Theory and Applications of

Satisfiability Testing SAT 2016 19th

International Conference Bordeaux
France July 5 8 2016 Proceedings 1st
Edition Nadia Creignou
Visit to download the full and correct content document:
https://textbookfull.com/product/theory-and-applications-of-satisfiability-testing-sat-20
16-19th-international-conference-bordeaux-france-july-5-8-2016-proceedings-1st-editi
on-nadia-creignou/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Business Information Systems 19th International

Conference BIS 2016 Leipzig Germany July 6 8 2016
Proceedings 1st Edition Witold Abramowicz

https://textbookfull.com/product/business-information-
systems-19th-international-conference-bis-2016-leipzig-germany-
july-6-8-2016-proceedings-1st-edition-witold-abramowicz/

Theory and Applications of Satisfiability Testing SAT

2020 23rd International Conference Alghero Italy July 3
10 2020 Proceedings Lecture Notes in Computer Science
12178 Luca Pulina (Editor)
https://textbookfull.com/product/theory-and-applications-of-
satisfiability-testing-sat-2020-23rd-international-conference-
alghero-italy-july-3-10-2020-proceedings-lecture-notes-in-
computer-science-12178-luca-pulina-editor/

Bioinformatics Research and Applications 12th

International Symposium ISBRA 2016 Minsk Belarus June 5
8 2016 Proceedings 1st Edition Anu Bourgeois

https://textbookfull.com/product/bioinformatics-research-and-
applications-12th-international-symposium-isbra-2016-minsk-
belarus-june-5-8-2016-proceedings-1st-edition-anu-bourgeois/

Principles and Practice of Constraint Programming 22nd

International Conference CP 2016 Toulouse France
September 5 9 2016 Proceedings 1st Edition Michel
Rueher (Eds.)
https://textbookfull.com/product/principles-and-practice-of-
constraint-programming-22nd-international-conference-
cp-2016-toulouse-france-september-5-9-2016-proceedings-1st-
Ad hoc Mobile and Wireless Networks 15th International
Conference ADHOC NOW 2016 Lille France July 4 6 2016
Proceedings 1st Edition Nathalie Mitton

https://textbookfull.com/product/ad-hoc-mobile-and-wireless-
networks-15th-international-conference-adhoc-now-2016-lille-
france-july-4-6-2016-proceedings-1st-edition-nathalie-mitton/

Software Engineering and Formal Methods 14th

International Conference SEFM 2016 Held as Part of STAF
2016 Vienna Austria July 4 8 2016 Proceedings 1st
Edition Rocco De Nicola
https://textbookfull.com/product/software-engineering-and-formal-
methods-14th-international-conference-sefm-2016-held-as-part-of-
staf-2016-vienna-austria-july-4-8-2016-proceedings-1st-edition-
rocco-de-nicola/

Modeling and Simulation in Engineering Economics and

Management International Conference MS 2016 Teruel
Spain July 4 5 2016 Proceedings 1st Edition Raúl León

https://textbookfull.com/product/modeling-and-simulation-in-
engineering-economics-and-management-international-conference-
ms-2016-teruel-spain-july-4-5-2016-proceedings-1st-edition-raul-
leon/

Engineering Applications of Neural Networks 17th

International Conference EANN 2016 Aberdeen UK
September 2 5 2016 Proceedings 1st Edition Chrisina
Jayne
https://textbookfull.com/product/engineering-applications-of-
neural-networks-17th-international-conference-eann-2016-aberdeen-
uk-september-2-5-2016-proceedings-1st-edition-chrisina-jayne/

Fundamental Approaches to Software Engineering 19th

International Conference FASE 2016 Held as Part of the
European Joint Conferences on Theory and Practice of
Software ETAPS 2016 Eindhoven The Netherlands April 2 8
2016 Proceedings 1st Edition Perdita Stevens
https://textbookfull.com/product/fundamental-approaches-to-
software-engineering-19th-international-conference-
fase-2016-held-as-part-of-the-european-joint-conferences-on-
 Nadia Creignou
Daniel Le Berre (Eds.)

Theory and Applications

LNCS 9710

of Satisfiability Testing –
SAT 2016
19th International Conference
Bordeaux, France, July 5–8, 2016
Proceedings

123
Lecture Notes in Computer Science 9710
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zürich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7407
Nadia Creignou Daniel Le Berre (Eds.)
•

Theory and Applications

of Satisfiability Testing –
SAT 2016
19th International Conference
Bordeaux, France, July 5–8, 2016
Proceedings

123
Editors
Nadia Creignou Daniel Le Berre
Aix-Marseille Université Université d’Artois
Marseille Lens
France France

ISSN 0302-9743 ISSN 1611-3349 (electronic)

Lecture Notes in Computer Science
ISBN 978-3-319-40969-6 ISBN 978-3-319-40970-2 (eBook)
DOI 10.1007/978-3-319-40970-2

Library of Congress Control Number: 2016941614

LNCS Sublibrary: SL1 – Theoretical Computer Science and General Issues

© Springer International Publishing Switzerland 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
 Preface

This volume contains the papers presented at the 19th International Conference on
Theory and Applications of Satisfiability Testing (SAT 2016) held during July 5–8,
2016, in Bordeaux, France. SAT 2016 was hosted by the Computer Science Laboratory
of Bordeaux (LaBRI).
The International Conference on Theory and Applications of Satisfiability Testing
(SAT) is the premier annual meeting for researchers focusing on the theory and
applications of the propositional satisfiability problem, broadly construed. Aside from
plain propositional satisfiability, the scope of the meeting includes Boolean opti-
mization (including MaxSAT and pseudo-Boolean (PB) constraints), quantified Boo-
lean formulas (QBF), satisfiability modulo theories (SMT), and constraint
programming (CP) for problems with clear connections to Boolean-level reasoning.
Many hard combinatorial problems can be tackled using SAT-based techniques,
including problems that arise in formal verification, artificial intelligence, operations
research, computational biology, cryptology, data mining, machine learning, mathe-
matics, etc. Indeed, the theoretical and practical advances in SAT research over the past
20 years have contributed to making SAT technology an indispensable tool in a variety
of domains.
SAT 2016 welcomed scientific contributions addressing different aspects of SAT
interpreted in a broad sense, including (but not restricted to) theoretical advances
(including exact algorithms, proof complexity, and other complexity issues), practical
search algorithms, knowledge compilation, implementation-level details of SAT sol-
vers and SAT-based systems, problem encodings and reformulations, applications
(including both novel applications domains and improvements to existing approaches),
as well as case studies and reports on findings based on rigorous experimentation.
A total of 70 papers were submitted this year distributed into 48 long papers, 13
short papers, and nine tool papers. The papers were reviewed by the Program Com-
mittee (33 members), with the help of 65 additional reviewers. Only one regular paper
was found by the Program Committee to be out of the scope for the conference. Each
of the remaining submissions was reviewed by at least three different reviewers.
A rebuttal period allowed the authors to provide a feedback to the reviewers. After that,
the discussion among the Program Committee took place. External reviewers sup-
porting the Program Committee were also invited to participate directly in the dis-
cussions for the papers they reviewed. This year, the authors received a meta-review,
summarizing the discussion that occurred after the rebuttal and the reasons of the final
recommendation. The final recommendation was to accept 31 submissions (22 long,
four short, and five tool papers) and to accept conditionally five additional papers. The
latter (four long and one short) eventually satisfied the conditions for acceptance.
In addition to presentations on the accepted papers, the scientific program of SAT
2016 included three invited talks:
VI Preface

– Phokion Kolaitis (University of California Santa Cruz, IBM, USA) “Coping with
Inconsistent Databases: Semantics, Algorithms, and Complexity”
– David Monniaux (VERIMAG University of Grenoble, CNRS, France) “Satisfia-
bility Testing, a Disruptive Technology in Program Verification”
– Torsten Schaub (University of Potsdam, Germany, EurAI sponsored) “From SAT to
ASP and Back!?”
As in previous years, SAT 2016 hosted various associated events, including four
workshops on July 4:
– 6th International Workshop on the Cross-Fertilization Between CSP and SAT
(CSPSAT 2016) organized by Yael Ben-Haim, Valentin Mayer-Eichberger, and
Yehuda Naveh
– “Graph Structure and Satisfiability Testing” organized by Simone Bova and Stefan
Mengel
– 7th Pragmatics of SAT International Workshop (PoS 2016) organized by Olivier
Roussel and Allen Van Gelder
– 4th International Workshop on Quantified Boolean Formulas (QBF 2016) organized
by Florian Lonsing and Martina Seidl
There were also four competitive events, which ran before the conference and whose
results were disclosed during the conference:
– MAXSAT evaluation organized by Josep Argelich, Chu Min Li, Felip Manyà and
Jordi Planes
– PB competition organized by Olivier Roussel
– QBF evaluation organized by Luca Pulina
– SAT competition organized by Marijn Heule, Matti Jarvisalo, and Tomas Baylo
Moreover, this year a full day of tutorials — “How to Solve My Problem with
SAT?” — was organized right after the conference, on July 9.
March 2016 was a terrible month for the SAT community. On March 12, Helmut
Veith, our esteemed colleague from TU Vienna, passed away at the age of 45. His work
on counter example guided abstraction refinement is widely used in the SAT com-
munity, especially in recent years to tackle QBF problems: A specific session on that
topic was organized during the conference. On March 13, Hilary Putnam, one of the
authors of the seminal “Davis and Putnam” procedure, central in current SAT research,
passed away at the age of 90. The first session on SAT solving was dedicated to his
memory. Our thoughts are with their families during this difficult time.
We would like to thank everyone who contributed to making SAT 2016 a success.
First and foremost we would like to thank the members of the Program Committee and
the additional reviewers for their careful and thorough work, without which it would
not have been possible for us to put together such an outstanding conference. We also
wish to thank all the authors who submitted their work for our consideration. We thank
the SAT Association chair Armin Biere, vice chair John Franco, and treasurer Hans
Kleine Büning for their help and advice in organizational matters. The EasyChair
conference systems provided invaluable assistance in coordinating the submission and
review process, in organizing the program, as well as in the assembly of these
 Preface VII

proceedings. We also thank the local organization team for their efforts with practical
aspects of local organization.
Finally, we gratefully thank the University of Bordeaux, Bordeaux INP, the Com-
puter Science Laboratory of Bordeaux (LaBRI), the GIS Albatros (Bordeaux), the
CNRS, the Laboratory of Fundamental Computer Science of Marseilles (LIF), the Lens
Computer Science Research Laboratory (CRIL), the European Association for Artifical
Intelligence (EurAI), the SAT association, the French-Speaking Constraints Associa-
tion (AFPC), Intel, RATP and Safe-River for financial and organizational support for
SAT 2016.

April 2016 Daniel Le Berre

Nadia Creignou
 Organization

Program Committee
Fahiem Bacchus University of Toronto, Canada
Yael Ben-Haim IBM Research, Israel
Olaf Beyersdorff University of Leeds, UK
Armin Biere Johannes Kepler University, Austria
Nikolaj Bjorner Microsoft Research, USA
Maria Luisa Bonet Universitat Politecnica de Catalunya, Spain
Sam Buss UCSD, USA
Nadia Creignou Aix-Marseille Université, LIF-CNRS, France
Uwe Egly TU Wien, Austria
John Franco University of Cincinnati, USA
Djamal Habet Aix-Marseille Université, LSIS-CNRS, France
Marijn Heule The University of Texas at Austin, USA
Holger Hoos University of British Columbia, Canada
Frank Hutter University of Freiburg, Germany
Mikolas Janota Microsoft Research, UK
Matti Järvisalo University of Helsinki, Finland
Hans Kleine Büning University of Paderborn, Germany
Daniel Le Berre Université d’Artois, CRIL-CNRS, France
Ines Lynce INESC-ID/IST, University of Lisbon, Portugal
Marco Maratea DIBRIS, University of Genoa, Italy
Joao Marques-Silva Faculty of Science, University of Lisbon, Portugal
Stefan Mengel CRIL-CNRS, France
Alexander Nadel Intel, Israel
Nina Narodytska Samsung Research America, USA
Jakob Nordström KTH Royal Institute of Technology, Sweden
Albert Oliveras Technical University of Catalonia, Spain
Roberto Sebastiani DISI, University of Trento, Italy
Martina Seidl Johannes Kepler University Linz, Austria
Yuping Shen Institute of Logic and Cognition, Sun Yat-sen University,
China
Laurent Simon Labri, Bordeaux Institute of Technology, France
Takehide Soh Information Science and Technology Center,
Kobe University, Japan
Stefan Szeider TU Wien, Austria
Allen Van Gelder University of California, Santa Cruz, USA
X Organization

Additional Reviewers

Abramé, André Kolaitis, Phokion

Aleksandrowicz, Gadi Kotthoff, Lars
Audemard, Gilles Kovásznai, Gergely
Banbara, Mutsunori Lauria, Massimo
Baud-Berthier, Guillaume Lettmann, Theodor
Bayless, Sam Lindauer, Marius
Berkholz, Christoph Lonsing, Florian
Blinkhorn, Joshua Malod, Guillaume
Bofill, Miquel Manquinho, Vasco
Bonacina, Ilario Marquis, Pierre
Cameron, Chris Martins, Ruben
Cao, Weiwei Meel, Kuldeep S.
Chew, Leroy Miksa, Mladen
de Rezende, Susanna F. Nabeshima, Hidetomo
Demri, Stéphane Naveh, Yehuda
Dodaro, Carmine Neves, Miguel
Eggensperger, Katharina Oetsch, Johannes
Elffers, Jan Panagiotou, Konstantinos
Falkner, Stefan Pich, Ján
Feng, Shiguang Prcovic, Nicolas
Fichte, Johannes Klaus Preiner, Mathias
Fröhlich, Andreas Rozier, Kristin Yvonne
Galesi, Nicola Ryvchin, Vadim
Ganesh, Vijay Slivovsky, Friedrich
Gange, Graeme Strichman, Ofer
Gaspers, Serge Trentin, Patrick
Griggio, Alberto Tveretina, Olga
Guthmann, Ofer Vinyals, Marc
Hermo, Montserrat Wang, Lingli
Hinde, Luke Widl, Magdalena
Ibanez-Garcia, Yazmin Angelica Wintersteiger, Christoph M.
Ignatiev, Alexey Yue, Weiya
Ivrii, Alexander
Invited Talks
 Coping with Inconsistent Databases:
Semantics, Algorithms, and Complexity

Phokion G. Kolaitis1,2
1
University of California Santa Cruz, Santa Cruz, USA
2
IBM Research – Almaden, San Jose, USA
kolaitis@cs.ucsc.edu

Abstract. Managing inconsistency in databases is a long-standing challenge.

The framework of database repairs provides a principled approach towards
coping with inconsistency in databases. Intuitively, a repair of an inconsistent
database is a consistent database that differs from the given inconsistent database
in a minimal way. Repair checking and consistent query answering are two
fundamental algorithmic problems arising in this context. The first of these two
problems asks whether, given two databases, one is a repair of the other. The
second asks whether, a query is true on every repair of a given inconsistent
database.
The aim of this talk is to give an overview of a body of results in this area
with emphasis on the computational complexity of repair checking and con-
sistent query answering, including the quest for dichotomy theorems. In addition
to presenting open problems, the last part of the talk will include a discussion
of the potential use of solvers in developing practical systems for consistent
query answering.

Keywords: Inconsistent databases
Database dependencies
Database repairs

Consistent query answering
Computational complexity

References
1. Afrati, F.N., Kolaitis, P.G.: Repair checking in inconsistent databases: algorithms and com-
plexity. In: 12th International Conference on Database Theory. ICDT 2009, St. Petersburg,
Russia, March 23–25, 2009, Proceedings, pp. 31–41 (2009)
2. Arenas, M., Bertossi, L.E., Chomicki, J.: Consistent query answers in inconsistent databases.
In: Proceedings of the Eighteenth ACM SIGACT-SIGMOD-SIGART Symposium on Prin-
ciples of Database Systems, May 31 – June 2, 1999, Philadelphia, Pennsylvania, USA,
pp. 68–79 (1999)
3. Arming, S., Pichler, R., Sallinger, E.: Complexity of repair checking and consistent query
answering. In: 19th International Conference on Database Theory. ICDT 2016, Bordeaux,
France, March 15–18, 2016, pp. 21:1–21:18 (2016)
4. Bertossi, L.E.: Database Repairing and Consistent Query Answering. Synthesis Lectures on
Data Management. Morgan & Claypool Publishers (2011)
5. ten Cate, B., Fontaine, G., Kolaitis, P.G.: On the data complexity of consistent query
answering. Theory Comput. Syst. 57(4), 843–891 (2015)
XIV P.G. Kolaitis

6. ten Cate, B., Halpert, R.L., Kolaitis, P.G.: Exchange-repairs: managing inconsistency in data
exchange. In: Kontchakov, R., Mugnier, M.L. (eds.) RR 2014. LNCS, vol. 8741, pp. 140–
156. Springer, Switzerland (2014)
7. ten Cate, B., Halpert, R.L., Kolaitis, P.G.: Practical query answering in data exchange under
inconsistency-tolerant semantics. In: Proceedings of the 19th International Conference on
Extending Database Technology. EDBT 2016, Bordeaux, France, March 15–16, 2016,
pp. 233–244 (2016)
8. Chomicki, J.: Consistent query answering: five easy pieces. In: 11th International Conference
on Database Theory - ICDT 2007, Barcelona, Spain, January 10–12, 2007, Proceedings,
pp. 1–17 (2007)
9. Chomicki, J., Marcinkowski, J.: Minimal-change integrity maintenance using tuple deletions.
Inf. Comput. 197(1–2), 90–121 (2005)
10. Chomicki, J., Marcinkowski, J.: Staworko, S.: Hippo: a system for computing consistent
answers to a class of SQL queries. In: Advances in Database Technology - EDBT 2004, 9th
International Conference on Extending Database Technology, Heraklion, Crete, Greece,
March 14–18, 2004, Proceedings, pp. 841–844 (2004)
11. Fagin, R., Kimelfeld, B., Kolaitis, P.G.: Dichotomies in the complexity of preferred repairs.
In: Proceedings of the 34th ACM Symposium on Principles of Database Systems. PODS
2015, Melbourne, Victoria, Australia, May 31 – June 4, 2015, pp. 3–15 (2015)
12. Fontaine, G.: Why is it hard to obtain a dichotomy for consistent query answering? ACM
Trans. Comput. Log. 16(1), 7:1–7:24 (2015)
13. Fuxman, A., Fazli, E., Miller, R.J.: ConQuer: Efficient management of inconsistent data-
bases. In: Proceedings of the ACM SIGMOD International Conference on Management of
Data, Baltimore, Maryland, USA, June 14–16, 2005, pp. 155–166 (2005)
14. Fuxman, A., Fuxman, D., Miller, R.J.: ConQuer: a system for efficient querying over
inconsistent databases. In: Proceedings of the 31st International Conference on Very Large
Data Bases, Trondheim, Norway, August 30 – September 2, 2005, pp. 1354–1357 (2005)
15. Kolaitis, P.G., Pema, E.: A dichotomy in the complexity of consistent query answering for
queries with two atoms. Inf. Process. Lett. 112(3), 77–85 (2012)
16. Kolaitis, P.G., Pema, E., Tan, W.: Efficient querying of inconsistent databases with binary
integer programming. PVLDB 6(6), 397–408 (2013)
17. Koutris, P., Wijsen, J.: The data complexity of consistent query answering for self-join-free
conjunctive queries under primary key constraints. In: Proceedings of the 34th ACM
Symposium on Principles of Database Systems. PODS 2015, Melbourne, Victoria, Australia,
May 31 – June 4, 2015, pp. 17–29 (2015)
18. Marileo, M.C., Bertossi, L.E.: The consistency extractor system: answer set programs for
consistent query answering in databases. Data Knowl. Eng. 69(6), 545–572 (2010)
19. Staworko, S., Chomicki, J., Marcinkowski, J.: Prioritized repairing and consistent query
answering in relational databases. Ann. Math. Artif. Intell. 64(2–3), 209–246 (2012)
20. Wijsen, J.: Certain conjunctive query answering in first-order logic. ACM Trans. Database
Syst. 37(2), 9 (2012)
21. Wijsen, J.: A survey of the data complexity of consistent query answering under key con-
straints. In: 8th International Symposium on Foundations of Information and Knowledge
Systems. FoIKS 2014, Bordeaux, France, March 3–7, 2014. Proceedings, pp. 62–78 (2014)
 From SAT to ASP and Back!?

Torsten Schaub1,2
1
University of Potsdam, Potsdam, Germany
2
INRIA Rennes, Rennes, France
torsten@cs.uni-potsdam.de

Answer Set Programming (ASP; [1–4]) provides an approach to declarative problem

solving that combines a rich yet simple modeling language with effective Boolean
constraint solving capacities. This makes ASP a model, ground, and solve paradigm, in
which a problem is expressed as a set of first-order rules, which are subsequently turned
into a propositional format by systematically replacing all variables, before finally the
models of the resulting propositional rules are computed. ASP is particularly suited for
modeling problems in the area of Knowledge Representation and Reasoning involving
incomplete, inconsistent, and changing information due to its nonmonotonic semantic
foundations. As such, it offers, in addition to satisfiability testing, various reasoning
modes, including different forms of model enumeration, intersection or unioning, as
well as multi-objective optimization. From a formal perspective, ASP allows for
solving all search problems in NP (and NPNP) in a uniform way, that is, by separating
problem encodings and instances. Hence, ASP is well-suited for solving hard combi-
natorial search (and optimization) problems. Interesting applications of ASP include
decision support systems for NASA shuttle controllers [5], industrial team-building [6],
music composition [7], natural language processing [8], package configuration [9],
phylogeneticics [10], robotics [11, 12], systems biology [13–15], timetabling [16], and
many more. The versatility of ASP is nicely reflected by the ASP solver clasp [17],
winning first places at various solver competitions, including ASP, MISC, PB, and
SAT. In fact, clasp is at the heart of the open source platform potassco.-
sourceforge.net. Potassco stands for the “Potsdam Answer Set Solving Col-
lection” [18] and has seen more than 145000 downloads world-wide since its inception
at the end of 2008.
The talk will start with a gentle introduction to ASP, while focusing on the com-
monalities and differences to SAT. It will discuss the different semantic foundations
and describe the impact of a modelling language along with off-the-shelf grounding
systems. Finally, it will highlight some resulting techniques, like meta-programming,
preference handling, heuristic constructs, and theory reasoning.

References
1. Gelfond, M., Lifschitz, V.: The stable model semantics for logic programming. In: Kowalski,
R., Bowen, K. (eds.) Proceedings of the Fifth International Conference and Symposium of
Logic Programming. ICLP1988, pp. 1070–1080. MIT Press (1988)
XVI T. Schaub

2. Baral, C.: Knowledge Representation, Reasoning and Declarative Problem Solving.

Cambridge University Press (2003)
3. Gebser, M., Kaminski, R., Kaufmann, B., Schaub, T.: Answer Set solving in practice.
Synthesis Lectures on Artificial Intelligence and Machine Learning. Morgan and Claypool
Publishers (2012)
4. Gelfond, M., Kahl, Y.: Knowledge Representation, Reasoning, and the Design of Intelligent
Agents: The Answer-Set Programming Approach. Cambridge University Press (2014)
5. Nogueira, M., Balduccini, M., Gelfond, M., Watson, R., Barry, M.: An A-prolog decision
support system for the space shuttle. In: Ramakrishnan, I. (ed.) PADL 2001. LNCS, vol.
1990, pp. 169–183. Springer, Berlin (2001)
6. Grasso, G., Iiritano, S., Leone, N., Lio, V., Ricca, F., Scalise, F.: An ASP-based system for
team-building in the Gioia-Tauro seaport. In: Carro, M., Peña, R. (eds.) PADL 2010. LNCS,
vol. 5937, pp. 40–42. Springer, Berlin (2010)
7. Boenn, G., Brain, M., de Vos, M., Fitch, J.: Automatic composition of melodic and harmonic
music by answer set programming. In: Garcia de la Banda, M., Pontelli, E. (eds.) ICLP 2008.
LNCS, vol. 5366, pp. 160–174. Springer, Berlin (2008)
8. Schwitter, R.: The jobs puzzle: taking on the challenge via controlled natural language
processing. Theory Pract. Logic Program. 13(4–5), 487–501 (2013)
9. Gebser, M., Kaminski, R., Schaub, T.: aspcud: A Linux package configuration tool based on
answer set programming. In: Drescher, C., Lynce, I., Treinen, R., (eds.) Proceedings of the
Second International Workshop on Logics for Component Configuration. LoCoCo 2011.
Electronic Proceedings in Theoretical Computer Science (EPTCS), vol. 65, pp. 12–25 (2011)
10. Brooks, D., Erdem, E., Erdogan, S., Minett, J., Ringe, D.: Inferring phylogenetic trees using
answer set programming. J. Autom. Reason. 39(4), 471–511 (2007)
11. Chen, X., Ji, J., Jiang, J., Jin, G., Wang, F., Xie, J.: Developing high-level cognitive
functions for service robots. In: van der Hoek, W., Kaminka, G., Lespérance, Y., Luck, M.,
Sen, S. (eds.) Proceedings of the Ninth International Conference on Autonomous Agents and
Multiagent Systems. AAMAS 2010, IFAAMAS, pp. 989–996 (2010)
12. Erdem, E., Haspalamutgil, K., Palaz, C., Patoglu, V., Uras, T.: Combining high-level causal
reasoning with low-level geometric reasoning and motion planning for robotic manipulation.
In: Proceedings of the IEEE International Conference on Robotics and Automation. ICRA
2011, pp. 4575–4581. IEEE (2011)
13. Erdem, E., Türe, F.: Efficient haplotype inference with answer set programming. In: Fox, D.,
Gomes, C. (eds.) Proceedings of the Twenty-Third National Conference on Artificial
Intelligence. AAAI 2008, pp. 436–441. AAAI Press (2008)
14. Gebser, M., Schaub, T., Thiele, S., Veber, P.: Detecting inconsistencies in large biological
networks with answer set programming. Theory Pract. Logic Program. 11(2–3), 323–360
(2011)
15. Gebser, M., Guziolowski, C., Ivanchev, M., Schaub, T., Siegel, A., Thiele, S., Veber, P.:
Repair and prediction (under inconsistency) in large biological networks with answer set
programming. In: Lin, F., Sattler, U. (eds.) Proceedings of the Twelfth International Con-
ference on Principles of Knowledge Representation and Reasoning. KR 2010, pp.497–507.
AAAI Press (2010)
16. Banbara, M., Soh, T., Tamura, N., Inoue, K., Schaub, T.: Answer set programming as a
modeling language for course timetabling. Theory Pract. Logic Program. 13(4–5), 783–798
(2013)
17. Gebser, M., Kaufmann, B., Schaub, T.: Conflict-driven answer set solving: from theory to
practice. Artif. Intell. 187–188, 52–89 (2012)
18. Gebser, M., Kaminski, R., Kaufmann, B., Ostrowski, M., Schaub, T., Schneider, M.:
Potassco: the Potsdam answer set solving collection. AI Commun. 24(2), 107–124 (2011)
 Satisfiability Testing, a Disruptive Technology
in Program Verification?

David Monniaux1,2
1
Université Grenoble Alpes, VERIMAG, 38000 Grenoble, France
2
CNRS, Verimag, 38000 Grenoble, France
David.Monniaux@imag.fr

Abstract. In the 2000s, progress in satisfiability testing shook automated and

assisted program verification. The advent of efficient satisfiability modulo theory
(SMT) solvers allowed new approaches: efficient testing and symbolic execu-
tion, new methods for generating inductive invariants, and more automated
assisted proof.

Program verification consists in proving properties of software, be them safety

(“whatever the program execution, some property is always satisfied”) or liveness
(“some action will always eventually happen”). Improvements in satisfiability testing
have allowed exciting combinations of exact decision procedures [12], often based on
Boolean satisfiability [2], with existing approaches.

1 Program Verification Before SMT

Traditionally, program verification (i) either relied heavily on the user tediously pro-
viding inductive invariants, ranking functions as well as proofs (ii) either simplified the
problems through abstractions that sometimes were sufficient to prove the property,
sometimes were not. Excessively coarse abstractions (e.g. one single interval of vari-
ation per program variable per location) could sometimes be refined by explicit par-
titioning [16], but its cost is exponential and thus it must be quickly limited.
Program verification was most often, and still is, considered too costly in human
and algorithmic terms, and thus in most practical cases, it was replaced by testing, with
test cases chosen by hand or through test case generation techniques. Fuzzing is a kind
of testing where variant of input files or protocol exchanges are randomly modified so
as to trigger bugs in parsers, which could be exploited as security vulnerabilities.
Both program verification and testing were transformed by the advent of a dis-
ruptive technology: satisfiability modulo theory (SMT), that is, efficient algorithms for
checking that a first-order logic formula over a given theory is satisfiable — e.g.
(x < y ∨ x ≥ y + 3) ∧ (x ≥ 0 ∧ x + y ≤ 5) is a first-order formula over linear real or
integer arithmetic.

The research leading to these results has received funding from the European Research Council
under the European Union’s Seventh Framework Programme (FP/2007-2013)/ERC Grant Agreement
nr. 306595 “STATOR”.
XVIII D. Monniaux

2 Testing

In bounded model checking [1], the program is unrolled up to a finite depth, and a first-
order formula is generated, whose solutions are the program traces that violate the
desired property before the given depth. In symbolic execution, sequences of program
statements are translated into a first-order formula, and the feasibility of tests into new
branches is checked by satisfiability testing; this approach has been successfully
applied to fuzzing, that is, searching for inputs that trigger security violations in file or
protocol parsers [8]. If symbolic execution is too costly, certain unknowns may be
chosen to have concrete values, while others are left symbolic, leading to concolic
execution.

3 Automatic Verification

Bounded model checking and symbolic execution cannot prove the safety of programs,
except in the rare case where there is a small constant bound on execution lengths.
Inductive invariant inference was also greatly transformed by the advent of satisfiability
modulo theory solvers. First, inference approaches designed to operate over a control-
flow graph and produce an invariant per control location were modified to traverse
loop-free program fragments (large blocks) encoded into first order formulas, whose
satisfiability is checked by SMT [7, 15]. Second, counterexample-guided abstraction
refinement approaches mine proofs of unreachability of errors through finite unrollings
for arguments that could become inductive, in particular by extraction of Craig
interpolants [14].
Research in SMT solving has strived to extend the class of formulas handled by
solvers [12]: from quantifier-free linear real arithmetic, solved by a combination of
constraint-driven clause learning (CDCL) SAT-solving [13] and exact-precision
simplex algorithm [5], and uninterpreted functions, solvers were extended to linear
integer arithmetic, nonlinear (polynomial) arithmetic, arrays, bitvector arithmetic,
character strings, data structures, and quantified formulas. Some of these combinations
are undecidable or have high lower bounds on their worst-case complexity [6], yet this
is not considered a major hindrance; practical efficiency is paramount.

4 Assisted Verification

In assisted proof, using tools such as Coq or Isabelle, the user traditionally has to
provide detailed arguments why the claimed theorem is true. Automation is tradi-
tionally limited. When the theorem, or a part of it, fits within a class decidable by SMT,
it is tempting to check it using SMT . It may however be unwise to blindly trust a SMT
solver, a complex piece of software likely to have bugs; and some assistants require all
 Satisfiability Testing, a Disruptive Technology in Program Verification? XIX

proofs to be broken down into basic steps checked by a proof checker. It is therefore
desirable that the solver provides an independently checkable proof witness; one
challenge is to keep the witness small enough to be manageable while keeping the
checker small [3, 11].

5 Challenges

Most extant SMT solvers are based on the DPLL(T) framework: a combination of a
CDCL SAT solver and a decision procedure for conjunctions. This framework has
known weaknesses: for instance, some industrially relevant families of formulas induce
exponential behavior [9]. There have been multiple proposals for solvers based on
other approaches [4, 10], but they are less mature. Quantifiers are still often difficult to
handle. Verification approaches based on Craig interpolation are often brittle; more
generally, reliance on random number generators sometimes results in unpredictable
behavior. Better collaboration between verification and SAT experts is needed to
overcome these challenges.

References
1. Biere, A.: Bounded model checking. In: Biere et al. [2], vol. 185, pp. 455–481 (2009)
2. Biere, A., Heule, M.J.H., van Maaren, H., Walsh, T. (eds.): Handbook of Satisfiability,
Frontiers in Artificial Intelligence and Applications, vol. 185. IOS Press (2009)
3. Böhme, S., Weber, T.: Fast LCF-style proof reconstruction for Z3. In: Kaufmann, M.,
Paulson, L.C. (eds.) ITP 2010. LNCS, vol. 6172, pp. 179–194. Springer, Berlin (2010)
4. Brain, M., D’Silva, V., Griggio, A., Haller, L., Kroening, D.: Deciding floating-point logic
with abstract conflict driven clause learning. Formal Methods Syst. Des. 45(2), 213–245
(2014)
5. Dutertre, B., de Moura, L.M.: Integrating simplex with DPLL(T). Sri-csl-06-01, SRI Inter-
national, Computer Science Laboratory (2006)
6. Fischer, M.J., Rabin, M.O.: Super-exponential complexity of presburger arithmetic. In:
Karp, R. (ed.) Complexity of Computation. SIAM–AMS Proceedings, pp. 27–42, no. 7.
American Mathematical Society (1974). citeseer.ist.psu.edu/fischer74superexponential.html
7. Gawlitza, T., Monniaux, D.: Invariant generation through strategy iteration in succinctly
represented control flow graphs. Logical Methods in Computer Science (2012)
8. Godefroid, P., Levin, M.Y., Molnar, D.: SAGE: Whitebox fuzzing for security testing.
Queue 10(1), 20:20–20:27 (2012)
9. Henry, J., Asavoae, M., Monniaux, D., Maiza, C.: How to compute worst-case execution
time by optimization modulo theory and a clever encoding of program semantics. In: Zhang,
Y., Kulkarni, P. (eds.) Languages, Compilers, Tools and Theory for Embedded Systems
(LCTES), pp. 43–52. ACM (2014)
10. Jovanović, D., de Moura, L.: Solving non-linear arithmetic. In: IJCAR (2012)
11. Keller, C.: Extended resolution as certificates for propositional logic. In: Blanchette, J.C.,
Urban, J. (eds.) Proof Exchange for Theorem Proving (PxTP). EPiC Series, vol. 14, pp. 96–
109. EasyChair (2013). http://www.easychair.org/publications/?page=117514525
XX D. Monniaux

12. Kroening, D., Strichman, O.: Decision Procedures. Springer (2008)

13. Marques-Silva, J.P., Lynce, I., Malik, S.: Conflict-driven clause learning SAT solvers. In:
Biere et al. [2], vol. 185, pp. 131–153 (2009)
14. McMillan, K.L.: Lazy abstraction with interpolants. In: Ball, T., Jones, R.B. (eds.) CAV
2006. LNCS, vol. 4144, pp. 123–136. Springer, Berlin (2006)
15. Monniaux, D., Gonnord, L.: Using bounded model checking to focus fixpoint iterations. In:
Yahav, E. (ed.) SAS 2011. LNCS, vol. 6887, pp. 369–385. Springer, Berlin (2011)
16. Rival, X., Mauborgne, L.: The trace partitioning abstract domain. ACM Trans. Program.
Lang. Syst. 29(5) (2007)
 Contents

Complexity

Parameterized Compilation Lower Bounds for Restricted CNF-Formulas . . . . 3

Stefan Mengel

Satisfiability via Smooth Pictures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Mateus de Oliveira Oliveira

Solution-Graphs of Boolean Formulas and Isomorphism . . . . . . . . . . . . . . . 29

Patrick Scharpfenecker and Jacobo Torán

Strong Backdoors for Default Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Johannes K. Fichte, Arne Meier, and Irina Schindler

The Normalized Autocorrelation Length of Random Max r-Sat Converges

in Probability to ð1  1=2r Þ=r. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Daniel Berend and Yochai Twitto

Tight Upper Bound on Splitting by Linear Combinations

for Pigeonhole Principle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Vsevolod Oparin

Satisfiability Solving

Extreme Cases in SAT Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Gilles Audemard and Laurent Simon

Improved Static Symmetry Breaking for SAT . . . . . . . . . . . . . . . . . . . . . . . 104

Jo Devriendt, Bart Bogaerts, Maurice Bruynooghe, and Marc Denecker

Learning Rate Based Branching Heuristic for SAT Solvers. . . . . . . . . . . . . . 123

Jia Hui Liang, Vijay Ganesh, Pascal Poupart, and Krzysztof Czarnecki

On the Hardness of SAT with Community Structure . . . . . . . . . . . . . . . . . . 141

Nathan Mull, Daniel J. Fremont, and Sanjit A. Seshia

Trade-offs Between Time and Memory in a Tighter Model

of CDCL SAT Solvers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Jan Elffers, Jan Johannsen, Massimo Lauria, Thomas Magnard,
Jakob Nordström, and Marc Vinyals
XXII Contents

Satisfiability Applications

A SAT Approach to Branchwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

Neha Lodha, Sebastian Ordyniak, and Stefan Szeider

Computing Maximum Unavoidable Subgraphs Using SAT Solvers . . . . . . . . 196

C.K. Cuong and M.J.H. Heule

Heuristic NPN Classification for Large Functions Using AIGs

and LEXSAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Mathias Soeken, Alan Mishchenko, Ana Petkovska, Baruch Sterin,
Paolo Ienne, Robert K. Brayton, and Giovanni De Micheli

Solving and Verifying the Boolean Pythagorean Triples Problem

via Cube-and-Conquer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Marijn J.H. Heule, Oliver Kullmann, and Victor W. Marek

Satisfiability Modulo Theory

Deciding Bit-Vector Formulas with mcSAT . . . . . . . . . . . . . . . . . . . . . . . . 249

Aleksandar Zeljić, Christoph M. Wintersteiger, and Philipp Rümmer

Solving Quantified Bit-Vector Formulas Using Binary Decision Diagrams . . . 267

Martin Jonáš and Jan Strejček

Speeding up the Constraint-Based Method in Difference Logic . . . . . . . . . . . 284

Lorenzo Candeago, Daniel Larraz, Albert Oliveras,
Enric Rodríguez-Carbonell, and Albert Rubio

Synthesis of Domain Specific CNF Encoders for Bit-Vector Solvers . . . . . . . 302

Jeevana Priya Inala, Rohit Singh, and Armando Solar-Lezama

Beyond SAT

Finding Finite Models in Multi-sorted First-Order Logic . . . . . . . . . . . . . . . 323

Giles Reger, Martin Suda, and Andrei Voronkov

MCS Extraction with Sublinear Oracle Queries . . . . . . . . . . . . . . . . . . . . . . 342

Carlos Mencía, Alexey Ignatiev, Alessandro Previti,
and Joao Marques-Silva

Predicate Elimination for Preprocessing in First-Order Theorem Proving . . . . 361

Zurab Khasidashvili and Konstantin Korovin
 Contents XXIII

Quantified Boolean Formula

Incremental Determinization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

Markus N. Rabe and Sanjit A. Seshia

Non-prenex QBF Solving Using Abstraction . . . . . . . . . . . . . . . . . . . . . . . 393

Leander Tentrup

On Q-Resolution and CDCL QBF Solving . . . . . . . . . . . . . . . . . . . . . . . . . 402

Mikoláš Janota

On Stronger Calculi for QBFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419

Uwe Egly

Q-Resolution with Generalized Axioms . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

Florian Lonsing, Uwe Egly, and Martina Seidl

2QBF: Challenges and Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

Valeriy Balabanov, Jie-Hong Roland Jiang, Christoph Scholl,
Alan Mishchenko, and Robert K. Brayton

Dependency QBF

Dependency Schemes for DQBF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473

Ralf Wimmer, Christoph Scholl, Karina Wimmer, and Bernd Becker

Lifting QBF Resolution Calculi to DQBF . . . . . . . . . . . . . . . . . . . . . . . . . 490

Olaf Beyersdorff, Leroy Chew, Renate A. Schmidt, and Martin Suda

Long Distance Q-Resolution with Dependency Schemes . . . . . . . . . . . . . . . 500

Tomáš Peitl, Friedrich Slivovsky, and Stefan Szeider

Tools

BEACON: An Efficient SAT-Based Tool for Debugging ELþ Ontologies . . . 521

M. Fareed Arif, Carlos Mencía, Alexey Ignatiev, Norbert Manthey,
Rafael Peñaloza, and Joao Marques-Silva

HordeQBF: A Modular and Massively Parallel QBF Solver . . . . . . . . . . . . . 531

Tomáš Balyo and Florian Lonsing

LMHS: A SAT-IP Hybrid MaxSAT Solver . . . . . . . . . . . . . . . . . . . . . . . . 539

Paul Saikko, Jeremias Berg, and Matti Järvisalo

OpenSMT2: An SMT Solver for Multi-core and Cloud Computing . . . . . . . . 547

Antti E.J. Hyvärinen, Matteo Marescotti, Leonardo Alt,
and Natasha Sharygina
XXIV Contents

SpyBug: Automated Bug Detection in the Configuration Space

of SAT Solvers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
Norbert Manthey and Marius Lindauer

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563

Complexity
 Parameterized Compilation Lower Bounds
for Restricted CNF-Formulas

Stefan Mengel(B)

CNRS, CRIL UMR 8188, Lens, France

mengel@cril.fr

Abstract. We show unconditional parameterized lower bounds in the

area of knowledge compilation, more specifically on the size of circuits in
decomposable negation normal form (DNNF) that encode CNF-formulas
restricted by several graph width measures. In particular, we show that
– there are CNF formulas of size n and modular incidence treewidth
k whose smallest DNNF-encoding has size nΩ(k) , and
– there are CNF formulas of size n and incidence neighborhood
√
diver-
sity k whose smallest DNNF-encoding has size nΩ( k) .
These results complement recent upper bounds for compiling CNF into
DNNF and strengthen—quantitatively and qualitatively—known condi-
tional lower bounds for cliquewidth. Moreover, they show that, unlike
for many graph problems, the parameters considered here behave signif-
icantly differently from treewidth.

1 Introduction
Knowledge compilation is a preprocessing regime that aims to translate or “com-
pile” knowledge bases, generally encoded as CNF formulas, into different repre-
sentations more convenient for a task at hand. The idea is that many queries
one would like to answer on the knowledge base, say clause entailment queries,
are intractable in CNF encoding, but tractable for other representations. When
there are many queries on the same knowledge base, as for example in product
configuration, it makes sense to invest into a costly preprocessing to change the
representation once in order to then speed up the queries and thus amortize the
time spent on the preprocessing.
One critical question when following this approach is the choice of the rep-
resentation that the knowledge is encoded into. In general, there is a trade-off
between the usefulness of a representation (which queries does it support effi-
ciently?) and succinctness (what is the size of the encoded knowledge base?). This
trade-off has been studied systematically [8], leading to a fine understanding of
the different representations. In particular, circuits in decomposable negation
normal form (short DNNF) [6] have been identified as a representation that is
more succinct than nearly all other representations while still allowing useful
queries. Consequently, DNNFs play a central role in knowledge compilation.
This paper should be seen as complementing the findings of [1]: In that paper,
algorithms compiling CNF formulas with restricted underlying graph structure

c Springer International Publishing Switzerland 2016
N. Creignou and D. Le Berre (Eds.): SAT 2016, LNCS 9710, pp. 3–12, 2016.
DOI: 10.1007/978-3-319-40970-2 1
4 S. Mengel

were presented, showing that popular graph width measures like treewidth and
cliquewidth can be used in knowledge compilation. More specifically, every CNF
formula of incidence treewidth k and size n can be compiled into a DNNF of size
2O(k) n. Moreover, if k is the incidence cliquewidth, the size bound on the encoding
becomes nO(k) . As has long been observed, 2O(k) n is of course far preferable to
nO(k) for nontrivial sizes of n—in fact, this is the main premise of the field of
parameterized complexity theory, see e.g. [13]. Consequently, the results of [1]
leave open the question if the algorithm for clique-width based compilation of
CNF formulas can be improved.
In fact, the paper [1] already gives a partial answer to this question, proving
that there is no compilation algorithm achieving fixed-parameter compilability,
i.e., a size bound of f (k)p(|F |) for a function f and a polynomial p. But unfortu-
nately this result is based on the plausible but rather non-standard complexity
assumption that not all problem in W[1] have FPT-size circuits. The result of
this paper is that this assumption is not necessary. We prove a lower bound of
|F |Ω(k) for formulas of modular incidence treewidth k where modular treewidth
is a restriction of cliquewidth proposed in [19]. It follows that √the result in [1]
is essentially tight. Moreover, we show a lower bound of |F |Ω( k) for formulas
of neighborhood diversity k [16]. This intuitively shows that all graph width
measures that are stable under adding modules, i.e., adding a new vertex that
has exactly the same neighborhood as an existing vertex, behave qualitatively
worse than treewidth for compilation into DNNFs.

Related work. Parameterized knowledge compilation was first introduced by

Chen [4] and has seen some recent renewed interest, see e.g. [5,9] for work on
conditional lower bounds. Unconditional lower bounds based on treewidth can
e.g. be found in [20,21], but they are only for different versions of branching
programs that are known to be less succinct than DNNF. Moreover, these lower
bounds fail for DNNFs as witnessed by the upper bounds of [1].
There is a long line of research using graph and hypergraph width measures
for problems related to propositional satisfiability, see e.g. the extensive discus-
sion in [3]. The paper [18] gave the first parameterized lower bounds on SAT
with respect to graph width measures, in particular cliquewidth. This result was
later improved to modular treewidth to complement an upper bound for model
counting [19] and very recently to neighborhood diversity [10], a width mea-
sure introduced in [16]. We remark that the latter result could be turned into a
conditional parameterized lower bound similar to that in [1] discussed above.
Our lower bounds strongly rely on the framework for DNNF lower bounds
proposed in [2] and communication theory lower bounds from [12], for more
details see Sect. 3.

2 Preliminaries

In the scope of this paper, a linear code C is the solution of a system of linear
equations Ax̄ = 0 over the boolean field F2 . The matrix A is called the parity-
 Parameterized Compilation Lower Bounds for Restricted CNF-Formulas 5

check matrix of C. The characteristic function fC is the boolean function that,

given a boolean string e, evaluates to 1 if and only if e is in C.
We use the notation [n] := {1, . . . , n} and [n1 , n2 ] := {n1 , n1 + 1, . . . , n2 }
to denote integer intervals. We use standard notations from graph theory and
assume the reader to have a basic background in in the area [11]. By N (v) we
denote the open neighborhood of a vertex in a graph.
We say that two vertices u, v in a graph G = (V, E) have the same neigh-
borhood type if and only if N (u) \ {v} = N (v) \ {u}. It can be shown that
having the same neighborhood type is an equivalence relation on V [16]. The
neighborhood diversity of G is defined to be the number of equivalence classes
of V with respect to neighborhood types.
A generalization of neighborhood diversity is modular treewidth which is
defined as follows: From a graph G we construct a new graph G by contracting
all vertices sharing a neighborhood type, i.e., from every equivalence class we
delete all vertices but one. The modular treewidth of G is then defined to be the
treewidth of G1 . Modular pathwith is defined in the obvious analogous way.
We assume basic familiarity with propositional logic and in particular CNF
formulas. We define the size of a CNF formula to be the overall number of
occurrences of literals, i.e., the sum of the sizes of the clauses where the size of
a clause is the number of literals in the clause. The incidence graph of a CNF
formula F has as vertices the variables and clauses of F and an edge between
every clause and the vertices contained in it. The projection of an assignment
a : X → {0, 1} to a set Z is the restriction of a to the variable set Z. This
definition generalizes to sets of assignments in the obvious way. Moreover, the
projection of a boolean function f on X to Z is defined as the boolean function
on Z whose satisfying assignments are those of f projected to Z. For the width
measures introduced above, we define the with of a formula to be that of its
incidence graph.

3 Statement of the Main Results and Preparation of the

Proof
We now state our main results. The first theorem shows that modular pathwith—
and thus also more general parameters like cliquewidth and modular treewidth—
do not allow fixed-parameter compilation to DNNF.
Theorem 1. For every k and for every n big enough there is a CNF formula F
of size at most n and modular pathwidth k such that any DNNF computing the
same function as F must have size nΩ(k) .
We also show lower bounds for neighborhood diversity that are nearly as
strong as those for modular pathwidth.
1
Note that the definition in [19] differs from the one we give here, but can easily be
seen to be equivalent for bipartite graphs and thus incidence graphs of CNF formulas.
We keep our definition to be more consistent with the definition of neighborhood
diversity.
6 S. Mengel

Theorem 2. For every k and for every n big enough there is a CNF formula F
of size polynomial in n and with neighborhood diversity
√
k such that any DNNF
Ω( k)
computing the same function as F must have size n .
At this point, the attentive reader may be a little concerned because we
promise to prove lower bounds for DNNF which we have not even defined in
the preliminaries. In fact, it is the main strength of the approach in [2] that
the definition and properties of DNNF are not necessary to show our lower
bounds, because we can reduce showing lower bounds on DNNF to a problem in
communication complexity. Since we will not use any properties of DNNF, we
have decided to leave out the definition for space reasons and refer to e.g. [8].
Here we will only use the following result.
Theorem 3 [2]. Let f be a function computed by a DNNF of size s. Then f
has a multi-partition rectangle cover of size s.
Now the reader might be a little puzzled about what multi-partition rectangle
covers of a function are. Since we will also only use them as a black box in our
proofs and do not rely on any of their properties, we have opted to leave out
their definition and refer to [12]
We will use a powerful theorem which follows directly from the results in [12].
Theorem 4 [12]. For every n ∈ N and every m ≤ n /32 there is a linear code
C with a m × n parity check matrix such that every multi-partition rectangle

cover of the characteristic function fC has size at least 14 2m .

4 Accepting Codes by CNF Formulas

In this section we will construct CNF formulas to accept linear codes. We will
first start with a naive encoding that will turn out to be of unbounded modular
treewidth and thus not directly helpful to us. We will then show how to change
the encoding in such a way that the modular treewidth and even the neighbor-
hood diversity are small and the size of the resulting CNF is small enough to
show meaningful lower bounds for encodings in DNNF with Theorem 4.

4.1 The Naive Approach

In this subsection, we show how we can check m linear equations on variables
x1 , . . . , xn efficiently by CNF. The idea is to simply consider one variable after
the other and remember the parity for the equations at hand. To this end, fix an
m × n matrix A = (aij ). We denote the resulting equations of the system Ax̄ = 0
by E1 , . . . , Em . For each equation Ei we introduce variables zij for j ∈ [n] which
intuitively remembers the parity of Ei up to seeing the variable xj .
We encode the computations for each Ei individually: Introduce constraints

ai,1 x1 = zi,1 , (1)

zi,j−1 + aij xj = zij . (2)
 Parameterized Compilation Lower Bounds for Restricted CNF-Formulas 7

Note that zi,n yields the parity for equation Ei which can then be checked for 0.
This yields a system whose accepted inputs projected to the xi are the code
words of the considered code. The constraints have all at most 3 variables, so
we can encode them into CNF easily.
Unfortunately, the resulting CNF can be shown to have high modular tree-
width, so it is not useful for our considerations. We will see how to reduce the
modular treewidth and the neighborhood diversity of the system without blowing
up the size of the resulting CNF-encoding too much.

4.2 Bounding Modular Treewidth

The idea for decreasing the modular treewidth is to not encode all constraints
on the parities individually but combine them into larger constraints. So fix n
and k and set m := k log(n). For each j, we will combine the constraints from
(1) and (2) for blocks of log(n) values of i into one. The resulting constraints are

R1 (x1 , z log(n)+1,1 , . . . , z(+1) log(n),1 ) := {(d1 , t log(n)+1,1 , . . . , t(+1) log(n),1 ) |

ai,1 d1 = ti,1 , i =  log(n) + 1, . . . , ( + 1) log(n)}
and
Rj (xi , z log(n)+1,j−1 , . . . , z(+1) log(n),j−1 , z log(n)+1,j−1 , . . . , z(+1) log(n),j−1 ) :=
{(di , t log(n)+1,j−1 , . . . , t(+1) log(n),j−1 , t log(n)+1,j , . . . , t(+1) log(n),j ) |
ti,j−1 + aij dj = ti,j , i =  log(n) + 1, . . . , ( + 1) log(n)}
for  = 0, . . . , k − 1.
Note that the constraints Rj have at most 2 log(n) + 1 boolean variables,
so we can encode them into CNF of quadratic size where every clause contains
all variables of Rj . Moreover, the Rj encode all previous constraints from (1)
and (2), so the assignments satisfying all Rj projected to the xi still are exactly
the code words of the code we consider. Call the resulting CNF F .
Claim. F has modular pathwidth at most 2k − 1.
Proof. Note that the clauses introduced when translating the constraint Rj into
CNF have by construction all the same set of variables. Thus these clauses have
the same neighborhood type, and we can for modular treewidth restrict to an
instance just having one clause for each Rj . We call the resulting vertex in the
incidence graph r,j . Next, observe that the variables z log(n)+i,j and z log(n)+i ,j
for i, i ∈ [log(n)] appear in exactly the same clauses. Thus these variables have
the same neighborhood type as well, so we can delete all but one of them, say
z log(n),j for  = 1, . . . , k. Call the resulting vertices in the incidence graph s,j .
The resulting graph G = (V, E) has
V ={xj , s,j , r,j | j ∈ [n],  ∈ [k]}.
E ={xj r,j | j ∈ [n],  ∈ [k]} ∪ {s,j−1 r,j | j ∈ [2, n],  ∈ [k]}
∪ {s,j r,j | j ∈ [n],  ∈ [k]}
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of Ninon
de l'Enclos and her century
 This ebook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this ebook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

Title: Ninon de l'Enclos and her century

Author: Mary C. Rowsell

Release date: October 24, 2023 [eBook #71953]

Language: English

Original publication: New York: Brentano's, 1910

Credits: Susan Skinner, John Campbell and the Online

Distributed Proofreading Team at https://www.pgdp.net
(This file was produced from images generously made
available by The Internet Archive)

*** START OF THE PROJECT GUTENBERG EBOOK NINON DE

L'ENCLOS AND HER CENTURY ***
TRANSCRIBER’S NOTE
Footnote anchors are denoted by [number], and the footnotes have been placed at
the end of the book.
New original cover art included with this eBook is granted to the public domain.
Some minor changes to the text are noted at the end of the book.
 NINON DE L’ENCLOS.
From an Original Picture given by herself to the
Countess of Sandwich and by the present Earl of
Sandwich to Mr Walpole 1757
 N I N O N D E L’ E N C L O S
AND HER CENTURY

BY

M. C. ROWSELL
AUTHOR OF
“THE FRIEND OF THE PEOPLE,” “TRAITOR OR PATRIOT,” “THORNDYKE
MANOR,” “MONSIEUR DE PARIS,” ETC. ETC.

WITH ILLUSTRATIONS

B R E N TA N O ’ S
NEW YORK

H U R S T & B L A C K E T T, L I M I T E D
LONDON
1910
Printed in Great Britain
 CONTENTS
PAGE
Chapter I 1
Birth—Parentage—“Arms and the Man”—A Vain Hope—Contraband Novels—A
Change of Educational System—Ninon’s Endowments—The Wrinkle—A Letter
to M. de L’Enclos and What Came of it—A Glorious Time—“Troublesome
Huguenots”—The Château at Loches, and a New Acquaintance—“When Greek
meets Greek”—The Prisoners—“Liberty”—The Shades of Night—Vagabonds?
or Two Young Gentlemen of Consequence?—Tired Out—A Dilemma—Ninon
Herself Again—Consolation.

Chapter II 14
Troublesome Huguenots—Madame de L’Enclos—An Escapade and Nurse
Madeleine—Their Majesties—The Hôtel Bourgogne—The End of the Adventure
—St Vincent de Paul and his Charities—Dying Paternal Counsel—Ninon’s New
Home—Duelling—Richelieu and the Times.

Chapter III 27
A Life-long Friend—St Evrémond’s Courtly Mot—Rabelais v. Petronius—Society
and the Salons—The Golden Days—The Man in Black.

Chapter IV 36
A “Delicious Person”—Voiture’s Jealousy—A Tardy Recognition—Coward
Conscience—A Protestant Pope—The Hôtel de Rambouillet—St Evrémond—
The Duel—Nurse Madeleine—Cloistral Seclusion and Jacques Callot—“Merry
Companions Every One”—and One in Particular.

Chapter V 51
An Excursion to Gentilly—“Uraniæ Sacrum”—César and Ruggieri—The rue
d’Enfer and the Capucins—Perditor—The Love-philtre—Seeing the Devil
—“Now You are Mine!”

Chapter VI 61
Nemesis—Ninon’s Theories—Wits and Beaux of the Salons—Found at Last
—“The Smart Set”—A Domestic Ménage—Scarron—The Fatal Carnival—The
Bond of Ninon—Corneille and The Cid—The Cardinal’s Jealousy—Enlarging
the Borders—Monsieur l’Abbé and the Capon Leg—The Grey Cardinal—A
Faithful Servant.

Chapter VII 81
Mélusine—Cinq-Mars—An Ill-advised Marriage—The Conspiracy—The
 Revenge—The Scaffold—A Cry from the Bastille—The Lady’s Man—“The
Cardinal’s Hangman”—Finis—Louis’s Evensong—A Little Oversight—The
King’s Nightcap—Mazarin—Ninon’s Hero.

Chapter VIII 91
“Loving like a Madman”—A Great Transformation—The Unjust Tax—Parted
Lovers—A Gay Court and A School for Scandal, and Mazarin’s Policy—The
Regent’s Caprices—The King’s Upholsterer’s Young Son—The Théâtre Illustre
—The Company of Monsieur and Molière.

Chapter IX 103
The Rift in the Lute—In the Vexin—The Miracle of the Gardener’s Cottage—
Italian Opera in Paris—Parted Lovers—“Ninum”—Scarron and Françoise
d’Aubigné—Treachery—A Journey to Naples—Masaniello—Renewing
Acquaintances—Mazarin’s Mandate.

Chapter X 115
The Fronde and Mazarin—A Brittany Manor—Borrowed Locks—The Flight to St
Germains—A Gouty Duke—Across the Channel—The Evil Genius—The
Scaffold at Whitehall—Starving in the Louvre—The Mazarinade—Poverty—
Condé’s Indignation—The Cannon of the Bastille—The Young King.

Chapter XI 124
Invalids in the rue des Tournelles—On the Battlements—“La Grande
Mademoiselle”—Casting Lots—The Sacrifice—The Bag of Gold—“Get Thee to
a Convent”—The Battle of the Sonnets—A Curl-paper—The Triumph and
Defeat of Bacchus—A Secret Door—Cross Questions and Crooked Answers—
The Youthful Autocrat.

Chapter XII 135

The Whirligig of Time, and an Old Friend—Going to the Fair—A Terrible
Experience—The Young Abbé—“The Brigands of La Trappe”—The New
Ordering—An Enduring Memory—The King over the Water—Unfulfilled
Aspirations—“Not Good-looking.”

Chapter XIII 144

Christina’s Modes and Robes—Encumbering Favour—A Comedy at the Petit-
Bourbon—The Liberty of the Queen and the Liberty of the Subject—Tears and
Absolutions—The Tragedy in the Galérie des Cerfs—Disillusions.

Chapter XIV 154

Les Précieuses Ridicules—Sappho and Le Grand Cyrus—The Poets of the
Latin Quarter—The Satire which Kills—A Lost Child—Periwigs and New Modes
—The Royal Marriage and a Grand Entry.
Chapter XV 163
Réunions—The Scarrons—The Fête at Vaux—The Little Old Man in the
Dressing-gown—Louise de la Vallière—How the Mice Play when the Cat’s
Away—“Pauvre Scarron”—An Atrocious Crime.

Chapter XVI 175

A Lettre de Cachet—Mazarin’s Dying Counsel—Madame Scarron Continues to
Receive—Fouquet’s Intentions and What Came of Them—The Squirrel and the
Snake—The Man in the Iron Mask—An Incommoding Admirer—“Calice cher, ou
le parfum n’est plus”—The Roses’ Sepulchre.

Chapter XVII 185

A Fashionable Water-cure Resort—M. de Roquelaure and his Friends—Louis le
Grand—“A Favourite with the Ladies”—The Broken Sword—A Billet-doux—La
Vallière and la Montespan—The Rebukes from the Pulpit—Putting to the Test—
Le Tartufe—The Triumphs of Molière—The Story of Clotilde.

Chapter XVIII 199

A Disastrous Wooing—Fénelon—“Mademoiselle de L’Enclos”—The Pride that
had a Fall—The Death of the Duchesse d’Orléans—Intrigue—The Sun-King
and the Shadows—The Clermont Scholar’s Crime—Monsieur de Montespan—
Tardy Indignation—The Encounter—The Filles Répenties—What the Cards
Foretold.

Chapter XIX 212

“In Durance Vile”—Molière’s Mot—The Malade Imaginaire—“Rogues and
Vagabonds”—The Passing of Molière—The Narrowing Circle—Fontenelle—Lulli
—Racine—The Little Marquis—A Tardy Pardon—The Charming Widow Scarron
—A Journey to the Vosges, and the Haunted Chamber.

Chapter XX 228
The Crime of Madame Tiquet—A Charming Little Hand—Aqua Toffana—The
Casket—A Devout Criminal—The Sinner and the Saint—Monsieur de Lauzun’s
Boots—“Sister Louise”—La Fontange—“Madame de Maintenant”—The Blanks
in the Circle—The Vatican Fishes and their Good Example—Piety at Versailles
—The Periwigs and the Paniers—Père la Chaise—A Dull Court—Monsieur de
St Evrémond’s Decision.

Chapter XXI 241

A Distinguished Salon—The Duke’s Homage—Quietism—The Disastrous Edict
—The Writing on the Window-pane—The Persecution of the Huguenots—The
Pamphleteers—The Story of Jean Larcher and The Ghost of M. Scarron—The
Two Policies.

251
Chapter XXII
Mademoiselle de L’Enclos’ Cercle—Madeleine de Scudéri—The Abbé Dubois
—“The French Calliope,” and the Romance of her Life—“Revenons à nos
Moutons”—A Resurrection?—Racine and his Detractors—“Esther”—Athalie and
St Cyr—Madame Guyon and the Quietists.

Chapter XXIII 263

A Grave Question—The Troublesome Brother-in-Law—“No Vocation”—The
Duke’s Choice—Peace for “La Grande Mademoiselle”—An Invitation to
Versailles—Behind the Arras—Between the Alternatives—D’Aubigné’s Shadow
—A Broken Friendship.

Chapter XXIV 275

The Falling of the Leaves—Gallican Rights—“The Eagle of Meaux”—Condé’s
Funeral Oration—The Abbé Gedouin’s Theory—A Bag of Bones—Marriage and
Sugar-plums—The Valour of Monsieur du Maine—The King’s Repentance—
The next Campaign—La Fontaine and Madame de Sablière—MM. de Port
Royal—The Fate of Madame Guyon—“Mademoiselle Balbien.”

Chapter XXV 288

The Melancholy King—The Portents of the Storm—The Ambition of Madame
Louise Quatorze—The Farrier of Provence—The Ghost in the Wood—Ninon’s
Objection—The King’s Conscience—A Dreary Court—Racine’s Slip of the
Tongue—The Passing of a Great Poet, and a Busy Pen Laid Down.

Chapter XXVI 301

Leaving the Old Home—“Wrinkles”—Young Years and Old Friends—“A Bad
Cook and a Little Bit of Hot Coal”—Voltaire—Irène—Making a Library—“Adieu,
Mes Amis”—The Man in Black.
 L I S T O F I L L U S T R AT I O N S

Ninon de L’Enclos Frontispiece

Cardinal Richelieu To face page 24

De la Rochefoucauld ” 48

Molière ” 100

St Evrémond ” 112

Ninon de L’Enclos ” 127

NINON DE L’ENCLOS
AND HER CENTURY
NINON DE L’ENCLOS
AND HER CENTURY
 CHAPTER I
Birth—Parentage—“Arms and the Man”—A Vain Hope—Contraband Novels—A
Change of Educational System—Ninon’s Endowments—The Wrinkle—A
Letter to M. de L’Enclos and What Came of it—A Glorious Time
—“Troublesome Huguenots”—The Château at Loches, and a New
Acquaintance—“When Greek meets Greek”—The Prisoners—“Liberty”—The
Shades of Night—Vagabonds? or Two Young Gentlemen of Consequence?—
Tired Out—A Dilemma—Ninon Herself Again—Consolation.

Anne de L’Enclos was born in Paris in 1615. She was the daughter
of Monsieur de L’Enclos, a gentleman of Touraine, and of his wife, a
member of the family of the Abra de Raconis of the Orléanois.
It would not be easy to find characteristics more diverse than
those distinguishing this pair. Their union was an alliance arranged
for them—a mariage de convenance. Diametrically opposite in
temperament, Monsieur was handsome and distinguished-looking;
while the face and figure of Madame were ordinary. She was
constitutionally timid, and intellectually narrow, devoted to
asceticism, and reserved in manner. She passed her time in
seclusion, dividing it between charitable works, the reading of pious
books, and attendance at Mass and the other services of the
Church. Monsieur de L’Enclos, on the other hand, was a votary of
every pleasure and delightful distraction the world could afford him.
Among them he counted duelling; he was a skilled swordsman, and
his rapier play was of the finest. A brave and gallant soldier, he had
served the royal cause during the later years of Henri IV., and so on
into the reign of Louis XIII. He was a bon vivant, and arms and
intrigue, which were as the breath of life to him, he sought after
wherever the choicest opportunities of those were likely to be found.
Notwithstanding, the rule of life-long bickering and mutual
reproach attending such ill-assorted unions, would seem to be
proved by its exception in the case of Ninon’s parents; since no
record of any such domestic strife stands against them. Bearing and
forbearing, they agreed to differ, and went their several ways—
Madame de L’Enclos undertaking the training and instruction of
Ninon in those earliest years, in the fond hope that there would be a
day when she should take the veil and become a nun. Before,
however, she attained to the years of as much discretion as she ever
possessed, she had arrived at the standpoint of the way she
intended to take of the life before her, which was to roll into years
that did not end until the dawning of the eighteenth century; and it in
no way included any such intention. So sturdily opposed to it,
indeed, was she, that it irresistibly suggests the possibility of her
being the inspiration of the old song—“Ninon wouldn’t be a nun”—
“I shan’t be a nun, I won’t be a nun,
I am so fond of pleasure that I won’t be a nun!”

For Ninon was her father’s child; almost all her inherited instincts
were from him. The endeavours of Madame de L’Enclos failed
disastrously. The monotony and rigid routine of the young girl’s life
repelled the bright, frank spirit, and drove it to opposite extreme,
resulting in sentiments of disgust for the pious observances of her
church; and taken there under compulsion day in, day out, she
usually contrived to substitute some plump little volume of romance,
or other light literature, at the function, for her Mass-book and
breviary, to while away the tedium.
In no very long time Monsieur de L’Enclos, noting the bent of his
daughter’s nature, himself took over her training. He carried it on, it
is scarcely necessary to say, upon a plane widely apart from the
mother’s. A man of refined intellect, he had studied the books and
philosophy of the renaissance of literature; and before Ninon was
eleven years old, while imbuing her with the love of reading such
books as the essays of Montaigne and the works of Charon, he
accustomed her to think and to reason for herself, an art of which
she very soon became a past-mistress, the result being an ardent
recognition of the law of liberty, and the Franciscan counsel of
perfection: “Fay ce qu’et voudray.” Ninon possessed an excellent gift
of tongues, cultivating it to the extent of acquiring fluently, Italian,
Spanish, and English, rendered the more easy of mastery from her
knowledge of Latin, which she so frequently quotes in her
correspondence.
Her love of music was great; she sang well, and was a proficient
on the lute, in which her father himself, a fine player, instructed her.
She conversed with facility, and doubtless took care to cultivate her
natural gifts in those days when the arts of conversation and
causerie were indispensable for shining in society, and she loved to
tell a good story; but she drew a distinct line at reciting. One day
when Mignard, the painter, deplored his handsome daughter’s
defective memory, she consoled him—“How fortunate you are,” she
said, “she cannot recite.”
The popular acceptation of Ninon de L’Enclos’ claims to celebrity
would appear to be her beauty, which she retained to almost the end
of her long life—a beauty that was notable; but it lay less in
perfection of the contours of her face, than in the glorious freshness
of her complexion, and the expression of her magnificent eyes, at
once vivacious and sympathetic, gentle and modest-glancing, yet
brilliant with voluptuous languor. Any defects of feature were
probably those which crowned their grace—and when as in the
matter of a slight wrinkle, which in advanced years she said had
rudely planted itself on her forehead, the courtly comment on this of
Monsieur de St Evrémond was to the effect that “Love had placed it
there to nestle in.” Her well-proportioned figure was a little above
middle height, and her dancing was infinitely graceful.
Provincial by descent, Mademoiselle de L’Enclos was a born
Parisian, in that word’s every sense. Her bright eyes first opened in a
small house lying within the shadows of Notre-Dame, the old Cité
itself, the heart of hearts of Paris, still at that time fair with green
spaces and leafy hedgerows, though these were to endure only a
few years longer. Her occasionally uttered wish that she had been
born a man, hardly calls for grave consideration. The desire to don
masculine garments and to ride and fence and shoot, and to indulge
generally in manly pursuits, occurred to her when she was still short
of twelve years old, by which time she was able to write well; and her
earliest epistolary correspondence included a letter addressed to her
father. It ran as follows:—

“My Very Honoured Father,—I am eleven years old. I am big and

strong; but I shall certainly fall ill, if I continue to assist at three masses
every day, especially on account of one performed by a great, gouty, fat
canon, who takes at least twelve minutes to get through the Epistle and
the Gospel, and whom the choir boys are obliged to put back again on
his feet after each genuflexion. I would as soon see one of the towers of
Notre Dame on the altar-steps; they would move quite as quickly, and
not keep me so long from breakfast. This is not at all cheering I can tell
you. In the interest of the health of your only child, it is time to put an
end to this state of things. But in what manner, you will ask, and how is it
to be set about? Nothing more simple. Let us suppose that instead of
me, Heaven had given you a son: I should have been brought up by you,
and not by my mother; already you would have begun to instruct me in
arms, and mounted me on horseback, which would have much better
pleased me than twiddling along the beads of a rosary to Aves, Paters,
and Credos. The present moment is the one for me to inform you that I
decide to be no longer a girl, and to become a boy.
“Will you therefore arrange to send for me to come to you, in order to
give me an education suitable to my new sex? I am with respect, my
very honoured father,—Your little
Ninon.”

This missive, which Ninon contrived to get posted without her

mother’s knowledge, met with her father’s hearty approval. No more
time was lost than it took to make her a handsome suit of clothes, of
the latest mode, the one bearing the palm for grace and
picturesqueness, far and away from all the fashions of men’s attire,
speaking for itself in the canvases of Vandyck; and Ninon stands
forth in the gallant bravery of silken doublet, with large loose sleeves
slashed to the shoulder; her collar a falling band of richest point lace;
the short velvet cloak hanging to the shoulder; the fringed breeches
meeting the wide-topped boots frilled about with fine lawn; the
plumed, broad-brimmed Flemish beaver hat, well-cocked to one side
upon the graceful head, covered with waves of dark hair falling to the
neck; gauntleted gloves of Spanish leather; her rapier hanging from
the richly-embroidered baldric crossing down from the right shoulder
—a picture that thrilled the heart of Monsieur de L’Enclos with
ecstasy; and when, splendidly mounted, she rode forth, ruffling it
gallantly beside him, he was the proud recipient of many a
compliment and encomium on the son of whose existence until now
nobody had been as much as aware.
These delightful days were destined, however, to come quickly to
an end. Fresh disturbances arose with the Huguenots of La Rochelle
and Loudun, and Monsieur de L’Enclos was summoned to join his
regiment. Ninon would doubtless have liked of all things to go with
him; but while this was impossible, she was spared the dreaded
alternative of the fat canon and the three Masses a day, by her father
accepting for her an invitation from his sister, the Baroness
Montaigu, who lived on her estate near Loches, on the borders of the
Indre. This lady, a widow and childless, had long been desirous of
making the acquaintance of her young niece, and on his way north-
west, Monsieur de L’Enclos left Ninon at the château. “And when we
have settled these wretched Huguenots,” said Monsieur de L’Enclos,
as he bade her farewell, and slipped a double louis into her hands, “I
will return for you.”
Madame de Montaigu was a charming lady, of the same spirited,
gay temperament as her brother. She received her niece with the
utmost kindness, and having been initiated into the girl’s whim for
playing the boy, she laughingly fell in with it, and addressed her with
the greatest gravity as “my pretty nephew,” introducing to her, a—
shall it be said?—another young gentleman, by name François de la
Rochefoucauld, Prince de Marsillac, the son of her intimate friend,
the Duchesse de la Rochefoucauld. The lad was a pupil at the
celebrated Jesuits’ College of La Flêche, founded by Henri IV., and
usually spent part of his holidays at the Loches château.
A year or two older than Ninon, Marsillac was a shy and retiring
boy, and at first rather shrank from his robustious new companion,
who, however, soon contrived to draw him out, putting him on his