Professional Documents
Culture Documents
Full Chapter Algebraic Perspectives On Substructural Logics Davide Fazio PDF
Full Chapter Algebraic Perspectives On Substructural Logics Davide Fazio PDF
https://textbookfull.com/product/a-general-algebraic-semantics-
for-sentential-logics-2nd-edition-josep-maria-font/
https://textbookfull.com/product/lectures-on-logarithmic-
algebraic-geometry-arthur-ogus/
https://textbookfull.com/product/algebraic-numbers-and-algebraic-
functions-first-edition-cohn/
Planning for Ecosystem Services in Cities Davide
Geneletti
https://textbookfull.com/product/planning-for-ecosystem-services-
in-cities-davide-geneletti/
https://textbookfull.com/product/acute-respiratory-distress-
syndrome-1st-edition-davide-chiumello-eds/
https://textbookfull.com/product/dynamic-stability-and-
bifurcation-in-nonconservative-mechanics-davide-bigoni/
https://textbookfull.com/product/reframing-institutional-logics-
substance-practice-and-history-1st-edition-alistair-mutch/
https://textbookfull.com/product/contemporary-perspectives-on-
relational-wellness-floriana-irtelli/
Trends in Logic 55
Davide Fazio
Antonio Ledda
Francesco Paoli Editors
Algebraic
Perspectives
on Substructural
Logics
Trends in Logic
Volume 55
TRENDS IN LOGIC
Studia Logica Library
VOLUME 55
Editor-in-Chief
Heinrich Wansing, Department of Philosophy, Ruhr University Bochum,
Bochum, Germany
Editorial Board
Arnon Avron, Department of Computer Science, University of Tel Aviv, Tel Aviv, Israel
Katalin Bimbó, Department of Philosophy, University of Alberta, Edmonton, AB, Canada
Giovanna Corsi, Department of Philosophy, University of Bologna, Bologna, Italy
Janusz Czelakowski, Institute of Mathematics and Informatics, University of Opole,
Opole, Poland
Roberto Giuntini, Department of Philosophy, University of Cagliari, Cagliari, Italy
Rajeev Goré, Australian National University, Canberra, ACT, Australia
Andreas Herzig, IRIT, University of Toulouse, Toulouse, France
Wesley Holliday, UC Berkeley, Lafayette, CA, USA
Andrzej Indrzejczak, Department of Logic, University of Lódz, Lódz, Poland
Daniele Mundici, Mathematics and Computer Science, University of Florence, Firenze, Italy
Sergei Odintsov, Sobolev Institute of Mathematics, Novosibirsk, Russia
Ewa Orlowska, Institute of Telecommunications, Warsaw, Poland
Peter Schroeder-Heister, Wilhelm-Schickard-Institut, Universität Tübingen, Tübingen,
Baden-Württemberg, Germany
Yde Venema, ILLC, Universiteit van Amsterdam,
Amsterdam, Noord-Holland, The Netherlands
Andreas Weiermann, Vakgroep Zuivere Wiskunde en Computeralgebra, University of Ghent,
Ghent, Belgium
Frank Wolter, Department of Computing, University of Liverpool, Liverpool, UK
Ming Xu, Department of Philosophy, Wuhan University, Wuhan, China
Jacek Malinowski, Institute of Philosophy and Sociology, Polish Academy of Sciences,
Warszawa, Poland
Assistant Editor
Daniel Skurt, Ruhr-University Bochum, Bochum, Germany
Founding Editor
Ryszard Wojcicki, Institute of Philosophy and Sociology, Polish Academy of Sciences,
Warsaw, Poland
The book series Trends in Logic covers essentially the same areas as the journal Studia Logica, that is,
contemporary formal logic and its applications and relations to other disciplines. The series aims at publishing
monographs and thematically coherent volumes dealing with important developments in logic and presenting
significant contributions to logical research.
Volumes of Trends in Logic may range from highly focused studies to presentations that make a subject
accessible to a broader scientific community or offer new perspectives for research. The series is open to
contributions devoted to topics ranging from algebraic logic, model theory, proof theory, philosophical logic,
non-classical logic, and logic in computer science to mathematical linguistics and formal epistemology. This
thematic spectrum is also reflected in the editorial board of Trends in Logic. Volumes may be devoted to
specific logical systems, particular methods and techniques, fundamental concepts, challenging open problems,
different approaches to logical consequence, combinations of logics, classes of algebras or other structures, or
interconnections between various logic-related domains. Authors interested in proposing a completed book or a
manuscript in progress or in conception can contact either christi.lue@springer.com or one of the Editors of the
Series.
Francesco Paoli
Editors
Algebraic Perspectives
on Substructural Logics
123
Editors
Davide Fazio Antonio Ledda
Dipartimento di Pedagogia Dipartimento di Pedagogia
Psicologia, Filosofia Psicologia, Filosofia
Università di Cagliari Università di Cagliari
Via Is Mirrionis, Cagliari, Italy Via Is Mirrionis, Cagliari, Italy
Francesco Paoli
Dipartimento di Pedagogia
Psicologia, Filosofia
Università di Cagliari
Via Is Mirrionis, Cagliari, Italy
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Acknowledgements
The editors gratefully acknowledge the following funding sources: Project “Per un’
estensione semantica della Logica Computazionale Quantistica—Impatto teorico e
ricadute implementative”, Regione Autonoma della Sardegna, (RAS: RASSR40341),
L.R. 7/2017, annualità 2017—Fondo di Sviluppo e Coesione (FSC) 2014–2020;
MIUR, within the projects PRIN 2017: “Logic and cognition. Theory, experiments,
and applications”, CUP: 2013YP4N3, and PRIN 2017: “Theory and applications of
resource sensitive logics”, CUP: 20173WKCM5 and gratefully acknowledge the
support of the Horizon 2020 program of the European Commission: SYSMICS
project, number: 689176, MSCA-RISE-2015. Finally, we thank the community of
researchers in the area of substructural logics and residuated structures, in particular
Hiroakira Ono and Constantine Tsinakis who played a decisive role in creating it and
in fostering the mutual collaboration among its members.
v
Contents
Editorial Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Davide Fazio, Antonio Ledda, and Francesco Paoli
Distributivity and Varlet Distributivity . . . . . . . . . . . . . . . . . . . . . . . . . 11
Paolo Aglianò
On Distributive Join Semilattices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Rodolfo C. Ertola-Biraben, Francesc Esteva, and Lluís Godo
Implication in Weakly and Dually Weakly Orthomodular Lattices . . . . 41
Ivan Chajda and Helmut Länger
Residuated Operators and Dedekind–MacNeille Completion . . . . . . . . . 57
Ivan Chajda, Helmut Länger, and Jan Paseka
PBZ*–Lattices: Ordinal and Horizontal Sums . . . . . . . . . . . . . . . . . . . . 73
Roberto Giuntini, Claudia Mureşan, and Francesco Paoli
EMV-Algebras—Extended MV-Algebras . . . . . . . . . . . . . . . . . . . . . . . . 107
Anatolij Dvurečenskij and Omid Zahiri
Quasi-Nelson; Or, Non-involutive Nelson Algebras . . . . . . . . . . . . . . . . 133
Umberto Rivieccio and Matthew Spinks
Hyperdoctrines and the Ontology of Stratified Semantics . . . . . . . . . . . 169
Shay Allen Logan
vii
Contributors
ix
x Contributors
Although an elective bond with algebra had been a characteristic trait of mathe-
matical logic since its inception—think of Boole’s work, to name the most obvious
example—the XXth century and the first two decades of the XXIst century witnessed
an increasing and ever more fruitful interplay between these two disciplines. The
trailblazing enquiries by Tarski, Lindenbaum, McKinsey and other authors into the
algebraic semantics of classical, intuitionistic and modal logics established bridges
between these provinces of logic and well-trodden algebraic territory, such as the
theories of Boolean algebras, Heyting algebras, and interior algebras [38, 48]. Later,
the birth of Abstract Algebraic Logic as an autonomous field of research, originating
from the confluence of the Polish logicians’ investigations into the theories of logical
matrices and consequence operators and the study of algebraisable logics and their
generalisations [6, 15, 22, 53], was only possible thanks to an extensive recourse to
universal algebra and lattice theory, as well as to the theories of the individual classes
of algebras acting as counterparts of the different logics under investigation.
Substructural logics [23, 39, 44, 46] are no exception to this trend. Here, too,
logic and algebra found fertile synergies after running on separate tracks for quite
some time. Substructural logics are usually defined via certain Gentzen-style sequent
calculi, or variants thereof, where the structural rules are suitably restricted or even
done away with—whence the “substructural” label. These logics did not emerge
together as a uniform corpus. Rather, this name started to enter common usage at
some point1 to refer to a largely heterogeneous family of pre-existing logics with
different origins and motivations:
• Lambek calculus was used since the late 1950s in mathematical linguistics and
especially in type-theoretic grammars [35];
• linear logic had been developed in the 1980s by J.Y. Girard in the context of
the proof theory for constructive logics, but presented interesting connections
with domain theory, the semantics of functional programming and other topics of
interest to computer scientists [26, 49];
• relevance logics had been introduced in the 1950s to solve paradoxes of material
implication and had obvious ramifications into the philosophy of logic and the
philosophy of language [1, 37];
• finally, fuzzy logics [5] had been intensively and fruitfully studied for some decades
before it was realised that they belonged to all intents and purposes to the sub-
structural family [9, 21, 41].
Observe that not all these logics were originally introduced by means of sequent
calculi: in some cases this kind of formulation was given only later, while in other
cases it is not even available now. Substructural logics are united by a series of “family
resemblances”, rather than by their falling under some common all-encompassing
definition. One such resemblance is the validity of some suitable generalisation of the
deduction-detachment equivalence in classical logic2 : ϕ · ψ entails χ iff ψ entails
ϕ\χ iff ϕ entails χ /ψ, for some (possibly non-idempotent and non-commutative)
conjunction · and some implications \, /. In any event, up until the 1990s substructural
logics were mainly investigated by syntactic methods, or by resorting to variants
of Kripke semantics [19, 43, 50]; with the exception of fuzzy logics, algebra was
occasionally employed but was far from being predominant.
Quite independently of these developments, in the 1930s, M. Ward and R.P. Dil-
worth [17, 51, 52] introduced residuated lattices as an abstraction from lattices of
ideals in rings. It is well-known that, given any ring with unit R and any two-sided
ideals I, J of R, upon defining
1 There is even something like an official birth date for the term “substructural logics”. This phrase
was coined by Kosta Došen on the occasion of the Tübingen conference on “Logics with restricted
structural rules” (7–8 October, 1990) co-organised with Peter Schroeder-Heister, which showcased
the Gotha of substructural research at the time.
2 In the commutative case the two implications always give a common result, denoted ϕ → ψ, and
the equivalence assumes the more familiar form: ϕ · ψ entails χ iff ϕ entails ψ → χ.
Editorial Introduction 3
n
I·J= ak bk : ak ∈ I, bk ∈ J, n 1 ;
k=1
I \J = {x ∈ R : I x ⊆ J } ;
J/I = {x ∈ R : x I ⊆ J } ,
such a product of ideals is a residuated operation with respect to set inclusion, with
the two divisions as residuals—namely, for any two-sided ideals I, J, K of R, we
have that:
I · J ⊆ K iff J ⊆ I \K iff I ⊆ K /J .
It soon became apparent that this umbrella notion is extremely wide-ranging and
can be appropriately tweaked so as to subsume classes of algebras of prime impor-
tance in algebra and elsewhere4 : lattice-ordered groups; Heyting algebras, hence
Boolean algebras as a special case; Chang’s MV-algebras [14]—and by this list we
have barely started to scratch the surface. At the same time, the structure theory of
residuated lattices is surprisingly robust and elegant, revealing as it does subtle and
profound connections between the structure theories of individual classes of algebras
falling under the concept [7, 8, 23, 39].
Yet, there is another aspect that is even more noteworthy in the present context.
The algebraic residuation equivalence looks amazingly similar to the deduction-
detachment equivalence previously mentioned in connection with substructural log-
ics. This is no coincidence. The operations in residuated lattices, indeed, can be
used to provide an algebraic semantics for connectives in substructural logics, with
product interpreting conjunction and the divisions interpreting implications. More
precisely, all the principal substructural logics turn out to be algebraisable with qua-
sivarieties of (expansions of) residuated lattices as equivalent algebraic semantics.
Thanks to the groundbreaking work by Hiroakira Ono, Constantine Tsinakis and
many other researchers, from the late 1990s onwards there has been a major revolu-
tion in the methods and approaches employed to investigate substructural logics [7,
24, 33, 42]. Not only has the study of residuated lattices grown into a very sophis-
ticated and mathematically advanced theory, but algebraic methods have been put
3 This is not the original definition of a residuated lattice given by Ward and Dilworth, but (essentially)
the definition due to Jipsen and Tsinakis [33] that is nowadays in current use. Observe that other
different usages of the term can be frequently encountered in the literature: for a comparison, see
e.g. [39].
4 In particular, to achieve this goal it may be necessary to expand the signature by an extra constant
to good use with increasing frequency to solve open problems concerning substruc-
tural logics, including decidability, interpolation, and completeness. As a matter of
fact, the algebraisability relation has proved to be beneficial in both directions: many
interesting problems concerning classes of residuated lattices have been solved using
proof-theoretic methods, i.e. by working on the associated logics first, and then by
translating back the outputs into the algebraic framework thanks to the bridge granted
by algebraisability [23, 39].
As we have just seen, the assortment of nonclassical logics that can be encompassed
under the substructural heading is impressive: it nearly exhausts the range of the
available alternatives to classical logic (including, as a bonus, classical logic itself
as a limiting case). There is a remarkable outlier, though. However much we may be
willing to stretch the meaning of “substructural”, quantum logics [16, 34] can hardly
be counted in this category. Here’s the reason why: in mainstream quantum logics,
there are no connectives of conjunction and implication that are related by something
like the deduction-detachment (residuation) equivalence. Although many quantum
logics are algebraisable, their equivalent algebraic semantics fail to be quasivarieties
of residuated lattices. As a consequence, quantum logics and substructural logics have
by and large evolved quite separately. In itself, this circumstance is lamentable—the
more interplay you have between two neighbouring fields, the more likely it is that
methods and concepts from either area can be exported into the other, providing
fresh insights into its open problems and contributing to its overall development.
Occasionally, productive exchanges between these areas have emerged; yet, they
all have fallen short of generating a common systematic perspective. We cite a few
examples below.
1. Despite the recurrent slogan “Quantum logics have no implication”, there have been
important studies on implications in orthomodular logic, or in weaker quantum
logics, highlighting inferential patterns to be found also in substructural logics
[31, 32, 45].
2. All the main quantum logics are non-distributive. So are many substructural logics.
Thus, it is not too surprising that proof-theoretic and semantic methods that work
well in the former context can also be adapted to the latter, or vice versa. An example
is phase semantics for linear logic [26], which displays evident similarities with the
Kripke-style semantics for orthologic independently found by Dishkant, Goldblatt
and Dalla Chiara [16, 18, 30].
3. There have been some attempts to introduce common generalisations of quantum
logics and substructural logics. One such effort is Sambin’s basic logic [2], defined
via a sequent calculus that combines the restrictions on structural rules typical
of substructural calculi with the restrictions on contexts characterising Gentzen
Editorial Introduction 5
systems for quantum logics [40, 47]. Basic logic is thus a common sublogic of
orthologic and of subexponential intuitionistic linear logic.
4. Fuzzy logics stand out among substructural logics because they can be given a
semantics in terms of residuated lattices over the [0, 1] real interval.5 In this respect,
they are close to the so-called unsharp quantum logics [16], which extend the
Birkhoff-von Neumann approach in so far as propositional formulas are assigned
meanings in more general structures than the lattices of closed subspaces of (equiv-
alently, of projection operators on) complex separable Hilbert spaces. The numer-
ous striking similarities between fuzzy logics and unsharp quantum logics have
been duly stressed in the literature [16, 20, 28, 36].
5. Finally, quantum structures have been used by Girard to interpret linear logical
proofs, obtaining a particular variant of denotational semantics for linear logic
[27].
Some recent results and approaches, however, hold promise to do better and to
provide a more comprehensive perspective that may be useful not only in establish-
ing a common framework subsuming both quantum and substructural logics, but
also in opening to researchers in either area the possibility of importing methods
and techniques “from across the street”. On the one hand, Chajda and Länger [10]
proved that the variety of orthomodular lattices is term-equivalent to a certain variety
of pointed left-residuated lattice-ordered groupoids. This suggests that the power-
ful techniques (e.g. nuclear retractions, quasi-completions) developed for residuated
lattice-ordered groupoids [13, 23, 25] can be generalised to the case where a single
residual is present. If so, it is not out of the question that there can be important
repercussions on the long-standing open problems as to whether orthomodular lat-
tices admit completions, or as to whether they have the finite model property.
Also, many quantum structures are not residuated, but they are operator residu-
ated, according to a concept developed in [11] and investigated in this volume in the
paper by Chajda and Paseka. This interesting generalisation of residuation has been
effectively employed with various uses, see e.g. [12].
5 There is an ongoing discussion on the correct definition of a fuzzy logic: see [3, 4, 29].
6 D. Fazio et al.
(JAIST) on 10–14 November 1999. Subsequent meetings were held at the JAIST
(2002), at the Jagiellonian University in Kraków (2006), at the JAIST again (2010)
and at the La Trobe University in Melbourne (2014).
AsubL Take 6 was hosted by the Department of Pedagogy, Psychology, Philosophy
of the University of Cagliari on June 11–13, 2018. Its programme included invited
talks by Hiroakira Ono, Constantine Tsinakis (cancelled), H.P. Sankappanavar, Ana-
tolij Dvurecenskij, Peter Schroeder-Heister and Tomasz Kowalski, as well as 36
contributed talks focussing on a wide range of different themes: residuated lattices
and their properties, individual classes of residuated lattices (including MV-algebras
and other classes of algebras of fuzzy logic), proof theory for substructural logics,
philosophical and game-theoretical interpretations of substructural logics, topics in
universal algebra and abstract algebraic logic with direct applications to substruc-
tural logics. Many talks were devoted to the connections between substructural and
quantum logics, or between residuated lattices and quantum structures, thus pro-
viding a stimulus for a greater interaction between the substructural and quantum
communities.
This volume includes some of the papers presented at the conference, representing
many of the topics discussed during the sessions.
Four crucial threads in the current research into substructural logics and residuated
lattices lie at the centre of the papers that have been collected into the present book.
1. Reducts of residuated lattices. Residuated lattices are algebraic structures with a
very rich signature. Many of their properties, however, rather than depending on
the presence of all the operation symbols, are due to more fundamental features of
certain reducts of theirs. Although quite a lot is known about these aspects, at least
for some particular classes of residuated lattices, practitioners are at work to obtain
further insights. The paper by Paolo Aglianò, “Distributivity and Varlet distribu-
tivity”, explores the relationships between distributivity, Varlet distributivity and
divisibility in residuated semilattices, with applications to some well-known sub-
classes such as hoops. The paper by Rodolfo C. Ertola-Biraben, Francesc Esteva,
and Lluis Godo, “On distributive join-semilattices”, conceives a notion of distribu-
tivity for join semilattices that is motivated by Gentzen’s disjunction elimination
rule in natural deduction, and carefully compares it to many other notions present in
the literature. Moreover, the authors investigate this notion of distributivity in join
semilattices with arrow, the algebraic structures corresponding to the disjunction-
implication fragment of intuitionistic logic.
2. Relationships between substructural logics and quantum logics. Some articles
underscore the relationships between substructural and quantum logics, or between
residuated lattices and quantum structures, about which much was said above. In
“Implication in weakly and dually weakly orthomodular lattices”, Ivan Chajda
Editorial Introduction 7
References
1. A.R. Anderson, N.D. Belnap Jr., Entailment: The Logic of Relevance and Necessity (Princeton
University Press, Princeton, 1975)
2. G. Battilotti, G. Sambin, Basic logic and the cube of its extensions, in Logic and Foundations
of Mathematics, ed. by A. Cantini et al. (Kluwer Academic Publishers, Dordrecht, 1999), pp.
165–186
3. L. Bĕhounek, In which sense is fuzzy logic a logic for vagueness, in Logics for Reasoning
about Preferences, Uncertainty, and Vagueness, ed. by T. Lukasiewicz et al. (2014), pp. 26–38
4. L. Bĕhounek, P. Cintula, Fuzzy logic as the logic of chains. Fuzzy Sets Syst. 157, 604–610
(2006)
5. L. Bĕhounek, P. Cintula, P. Hàjek, Introduction to mathematical fuzzy logic, in Handbook of
Mathematical Fuzzy Logic, ed. by P. Cintula et al., vol. 1 (College Publication, London, 2011)
6. W.J. Blok, D. Pigozzi, Algebraizable Logics. Memoirs of the American Mathematical Society
(American Mathematical Society, Providence, 1989)
7. K. Blount, C. Tsinakis, The structure of residuated lattices. Int. J. Algebra Comput. 13(4),
437–461 (2003)
8. M. Botur, J. Kühr, L. Liu, C. Tsinakis, The Conrad program: from -groups to algebras of
logic. J. Algebra 450, 173–203 (2016)
9. E. Casari, Comparative logics and Abelian -groups, in Logic Colloquium’88, ed. by R. Ferro
et al. (North Holland, Amsterdam, 1989), pp. 161–190
10. I. Chajda, H. Länger, Orthomodular lattices can be converted into left-residuated -groupoids.
Miskolc Math. Notes 18, 685–689 (2017)
11. I. Chajda, H. Länger, Residuated operators in complemented posets. Asian-Eur. J. Math. 11,
1850097 (2018)
12. I. Chajda, H. Länger, Residuation in modular lattices and posets. Asian-Eur. J. Math. 12(2),
1950092 (2019)
13. A. Ciabattoni, N. Galatos, K. Terui, Algebraic proof theory for substructural logics: cut-
elimination and completions. Ann. Pure Appl. Log. 163, 266–290 (2012)
14. R. Cignoli, I.M.L. D’Ottaviano, D. Mundici, Algebraic Foundations of Many-Valued Reason-
ing. Trends in Logic—Studia Logica Library (Kluwer Academic Publishers, Dordrecht, 1999)
15. J. Czelakowski, Protoalgebraic Logics. Trends in Logic–Studia Logica Library, vol. 10 (Kluwer
Academic Publishers, Dordrecht, 2001)
16. M.L. Dalla Chiara, R. Giuntini, R. Greechie, Reasoning in Quantum Theory. Sharp and Unsharp
Quantum Logics. Trends in Logic—Studia Logica Library (Kluwer Academic Publishers, Dor-
drecht, 2004)
17. R.P. Dilworth, Non-commutative residuated lattices. Trans. Am. Math. Soc. 46, 426–444 (1939)
18. H. Dishkant, Semantics of the minimal logic of quantum mechanics. Stud. Log. 30, 17–29
(1972)
19. K. Dośen, Sequent systems and groupoid models, I. Stud. Log. 47, 353–385 (1988)
20. A. Dvurečenskij, S. Pulmannová, New Trends in Quantum Structures (Kluwer Academic Pub-
lishers/Ister Science, Dordrecht/Bratislava, 2000)
21. F. Esteva, L. Godo, A. García-Cerdaña, On the hierarchy of t-norm based residuated fuzzy
logics, in Beyond Two: Theory and Applications of Multiple-Valued Logic, ed. by M.C. Fitting,
E. Orlowska (Physica, Heidelberg, 2003), pp. 251–272
22. J.M. Font, Abstract Algebraic Logic: An Introductory Textbook (College Publications, London,
2016)
23. N. Galatos, P. Jipsen, T. Kowalski, H. Ono, Residuated Lattices: An Algebraic Glimpse at Sub-
structural Logics. Studies in Logic and the Foundations of Mathematics (Elsevier, Amsterdam,
2007)
24. N. Galatos, H. Ono, Algebraization, parametrized local deduction theorem and interpolation
for substructural logics over FL. Stud. Log. 83, 279–308 (2006)
25. J. Gil Férez, L. Spada, C. Tsinakis, H. Zhou, Join completions of partially ordered algebras.
Annals. Pure. Appl. Logic. 171(10), 102842 (2020)
Editorial Introduction 9
26. J.-Y. Girard, Linear logic. Theor. Comput. Sci. 50, 1–102 (1987)
27. J.-Y. Girard, Between logic and quantic: a tract, in Linear Logic in Computer Science, ed. by
P. Ruet et al. (Cambridge University Press, Cambridge, 2004), pp. 346–390
28. R. Giuntini, Quantum MV algebras. Stud. Log. 56, 393–417 (1996)
29. R. Giuntini, F. Paoli, H. Freytes, A. Ledda, G. Sergioli, What is fuzzy logic - and why it matters
to us, in On Fuzziness: A Homage to Lotfi A. Zadeh (Springer, Berlin, 2013), pp. 211–215
30. R. Goldblatt, Semantical analysis of orthologic. J. Philos. Log. 3, 19–35 (1974)
31. G.M. Hardegree, The conditional in quantum logic, in Logic and Probability in Quantum
Mechanics, ed. by P. Suppes (Reidel, Dordrecht, 1976), pp. 55–72
32. G.M. Hardegree, Quasi-implication algebras I. Algebra Univers. 12, 30–47 (1981)
33. P. Jipsen, C. Tsinakis, A survey of residuated lattices, in Ordered Algebraic Structures, ed. by
J. Martinez (Kluwer, Dordrecht, 2002), pp. 19–56
34. G. Kalmbach, Orthomodular Lattices (Academic, New York, 1983)
35. J. Lambek, The mathematics of sentence structure. Am. Math. Mon. 65, 154–170 (1958)
36. A. Ledda, M. Konig, F. Paoli, R. Giuntini, MV algebras and quantum computation. Stud. Log.
82(2), 245–270 (2006)
37. E. Mares, Relevant Logic (Cambridge University Press, Cambridge, 2004)
38. J.C.C. McKinsey, A. Tarski, On closed elements in closure algebras. Ann. Math. 47(2), 122–162
(1946)
39. G. Metcalfe, F. Paoli, C. Tsinakis, Ordered algebras and logic, in Uncertainty and Rationality,
ed. by H. Hosni, F. Montagna. Publications of the Scuola Normale Superiore di Pisa, vol. 10
(2010), pp. 1–85
40. H. Nishimura, Sequential method in quantum logic. J. Symb. Log. 45, 339–352 (1980)
41. H. Ono, Structural rules and a logical hierarchy, in Mathematical Logic, ed. by P.P. Petkov
(Plenum, New York, 1990), pp. 95–104
42. H. Ono, Substructural logics and residuated lattices: an introduction, in 50 Years of Studia
Logica, ed. by V.F. Hendricks, J. Malinowski. Trends in Logic—Studia Logica Library, vol.
20 (Kluwer Academic Publishers, Dordrecht, 2003), pp. 177–212
43. H. Ono, Y. Komori, Logics without the contraction rule. J. Symb. Log. 50, 169–201 (1985)
44. F. Paoli, Substructural Logics: A Primer. Trends in Logic—Studia Logica Library, vol. 13
(Kluwer Academic Publishers, Dordrecht, 2002)
45. R. Piziak, Orthomodular lattices as implication algebras. J. Philos. Log. 3, 413–418 (1974)
46. G. Restall, An Introduction to Substructural Logics (Routledge, London, 2000)
47. J. Schülte Monting, Cut elimination and word problem for varieties of lattices. Algebra Univers.
12, 290–321 (1981)
48. A. Tarski, Foundations of the calculus of systems (Engl. transl.), in Logic, Semantics, Meta-
mathematics (Oxford University Press, New York, 1956), pp. 342–383
49. A.S. Troelstra, Lectures on Linear Logic. CSLI Lecture Notes, Stanford (1992)
50. H. Wansing, The Logic of Information Structures. Lecture Notes in Artificial Intelligence, vol.
681 (Springer, Berlin, 1993)
51. M. Ward, R.P. Dilworth, Residuated lattices. Proc. Natl. Acad. Sci. 24, 162–164 (1938)
52. M. Ward, R.P. Dilworth, Residuated lattices. Trans. Am. Math. Soc. 45, 335–354 (1939)
53. R. Wojcicki, Theory of Logical Calculi (Kluwer Academic Publishers, Dordrecht, 1988)
Distributivity and Varlet Distributivity
Paolo Aglianò
Abstract In this note we study the relationships between three properties of resid-
uated (meet) semilattices, i.e.: (1) being divisible, (2) being distributive (3) being
v-distributive.
1 Preliminaries
In this note we study the relationships between three properties of residuated (meet)
semilattices (whose definition is found below), i.e.: (1) being divisible, (2) being
distributive, (3) being v-distributive. While the connection between the first two
properties has been already explored in [1], the third deals with a different aspect
worth investigating.
A semilattice ordered residuated monoid (shortly a residuated semilattice) is
an algebra A = A, ∧, ·, /, \, 1 where
1. A, ∧ is a semilattice;
2. A, ·, 1 is a monoid;
3. / and \ are the left and right residuation w.r.t. ·.
In other words, for a, b, c ∈ A
This implies at once that a ≤ b if and only if a\b ≥ 1, if and only if b/a ≥ 1.
Residuated semilattices are clearly related to residuated lattices introduced in [5] and
investigated at length in many other papers (see [11] and the bibliography therein);
residuated lattices have a very rich structure that is largely inherited by residuated
semilattices. Residuated semilattices form a variety, whose axiomatization can be
easily derived from the existing ones for residuated lattices (see again [5] or [13]).
P. Aglianò (B)
DIISM, University of Siena, Siena, Italy
e-mail: agliano@live.com
Here we will simply state all the equations we need and that can be easily proved
using the above definition.
To save space it is convenient to define the truncated residuations
The operations / and /1 are monotonic in the first argument and antimonotonic in the
second argument; the operations \ and 1\ are monotonic in the second argument and
antimonotonic in the first argument. Moreover:
Proof Almost all the equations are straightforward consequences of the definition;
however the derivation of (10) is less direct, so we will prove one half of it, just to
illustrate the technique. Let a, b ∈ A; if in (7) we set x = a\b, y = b then
a\b ≤ (b/(a\b))\b.
The following lemma can be easily deduced from the results in [11]; we include a
proof for reader’s convenience.
Lemma 2 For a residuated semilattice A the following are equivalent:
1. A is divisible;
2. if a, b ∈ A and a ≤ b, then (a/b)b = a = b(b\a);
3. if a, b ∈ A, then (a/1 b)b = a ∧ b = b(b 1\a).
Proof Assume (1) and let a ≤ b; then there is a c with a = cb which implies c ≤ a/b.
Thus a = cb ≤ (a/b)b; but the converse is always true so equality holds. A similar
argument shows that b(b\a) = a and (2) holds.
Assume (2) and note that for any u ≥ 1 we have (1/u)u = 1; if also u 2 = u then
u = 1 · u = (1/u)u 2 = (1/u)u = 1.
Since for any a ∈ A a/a is idempotent and greater than 1, it follows that a/a = 1.
Now since a ∧ b ≤ b, from (2) follows that
and the other equality follows by using the corresponding properties of \ and 1\.
Finally that (3) implies (4) is obvious.
While the concept of being distributive is mainly used for lattices, it is possible
to define an analogous concept for semilattices: a semilattice L is distributive if,
for all a, b, c ∈ L, if a ≥ b ∧ c, then there exist b , c ∈ L with b ≤ b , c ≤ c and
b ∧ c = a. This is a real generalization: if a lattice is distributive as a semilattice, then
it is a distributive lattice since it must omit the two five-element lattices M3 and N5
(see [12] for an extended discussion). Distributivity is not inherited by subalgebras
or homomorphic images; it is however inherited by retracts [12] and that can be
used to show that the direct product of distributive residuated semilattices is again
distributive.
Theorem 1 ([1]) Any integral and divisible residuated semilattice is distributive.
In a very old paper [14] J. Varlet considered a different kind of distributivity.
He stated the property for residuated partially ordered groupoids (pogroupoids). A
residuated pogroupoid A = A, ·, /, \, ≤ is v-distributive (short for Varlet distribu-
tive) if for any a, b, c ∈ A if bc ≤ a then there are b ≥ b and c ≥ c with b c = a;
a v-distributive residuated semilattice is just one in which its residuated pogroupoid
reduct is v-distributive. We note in passing that this property (and related ones)
has been considered for quantum structures under the name of Riesz decomposition
property, see [8]. J. Varlet [14] characterized v-distributive pogroupoids by means
of equations; therefore is no surprise that v-distributive residuated semilattices form
a variety.
14 P. Aglianò
Proof It is very easy to show the 2. and 3. are equivalent, just using the elementary
properties of residuation; so we will show that 1. and 2. are equivalent. Suppose that
2. holds and let a, b, c ∈ A with ab ≤ c; then b ≤ a\c and from Lemma 1(7) we get
a ≤ c/(a\c).
By 2.
(c/(a\c))(a\c) = c
so A is v-distributive.
Conversely suppose that A is v-distributive; if a, b, c ∈ A we observe that
(c/(a\c))(a\c) ≤ c
c = uv = (c/(a\c))v
and by residuation v ≤ (c/(a\c))\c = a\c by Lemma 1(10). Hence v = a\c and (2)
holds.
2 The Relationships
• if a ≥ 1, then a is invertible.
Let G(A) the set of invertible elements of A and I(A) the set of integral elements
of A. Observe that G(A) ∩ I(A) = {1}, since if a ∈ G(A), then a(1/a) = 1 and if
a ∈ I(A), then 1/a = 1. Moreover I(A) is an integral subalgebra of A, while G(A) is
closed under {/, \, ·} but not in general under meets. If G(A) = A, then it is obvious
that A is an -group where a/b = ab−1 and a\b = a −1 b, i.e. a divisible residuated
lattice that in general is not integral. However, any divisible residuated lattice is a
distributive lattice [9] so it is distributive as a semilattice. Moreover for any a, b ∈ A
(b/(a\b))(a\b) = b
and A is v-distributive.
If A has a bottom ⊥ then ⊥/⊥ is clearly a top element. If A has a top element
, then 1 ≤ implies ≤ and so equality holds; however from (the proof of)
Lemma 2 a divisible residuated semilattice cannot have idempotent elements larger
than 1 except for 1 itself. So if such a semilattice has a top, then it must be 1, i.e.
the semilattice must be integral so 1. applies. If A is finite then it has a top (and a
bottom) so the previous case applies.
16 P. Aglianò
To prove 4. we observe that if the least upper bound of any two elements exists in A,
then A is the reduct of a residuated lattice. This is not entirely obvious, in that we have
to prove that the join behaves correctly w.r.t. the residuals, but it is true nevertheless
(see for instance [5]). So let A the divisible residuated lattice of which A is a reduct;
then G(A) = G(A ) and I (A) = I (A ); however, since A is a divisible residuated
lattice, G(A ) is an -group and A ∼ = G(A ) × I (A ) ([9], Theorem 5.2). So A is
a direct product of two v-distributive residuated lattices and hence is v-distributive.
Since v-distributivity is characterized by an equation involving only operations in
the type of A, the latter is v-distributive as well. Finally if A is representable, then
by the above observation is (term equivalent to) a residuated lattice; since it is also
divisible the previous argument applies.
(x → y) ∨ (y → x) ≈ 1.
Distributivity and Varlet Distributivity 17
So we have only to check that if A satisfies that equation, then so does A∗ , but the
verification is straightforward. Moreover:
Proof Let A∗0 , A∗1 be the subsets of A∗ whose first coordinate is 0 or 1, respectively.
Then A∗0 ∪ A∗1 = A and for all x ∈ A∗0 , y ∈ A∗1 is x ≤ y. Let then x, y, z ∈ A∗ with
y ∧ z ≤ x; if x, y, z ∈ A∗1 then there is nothing to prove. The same if x, y, z ∈ A∗0
since the ordering there is the dual of the ordering in A and the dual of a distributive
lattice is still distributive.
Suppose then that x ∈ A∗1 and z ∈ A∗0 ; then for all y ∈ A∗ , y ∧ z ≤ x. However
x ∨ y ≥ y, x ≥ z and (x ∨ y) ∧ x = x, so distributivity holds. There are no other
distinct possibilities, hence the proof is complete.
(x → y) → y ≈ (y → x) → x;
Lemma 4 ([3]) Let A be a basic hoop; then A∗ is a basic hoop if and only if A is
cancellative.
It follows that we can obtain a non divisible residuated lattice by rotating any non
cancellative basic hoop, but:
((1, a → 0, b) → 0, b) → ((1, a → 0, b) = (0, ab → 0, b) → 0, ab
= 1, b → ab0, ab
= 0, (b → ab) → ab
= 0, b ∨ ab = 0, b,
where we used that A∗ is a Wajsberg hoop in the second coordinate. The other case
is x = 0, u, y = 0, v with u < v; then
((0, u → 0, v) → 0, v) → (0, u → 0, v) = (1, v → u → 0, v) → 1, v → u
= 0, (v → u)v1, v → u
= 0, (v → u) → (v → u)v
= 0, (v → u) → u = 0, v
Then it is easily checked that [0, 1] becomes a totally ordered hoop, called a product
hoop; if we rotate it, then for any x ∈ [0, 1)
(x → 0x) → 0x = (x → 0) → 0 = 0 → 0 = 1 = x
so its disconnected rotation cannot be v-distributive. The reader can check that to
obtain a non representable example we can apply the same procedure to any gener-
alized Skolem lattice as described in [6].
Our last example will be a non distributive but v-distributive residuated semilattice.
Such a semilattice cannot be a divisible lattice o a divisible and integral semilattice (by
[1]) or representable (since the underlying semilattice structure is a subdirect product
of chains, hence distributive). However it can be in principle integral and commutative
and our example is exactly of this kind. The universe of A is {0, a, b, c, d, 1} and the
operation tables are:
Distributivity and Varlet Distributivity 19
c
a
b
· 0abcd 1 → 0ab cd 1
0 00000 1 0 111 11 1
a 0a00a a a c1c c1 1
b 00bbb b b aa1 11 1
c 00bbb c c aad 11 1
d 0abbd d d 0ac c1 1
1 0abcd 1 1 0ab cd 1
The reader may check that A is integral and v-distributive but the order structure
is the one in Fig. 1. Hence A is an integral v-distributive residuated semilattice that
is not distributive; thus by Theorem 1 it cannot be divisible.
We close this section with an observation. Let V be the variety of integral, com-
mutative and v-distributive residuated semilattices. By Theorem 2 V is axiomatized,
modulo integral and commutative residuated semilattices, by the single equation
((x → y) → y)(x → y) ≈ y.
By Lemma 2, V contains the variety H of hoops and by the example above it contains
it properly. By some preliminary results, too raw to be reported here, it seems that V
inherits some interesting properties from the variety of hoops. This is a path that we
intend to follow in future investigations.
Acknowledgements We thank Peter Jipsen for drawing our attention to the paper [14], thus orig-
inating this investigation. We thank also the anonymous referee who caught a mistake in the first
version and forced us to write a much better paper.
20 P. Aglianò
References
1. P. Aglianò, A short note on divisible residuated semilattices, 2019, To appear in Soft Computing
2. P. Aglianò, I. Ferreirim, F. Montagna, Basic hoops: an algebraic study of continuous t-norms.
Studia Logica 87, 73–98 (2007)
3. P. Aglianò, S. Ugolini, MTL-algebras as rotations of basic hoops. J. Logic Comput. (2019).
https://doi.org/10.1093/logcom/exz005
4. W.J. Blok, I. Ferreirim, On the structure of hoops. Algebra Univ. 43, 233–257 (2000)
5. K. Blount, C. Tsinakis, The structure of residuated lattices. Internat. J. Algebra Comput. 13(4),
437–461 (2003)
6. M. Busaniche, R. Cignoli, Constructive logic with strong negation as a substructural logic. J.
Logic Comput. 20, 761–793 (2008)
7. M. Busaniche, M. Marcos, S. Ugolini, Representation by triples of algebras with an MV-retract.
Fuzzy Sets Syst. (2018). https://doi.org/10.1016/j.fss.2018.10.024
8. A. Dvurečenskij, T. Vetterlein, Pseudoeffect algebras I. Basic properties. Inter. J. Theor. Phys.
40, 685–701 (2001)
9. N. Galatos, C. Tsinakis, Generalized MV-algebras. J. Algebra 283, 254–291 (2005)
10. S. Jenei, On the structure of rotation-invariant semigroups. Arch. Math. Logic 42, 489–514
(2003)
11. P. Jipsen, C. Tsinakis, A survey of residuated lattices, in Ordered Algebraic Strucures, ed. by
J. Martinez (Kluwer Academic Publisher, Dordrecht, 1982), pp. 19–56
12. J. Rhodes, Modular and distributive semilattices. Trans. Am. Math. Soc. 201, 31–41 (1975)
13. C. Van Alten, Representable Biresiduated Lattices. J. Algebra 247, 672–691 (2002)
14. J. Varlet, On distributive residuated groupoids. Semigroup Forum 6, 80–85 (1973)
On Distributive Join Semilattices
1 Introduction
Different notions of distributivity for semilattices have been proposed in the literature
as a generalization of the usual distributive property for lattices. As far as we know,
notions of distributivity for semilattices have been given, in chronological order, by
Grätzer and Schmidt [9] in 1962, by Katriňák [12] in 1968, by Balbes [1] in 1969,
by Schein [16] in 1972, by Hickman [11] in 1984, and by Larmerová and Rachůnek
[14] in 1988. Following the names of its authors, we will use the terminology GS-,
K-, B-, Sn -, H-, and LR-distributivity, respectively.
In this paper, motivated by Gentzen’s disjunction elimination rule in his Natural
Deduction calculus, and reading inequalities with meet in a natural way, we conceive
another notion of distributivity for join semilattices, that we call ND-distributivity.
We aim to find out whether it is equivalent to any of the notions already present in
the literature. In doing so, we also compare the different notions of distributivity for
R. C. Ertola-Biraben (B)
CLE-UNICAMP, Campinas, SP 13083-859, Brazil
e-mail: rcertola@cle.unicamp.br
F. Esteva · L. Godo
IIIA-CSIC, 08193 Belaterra, Spain
e-mail: esteva@iiia.csic.es
L. Godo
e-mail: godo@iiia.csic.es
Alle mir bekannten Exemplare 1 von den Sangi Inseln weichen durch
den weniger behaarten Schwanz und die wenig behaarten Tarsen
von dem gleich grossen T. fuscus ab, sie nähern sich also darin der
Philippinen-Form mit ihrem ganz spärlich und kurz behaarten (und
unbeschuppten) Schwanz und ihren so gut wie nackten Tarsen,
während fuscus gut behaarte Tarsen und einen sehr stark behaarten
Schwanz hat. Die langen und dunklen Haare des Schwanzes
reichen bei sangirensis proximal nicht so weit und die Haare sind
kürzer. Die Beschuppung ist dieselbe wie bei fuscus. Das Museum
besitzt ein Exemplar von Siao und eins von Gross Sangi, das
Berliner, Wiener und Braunschweiger je eins von Gross Sangi
(erstere 4 aus meinen Sammlungen, letzteres von Dr. P l a t e n ) mit
denselben Charakteren, 2 das Leidener (Cat. XI, 81 1892) eins von
„Sangi“, von dem Dr. J e n t i n k so freundlich war mir mitzutheilen,
dass der Schwanz und der Tarsus weniger behaart seien als bei
Celebes Exemplaren. Es liegt hierin also eine insulare Abweichung
und eine Hinneigung zur Philippinen Form. Ich hoffe später eine
Abbildung der Art geben zu können.
Tafel IV
Ich beschrieb diese Art Abh. Mus. Dresden 1894/5 Nr. 1 und habe
dem Gesagten wenig hinzuzufügen, da die Abbildung in n. Gr. zur
weiteren Erkennung der Merkmale genügen dürfte. Nur über die
Behaarung des Schwanzes möchte ich noch einige Worte sagen, da
diese, der Natur der Sache nach, in der Abbildung nicht deutlich
genug wiedergegeben werden konnte. Die proximalen ¾ des
Schwanzes sind fast nackt, nur mit spärlich und einzeln stehenden,
kaum 1 mm langen weissen Härchen besetzt; am distalen Viertel
werden sie allmählich bis 3 mm lang und an den distalen 4
Centimetern stehen sie eng aneinander und sind bräunlich gefärbt.
— Das Museum erhielt inzwischen 2 weitere Exemplare von den
Philippinen, und zwar noch eins von Samar durch Dr.
S c h a d e n b e r g und eins von Nord Mindanao durch Dr. R i z a l .
Das rothbraune Gesicht und überhaupt die braunere Farbe ist bei
allen auffallend, und sie sind hierdurch zusammen mit den fast
nackten Tarsen und dem wenig behaarten Schwanze leicht von
anderen Tarsiern zu unterscheiden.
[Inhalt]
9. Tarsius spectrum (Pall.)
Tafel V und VI
Ein altes Männchen und ein junges Weibchen sind auf Tafel V in ⅕–
⅙ n. Gr. abgebildet.
Da mir das Exemplar 2310 in Spiritus zukam, so liess ich die (linke)
Vola und Planta, ihres bemerkenswerthen Oberflächenreliefs wegen,
photographiren und bilde sie Tafel VI Figur 2 und 3 in n. Gr. ab. Ein
auffallender Unterschied mit anderen von mir daraufhin untersuchten
Paradoxuri besteht darin, dass die Tastballen bei P. musschenbroeki
glatt, bei jenen gefeldert sind.