.
Nabídka témat bakalářských a diplomových prací:
TemataPraci13.pdf
Publications
BibTeX file
BibTeX entries can be taken from sv.bib. Note that the file
contains a preamble with some strings and macros.
Book
- Logika: neúplnost, složitost a nutnost
(Logic: Incompleteness, Complexity, and Necessity),
Academia Praha, 2002.
464 pages,
section 5.2 about Gödel-Dummett fuzzy logic was written by Petr Hájek. See the
"Book" tab above.
Some workshop or conference presentations
-
Logika, Gödel, neúplnost (Logic, Gödel, Incompleteness).
A talk at Friday meetings of Sysifos,
February 23th, 2018. A videorecord created by Ludvík Hájek is available on
YouTube.
-
Kurt Gödel: úplnost a neúplnost (Kurt Gödel: Completeness and
Incompleteness). A talk at Hvězdárna and planetárium Brno (Brno Observatory and Planetarium),
March 9th, 2017. A videorecord created by Josef Forman is available on
YouTube.
-
Rosser Sentences and Rosser Logics.
A talk at the Prague gathering of logicians,
Jan 2017.
-
On Purely implicational Fragments of Intuitionistic
Propositional Logic. A talk at the Logica '11 International Conference,
Hejnice, Czech Republic, June 2011. [pdf4]
Papers
- Do We Need Recursion? In I. Sedlár and M. Blicha, eds.,
The Logica Yearbook 2019,
pp. 193–204. College Publications, London, 2020. [pdf][talk]
- Modern Czech Logic:
Vopěnka and Hájek, History and Background. Journal of Applied Logics — IfCoLoG Journal
of Logics and their Applications, 5(6):1261–1271, 2018. [pdf]
- Předmluva: o logice 20. století (Preface: On the 20th Century Logic).
In L. Dostálová, ed., Kurt Gödel: úplnost a neúplnost (Kurt Gödel: Completeness and Incompleteness),
pp. 13–53. OPS in collaboration with University of West Bohemia Publishing,
Plzeň, 2015. ISBN 978-80-87269-39-8. In Czech. [pdf]
- On Strong Fragments of Peano Arithmetic.
In P. Arazim and M. Dančák, eds.,
The Logica Yearbook 2014,
pp. 281–291. College Publications, London, 2015. [pdf][talk]
- Výuka logiky bez logických pojmů (Teaching logic without the use
of logical notions). In L. Dostálová and K. Šebela, eds.,
Organon VII — Nihil novi, pp. 49–57. University of West Bohemia Publishing,
Plzeň, 2011. ISBN 978-80-261-0035-5. In Czech.
[pdf][pdf2]
- Infinite natural numbers: an unwanted phenomenon, or a useful concept?
In M. Peliš and V. Punčochář, eds.,
The Logica Yearbook 2010,
pp. 283–294. College Publications, London, 2011. [pdf][pdf2][talk]
- Decision Problems of Some Intermediate Logics and Their Fragments. In M. Peliš, ed.,
The Logica Yearbook 2009,
pp. 297–307. College Publications, London, 2010. [pdf][pdf2][talk]
- Relatives of Robinson Arithmetic. In M. Peliš, ed.,
The Logica Yearbook 2008,
pp. 253–263. College Publications, London, 2009. [pdf][pdf2][talk]
-
On Interpretability in the Theory of Concatenation.
Notre Dame Journal of Formal Logic, 50(1):87–95, 2009.
[pdf][pdf2]
- Weak Theories and Essential Incompleteness. In M. Peliš, ed.,
The Logica Yearbook 2007, pp. 213–224. Filosofia, Praha, 2008.
[pdf][pdf2][talk]
-
An Interpretation of Robinson Arithmetic in its Grzegorczyk's Weaker Variant.
Fundamenta Informaticae, 81(1–3):347–354, 2007.
[pdf][pdf2]
- Gödel-Dummett Predicate Logics and Axioms of Prenexability.
In O. Tomala and R. Honzík, eds.,
The Logica Yearbook 2006, pp. 251–260. Filosofia, Praha, 2007.
[pdf][pdf2][talk]
- On Modal Systems with Rosser Modalities. In M. Bílková and O. Tomala, eds.,
The Logica Yearbook 2005, pp. 203–214. Filosofia, Praha, 2006.
[pdf][pdf2][talk]
-
On Sequent Calculi for
Intuitionistic Propositional Logic.
Comm. Math. Univ. Carolinae, 47(1):159–173, 2006.
[pdf][pdf2]
-
With Blanka Kozlíkova, On
Interplay of Quantifiers in Gödel-Dummett Fuzzy Logics.
Archive for Math. Logic, 45(5):569–580, 2006.
[pdf][pdf2]
-
Note on
Inter-Expressibility of Logical Connectives in Finitely-Valued Gödel-Dummett Logics.
Soft
Computing, 10(7):629–630, 2006.
[pdf][pdf2]
-
Gödlov prvi izrek o nepolnosti.
Obzornik za matematiko in fyziko, 51(4):113–121, 2004.
-
The limit lemma
in fragments of arithmetic. Comm.
Math. Univ. Carolinae, 44(3):565–568, 2003.
[pdf][pdf2]
-
A
note on arithmetical completeness of theories with Rosser modalities.
In K. Bendová and P. Jirků, eds., Miscellanea logica V,
pp. 123–132. Karolinum, Praha, 2003. ISBN 80-246-0799-9.
-
On
structural proofs of Gödel First Incompleteness Theorem.
In K. Bendová and P. Jirků, eds., Miscellanea logica V,
pp. 115–122. Karolinum, Praha, 2003. ISBN 80-246-0799-9.
-
The decision problem of
provability logic with only one atom.
Archive for Math. Logic, 42(8):763–768, 2003.
Errata: The final paragraph, just before the list of references, contains an unpleasant
misprint: 'added in January 2002', should be 'added in January 2003'. My apologies!
[pdf][pdf2]
-
On the polynomial-space
completeness of intuitionistic propositional logic.
Archive for Math. Logic, 42(7):711–716, 2003.
[pdf][pdf2]
-
With K. Bendová,
On Inter-Expressibility
of Logical Connectives in Gödel Fuzzy Logic.
Soft
Computing, 4(2):103–105, 2000.
[pdf][pdf2]
-
On Provability Logic.
Nordic
Journal of Philosophical Logic, 4(2):95–116, 2000.
[pdf][pdf2]
-
Arithmetical
Classification of the Set of All Provably Recursive Functions.
Comm. Math. Univ. Carolinae, 40(4):631–634, 1999.
[pdf][pdf2]
-
Poznamka
o vztahu vyplyvani v klasicke predikatove logice (A Note on
the Consequence Relation in Classical Predicate Logic). In
P. Jirků and V. Švejdar, eds., Miscellanea
Logica II. Karolinum, Praha, 1999. In Czech.
-
Logika
v aritmetice. In P. Jirků and V. Švejdar, eds.,
Miscellanea Logica I. Karolinum, Praha, 1998. In Czech.
-
With P. Hájek, Matematicka logika (Mathematical Logic).
Praha, November 1994. Lecture notes, electronic version only, in Czech.
-
Some Independence Results in Interpretability Logic.
Studia Logica, 50(1):29–38, 1991.
- With P. Hájek,
A Note on the Normal Form of Closed Formulas
of Interpretability Logic. Studia Logica, 50(1):25–28, 1991.
-
Modal Analysis of Generalized Rosser Sentences. J. Symb. Logic, 48(4):986–999, 1983.
[JSTOR]
-
A Sentence that is Difficult to Interpret.
Comm. Math. Univ. Carolinae,
22(4):661–666, 1981.
-
Degrees of Interpretability.
Comm. Math. Univ. Carolinae,
19(4):789–813, 1978.
Logic: Incompleteness, Complexity, and Necessity (Academia, 2002)
I do not think logic is a field where people are taught how to think logically.
Instead, I consider it a theory about deductive thinking for people who already
are able to think logically, usually because they have some experience with
academic mathematics.
My ambition was to have more than a list of definitions, theorems, and proofs; I am
trying to include comments and explanations, and I believe that the reader should be led
from some questions to some solutions.
I am grateful to Petr Hájek, who was my teacher and who gave one of the initial
impulses for writing this book. He actually wrote only a small part of the text, but
much larger part and perhaps everything was inspired by him.
As a result of an agreement with the Academia publishing company,
a pdf version of the book is now available
(full text!) for study purposes (only!). When downloading the file, it is recommended
to download also Errata
and save both files to the same folder on local computer. That will activate the links
in the text marked by big red 'E'.
Redistribution and modifications of the file are not allowed! Printing of
(any part of) the document is not allowed!
Po domluvě s nakladatelstvím Academia je nyní
plný text knihy ve formátu pdf
přístupný pro studium (pouze!). Při kopírování souboru na lokální
počítač je doporučeno stáhnout také Errata
a umístit oba soubory do téže složky. Soubor nesmí být dále šířen
ani nesmí být modifikován! Žádná jeho část nesmí být tištěna!
Click on the blue word 'Academia' on the first page of the book to obtain
further information about their bookstores and other books they publish.
Comments and additions to Errata are welcome.
Modal and non-classical logics
Exams
Ke zkoušce si prosím připravte seznam šesti cvičení ze
seznamu cvModL.pdf, která dovedete řešit. Na poslední přednášce
si tento požadavek upřesníme.
Course contents in Fall 2022/23 (Czech)
- 5. ledna 2023
- Podrámce a disjunktní sjednocení rámců jako speciální případy bisimulace mezi rámci.
Charakteristická třída libovolné logiky je uzavřena na podrámce, disjunktní sjednocení
a p-morfní obrazy.
- 22. prosince 2022
- Cresswellův model. Logika K + H je slabší než GL.
- 15. prosince 2022
- Charakteristická třída logiky GL je třída všech tranzitivních obráceně fundovaných rámců.
Táž třída je i charakteristickou třídou pro schéma H.
- 8. prosince 2022
- Logika S5 a její charakteristická třída. Schéma 4 je dokazatelné v S5. V logice S4
platí, že z dvojice sousedících a stejných modalit lze jednu ekvivalentně vypustit.
V S5 platí, že ze dvou sousedících libovolných modalit lze první ekvivalentně vypustit.
Logika GL. Schéma 4 je dokazatelné v GL.
- 1. prosince 2022
- Schémata 4, T, B a 5 a jejich charakteristické třídy. Logika S4.
- 24. listopadu 2022
- Logika K4, příklad důkazu v ní, její gentzenovský kalkulus. Charakteristické třídy logik.
Domácí cvičení. 1. Dokažte, že hilbertovský a gentzenovský kalkulus se navzájem simulují. Takže
jsou spolu ekvivalentní.
2. Nalezněte charakteristickou třídu logiky, která vznikne přidáním
schématu A → □◇A k logice K.
3. Nalezněte charakteristickou třídu logiky, která vznikne přidáním
formule ¬□⊥ → ¬□¬□⊥
ke K.
- 10. listopadu 2022
- Logika K je rozhodnutelná. Pro oba její kalkuly platí věta o úplnosti. Pro gentzenovský kalkulus
platí i věta o eliminovatelnosti řezů. Vlastnost konečných modelů: každá nedokazatelná formule má
kripkovský protipříklad, který je konečným antireflexivním stromem.
- 3. listopadu 2022
- Gentzenovský kalkulus pro logiku K. Vzájemná simulovatelnost gentzenovského a hilbertovského kalkulu.
Podmodel kripkovského modelu generovaný prvkem.
- 27. října 2022
- Zbývající pravidla kalkulu GK. Subformula property. Pro kalkulus GK platí věta o korektnosti, věta o
úplnosti a věta o eliminovatelnosti řezů.
Domácí cvičení. Navrhněte vhodná pravidla pro spojku ≡.
- 20. října 2022
- Dokazatelnost a nedokazatelnost v logice K. Gentzenovský kalkulus GK pro klasickou výrokovou logiku.
- 13. října 2022
- Splněnost formule ve vrcholu kripkovského modelu, platnost formule v modelu.
Schéma □A → □□A platí
v každém modelu, jehož rámec je tranzitivní (ale i v některých jiných modelech).
Logika K a její hilbertovský kalkulus.
Domácí cvičení. S užitím faktu, že všechny tautologie jsou dokazatelné, zdůvodněte, že
□A&□B → □(A&B)
je v K dokazatelná pro každou volbu formule A.
- 6. října 2022
- Opakování klasické výrokové logiky: logické spojky, tautologie, tautologický důsledek, věta o kompaktnosti.
Modality nutnosti a možnosti. Modální výrokové formule. Různé modální principy. Například
□A → A se tradičně nazývá princip T.
Dále □(□A → A) → □A
je Löbův axiom, kdežto □A → A / A
je Löbovo pravidlo.
Domácí cvičení. Uvažujte rámec se čtyřmi vrcholy 1, 2, 3 a 4, v němž z 1 jsou dosažitelné
vrcholy 3 a 4, z 2 je dosažitelný 1, z 4 jsou dosažitelné 4 a 2.
Atom p je splněn ve vrcholech 1 a 2.
Ve kterých vrcholech je splněna formule ◇□◇p?
Properties of axiomatic theories
Exercises
Please pay attention to the list of exercises.
Here are the same exercises in English.
Course contents in Fall 2023/24
- Nov 30th, 2023
- Examples of axiomatizable classes and classes that are not axiomatizable. The complement of a finitely
axiomatizable class C, i.e. the class of all structures for the given language that are
not in C, is again finitely axiomatizable. Definable sets and definable elements of a
structure.
Homework. Exercises 10–13 are not so difficult and they are partly already solved.
14 and 17 are hopefully illustrative but also not very difficult. 15 and 16 are more tricky.
Please pick some of these and send a solution or questions concerning them.
- Nov 28th, 2023
- General Łoś-Vaught test: if T is consistent, has no finite models and is κ-categorical
for some uncountable κ, then it is complete. SUCC is an example of a theory to which this
theorem is applicable. Axiomatizable and finitely axiomatizable classes of structures. Example: if L
is an arbitrary language, then the class of all finite structures for L is not
axiomatizable. This is an almost immediate consequence of the theorem saying that if T has
unboundedly large finite models, then it also as infinite models.
- Nov 23rd, 2023
- Review of semantic results and their consequences: Weak Löwenheim-Skolem theorem, Weak Łoś-Vaught
test (we already discussed it on Tuesday), the compactness theorem. If T has unboundedly large
finite models, then it also has infinite models. It T has infinite models, then it has models
of any cardinality κ ≥ |L| + ℵ0. A consequence
is that Th(〈N,+,0,s〉) has models of any infinite cardinality. It is clear that
the order type of an uncountable model of this theory
is ω + (ω* + ω) · λ
where λ is a dense linear order. It remains to prove the existence of countable nonstandard
models of this theory.
- Nov 21st, 2023
- Vaught (Łoś-Vaught) test: if T is consistent, its language is at most countable and every
two at most countable models of T are isomorphic, then T is complete. There are not so many theories
to which this test is applicable. One example is DNO, the theory of dense linear order with no endpoints.
- Nov 16th, 2023
- Quantifier elimination for the theory DOS. Other classroom examples of structures that
admit quantifier elimination (i.e., from which a theory admitting quantifier elimination can be
spotted out) are 〈R,<〉 and 〈R,+,0,1〉.
More difficult examples, important in the history of logic, are as follows.
In 1929 Presburger showed that 〈N,+,0,s〉 admits quantifier elimination. The theory of
this structure is called Presburger arithmetic. Later Tarski showed the same about the
structure ℝ of the reals (with both addition and multiplication). For more about the history
(of quantifier elimination) see the paper Doner-Hodges
and Tarski's bibliography by Givant. They cite Tarski's book published in 1948 but they
also seem to suggest that Tarski obtained this result already in 1930. Probably not: the
same Wilfred Hodges speaks
here about the year 1949.
Homework. Exercise 19.
- Nov 14th, 2023
- DO, the theory of discrete linear order. The theory DOS is a conservative extension
of DO. In fact, it is an extension by definitions. It is also a conservative extension of SUCC.
Some notation for types of linear order: ω and η are shorthands for
the types of 〈N,<〉 and 〈Q,<〉 respectively. Examples of models
of DO are ω, ω + ω* + ω, and
also ω + (ω* + ω) · λ where
λ is any linear order.
- Nov 9th, 2023
- The remaining part of the proof of quantifier elimination for the theory SUCC. This theory
is complete. It is clear that any two models of a complete theory are elementarily equivalent.
That is, they do not differ in the validity of any sentence. In particular,
〈N,0,s〉 and 〈N,0,s〉 + 〈Z,s〉
are elementarily equivalent. There is no formula in the language {0,S} saying that
“the number x is reachable from zero by finitely many steps of the successor function”.
Contrary to what was said today, the book
does not contain a brief proof of the quantifier elimination for SUCC, my apologies. It discusses a closely
related theory DOS. This is a plan for the next talk.
- Nov 7th, 2023
- Three lemmas concerning the theory Q where the third of them says that every open sentence in
the arithmetic language is either provable or refutable in Q. The same three lemmas clearly hold
also for the theory SUCC. The theory SUCC admits quantifier elimination: every formula φ
in the language {0,S} is SUCC-equivalent to a quantifier-free formula containing
no other free variables than those occurring free in φ. To show this, it is sufficient to
treat formulas of the form ∃xψ where ψ is a clause (a conjunction of literals).
- Nov 2nd, 2023
- The axioms Q1–Q9 of Robinson arithmetic Q. Facts about ≤ and <
provable and unprovable in Q. A simple model with two nonstandard elements shows that the
properties of operations (like the commutativity and distributivity) and of order (like the transitivity)
are usually unprovable in Q. Robinson arithmetic is not a conservative extension of Q1–Q5.
- Oct 31st, 2023
- Extensions and conservative extensions of a theory. Q1–Q5 is a conservative extension
of Q1–Q3. The axioms Q8 and Q9 concerning the unstrict and strict order. Example
sentences provable and unprovable in Q1–Q5,Q8,Q9. A special case of conservative extension
is extension by definitions. Q8 and Q9 are definitions in this sense, and Q1–Q5,Q8,Q9
is thus a conservative extension of Q1–Q5 (and of Q1–Q3).
- Oct 26th, 2023
- Aspects of the completeness theorem: the strong completeness theorem is in fact a conjunction
of two claims, the completeness theorem (no adjective) and the compactness theorem. The fact that
the Zermelo-Fraenkel set theory, if it is consistent, also has countable models is called Skolem's
paradox. However, there is nothing paradoxical about it. The axioms Q4 and Q5. Example
sentences provable from Q1–Q5.
Homework. Show that every model of Q1–Q3 has an expansion that is a model of Q1–Q5.
Hint. We know how each model of Q1–Q3 looks like: the standard part, which is isomorphic
to 〈N,0,s〉, plus any number of finite loops of arbitrary sizes, plus any number
of chains isomorphic to 〈Z,s〉.
- Oct 24th, 2023
- Many different models for various theories, some notation for them
(like 〈N,0,s〉 + 〈Z,s〉 + 〈Z,s〉).
The theory SUCC. The strong completeness theorem and the Löwenheim-Skolem theorem.
- Oct 19th, 2023
- Standard and nonstandard elements of structures in which at least Q1 and Q2 are valid.
Standard and nonstandard models. Logically valid formulas. The consequence relation.
Contradictory and consistent theories.
Homework. No homework is officially assigned. However, all exercises 1–11 are solvable now.
Anyone who can solve Exercise 4 obviously mastered the definition of the consequence relation.
- Oct 17th, 2023
- Tarski's definition. Lines concerning connectives imply that all tautologies are valid in every
structure. The soundness of the generalization rules. The strong soundness theorem as a tool for
proving unprovability.
Homework. Exercise 9.
- Oct 12th, 2023
- The deduction theorem. A review of semantic notions: valuation of variables in a structure,
formula satisfied by a valuation in a structure, formula valid in a structure. The soundness theorem.
- Oct 10th, 2023
- A survey of Hilbert-style calculus HKe. A formal proof
of ∃x(P(x) → ∀yP(y))
in the theory with no axioms and
of ∀x∃y(S(y) ≠ x) in the
theory having Q1 and Q2 as axioms.
Homework. show that also
∃x(∃yP(y) → P(x))
has a formal proof.
- Oct 5th, 2023
- More examples of informal and formal proofs. The generalization rules. Numerals.
Homework. Construct formal proofs of some (not necessarily all) formulas in Exercise 1.
- Oct 3rd, 2023
- A review of logical syntax: from terms to sentences and axiomatic theories.
Terms, formulas, axiomatic theories. A sample theory having Q1 and Q2 as axioms.
Homework. Think about an informal proof of the sentence numbered (iii).
Arithmetics and algorithms
Please pay attention to the list of exercises.
Exams (Czech)
Ke zkoušce si prosím připravte
(přineste) řešení patnácti cvičení (i s vysvětlením
matematického či algoritmického pozadí). Zkouška pak pokračuje prodiskutováním
některého důkazu. V SISu budou vypsány nějaké termíny na pokud možno všechny úterky
do konce června s výjimkou týdne od 20.6. Později pokud možno i v září, ale počítejte prosím s určitou
improvizací a ne s úplně každým týdnem. Všechny termíny lze využívat i jako konzultace a mailem se lze
domluvit i na jiné dny než ty vypsané v SISu, většinou na
Po a St navečer nebo Út celkem kdykoliv.
Course contents in Spring 2022/23 (Czech)
- 11. května 2023
- Když p je prvočíslo a k je dělitel čísla p - 1,
pak v grupě Φ(p) je přesně φ(k) prvků řádu k. Z toho
mimo jiné plyne, že ve Φ(p) existují prvky maximálního možného
řádu p - 1, neboli že Φ(p) je cyklická grupa.
- 4. května 2023
- Řády prvků v grupách. Řád libovolného a ∈ Φ(n)
musí být dělitelem čísla φ(n). Známé tvrzení, že polynom stupně k, který není
identicky nulový, má nejvýše k různých kořenů, platí ve všech oborech integrity.
- 27. dubna 2023
- Kryptografická metoda RSA. Mersennova čísla a Mersennova prvočísla.
- 20. dubna 2023
- Počítání s modulárními reprezentacemi. Čínská zbytková věta.
- 13. dubna 2023
- Umocňování xy v ℤm, chápané jako funkce tří proměnných,
je polynomiálně počitatelné. Pseudoprvočísla.
- 30. března 2023
- Bézoutova věta jako důsledek korektnosti čtyřsloupcového Eukleidova algoritmu. Některé její důsledky:
ekvivalentní podmínka pro nesoudělnost nenulových čísel a a b (tj.
podmínka ∀z(a ∣ bz → a ∣ z))
a fakt, že když prvočíslo je dělitelem součinu xy, musí být dělitelem některého
z čísel x a y (toto ponecháno jako cvičení).
- 23. března 2023
- Úloha určit (z a a m), zda a je
v ℤm invertibilní, a také úloha nalézt inverzní prvek
k invertibilnímu a jsou polynomiálně počitatelné.
Nesoudělnost, prvočísla (v oborech integrity). Čtyřsloupcové rozšíření Eukleidova algoritmu.
Domácí úkol: cvičení 5 (cvičení 1-4 jsou také řešitelná nebo již vyřešená).
- 16. března 2023
- Definice největšího společného dělitele, úloha GCD.
Eukleidův algoritmus a jeho časové nároky.
- 9. března 2023
- Eulerova grupa (nějakého přirozeného čísla). Tělesa a obory integrity. Dělitelnost v oborech integrity.
Cvičení. Dokažte, že relace asociativnosti je ekvivalence.
- 2. března 2023
- Komutativní monoidy a komutativní (čili Abelovy) grupy. Invertibilní prvky monoidu tvoří grupu.
Věta (možná Eulerova) o umocňování v konečných komutativních grupách. Okruhy a počítání s opačnými
prvky v nich. Struktury ℤ, ℚ, ℤm a ℚ[x]
jako příklady okruhů.
- 23. února 2023
- Pojem polynomiálního algoritmu. Rozdíl mezi úlohami Factoring
a Primes. Monoidy a invertibilní prvky v nich.
Cvičení. Je-li na množině {0,. .,14}
operace ∗ definována předpisem x∗y = Mod(xy,15),
které prvky jsou invertibilní?
- 16. února 2023
- Několik všeobecně matematických pojmů: přirozená, celá a racionální čísla, dělení se zbytkem, relace
a některé jejich vlastnosti (reflexivita, tranzitivita). Logické symboly (spojky a kvantifikátory).
Úlohy a algoritmy. Úlohy Mult a Primes, a algoritmy,
které je počítají. Časové nároky algoritmů. Například školní algoritmus pro
úlohu Mult počítá v kvadratickém čase.
Huge number calculator
The calculator BNCalc.pdf is implemented
as a (set of scripts behind a) pdf document and can be used for operating with
really huge integers. Note that there are some undocumented key combinations
that make it easier to repeat the given operation.
The "result" text field in the "Factoring" page, marked by an arrow in this
image, accepts Shift‑M (or Shift‑B with the same
meaning).
Also, the result text field on the preceding page accepts Shift‑M and Shift‑B.
Incompleteness and Gödel Theorems
Here is an (updated May 19th, 2023) list
of exercises.
Exams (Czech)
Please bring a list of 14 exercises whose solution you are able to present.
Course contents in Spring 2022/23
- May 11th, 2023
- Completeness of the propositional calculus GJ. Two by-products of the proof:
the cut-elimination theorem and the finite model property. Kripke semantics of intuitionistic
predicate logic.
- May 9th, 2023
- The relationship between classical and intuitionistic logic: Glivenko's theorem. The disjunction
property. The calculus GJ.
- May 4th, 2023
- Constructing Kripke counterexamples. Every intuitionistic tautology is a (classical) tautology.
- May 2nd, 2023
- Kripke semantics for intuitionistic propositional logic.
- Apr 27th, 2023
- Essentially undecidable and essentially incomplete theories. Motivating considerations concerning
intuitionistic logic.
- Apr 25th, 2023
- Rosser's theorem. Craig's trick.
- Apr 20th, 2023
- Some connections of Gödel's 2nd theorem: Hilbert's program, uniqueness of fixed points, the variety of
formulas τ(z) defining the same theory.
- Apr 18th, 2023
- Gödel's second incompleteness theorem. Its formalization in PA.
- Apr 13th, 2023
- The self-reference theorem. Th(ℕ) is a non-arithmetical set. Gödel's self-referential sentence.
- Apr 11th, 2023
- Löb's derivability conditions. An ad hoc semantics showing that some consistency statements
are provable in PA.
- Apr 4th, 2023
- Facts about provability (the deduction theorem etc.) that are provable in PA. The formalized
completeness theorem.
- Mar 30th, 2023
- Formalization of provability and consistency in PA.
- Mar 28th, 2023
- Using balanced strings in some cases, the basic syntactic notions (variables, terms, formulas, substitutability,
the substitution function etc.) can be formalized in PA.
- Mar 23rd, 2023
- String functions, balanced strings.
- Mar 21st, 2023
- Partial recursive functions are closed under primitive recursion and course-of-values recursion.
An infinite subset of N is RE if and only if it is the range of some one-to-one function.
- Mar 16th, 2023
- Partial recursive functions are closed under composition and minimization.
- Mar 14th, 2023
- Some connections between RE sets and partial recursive functions: a set is a projection
of a Δ0 relation (or a projection of a recursive relation) if and only
if it is the domain of some partial recursive functions, and a nonempty subset of N is RE
if and only if it is the range of some recursive function.
- Mar 9th, 2023
- Recursively enumerable sets, partially recursive and recursive functions. The collection schema (with
some additions) can be used to construct an alternative axiomatization of PA.
- Mar 7th, 2023
- Further functions formalizable in PA: the number of bits
in the binary expansion of x (its graph is Δ0(PA)) and the number of primes
less than x (its graph is certainly Σ1(PA), but it is not known to
be Δ0(PA)).
- Mar 2nd, 2023
- Exponentiation can be formalized in PA. The formula y = zx
is Δ0(PA).
- Feb 28th, 2023
- The fact that irreducible numbers are the same as prime numbers is provable in PA. Properties of
powers of two are provable in PA as well. For example, the product of any two powers of two is again a power of two.
- Feb 23th, 2023
- The infinitude of irreducible numbers, and also Bézout's theorem are provable in PA.
- Feb 21th, 2023
- The least number principle (LNP) is provable in PA. More exactly,
over (Q + ∀x(x<S(x))) it is equivalent to
the induction schema. However, these two schemas are not equivalent over Q itself.
- Feb 16th, 2023
- Peano arithmetic PA. Using the schema Ind to prove the expected properties of operations and the
two order relations.
- Feb 14th, 2023
- Robinson arithmetic and its properties.
Chapters in Classical Logic
About
A seminar partly based on students' presentations, devoted to various
topics in metamathematics, non-classical logics, and proof theory.
The topics usually include: Provability Logic,
independence of the Paris-Harrington combinatorial
principle over Peano arithmetic,
relations between ZF, GB and the Kelly-Morse set theories,
upper and lower bounds for cut-elimination in
classical predicate logic, decision procedures
for intuitionistic propositional logic.
Here is the list of topics (in Czech) in the academic year 2022-23.
SIS: ALG500003.
Exams (Czech)
Ke zkoušce je třeba zvládnout témata dle vlastního výběru, která pokrývají
13 dvouhodinovek. Vlastní přednáška se do počtu 13 počítá trojnásobně.
Course contents in 2022/23 (Czech)
- Nezávislost Paris-Harringtonova principu na PA (D. Roubínek, 2. a 9.5.)
- Poslední kapitola v knize Handbook of Mathematical Logic.
- Intuicionistická predikátová logika (B. Heřmanová, 3.5 týdne od 4.4.)
- Sémantika intuicionistické predikátové logiky. Tři prenexní implikace, které nejsou intuicionisticky
logicky platné. Schéma DNS. Chování rovnosti, mimolehlost, axiomy stability.
- Kombinatorické principy nezávislé na PA (D. Roubínek, 4 týdny od 7.3.23)
- Kayova teorie PA-. Ramseyova věta. Důkaz nekonečné Ramseyovy věty převedením
na Königovo lemma.
- Hyperprosté množiny a nezávislá axiomatizovatelnost (J. Urbánek, 5 týdnů od 22.12.)
- Ekvivalentní definice hyperprosté množiny. Rekurzívně axiomatizovatelná teorie je nezávisle
axiomatizovatelná, právě když má
axiomatizaci { αi ; i ∈ N },
jejíž markant (tj. množina všech n takových, že αn+1 je dokazatelná
z konjunkce všech αi
pro i ≤ n) není hypersimple. To je právě tehdy,
když každá axiomatizace má tuto vlastnost.
- Aritmetická úplnost logiky dokazatelnosti (V. Švejdar, 5 týdnů od 15.11.)
- Varianty axiomatizace logiky GL. Logika GLS, která není normální, je kandidátem na
logiku, která axiomatizuje množinu všech ℕ-tautologií. GLS-modely. Kripkovská úplnost
logiky GLS. Solovayův důkaz aritmetické úplnosti obou logik.
- Metamatematika teorií množin (M. Blahynka, 3 týdny od 25.10.22)
- Gödelova-Bernaysova teorie množin GB a její konečná axiomatizovatelnost. Pravdivostní relace
a obsaditelná přirozená čísla. V GB není dokazatelná indukce pro všechny její formule, a je
nad ZF konzervativní vůči množinovým sentencím.
- Autoreference (V. Švejdar, 3 týdny)
- Aritmetizace dokazatelnosti. Löbovy podmínky. Gödelova, Henkinova a Löbova autoreferenční rovnice.
Přinejmenším první dvě mají jednoznačně určené řešení.
Interpretation of Gödel incompleteness theorems
About
A seminar based on students' presentations, devoted to reading papers
mainly by Smorynski and Detlefsen.
Here is a course announcement.
SIS: ALGV19006.
Course contents in 2023/24
- Další plány
- Timea Cochová: něco (cokoliv) na téma Gentzen, Adam Fürstenzeller: Smorynského
článek o Hilbertově programu.
- L. E. J. Brouwer: Life, Art and Mysticism (Šimon Pošta, od 16.11.)
- Detlefsenův článek Brouwerian intuitionism je teď volný.
- Detlefsenův článek z roku 79 (Timotej Šujan, od 2.10, 1.5 přednášky)
- Abstrakt (poskytnutý přednášejícím): Detlefsen v první části článku kritizuje takovou skeptickou
interpretaci (SI) Gödelovy druhé věty (G2), která by se snažila dojít ke skeptickému závěru, že víru
v konzistenci teorie T nelze odůvodnit důkazem sentence Con(T). Tvrdí, že tento
závěr z G2 žádným zřejmým způsobem neplyne, a tedy je potřeba dokázat určité pomocné předpoklady
SUP1 a SUP2. Následně se snaží ukázat, že platnost SUP1 vyvrací původní SI —
v jeho argumentu jsou ústřední pojmy jako reflexivní teorie
či (jeho vlastní pojem) epistemicky kompaktní teorie. V druhé části článku se snaží upravit SI tak,
aby unikla ze spárů jeho kritiky, avšak následně i takovou upravenou SI odmítá jako nepodloženou.
V třetí části krátce odmítá jako nepodloženou Resnikovu interpretaci G2 v kontextu
“nejsilnějšího/nejsložitějšího systému, který jsme ještě ochotní vážně zkoumat”, a to i kvůli
nejasnému popisu takového systému. V závěrečné části se pak snaží zachránit Hilbertův program tak, že
představuje tzv. omezené ω-pravidlo, které je dle něj v souladu s finitními metodami, které
Hilbertův program vyžaduje, a tvrdí, že např. Con(PA) je dokazatelné z PA+(ω-pravidlo).
- Gödelův článek z roku 1931 (L. Clowez, od 19.10., 2.5 přednášky)
- Kódování logické syntaxe, které je založeno na prvočíselných rozkladech přirozených čísel.
Definice kalkulu poměrně podobná dnešní definici kalkulu hilbertovského. Gödel pracuje s axiomatickou
teorií, která odráží takzvanou teorii typů. Viditelný rozdíl vůči dnešním přístupům je to, že nepracuje
s nějakým předem fixovaným jazykem. Pouze trvá na tom, že k dispozici jsou symboly pro nulu a následnickou
funkci. Článek pracuje s “rekurzívními” funkcemi, ale definici se shoduje s tím, čemu
se dnes říká primitivně rekurzívní funkce (dnes víme, že rekurzívní funkce lze ztotožnit s funkcemi
algoritmicky počitatelnými a je jich víc než primitivně rekurzívních). Jako celkem věrohodná tedy vypadá
historická hypotéza, že kromě vět o neúplnosti je článek důležitý také tím, že v něm byly zavedeny
primitivně rekurzívní funkce.
- Dnešní pohled na Gödelovy věty (V. Švejdar, 2 týdny)
- Trochu životopisných údajů ke Gödelovi. Ingredience důkazu jeho vět o neúplnosti: kódování syntaxe,
numerály, definovatelné množiny, Σ-formule a věta o Σ-úplnosti, predikát dokazatelnosti, věta
o autoreferenci.
Other activities
TeX/LaTeX, pdf, ...
FBUsersSetup
The Firebird SQL server comes with a command line utility
gsec
for manipulating database users. Using it, one can add, delete, or modify server's
users, where a modification practically means a password change.
FBUsersSetup.exe is written as an Inno Setup Windows
installer*)Unlike normal installers, it does not
write anything to the registry and does not leave any trace on your computer. So no
deinstallation is provided/needed.
and it basically is a gui wrapper for the gsec
utility.
As such, it has little additional functionality in comparison to gsec
.
However, it is easy to use, it can modify users on any computer in or outside the local network,
it never gets confused if there are parallel instances of Firebird running on the local computer,
and it can detect the properties of the local Firebird(s).
The utility is rather self-explanatory, the essential dialog pages being
- the selection
of a db server

On the local computer, TIGER, there are two parallel instances of Firebird.
One communicates through the standard port 3050, the other through port fb25 that translates
to 3060 in the services
file. The user can click on one of them in the listbox,
or say something else. Assume that, since he wants to modify users on the computer JUPITER where
Firebird communicates through port 3060, he deleted the text "TIGER" in the upper text field
and typed "JUPITER/3060" instead. In the two bottom text fields he has to specify the admin login.
In most cases, the admin would be SYSDBA. However, in Firebird 2.0 and later any user can
change his own password.
, and then
- a user action
The version of Firebird addressed as JUPITER/3060 is 2.5.2. Besides SYSDBA, there
are three other database users on that server. Since database user's first and last names are hardly ever needed,
FBUsersSetup somewhat misuses these data and codes the current time into them: JOSEPH and VIOLA
were inserted on Apr 14th, 2013. JASMIN was not inserted using FBUsersSetup and so her
first and last names are unknown. A new database user ADAM is now about to be inserted.
on that server.
DBUSubst
The utility DBUSubst.exe
can substitute one Firebird database user for
another. A substitution of user M for existing user D in a database
means that M becomes an owner of all database objects that user D owned,
he inherits all privileges that were granted to D, and he becomes a grantor
of all privileges that D granted. After being substituted out, user D
grants and is granted nothing, and owns no database object. If D was
an owner of a database then M becomes a new database owner.
Thus user substitution is a more general operation than database
ownership change.
Similarly as FBUsersSetup, DBUSubst.exe is written as a (harmless for
your computer) Inno Setup Windows installer. It basically is a wrapper for
the isql
command line utility and
some sql scripts. The scripts are available in DBUSubst.zip.
They are documented and thus contain a detailed information about the
substitution process. The document DBUSubst.pdf
contains some introductory information about the user privileges and some Firebird internals.
Two essential dialog pages of DBUSubst.exe are
- the database connection

Our database zoodata.fdb resides on a computer orel where
(the relevant instance of) Firebird communicates thru port 3060;
orel can be a computer in the local network or a
computer visible on internet. Of course, if it is not the local computer
on which DBUSubst is run, the browse button cannot be used; the path to zoodata.fdb
has to be typed.
When developing the database, we accessed it as a user LION.
This user is an owner of the database.
We also have another user VISITOR who is designed to have some restricted
access. Now when the database is about to be
deployed, it appears that the name LION was not wisely selected and that the
owner name should be ZOODBA. So we are about to substitute ZOODBA for LION.
ZOODBA already exists on the server in question, i.e. orel/3060.
with the NEW user's login, and then
- the login of
the dead
user

Since we were allowed to proceed to this dialog window, the login of
ZOODBA was verified against Firebird on orel/3060. It was also detected
that the version of this Firebird server is 2.5.1 and that the owner
of the database is LION, see the blue text labels in the bottom.
We are now selecting LION as the user who will be substituted by ZOODBA,
and we have to specify his password. The password of VISITOR is immaterial
for the current substitution process since VISITOR is not
a grantor of anything. However, if there were other users who
granted something to LION, further dialog windows asking about their
passwords would follow.
, i.e. the user who is to disappear.