Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

Formal Techniques for Distributed Objects

Components and Systems 39th IFIP WG 6 1
International Conference FORTE 2019 Held as Part
of the 14th International Federated Conference on
Distributed Computing Techniques DisCoTec 2019
https://textbookfull.com/product/formal-techniques-for-
Kongens Lyngby Denm Jorge A. Pérez
distributed-objects-components-and-systems-39th-ifip-
wg-6-1-international-conference-forte-2019-held-as-part-of-
the-14th-international-federated-conference-on-distributed-
computi/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Jorge A. Pérez
Nobuko Yoshida (Eds.)

Formal Techniques
LNCS 11535

for Distributed Objects,

Components, and Systems
39th IFIP WG 6.1 International Conference, FORTE 2019
Held as Part of the 14th International Federated Conference
on Distributed Computing Techniques, DisCoTec 2019
Kongens Lyngby, Denmark, June 17–21, 2019, Proceedings

123
Lecture Notes in Computer Science 11535
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Editorial Board Members

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, Zurich, 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, Chennai, 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
More information about this series at http://www.springer.com/series/7408
Jorge A. Pérez Nobuko Yoshida (Eds.)
•

Formal Techniques
for Distributed Objects,
Components, and Systems
39th IFIP WG 6.1 International Conference, FORTE 2019
Held as Part of the 14th International Federated Conference
on Distributed Computing Techniques, DisCoTec 2019
Kongens Lyngby, Denmark, June 17–21, 2019
Proceedings

123
Editors
Jorge A. Pérez Nobuko Yoshida
University of Groningen Imperial College London
Groningen, The Netherlands London, UK

ISSN 0302-9743 ISSN 1611-3349 (electronic)

Lecture Notes in Computer Science
ISBN 978-3-030-21758-7 ISBN 978-3-030-21759-4 (eBook)
https://doi.org/10.1007/978-3-030-21759-4
LNCS Sublibrary: SL2 – Programming and Software Engineering

© IFIP International Federation for Information Processing 2019

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, expressed or implied, with respect to the material contained herein or for any errors or
omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
 Foreword

The 14th International Federated Conference on Distributed Computing Techniques

(DisCoTec) took place in Kongens Lyngby, Denmark, during June 17–21, 2019. It was
organized by the Department of Applied Mathematics and Computer Science at the
Technical University of Denmark.
The DisCoTec series is one of the major events sponsored by the International
Federation for Information Processing (IFIP). It comprised three conferences:
– COORDINATION, the IFIP WG 6.1 21st International Conference on Coordina-
tion Models and Languages
– DAIS, the IFIP WG 6.1 19th International Conference on Distributed Applications
and Interoperable Systems
– FORTE, the IFIP WG 6.1 39th International Conference on Formal Techniques for
Distributed Objects, Components and Systems
Together, these conferences cover a broad spectrum of distributed computing
subjects, ranging from theoretical foundations and formal description techniques to
systems research issues.
In addition to the individual sessions of each conference, the event included several
plenary sessions that gathered attendants from the three conferences. This year, the
general chair and the DisCoTec Steering Committee joined the three DisCoTec con-
ferences in the selection and nomination of the plenary keynote speakers, whose
number was accordingly increased from the traditional three to five. The five keynote
speakers and the title of their talks are listed below:
– Prof. David Basin (ETH Zürich, Switzerland) – “Security Protocols: Model
Checking Standards”
– Dr. Anne-Marie Kermarrec (Inria Rennes, France) – “Making Sense of Fast Big
Data”
– Prof. Marta Kwiatkowska (University of Oxford, UK) – “Versatile Quantitative
Modelling: Verification, Synthesis and Data Inference for Cyber-Physical Systems”
– Prof. Silvio Micali (MIT, USA)—“ALGORAND—The Distributed Ledger for the
Borderless Economy”
– Prof. Martin Wirsing (LMU, Germany) – “Toward Formally Designing Collective
Adaptive Systems”
As is traditional in DisCoTec, an additional joint session with the best papers from
each conference was organized. The best papers were:
– “Representing Dependencies in Event Structures” by G. Michele Pinna
(Coordination)
– “FOUGERE: User-Centric Location Privacy in Mobile Crowdsourcing Apps” by
Lakhdar Meftah, Romain Rouvoy and Isabelle Chrisment (DAIS)
vi Foreword

– “Psi-Calculi Revisited: Connectivity and Compositionality” by Johannes Åman

Pohjola (FORTE)
Associated with the federated event were also two satellite events that took place:
– ICE, the 12th International Workshop on Interaction and Concurrency Experience
– DisCoRail, the First International Workshop on Distributed Computing in Future
Railway Systems
I would like to thank the Program Committee chairs of the different events for their
help and cooperation during the preparation of the conference, and the Steering
Committee and Advisory Boards of DisCoTec and their conferences for their guidance
and support. The organization of DisCoTec 2019 was only possible thanks to the
dedicated work of the Organizing Committee, including Francisco “Kiko” Fernández
Reyes and Francesco Tiezzi (publicity chairs), Maurice ter Beek, Valerio Schiavoni,
and Andrea Vandin (workshop chairs), Ann-Cathrin Dunker (logistics and finances), as
well as all the students and colleagues who volunteered their time to help. Finally, I
would like to thank IFIP WG 6.1 for sponsoring this event, Springer’s Lecture Notes in
Computer Science team for their support and sponsorship, EasyChair for providing the
reviewing infrastructure, the Nordic IoT Hub for their sponsorship, and the Technical
University of Denmark for providing meeting rooms and additional support.

June 2019 Alberto Lluch Lafuente

 Preface

This volume contains the papers presented at FORTE 2019: the 39th IFIP WG 6.1
International Conference on Formal Techniques for Distributed Objects, Components,
and Systems. FORTE 2019 was held as one of three main conferences of the 14th
International Federated Conference on Distributed Computing Techniques (DisCoTec),
during June 17–21, 2019 in Lyngby, Denmark.
FORTE is a well-established forum for fundamental research on theory, models,
tools, and applications for distributed systems, with special interest in:
– Software quality, reliability, availability, and safety
– Security, privacy, and trust in distributed and/or communicating systems
– Service-oriented, ubiquitous, and cloud computing systems
– Component- and model-based design
– Object technology, modularity, software adaptation
– Self-stabilization and self-healing/organizing
– Verification, validation, formal analysis, and testing of the above
The Program Committee received a total of 42 quality submissions, written by
authors from 21 different countries. Of these, 18 papers were selected for inclusion in
the scientific program: 15 full papers, one short paper, and two “journal first” papers—a
new submission category we introduced this year. Each submission was reviewed by at
least three Program Committee members with the help of external reviewers in selected
cases. There was one submission with which both of us declared ourselves in conflict;
Uwe Nestmann kindly agreed to oversee and lead the discussion for this submission,
which was eventually accepted.
The selection of accepted submissions was based on electronic discussions via the
EasyChair conference management system. Toward the end of this electronic discus-
sion, there was a two-day physical meeting in which we discussed the referee reports
for each submission with the relevant Program Committee members. We found this
combination of electronic and physical discussion highly effective.
As program chairs, we actively contributed to the selection of the five keynote
speakers of DisCoTec 2019:
– Prof. David Basin (ETH Zürich, Switzerland)
– Dr. Anne-Marie Kermarrec (Inria Rennes, France)
– Prof. Marta Kwiatkowska (University of Oxford, UK)
– Prof. Silvio Micali (MIT, USA)
– Prof. Martin Wirsing (LMU, Germany)
We are most grateful to Prof. Basin for accepting our invitation as FORTE-related
keynote speaker. This volume includes the abstract of his keynote talk: “Security
Protocols: Model Checking Standards.”
viii Preface

As is traditional in DisCoTec, a joint session with the best papers from each main
conference was organized. The best paper of FORTE 2019 was “Psi-Calculi Revisited:
Connectivity and Compositionality” by Johannes Åman Pohjola (Data61/CSIRO,
University of New South Wales, Australia).
We wish to thank all the authors of submitted papers, all the members of the
Program Committee for their thorough evaluations of the submissions, and the 26
external reviewers who assisted the evaluation process. We are also indebted to the
Steering Committee of FORTE for their advice and suggestions. Last but not least, we
thank the DisCoTec general chair, Alberto Lluch Lafuente, and his organization team
for their hard, effective work on providing an excellent environment for FORTE 2019
and all other conferences and workshops.

April 2019 Jorge A. Pérez

Nobuko Yoshida
 Organization

Program Committee
Samik Basu Iowa State University, USA
Annette Bieniusa University of Kaiserslautern, Germany
Stefano Calzavara Università Ca’ Foscari Venezia, Italy
Natalia Chechina Bournemouth University, UK
Mila Dalla Preda University of Verona, Italy
Rayna Dimitrova University of Leicester, UK
Patrick Eugster University of Lugano (USI), Switzerland
Ichiro Hasuo National Institute of Informatics, Japan
Thomas Hildebrandt University of Copenhagen, Denmark
Sophia Knight University of Minnesota, USA
Etienne Lozes I3S, University of Nice and CNRS, France
Emanuela Merelli University of Camerino, Italy
Roland Meyer TU Braunschweig, Germany
Uwe Nestmann TU Berlin, Germany
Gustavo Petri IRIF, Paris Diderot, Paris 7, France
Jorge A. Pérez University of Groningen, The Netherlands
Willard Rafnsson IT University of Copenhagen, Denmark
Anne Remke WWU Münster, Germany
Guido Salvaneschi TU Darmstadt, Germany
Cesar Sanchez IMDEA Software Institute, Spain
Ana Sokolova University of Salzburg, Austria
Alexander J. Summers ETH Zurich, Switzerland
Peter Thiemann Universität Freiburg, Germany
Jaco van de Pol Aarhus University, Denmark
Tim Willemse Eindhoven University of Technology, The Netherlands
Nobuko Yoshida Imperial College London, UK
Lukasz Ziarek SUNY Buffalo, USA

Additional Reviewers

Aldini, Alessandro Keiren, Jeroen Sasse, Ralf

Alvim, Mario S. Madiot, Jean-Marie Savvides, Savvas
Åman Pohjola, Johannes Maestri, Stefano Schweizer, Sebastian
Back, Christoffer Olling Maubert, Bastien Sedwards, Sean
Chini, Peter Menikkumbura, Danushka Tesei, Luca
Courtieu, Pierre Neumann, Elisabeth Wolff, Sebastian
Dubut, Jérémy Otoni, Rodrigo Yamada, Akihisa
Francalanza, Adrian Pilch, Carina Zeller, Peter
Inverso, Omar Radanne, Gabriel
 Security Protocols: Model Checking Standards
(Invited Talk)

David Basin

Department of Computer Science, ETH Zurich, Switzerland

The design of security protocols is typically approached as an art, rather than a science,
and often with disastrous consequences. But this need not be so! I have been working
for ca. 20 years on foundations, methods, and tools, both for developing protocols that
are correct by construction [9, 10] and for the post-hoc verification of existing designs
[1–4, 8]. In this talk I will introduce my work in this area and describe my experience
analyzing, improving, and contributing to different industry standards, both existing
and upcoming [5–7].

References
1. Basin, D.: Lazy infinite-state analysis of security protocols. In: Secure Networking — CQRE
[Secure] 1999. CQRE. LNCS, vol. 1740, pp. 30–42. Springer, Heidelberg (1999)
2. Basin, D., Cremers, C., Dreier, J., Sasse, R.: Symbolically analyzing security protocols using
tamarin. SIGLOG News 4(4), 19–30 (2017). https://doi.org/10.1145/3157831.3157835
3. Basin, D., Cremers, C., Meadows, C.: Model checking security protocols. In: Clarke, E.,
Henzinger, T., Veith, H., Bloem, R. (eds.) Handbook of Model Checking, pp. 727–762.
Springer, Cham (2018)
4. Basin, D., Mödersheim, S., Viganò, L.: OFMC: a symbolic model checker for security
protocols. Int. J. Inf. Secur. 4(3), 181–208 (2005). published online December 2004
5. Basin, D.A., Cremers, C., Meier, S.: Provably repairing the ISO/IEC 9798 standard for entity
authentication. J. Comput. Secur. 21(6), 817–846 (2013)
6. Basin, D.A., Cremers, C.J.F., Miyazaki, K., Radomirovic, S., Watanabe, D.: Improving the
security of cryptographic protocol standards. IEEE Secur. Priv. 13(3), 24–31 (2015). https://
doi.org/10.1109/MSP.2013.162, http://dx.doi.org/10.1109/MSP.2013.162
7. Basin, D.A., Dreier, J., Hirschi, L., Radomirovic, S., Sasse, R., Stettler, V.: Formal analysis of
5G authentication. CoRR abs/1806.10360 (2018). http://arxiv.org/abs/1806.10360
8. Schmidt, B., Meier, S., Cremers, C., Basin, D.: Automated analysis of Diffie-Hellman pro-
tocols and advanced security properties. In: Proceedings of the 25th IEEE Computer Security
Foundations Symposium (CSF), pp. 78–94 (2012)
9. Sprenger, C., Basin, D.: Refining key establishment. In: Proceedings of the 25th IEEE
Computer Security Foundations Symposium (CSF), pp. 230–246 (2012)
10. Sprenger, C., Basin, D.: Refining security protocols. J. Comput. Secur. 26(1), 71–120
(2018). https://doi.org/10.3233/JCS-16814, http://dx.doi.org/10.3233/JCS-16814
 Contents

Full Papers

Psi-Calculi Revisited: Connectivity and Compositionality . . . . . . . . . . . . . . . 3

Johannes Åman Pohjola

Squeezing Streams and Composition of Self-stabilizing Algorithms . . . . . . . . 21

Karine Altisen, Pierre Corbineau, and Stéphane Devismes

Parametric Updates in Parametric Timed Automata . . . . . . . . . . . . . . . . . . . 39

Étienne André, Didier Lime, and Mathias Ramparison

Parametric Statistical Model Checking of UAV Flight Plan . . . . . . . . . . . . . 57

Ran Bao, Christian Attiogbe, Benoît Delahaye, Paulin Fournier,
and Didier Lime

Only Connect, Securely . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Chandrika Bhardwaj and Sanjiva Prasad

Output-Sensitive Information Flow Analysis . . . . . . . . . . . . . . . . . . . . . . . . 93

Cristian Ene, Laurent Mounier, and Marie-Laure Potet

Component-aware Input-Output Conformance. . . . . . . . . . . . . . . . . . . . . . . 111

Alexander Graf-Brill and Holger Hermanns

Declarative Choreographies and Liveness . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Thomas T. Hildebrandt, Tijs Slaats, Hugo A. López, Søren Debois,
and Marco Carbone

Model Checking HPnGs in Multiple Dimensions: Representing State Sets

as Convex Polytopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Jannik Hüls and Anne Remke

Causal-Consistent Replay Debugging for Message Passing Programs. . . . . . . 167

Ivan Lanese, Adrián Palacios, and Germán Vidal

Correct and Efficient Antichain Algorithms for Refinement Checking . . . . . . 185

Maurice Laveaux, Jan Friso Groote, and Tim A. C. Willemse

Towards Verified Blockchain Architectures: A Case Study on Interactive

Architecture Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Diego Marmsoler
xiv Contents

Unfolding-Based Dynamic Partial Order Reduction of Asynchronous

Distributed Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
The Anh Pham, Thierry Jéron, and Martin Quinson

Encapsulation and Sharing in Dynamic Software Architectures:

The Hypercell Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Jean-Bernard Stefani and Martin Vassor

Decentralized Real-Time Safety Verification for Distributed

Cyber-Physical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Hoang-Dung Tran, Luan Viet Nguyen, Patrick Musau, Weiming Xiang,
and Taylor T. Johnson

Short and “Journal First” Papers

On Certifying Distributed Algorithms: Problem of Local Correctness. . . . . . . 281

Kim Völlinger

On a Higher-Order Calculus of Computational Fields . . . . . . . . . . . . . . . . . 289

Giorgio Audrito, Mirko Viroli, Ferruccio Damiani, Danilo Pianini,
and Jacob Beal

Semantically Sound Analysis of Content Security Policies . . . . . . . . . . . . . . 293

Stefano Calzavara, Alvise Rabitti, and Michele Bugliesi

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299

Full Papers
 Psi-Calculi Revisited: Connectivity
and Compositionality

Johannes Åman Pohjola1,2(B)

1
Data61/CSIRO, Sydney, Australia
[email protected]
2
University of New South Wales, Sydney, Australia

Abstract. Psi-calculi is a parametric framework for process calculi sim-

ilar to popular pi-calculus extensions such as the explicit fusion calcu-
lus, the applied pi-calculus and the spi calculus. Mechanised proofs of
standard algebraic and congruence properties of bisimilarity apply to all
calculi within the framework.
A limitation of psi-calculi is that communication channels must be
symmetric and transitive. In this paper, we give a new operational seman-
tics to psi-calculi that allows us to lift these restrictions and simplify
some of the proofs. The key technical innovation is to annotate transi-
tions with a provenance—a description of the scope and channel they
originate from.
We give mechanised proofs that our extension is conservative, and
that the standard algebraic and congruence properties of bisimilarity are
maintained. We show correspondence with a reduction semantics and
barbed bisimulation. We show how a pi-calculus with preorders that was
previously beyond the scope of psi-calculi can be captured, and how to
encode mixed choice under very strong quality criteria.

Keywords: Process algebra · Psi-calculi · Nominal logic ·

Interactive theorem proving · Bisimulation

1 Introduction

This paper is mainly concerned with channel connectivity, by which we mean the
relationship that describes which input channels are connected to which output
channels in a setting with message-passing concurrency. In the pi-calculus [18],
channel connectivity is syntactic identity: in the process

a(x).P | b y.Q

where one parallel component is waiting to receive on channel a and the other
is waiting to send on channel b, interaction is possible only if a = b.

c IFIP International Federation for Information Processing 2019

Published by Springer Nature Switzerland AG 2019
J. A. Pérez and N. Yoshida (Eds.): FORTE 2019, LNCS 11535, pp. 3–20, 2019.
https://doi.org/10.1007/978-3-030-21759-4_1
4 J. Åman Pohjola

Variants of the pi-calculus may have more interesting channel connectivity.

The explicit fusion calculus pi-F [9] extends the pi-calculus with a primitive for
fusing names; once fused, they are treated as being for all purposes one and the
same. Channel connectivity is then given by the equivalence closure of the name
fusions. For example, if we extend the above example with the fusion (a = b)

a(x).P | b y.Q | (a = b)
then communication is possible. Other examples may be found in e.g. calculi for
wireless communication [19], where channel connectivity can be used to directly
model the network’s topology.
Psi-calculi [2] is a family of applied process calculi, where standard meta-
theoretical results, such as the algebraic laws and congruence properties of bisim-
ulation, have been established once and for all through mechanised proofs [3] for
all members of the family. Psi-calculi generalises e.g. the pi-calculus and the
explicit fusion calculus in several ways. In place of atomic names it allows chan-
nels and messages to be taken from an (almost) freely chosen term language.
In place of fusions, it admits the formulas of an (almost) freely chosen logic as
first-class processes. Channel connectivity is determined by judgements of said
logic, with one restriction: the connectivity thus induced must be symmetric and
transitive.
The main contribution of the present paper is a new way to define the seman-
tics of psi-calculi that lets us lift this restriction, without sacrificing any of the
algebraic laws and compositionality properties. It is worth noting that this was
previously believed to be impossible: Bengtson et al. [2, p. 14] even offer coun-
terexamples to the effect that without symmetry and transitivity, scope extension
is unsound. However, a close reading reveals that these counterexamples apply
only to their particular choice of labelled semantics, and do not rule out the
possibility that the counterexamples could be invalidated by a rephrasing of the
labelled semantics such as ours.
The price we pay for this increased generality is more complicated transition
labels: we decorate input and output labels with a provenance that keeps track
of which prefix a transition originates from. The idea is that if I am an input
label and you are an output label, we can communicate if my subject is your
provenance, and vice versa. This is offset by other simplifications of the semantics
and associated proofs that provenances enable.

Contributions. This paper makes the following specific technical contributions:

– We define a new semantics of psi-calculi that lifts the requirement that chan-
nel connectivity must be symmetric and transitive, using the novel technical
device of provenances (Sect. 2).
– We prove that strong bisimulation is a congruence and satisfies the usual alge-
braic laws such as scope extension. Interestingly, provenances can be ignored
for the purpose of bisimulation. These proofs are machine-checked1 in Nomi-
nal Isabelle [24] (Sect. 3.1).
1
Isabelle proofs are available at https://github.com/IlmariReissumies/newpsi.
 Psi-Calculi Revisited: Connectivity and Compositionality 5

– We prove, again using Nominal Isabelle, that this paper’s developments con-
stitute a conservative extension of the original psi-calculi (Sect. 3.2).
– We further validate our semantics by defining a reduction semantics and
strong barbed congruence, and showing that they agree with their labelled
counterparts (Sect. 3.2).
– We capture a pi-calculus with preorders by Hirschkoff et al. [11], that was
previously beyond the scope of psi-calculi because of its non-transitive channel
connectivity. The bisimilarity we obtain turns out to coincide with that of
Hirschkoff et al. (Sect. 4.1).
– We exploit non-transitive connectivity to show that mixed choice is a derived
operator of psi-calculi in a very strong sense: its encoding is fully abstract
and satisfies strong operational correspondence (Sect. 4.2).

For lack of space we elide proofs; please see the associated technical report [1].

2 Definitions
This section introduces core definitions such as syntax and semantics. Many
definitions are shared with the original presentation of psi-calculi, so this section
also functions as a recapitulation of [2]. We will highlight the places where the
two differ.
We assume a countable set of names N ranged over by a, b, c, . . . , x, y, z. A
nominal set [8] is a set equipped with a permutation action ·; intuitively, if X ∈ X
and X is a nominal set, then (x y) · X, which denotes X with all occurrences
of the name x swapped for y and vice versa, is also an element of X. n(X) (the
support of X) is, intuitively, the set of names such that swapping them changes
X. We write a#X (“a is fresh in X) for a ∈ / n(X). A nominal set X has finite
support if for every X ∈ X, n(X) is finite. A function symbol f is equivariant if
p · f (x) = f (p · x); this generalises to n-ary function symbols in the obvious way.
Whenever we define inductive syntax with names, it is implicitly quotiented by
permutation of bound names, so e.g. (νx)ax = (νy)ay if x, y#a.
Psi-calculi is parameterised on an arbitrary term language and a logic of
environmental assertions:
.
Definition 1 (Parameters). A psi-calculus is a 7-tuple (T, A, C, , ⊗, 1, →)
with three finitely supported nominal sets:

1. T, the terms, ranged over by M, N, K, L, T ;

2. A, the assertions, ranged over by Ψ ; and
3. C, the conditions, ranged over by ϕ.

We assume each of the above is equipped with a substitution function [ := ] that

substitutes (sequences of ) terms for names. The remaining three parameters are
equivariant function symbols written in infix:
 : A × C ⇒ bool (entailment) ⊗ : A × A ⇒ A (composition)
.
1:A (unit) → : T × T ⇒ C (channel connectivity)
6 J. Åman Pohjola

.
Intuitively, M → K means the prefix M can send a message to the prefix K. The
substitution functions must satisfy certain natural criteria wrt. their treatment
of names; see [2] for the details.

Definition 2 (Static equivalence). Two assertions Ψ, Ψ  are statically equiv-

alent, written Ψ Ψ  , if ∀ϕ. Ψ  ϕ ⇔ Ψ   ϕ.

Definition 3 (Valid parameters). A psi-calculus is valid if (A/ , ⊗, 1) form

an abelian monoid.

Note that since the abelian monoid is closed, static equivalence is preserved
by composition. Henceforth we will only consider valid psi-calculi. The original
. .
presentation of psi-calculi had ↔ for channel equivalence in place of our →,
.
and required that channel equivalence be symmetric (formally, Ψ  M ↔ K iff
.
Ψ  K ↔ M ) and transitive.

Definition 4 (Process syntax). The processes (or agents) P, ranged over by

P, Q, R, are inductively defined by the grammar
P := 0 (nil) Ψ  (assertion)
M N.P (output) M (λx)N.P (input)
 : P
case ϕ (case) P |Q (parallel composition)
(νx)P (restriction) !P (replication)
A process is assertion guarded ( guarded for short) if all assertions occur under-
neath an input or output prefix. We require that in !P , P is guarded; that in
 : P, all P are guarded; and that in M (λ
case ϕ x)N . P it holds that x ⊆ n(N ).
We will use PG , QG to range over guarded processes.

Restriction, replication and parallel composition are standard. M N.P is a

process ready to send the message N on channel M , and then continue as P .
Similarly, M (λ x)N.P is a process ready to receive a message on channel M
that matches the pattern (λ x)N . The process Ψ  asserts a fact Ψ about the
environment. Intuitively, Ψ  | P means that P executes in an environment
where all conditions entailed by Ψ hold. P may itself contain assertions that
add or retract conditions. Environments can evolve dynamically: as a process
reduces, assertions may become unguarded and thus added to the environment.
case ϕ : P is a process that may act as any Pi whose guard ϕi is entailed by
the environment. For discussion of why replication and case must be guarded we
refer to [2,15].
The assertion environment of a process is described by its frame:

Definition 5 (Frames). The frame of P , written F(P ) = (νbP )ΨP where bP
bind into ΨP , is defined as

F(Ψ ) = (ν)Ψ F(P | Q) = F(P ) ⊗ F(Q) F((νx)P ) = (νx)F(P )

F(P ) = 1 otherwise
 Psi-Calculi Revisited: Connectivity and Compositionality 7

where name-binding and composition of frames is defined as (νx)(νbP )ΨP =

(νx, bP )ΨP , and, if bP #bQ , ΨQ and bQ #ΨP ,

(νbP )ΨP ⊗ (νbQ )ΨQ = (νbP , bQ )ΨP ⊗ ΨQ .

We extend entailment to frames as follows: F(P )  ϕ holds if, for some bP , ΨP
such that F(P ) = (νbP )ΨP and bP #ϕ, ΨP  ϕ. The freshness side-condition
bP #ϕ is important because it allows assertions to be used for representing local
state. By default, the assertion environment is effectively a form of global non-
monotonic state, which is not always appropriate for modelling distributed pro-
cesses. With ν-binding we recover locality by writing e.g. (νx)(x = M  | P ) for
a process P with a local variable x.
The notion of provenance is the main novelty of our semantics. It is the key
technical device used to make our semantics compositional:

Definition 6 (Provenances). The provenances Π, ranged over by π, are

either ⊥ or of form (ν
x; y)M , where M is a term, and x
, y bind into M .

We write M for (ν; )M . When x , y#x , y and x

#
y , we interpret the expres-
x; y)(ν x ; y )M as (ν
sion (ν   x x ; y y )M . Furthermore, we identify (ν
  x; y)⊥ and
⊥. Let π ↓ denote the result of moving all binders from the outermost binding
sequence to the innermost; that is, (ν x; y)M ↓= (ν; x , y)M . Similarly, π ↓ z
denotes the result of inserting z at the end of the outermost binding sequence:
formally, (νx; y)M ↓ z = (ν x, z; y)M .
Intuitively, a provenance describes the origin of an input or output transition.
For example, if an output transition is annotated with (ν x; y)M , the sender is
an output prefix with subject M that occurs underneath the ν-binders x , y.
For technical reasons, these binders are partitioned into two distinct sequences.
The intention is that x  are the frame binders, while y contains binders that
occur underneath case and replication; these are not part of the frame, but may
nonetheless bind into M . We prefer to keep them separate because the x  binders
are used for deriving  judgements, but y are not (cf. Definition 5).

Definition 7 (Labels). The labels L, ranged over by α, β, are:

M (ν
x)N (output) M N (input) τ (silent)

The bound names of α, written bn(alpha), is x  if α = M (ν

x)N and  other-
wise. The subject of α, written subj(α), is M if α = M (ν x)N or α = M N .
Analogously, the object of α, written obj(α), is N if α = M (ν
x)N or α = M N .

While the provenance describes the origin of a transition, a label describes

how it can interact. For example, a transition labelled with M N indicates readi-
ness to receive a message N from an output prefix with subject M .
8 J. Åman Pohjola

Table 1. Structured operational semantics. A symmetric version of Com is elided. In

the rule Com we assume that F (P ) = (ν bP )ΨP and F (Q) = (ν bQ )ΨQ where  bP is
fresh for Ψ and Q, x is fresh for Ψ, ΨQ , P , and bQ , y are similarly fresh. In rule ParL
we assume that F (Q) = (ν bQ )ΨQ where  bQ is fresh for Ψ, P, π and α. ParR has the
same freshness conditions but with the roles of P, Q swapped. In Open the expression
ã ∪ {b} means the sequence ã with b inserted anywhere.

. .
Ψ K→M Ψ M →K
In 
Out
K N [
y :=L] KN
Ψ  M (λ
y )N . P −−−−−−−→ P [ 
y := L] Ψ  M N . P −−→ P
M M

α
→ P
ΨQ ⊗ Ψ  P −
π
ParL α bn(α)#Q
Ψ  P | Q −−−→ P  | Q
π↓
bQ

α
→ Q
ΨP ⊗ Ψ  Q −
π
ParR α bn(α)#P
Ψ  P | Q −−−−→ P | Q
(ν
bP )π

M (ν
a)N KN
ΨQ ⊗ Ψ  P −−−−−−→ P  ΨP ⊗ Ψ  Q −−−−−−→ Q
(ν
bP ;
x)K (ν
bQ ;
y )M
Com τ 
a#Q
Ψ  P |Q − a)(P  | Q )
→ (ν
⊥

α α
→ P
Ψ  Pi − Ψ  ϕi → P
Ψ  P −
π π
Case α
Scope α b#α, Ψ
 : P −→ P 
Ψ  case ϕ Ψ  (νb)P −−−→ (νb)P 
π↓ (νb)π

M (ν
a)N
Ψ  P −−−−−→ P  b#a, Ψ, M → P
Ψ  P | !P −
α
π
Open Rep
π
M (ν
a∪{b})N b ∈ n(N ) α
Ψ  (νb)P −−−−−−−−→ P  Ψ  !P −→ P 
(νb)π π↓

Definition 8 (Operational semantics). The transition relation −→⊆ A ×

α
P × L × Π × P is inductively defined by the rules in Table 1. We write Ψ  P −
→
π
  
P for (Ψ, P, α, π, P ) ∈−→. In transitions, bn(α) binds into obj(α) and P .

The operational semantics differs from [2] mainly by the inclusion of prove-
nances: anything underneath the transition arrows is novel.
The Out rule states that in an environment where M is connected to K, the
prefix M N may send a message N from M to K. The In rule is dual to Out,
but also features pattern-matching. If the message is an instance of the pattern,
as witnessed by a substitution, that subtitution is applied to the continuation P .
 Psi-Calculi Revisited: Connectivity and Compositionality 9

In the Com rule, we see how provenances are used to determine when two
processes can interact. Specifically, a communication between P and Q can be
derived if P can send a message to M using prefix K, and if Q can receive a
message from K using prefix M . Because names occuring in M and K may
be local to P and Q respectively, we must be careful not to conflate the local
names of one with the other; this is why the provenance records all binding
names that occur above M, K in the process syntax. Note that even though we
identify frames and provenances up-to alpha, the Com rule insists that we con-
sider alpha-variants such that the frame binders and the outermost provenance
binders coincide. This ensures that the K on Q’s label really is the same as the
K in the provenance.
It is instructive to compare our Com rule with the original:

M (ν
a)N
ΨQ ⊗ Ψ  P −−−−−−→ P 
KN .
ΨP ⊗ Ψ  Q −−−→ Q Ψ ⊗ ΨP ⊗ ΨQ  M ↔ K
Com-Old 
a#Q
τ
a)(P  | Q )
Ψ  P | Q −→ (ν

where F(P ) = (νbP )ΨP and F(Q) = (νbQ )ΨQ and bP #Ψ, bQ , Q, M, P and
bQ #Ψ, bQ , Q, K, P . Here we have no way of knowing if M and K are able to syn-
chronise other than making a channel equivalence judgement. Hence any deriva-
tion involving Com-Old makes three channel equivalence judgements: once each
in In, Out and Com-Old. With Com we only make one—or more accurately,
we make the exact same judgement twice, in In resp. Out. Eliminating the
redundant judgements is crucial: the reason Com-Old needs associativity and
commutativity is to stitch these three judgements together, particularly when
one end of a communication is swapped for a bisimilar process that allows the
same interaction via different prefixes.
Note also that Com has fewer freshness side-conditions. A particularly unin-
tuitive aspect of Com-Old is that it requires bP #M and bQ #K, but not bP #K
and bQ #M : we would expect that all bound names can be chosen to be distinct
from all free names, but adding the missing freshness conditions makes scope
extension unsound [14, pp. 56–57]. With Com, it becomes clear why: because
bQ binds into M .
All the other rules can fire independently of what the provenance of the
premise is. They manipulate the provenance, but only for bookkeeping purposes:
α
in order for the Com rule to be sound, we maintain the invariant that if Ψ  P −
→
π
P  , the outer binders of π are precisely the binders of F(P ). Otherwise, the rules
are exactly the same as in the original psi-calculi.
The reader may notice a curious asymmetry between the treatment of prove-
nance binders in the ParL and ParR rules. This is to ensure that the order of
the provenance binders coincides with the order of the frame binders, and in the
frame F(P | Q), the binders of P occur syntactically outside the binders of Q
(cf. Definition 5).
10 J. Åman Pohjola

3 Meta-theory
In this section, we will derive the standard algebraic and congruence laws of
strong bisimulation, develop an alternative formulation of strong bisimulation
in terms of a reduction relation and barbed congruence, and show that our
extension of psi-calculi is conservative. While weak equivalences are beyond the
scope of the present paper, we believe it is possible (if tedious) to adapt the
results about weak bisimilarity from [15] to our setting.

3.1 Bisimulation
α α
We write Ψ  P −→ P  as shorthand for ∃π. Ψ  P − → P  . Bisimulation is
π
then defined exactly as in the original psi-calculi:
Definition 9 (Strong bisimulation). A symmetric relation R ⊆ A × P × P
is a strong bisimulation iff for every (Ψ, P, Q) ∈ R
1. Ψ ⊗ F(P ) Ψ ⊗ F(Q) (static equivalence)
2. ∀Ψ  .(Ψ ⊗ Ψ  , P, Q) ∈ R (extension of arbitrary assertion)
α α
3. If Ψ  P −→ P  and bn(α)#Ψ, Q, then there exists Q such that Ψ  Q −→
Q and (Ψ, P  , Q ) ∈ R (simulation)
. .
We let bisimilarity ∼ be the largest bisimulation. We write P ∼Ψ Q to mean
. . .
(Ψ, P, Q) ∈ ∼, and P ∼ Q for P ∼1 Q.
Clause 3 is the same as for pi-calculus bisimulation. Clause 1 requires that two
bisimilar processes expose statically equivalent assertion environments. Clause 2
states that if two processes are bisimilar in an environment, they must be bisim-
ilar in every extension of that environment. Without this clause, bisimulation is
not preserved by parallel composition.
This definition might raise some red flags for the experienced concurrency
theorist. We allow the matching transition from Q to have any provenance,
irrespectively of what P ’s provenance is. Hence the Com rule uses information
that is ignored for the purposes of bisimulation, which in most cases would result
in a bisimilarity that is not preserved by the parallel operator.
Before showing that bisimilarity is nonetheless compositional, we will argue
that bisimilarity would be too strong if Clause 4 required transitions with match-
ing provenances. Consider two distinct terms M, N that are connected to the
. .
same channels; that is, for all Ψ, K we have Ψ  M → K iff Ψ  N → K. We
would expect M .0 and N .0 to be bisimilar because they offer the same interac-
tion possibilities. With our definition, they are. But if bisimulation cared about
provenance they would be distinguished, because transitions originating from
M .0 will have provenance M while those from N .0 will have N .
The key intuition is that what matters is not which provenance a transition
has, but which channels the provenance is connected to. The latter is preserved
by Clause 3, as this key technical lemma—formally proven in Isabelle, by a
routine induction—hints at:
 Psi-Calculi Revisited: Connectivity and Compositionality 11

Lemma 1. (Find connected provenance)

MN
1. If Ψ  P −−−→ P  and C is finitely supported, then there exists bP , ΨP , x
, K
π
such that F(P ) = (νbP )ΨP and π = (νbP ; x
)K and bP #Ψ, P, M, N, P  , C, x

 .
#Ψ, P, N, P , C and Ψ ⊗ ΨP  M → K.
and x
2. A similar property for output transitions (elided).
In words, the provenance of a transition is always connected to its subject, and
the frame binders can always be chosen sufficiently fresh for any context. This
simplifies the proof that bisimilarity is preserved by parallel: in the original psi-
calculi, one of the more challenging aspects of this proof is finding sufficiently
fresh subjects to use in the Com-Old rule, and then using associativity and
symmetry to connect them (cf. [2, Lemma 5.11]). By Lemma 1 we already have
a sufficiently fresh subject: our communication partner’s provenance.
Theorem 1 (Congruence properties of strong bisimulation).
. .
1. P ∼Ψ Q ⇒ P | R ∼Ψ Q | R
. .
2. P ∼Ψ Q ⇒ (νx)P ∼Ψ (νx)Q if x#Ψ
. .
3. PG ∼Ψ QG ⇒ !PG ∼Ψ !QG
. .
4. ∀i.Pi ∼Ψ Qi ⇒ case ϕ  : P ∼Ψ case ϕ  if P, Q
:Q  are guarded
. .
5. P ∼Ψ Q ⇒ M N.P ∼Ψ M N.Q
Theorem 2 (Algebraic laws of strong bisimulation).
. . . .
P ∼Ψ P | 0 P | (Q | R) ∼Ψ (P | Q) | R P | Q ∼Ψ Q | P (νa)0 ∼Ψ 0
. .
P | (νa)Q ∼Ψ (νa)(P | Q) if a#P M N.(νa)P ∼Ψ (νa)M N.P if a#M, N
. .
x)N.(νa)P ∼Ψ (νa)M (λ
M (λ x)N.P if a#
x, M, N !P ∼Ψ P | !P

 ∼ . .
case ϕ
 : (νa)P  : P if a#ϕ
Ψ (νa)case ϕ  (νa)(νb)P ∼Ψ (νb)(νa)P
The proofs of Theorems 1 and 2 have been mechanised in Nominal Isabelle.
Note that bisimilarity is not preserved by input, for the same reasons as the
pi-calculus. As in the pi-calculus, we can define bisimulation congruence as the
substitution closure of bisimilarity, and thus obtain a true congruence which
satisfies all the algebraic laws above. We have verified this in Nominal Isabelle,
following [2].
The fact that bisimilarity is compositional yet ignores provenances suggests
that the semantics could be reformulated without provenance annotations on
labels. To achieve this, what is needed is a side-condition S for the Com rule
which, given an input and an output with subjects M, K, determines if the input
transition could have been derived from prefix K, and vice versa:
M (ν
a)N KN
ΨQ ⊗ Ψ  P −−−−−−→ P  ΨP ⊗ Ψ  Q −−−→ Q S
τ
a)(P  | Q )
Ψ  P | Q −→ (ν
12 J. Åman Pohjola

But we already have such an S: the semantics with provenances! So we can let

M (ν
a)N KN
S = ΨQ ⊗ Ψ  P −−−−−−→ P  ∧ ΨP ⊗ Ψ  Q −−−−−−→ Q
(ν
bP ;
x)K (ν
bQ ;
y )M

Of course, this definition is not satisfactory: the provenances are still there,
just swept under the carpet. Worse, we significantly complicate the definitions
by effectively introducing a stratified semantics. Thus the interesting question is
not whether such an S exists (it does), but whether S can be formulated in a way
that is significantly simpler than the semantics with provenances. The author
believes the answer is negative: S is a property about the roots of the proof trees
used to derive the transitions from P and Q. The provenance records just enough
information about the proof trees to show that M and K are connected; with
no provenances, it is not clear how this information could be obtained without
essentially reconstructing the proof tree.

3.2 Validation

We have defined semantics and bisimulation, and showed that bisimilarity satis-
fies the expected laws. But how do we know that they are the right semantics, and
the right bisimilarity? This section provides two answers to this question. First,
we show that our developments constitute a conservative extension of the origi-
nal psi-calculi. Second, we define a reduction semantics and barbed bisimulation
that are in agreement with our (labelled) semantics and (labelled) bisimilarity.
.
Let −→o and ∼o denote semantics and bisimilarity as defined by Bengtson
et al. [2], i.e., without provenances and with the Com-Old rule discussed in
Sect. 2. The following result has been mechanised in Nominal Isabelle:
.
Theorem 3 (Conservativity). When → is symmetric and transitive we have
. .
∼o = ∼ and −→o = −→.

Our reduction semantics departs from standard designs [4,17] by relying on

reduction contexts [7] instead of structural rules, for two reasons. First, standard
formulations tend to include rules like these:

P −→ P 
P | Q −→ P  | Q α.P + Q | α.R + S −→ P | R

A parallel rule like the above would be unsound because Q might contain asser-
tions that retract some conditions needed to derive P ’s reduction. The reduction
axiom assumes prefix-guarded choice. We want our semantics to apply to the full
calculus, without limiting the syntax to prefix-guarded case statements.
But first, a few auxiliary definitions. The reduction contexts are the contexts
in which communicating processes may occur:
 Psi-Calculi Revisited: Connectivity and Compositionality 13

Table 2. Reduction semantics. Here Ψ abbreviates the composition Ψ1 ⊗ Ψ2 ⊗ . . . , and

 abbreviates the parallel composition Ψ1  | Ψ2  | . . . —for empty sequences they
Ψ
are taken to be 1 and 0 respectively.

P ≡Q Q −→ Q Q ≡ P  P −→ Q
Struct Scope
P −→ P  (νa)P −→ (νa)Q

.
Ψ  M → N x := T]
K = L[ ∀ϕ ∈ conds(C). Ψ  ϕ
Ctxt
Ψ | C[M K.P, N (λx)L.Q] −→ Ψ | P | Q[x := T ] | ppr(C)

Definition 10 (Reduction contexts). The reduction contexts, ranged over

by C, are generated by the grammar

C := PG (process) [] (hole)
C | C (parallel)    
 : PG [] ϕ : C [] ϕ : QG (case)
case ϕ

Let H(C) denote the number of holes in C. C[P G ] denotes the process that results
from filling each hole of C with the corresponding element of P G , where holes
are numbered from left to right; if H(C) = |P
G |, C[PG ] is undefined.

We let structural congruence ≡ be the smallest equivalence relation on pro-

cesses derivable using Theorems 1 and 2. The conditions conds(C) and parallel
processes ppr(C) of a context C are, respectively, the conditions in C that guard
the holes, and the processes of C that are parallel to the holes:

ppr(PG ) = PG ppr([ ]) = 0 ppr(C1 | C2 ) = ppr(C1 ) | ppr(C2 )

 : P
ppr(case ϕ   
G [] ϕ : C [] ϕ : QG ) = ppr(C) conds(PG ) = ∅

conds([ ]) = ∅ conds(C1 | C2 ) = conds(C1 ) ∪ conds(C2 )

 : P
conds(case ϕ    
G [] ϕ : C [] ϕ : QG ) = {ϕ } ∪ conds(C)

Definition 11 (Reduction semantics). The reduction relation −→ ⊆ P × P

is defined inductively by the rules of Table 2.

In words, Ctxt states that if an input and output prefix occur in a reduc-
tion context, we may derive a reduction if the following holds: the prefixes are
connected in the current assertion environment, the message matches the input
pattern, and all conditions guarding the prefixes are entailed by the environ-
ment. The ppr(C) in the reduct makes sure any processes in parallel to the holes
are preserved.
14 J. Åman Pohjola

τ
Theorem 4. P −→ P  iff there is P  such that 1  P −→ P  and P  ≡ P 

For barbed bisimulation, we need to define what the observables are, and
what contexts an observer may use. We follow previous work by Johansson et
al. [15] on weak barbed bisimilarity for the original psi-calculi on both counts.
First, we take the barbs to be the output labels a process can exhibit: we define
M (ν
a)N
P ↓M (νa)N (P exposes M (ν a)N ) to mean ∃P  . 1  P −−−−−→ P  . We write

τ
P ↓M for ∃ a, N.P ↓M (ν A)N
 , and P ⇓α for P −→ ↓α . Second, we let observers
use static contexts, i.e. ones built from parallel and restriction.
.
Definition 12 (Barbed bisimilarity). Barbed bisimilarity, written ∼ , is
barb
.
the largest equivalence on processes such that P ∼ Q implies
barb

1. If P ↓M (νa)N and 
a#Q then Q ↓M (νa)N (barb similarity)
.
2. If P −→ P  then there exists Q such that Q −→ Q and P  ∼ Q
barb
(reduction simulation)
.
a)(P | R) ∼ (ν
3. (ν a)(Q | R) (closure under static contexts)
barb

Our proof that barbed and labelled bisimilarity coincides only considers psi-
calculi with a certain minimum of sanity and expressiveness. This rules out some
degenerate cases: psi-calculi where there are messages that can be sent but not
received, and psi-calculi where no transitions whatsoever are possible.

Definition 13. A psi-calculus is observational if:

.
1. For all P there are MP , KP such that F(P )  MP → KP and not P ⇓Kp .
2. If N = (x y) · M and y#M and x
, y are distinct then M [
x := y] = N .

The first clause means that no process can exhaust the set of barbs. Hence
observing contexts can signal success or failure without interference from the
process under observation. For example, in the pi-calculus MP , KP can be any
name x such that x#P . The second clause states that for swapping of distinct
names, substitution and permutation have the same behaviour. Any standard
definition of simultaneous substitution should satisfy this requirement. These
assumptions are present, explicitly or implicitly, in the work of Johansson et
al. [15]. Ours are given a slightly weaker formulation.
We can now state the main result of this section:
. .
Theorem 5. In all observational psi-calculi, P ∼ Q iff P ∼1 Q.
barb

4 Expressiveness

In this section, we study two examples of the expressiveness gained by dropping

symmetry and transitivity.
 Psi-Calculi Revisited: Connectivity and Compositionality 15

4.1 Pi-Calculus with Preorders

Recall that pi-F [25] extends the pi-calculus with name equalities (x = y) as
first-class processes. Communication in pi-F gives rise to equalities rather than
substitutions, so e.g. xy.P | xz.Q reduces to y = z | P | Q: the input and output
objects are fused. Hirschkoff et al. [11] observed that fusion and subtyping are
fundamentally incompatible, and propose a generalisation of pi-F called the pi-
calculus with preorders or πP to resolve the issue.
We are interested in πP because its channel connectivity is not transitive.
The equalities of pi-F are replaced with arcs a/b (“a is above b”) which act
as one-way fusions: anything that can be done with b can be done with a, but
not the other way around. The effect of a communication is to create an arc
with the output subject above the input subject, so x(y).P | x(z).Q reduces to
(νxy)(z/y | P | Q). We write ≺ for the reflexive and transitive closure of the
“is above” relation. Two names x, y are considered joinable for the purposes of
synchronisation if some name z is above both of them: formally, we write x  y
for ∃z.x ≺ z ∧ y ≺ z.
Hirschkoff et al. conclude by saying that “[it] could also be interesting to study
the representation of πP into Psi-calculi. This may not be immediate because
the latter make use of on an equivalence relation on channels, while the former
uses a preorder” [11, p. 387]. Having lifted the constraint that channels form
an equivalence relation, we happily accept the challenge. We write ΨP for the
psi-calculus we use to embed πP . We follow the presentation of πP from [12,13],
where the behavioural theory is most developed.

Definition 14. The psi-calculus ΨP is defined with the following parameters:

TN C  {x ≺ y : x, y ∈ N } ∪ {x  y : x, y ∈ N }

A  Pfin ({x ≺ y : x, y ∈ N }) 1  {} ⊗ ∪
.
→    the relation denoted  in [13].

The prefix operators of πP are different from those of psi-calculi: objects are
always bound, communication gives rise to an arc rather than a substitution, and
a conditional silent prefix [ϕ]τ.P is included.2 These are encodable as follows:

Definition 15 (Encoding of prefixes). The encoding   from πP to ΨP is

homomorphic on all operators except prefixes and arcs, where it is defined by

a/b = b ≺ a a(y).P  = (νxy)(ax.(x ≺ y | P ) where x#y, P

a(y).P  = (νy)(a(λx)x.(y ≺ x | P )) where x#y, P

[ϕ]τ.P  = case ϕ : (νx)(x(λx)x.0 | xx.P ) where x#P

2
We ignore protected prefixes because they are redundant, cf. Remark 1 of [12].
16 J. Åman Pohjola

This embedding of πP in psi-calculi comes with a notion of bisimilarity per

Definition 9. We show that it coincides with the labelled bisimilarity for πP
(written ∼) introduced in [12,13].
.
Theorem 6. P ∼ Q iff P  ∼ Q
Thus our encoding validates the behavioural theory of πP by connecting it
to our fully mechanised proofs, while also showing that a substantially different
design of the LTS yields the same bisimilarity. We will briefly compare these
designs. While we do rewriting of subjects in the prefix rules, Hirschkoff et al.
instead use relabelling rules like this one (mildly edited to match our notation):

a(x)
P −−−→ P  F(P )  a ≺ b
b(x)
P −−→ P 

An advantage of this rule is that it allows input and output labels to be as

simple as pi-calculus labels. A comparative disadvantage is that it is not syntax-
directed, and that the LTS has more rules in total. Note that this rule would
not be a viable alternative to provenances in psi-calculi: since it can be applied
more than once in a derivation, its inclusion assumes that the channels form a
preorder wrt. connectivity.
πP also has labels [ϕ]τ , meaning that a silent transition is allowed in envi-
ronments where ϕ is true. A rule for rewriting ϕ to a weaker condition, similar
to the above rule for subject rewriting, is included. Psi-calculi does not need this
because the Par rules take the assertion environment into account. πP transi-
[ϕ]τ τ
tions of kind P −−→ P  correspond to ΨP transitions of kind {ϕ}  P −→ P  .
Interestingly, the analogous full abstraction result fails to hold for the embed-
ding of pi-F in psi-calculi by Bengtson et al. [2], because outputs that emit dis-
tinct but fused names are distinguished by psi-calculus bisimilarity. This issue
does not arise here because πP objects are always bound; however, we believe the
encoding of Bengtson et al. can be made fully abstract by encoding free output
with bound output, exploiting the pi-F law a y.Q ∼ a(x)(Q | x = y).

4.2 Mixed Choice

This section will argue that because we allow non-transitive channel connectivity,
the case operator of psi-calculi becomes superfluous. The formal results here will
focus on encoding the special case of mixed choice. We will then briefly discuss
how to generalise these results to the full case operator.
Choice, written P + Q, is a process that behaves as either P or Q. In psi-
calculi we consider P + Q to abbreviate case  : P []  : Q for some condition
 that is always entailed. Mixed choice means that in P + Q, P and Q must
be prefix-guarded; that is, the outermost operators of P, Q must be input or
output prefixes. In particular, mixed choice allows choice between an input and
an output. There is a straightforward generalisation to n-ary sums that, in order
to simplify the presentation, we will not consider here.
 Psi-Calculi Revisited: Connectivity and Compositionality 17

.
Fix a psi-calculus P = (T, A, C, , ⊗, 1, →) with mixed choice; this will be
our source language. We will construct a target psi-calculus and an encoding such
that the target terms make no use of the case operator. The target language
E(P) adds to T the ability to tag a term M with a name x; we write Mx for the
tagged term. We write αx for tagging the subject of the prefix α with x. Tags
are used to uniquely identify which choice statement a prefix is a summand of.
As the assertions of E(P) we use A × Pfin (N ), where Pfin (N ) are the disabled
tags.
The encoding   from P to E(P) is homomorphic on all operators except
assertion and choice, where it is defined as follows:

Ψ  = (Ψ, ∅) α.P +β.Q = (νx)(αx .(P  | (1, {x}) | βx .(Q | (1, {x})

where x#α, β, P, Q. If we disregard the tag x, we see that the encoding sim-
ply offers up both summands in parallel. This clearly allows all behaviours of
α.P + β.Q, but there are two additional behaviours we must prevent: (1) com-
munication between the summands, and (2) lingering summands firing after the
other branch has already been taken. The tagging mechanism prevents both, as
a consequence of how we define channel equivalence on tagged terms in E(P):
. .
(Ψ, N)  Mx → Ny if Ψ  M → N and x = y and x, y ∈
/N

That is, tagged channels are connected if the underlying channel is connected. To
prevent (1) we require the tags to be different, and to prevent (2) we require that
the tags are not disabled. Note that this channel connectivity is not transitive,
not reflexive, and not monotonic wrt. assertion composition—not even if the
source language connectivity is.

Theorem 7 (Correctness of choice encoding).

α α
1. If Ψ  P −→ P  then there is P  such that (Ψ, ∅)  P  −→ P  and
.
P  ∼(Ψ,∅) P  .
α α
2. If (Ψ, ∅)  P  −→ P  then there is P  such that Ψ  P −−→
⊥
P  and
 . 
P ∼(Ψ,∅) P .
. .
3. P ∼1 Q iff P  ∼(1,∅) Q.

Here α⊥ denote the label α with all tags removed. It is immediate from Theo-
rem 7 and the definition of   that our encoding also satisfies the other standard
quality criteria [10]: it is compositional, it is name invariant, and it preserves and
reflects barbs and divergence.
In the original psi-calculi, our target language is invalid because of non-
transitive connectivity. If we remove the requirement that tags are distinct, and
only allow separate choice (where either both summands are inputs or both
summands are outputs), the encoding is correct for the original psi-calculi.
These results generalise in a straightforward way to mixed Case state-
ments case ϕ1 : α.P [] ϕ2 : β.Q by additionally tagging terms with a condition,
i.e. Mx,ϕ1 , that must be entailed in order to derive connectivity judgements
Exploring the Variety of Random
Documents with Different Content
yet at the same time convey lastingly any high order of serious
emotion or effect. The great men doing serious work in sculpture will
never find it necessary to go beyond the law of nature for the
architectonic basis of their expression. Faulty or arbitrary proportion
in handling the human figure is unnecessary; it is of no real help to
the artist, and no more desired by him than is the liberty of 16 lines
and ballad measure, by the sonneteer expert in the Petrarchan form
and rhyme of 14 pentameter verses. The real matter to be dealt with
in respect of peculiarities of site and circumstances lies within the
sphere of the artistic capacity, and is at once more easy and more
difficult than any wooden process of mis-handling the proportions of
the figure. It is at issue in the legend of the Byzantine writer, Tzetzes,
to which reference is made, wherein it is said that Pheidias and
Alcamenes competed on one occasion with rival figures of Athene,
but the explanation given of the reason why the work of Pheidias was
admired and preferred at the site, is, I venture to say, the wrong one,
in as far as it presupposes abnormal proportions in the successful
statue. To the author’s mind, no doubt, something profound and
abstruse was necessary in order to explain such a triumph, and the
idea that Pheidias was deeply versed in what must then have been
the occult mysteries of optics and geometry, fitted the need and was
pleasant to the love of the marvellous.
‘In such a case, Pheidias would certainly, with the intuitive artistic
sense and experience of a master, handle the style, composition,
lights and shadows, mass, line and silhouette of his work in relation
to its size, and the average height and distance from which it was to
be viewed. It might be finished highly in respect of surface, or left
moderately rough, a condition of little consequence compared with
the factors enumerated above. It would be made readable and
expressive, but there would be no modification of the sacred
proportions of the figure; no trace of allowance in order that “the
upper part which is further off from the eye should appear to be in
proportion when compared with the lower, which is nearer.” That
artists should appear to give up natural truth in their images for
considerations of abstract beauty, was grateful to the mind of Plato,
but is only another proof of the soaring qualities of the White Horse
in the Human Chariot!
 ‘Outside of a somewhat conscious effort towards the decorative in
form and towards the effective articulation of parts, I find little in the
work of Donatello to justify his being specially singled out as
illustrating those principles of the modification of true proportions
for sculpture in relation to the exigencies of site. The statues on the
Campanile need not, I imagine, be taken too seriously as exhibitions
of Donatello’s most careful judgement. Compared with such works of
his as we may feel at liberty to believe personal, they are rude and ill-
considered in design and execution. There is in the bones, mass, and
arrangement of the work very probably something of Donatello, but
in the detail and execution there is little or nothing of the hand that
did the Christ of S. Antonio of Padua, the bronze David of the
Bargello, or the bust of Niccolo da Uzzano.’

WAXEN EFFIGIES AND MEDALLIONS.

[§ 43, Polychrome Wax Effigies, ante, p. 149.]

Wax has been used from the time of the ancients as a modelling
material, both in connection with casting in bronze, and with the
making of small studies for reproduction in more permanent
materials. The production of a plastic work in wax intended to
remain as the finished expression of the artist’s idea is of course a
different matter. Among the Greeks, Lysistratos, the brother of
Lysippos, about the time of Alexander the Great, introduced the
practice of taking plaster moulds from the life, and then making casts
from them in wax. These he may have coloured, for the use of colour,
at any rate on terra cotta, was at the time universal, and in this way
have produced waxen effigies. (Hominis autem imaginem gypso e
facie ipsa primus omnium expressit ceraque in eam formam gypsi
infusa emendare instituit Lysistratus Sicyonius frater Lysippi. Plin.
Hist. Nat., XXXV, 153). Busts in coloured wax of departed ancestors
were kept by the Romans of position in the atria of their houses, and
the funereal use of the wax effigy can be followed from classical times
to those comparatively modern, for in Westminster Abbey can still be
seen the waxen effigies of Queen Elizabeth, Charles II, and other
sovereigns and nobles of the sixteenth and seventeenth centuries.
These, like the modern wax-works of popular exhibitions, are hardly
productions of art. What Vasari writes of is a highly refined and
artistic kind of work, that was practised in Italy from the early part of
the sixteenth century, and spread to France, Germany, and England
in each of which countries there were well-known executants in the
seventeenth or eighteenth centuries. The Connoisseur of March,
1904, contained an article on the chief of these.
Though modelled effigies in wax of a thoroughly artistic kind were
executed of or near the size of life and in the round, as may be seen in
the Italian waxen bust of a girl in the Musée Wicar at Lille, that has
been ascribed to Raphael, yet as a rule the execution was in
miniature and in relief. Specimens of this form of the work are to be
seen in the British Museum, in the Wallace Collection, and at South
Kensington.
In the Proceedings of the Huguenot Society of London, III, 4, there
is an article on the Gossets, a Huguenot family, some members of
which practised the art in England from the early part of the
eighteenth century, and a recipe for colouring the wax is there
quoted which it may be interesting to compare with that given by
Vasari. ‘To two ounces of flake white (the biacca of Vasari) add three
of Venice turpentine, if it be in summer, and four in winter, with
sufficient vermilion (cinabrio) to give it a pinkish tint. Grind these
together on a stone with a muller; then put them into a pound of fine
white wax, such as is used for making candles: this is molten ready in
an earthen pipkin. Turn them round over the fire for some time.
When thoroughly mixed the composition should be immediately
removed and poured into dishes previously wetted to prevent the
wax from sticking to them.’
This refers to the preparation of a self-coloured wax which may be
prepared of a flesh tint, or of a creamy white, or of any other desired
hue like those Vasari enumerates. The portraits in wax referred to in
our museums are sometimes in self-coloured material of this kind,
but at other times are coloured polychromatically in all their details.
This is the technique referred to by Vasari in § 43 as having been
introduced by certain ‘modern masters.’ In Opere, IV, 436 he refers to
one Pastorino of Siena as having acquired great celebrity for wax
portraits, and as having ‘invented a composition which is capable of
reproducing the hair, beard and skin, in the most natural manner. It
would take me too long’ he continues ‘to enumerate all the artists
who model wax portraits, for now-a-days there is scarcely a jeweller
who does not occupy himself with such work.’ This last remark is
significant, for one feature of these polychrome medallions is the
introduction of real stones, seed pearls, gold rings, and the like, in
connection with the modelled wax, so that collectors used to style the
works ‘Italian sixteenth century jewelled waxes.’ A portrait bust in
the Salting collection, shown on loan at South Kensington,
representing Elizabeth of France, wife of Philip II, is a good
specimen of the technique. The lady wears a jewelled hair net set
with real red and green stones, and a necklet of seed pearls. In her
ear is a ring of thin gold wire. The flesh parts are naturally coloured,
the hair is auburn, the bodice black, and there are two white feathers
in the headdress. We should gather from Vasari’s words in § 43 that
works of the kind were built up of waxes variously coloured in the
mass, and a close examination of extant specimens clearly shows that
this was the case. Local tints such as the red of the lips, etc., were
added with pigment.
The best modern notice of wax modelling in these forms is that
contained in Propert’s History of Miniature Art, Lond. 1887, chapter
xii, but little is said there of the technique. It should be noticed that
the medallion in coloured wax as a form of art has been revived with
considerable success in our own time and country by the Misses
Casella and others. The artists just named consider that it would be
impossible to finish work on the usual small scale in coloured waxes
alone, without touches of pigment added with the brush. It would be
interesting in this connection to know what were the exact processes
of painting in wax used by the ancients. Paintings, which must have
been on a small scale because they were on a ground of ivory, were
executed in coloured waxes laid on by the ‘cestrum’ (Pliny, Hist.
Nat., XXXV, 147), which is usually described as a sort of spatula,
something like one of the steel tools used by artists for finishing
figures in plaster. However the substance was applied, the whole
process was apparently carried out in the coloured waxes. There
must have been some similarity between this technique and that of
the wax medallions of Renaissance and modern times.

PROPORTIONATE ENLARGEMENT.
[§ 48, Transference of the full-sized Model to the Marble Block, ante,
p. 151.]
‘To enlarge the figure proportionately in the marble.’ Vasari has
said, ante, p. 150, that the model is to be the full size of the marble so
that there would be no question of enlargement but only of
accurately copying the form of the model in the new material. For
this mechanical aids are invoked, the latest and most elaborate of
which is the ‘pointing machine’ now in common use. The appliances
in Vasari’s time were much simpler. Cellini, in his Trattato sopra la
Scultura, describes the mechanical arrangements he made for
enlarging a model to the size of a proposed colossal effigy, and the
principle is the same whether there is to be enlargement or exact
reproduction.
The model, and a block roughly trimmed by rule of thumb to the
size and shape required, but of course somewhat larger than will
ultimately be needed, are placed side by side on tables of exactly the
same form and dimensions. About the model is set up a sort of
framework simple or elaborate, according to the character of the
piece, and a framework precisely similar in all respects is disposed
about the block. A measurement is then taken from one or more
points on the framework to a point on the model, and from a point or
points similarly situated on the other framework, and in the same
relative direction, a similar measurement is led towards the block. As
this is ex hypothesi a little larger than the model, the full
measurement cannot be taken until some of the superfluous marble
has been removed by suitable tools. When this is done a point can be
established on the block exactly corresponding to the point already
fixed on the model. This process can be repeated as often as is
necessary until all the important or salient points on the model have
been successively established on the marble block, which will
ultimately have approached so nearly to the exact similitude of the
model, that the artist can finish it by the eye.
The nature of what has been termed the framework, from which all
the measurements are taken, may vary. Cellini, on the occasion
referred to, surrounded his model with a sort of skeleton of a cubical
box, from the sides and corners of which he measured. In the
Encyclopédie of Diderot and d’Alembert, of the middle of the
eighteenth century, similar square frames, like those used as
stretchers for canvases, are suspended horizontally over model and
block, and plumb lines are hung from the corners, so that skeleton
cubes are established, which would answer the same purpose as
Cellini’s box. See Plate X, A. The arrangement contemplated by
Vasari was somewhat simpler. He does not establish a complete
hollow cube about his model and his block, but is apparently
satisfied with erecting perpendiculars beside each, from which the
measures would be led. The carpenter’s square (squadra) he has in
mind consists of two straight legs joined together at right angles. If
one leg be laid horizontally along the table the one at right angles to
it will be vertical, and from this the measurements are taken. In the
treatise on Sculpture by Leon Battista Alberti there is an elaborate
description of a device he invented for the purpose in view, and one
of his editors has illustrated this by a drawing reproduced here in
Plate X, B. The device explains itself, and any number of similar
contrivances could be employed.

THE USE OF FULL-SIZED MODELS.

[§ 49, Danger of Dispensing with the Full-sized Model, ante, p. 151.]

The question here is of the possibility of dispensing altogether with
a full-sized clay model, and proceeding at once to attack the marble
with the guidance only of the small original sketch. In modern times
this is practically never done, but it was the universal practice of the
Greek sculptors at any rate down to the later periods of Hellenic art.
These remarks of Vasari come just at the time of the change from the
ancient to the modern technique, for we shall see that Donatello in
the fifteenth century worked according to the simpler ancient
method, while Michelangelo in the sixteenth after beginning in the
same fashion finally settled down to the use of the full model, which
has ever since remained de rigeur.
 Plate X

DIAGRAM to illustrate Alberti’s method of

measurement

INTERIOR OF A SCULPTOR’S STUDIO IN THE

EIGHTEENTH CENTURY
 With illustrations of methods of measurement

Fig. 11.—Two views of unfinished Greek marble

statue blocked out on the ancient system. In
quarries on Mount Pentelicus, Athens.

The technique of the Greeks furnished the subject for an article by

Professor Ernest Gardner in the 14th volume of the Journal of
Hellenic Studies. He shows there by a comparison of unfinished
works that the Greek sculptors attacked the marble directly, and
proceeded apparently on the following method. Having obtained a
block about the size and shape required they set it up before them as
if in a front view, and then hewed away at the two sides till they had
brought the contour of these to the exact lines required for the
finished work. They then passed round through a right angle to the
side, and treated in a similar fashion the front and back of the block,
bringing these to the shape of the front and back of the desired
figure. The block would then, when looked at from the front or back
or from the sides, present the required outlines, but the section of it
would still be square in every part—there would be no rounding off.
The sketches, Fig. 11, show two views of a figure blocked out in this
fashion by an ancient Attic sculptor. It was found in old marble
workings on Mount Pentelicus, and is preserved at the modern
marble quarry at the back of that mountain. We owe the use of the
photographs employed to the courtesy of M. Georges Nicole, of
Geneva. They were published in the volume entitled Mélanges
Nicole, Geneva, 1905, in connection with an article on the figure by
the archaeologist just named. The next process was to cut away these
corners and with the guidance of the already established contours
gradually bring the whole into the required shape. A small model
may in every case be presupposed and there must have been some
system of measurement. Indeed on some antiques, as on a crouching
Venus in the gallery leading to the Venus of Milo in the Louvre, there
are still to be seen the knobs (puntelli) to which measurements were
taken during the progress of the work. Of the use of full-sized clay
models there is in Greece no evidence at all, until the late period of
the first century B.C., when we are told of Pasiteles, a very painstaking
sculptor of a decadent epoch, that he never executed a work without
first modelling it (nihil unquam fecit antequam finxit). This no doubt
implies a full-sized model in clay, for a small sketch would not be
mentioned as it is a matter of course.
The practice of the Italians is described by Cellini in words which
are important enough to quote. They are from the fourth chapter of
his treatise on Sculpture. ‘Now although many excellent masters of
assured technique have boldly attacked the marble with their tools,
as soon as they had carved the little model to completion, yet at the
end they have found themselves but little satisfied with their work.
For, to speak only of the best of the moderns, Donatello adopted this
method in his works; and another example is Michelangelo, who had
experience of both the methods, that is to say, of carving statues alike
from the small model and the big, and at the end, convinced of their
respective advantages and disadvantages, adopted the second
method (of the full-sized model). And this I saw myself at Florence
when he was working in the sacristy of S. Lorenzo (on the Medici
tombs).’ As regards Michelangelo’s early practice, Vasari records in
his Life that he carved the colossal marble ‘David’ with the sole aid of
a small wax model, according to Vasari one of those now preserved
in the Casa Buonarroti at Florence. This was in 1504. The Medici
tombs date twenty years later.
In connection with the direct practice of Donatello, it is worth
while referring to some words of Francesco Bocchi, a rhetorical
eulogist of the arts and artists of his native Florence, who wrote in
1571 a literary effusion on the sculptor’s St. George. He notes in his
introduction that Donatello was accustomed to compose his marble
figures compactly and to avoid projecting hands and arms, while for
his effigies in bronze he used much greater freedom in action. The
difference is really one of material, and Donatello’s practice of
working directly on the marble would necessarily involve this
restraint in composition. Anyone accustomed to deal with marble
blocks as vehicles of artistic expression, would avoid unnecessary
projections as these cause great waste of material and expenditure of
time. When plastering clay or wax on a flexible armature this
consideration is not present, and modelled figures will naturally be
freer in action than carved ones. As will presently be seen, certain
marked differences in the treatment of relief sculpture depend on
these same considerations of material and technique.
In direct work on the marble there is of course always the danger
of the sort of miscalculation that Vasari goes on to notice. Greek
figures sometimes show variations from correct proportions, for
example, the left thigh of the Venus of Milo is too short, but the
errors are not such as to destroy the effect of the works. Greek work
in marble shows a marvellously intimate knowledge on the part of
the carver of his material as well as a clear conception of what he was
aiming at. Even Michelangelo yields in this respect to the ancients,
for though no one was ever more thoroughly a master of the carver’s
technique, he made serious mistakes in calculating proportions, as in
the ‘Slave’ of the Louvre, where he has not left enough marble for the
leg of the figure. Moderns generally have not the ease which tradition
and practice gave to the Greek sculptors, and the full-sized model is
now a necessary precaution.

ITALIAN AND GREEK RELIEFS.

[§ 52, Pictorial or Perspective Reliefs, ante, p. 154.]

 Vasari ascribes comprehensively to the ‘ancients’ the invention of
the pictorial or perspectively treated relief, which was not in use in
mediaeval times, but came into vogue in the early years of the
fifteenth century. The first conspicuous instance of its employment
was in the models by Ghiberti for the second set of gates for the
Baptistry at Florence begun in 1425, but as these gates were not
finally completed till 1452, other artists had in the meantime
produced works in the same style. Donatello’s bronze relief of the
beheading of John the Baptist, on the font at Sienna, was completed
in 1427 and shows the same treatment in a modified form. It is a
treatment often called pictorial as it aims at effects of distance, with
receding planes and objects made smaller according to their
supposed distance from the foreground. The style has been
sufficiently criticized, and it is generally agreed that it represents a
defiance of the barriers fixed by the nature of things between
painting and sculpture. It depended mainly however not on the
influence of painting but on the study of perspective, which
Brunelleschi brought into vogue somewhere about the year 1420.
Brunelleschi’s perspectives, or those which he inspired, were worked
out in inlaid woods, or tarsia work, see postea, p. 303 f., and
exhibited elaborate architectural compositions crowded with
receding lines. These compositions were adopted by Donatello and
others for the backgrounds of their figure reliefs, and Ghiberti filled
his nearer planes with a crowd of figures represented as Vasari
describes according to the laws of linear, and so far as the material
permitted, of aerial perspective.
The question of the amount of warrant for this in antique practice
as a whole calls attention to an interesting moment in the general
history of relief sculpture. This has been dealt with recently by Franz
Wickhoff, in the Essay contributed by him to the publication of the
Vienna ‘Genesis,’ and issued in an English translation by Mrs Strong
under the title Roman Art (London, 1900), and also in Mrs Strong’s
own Roman Sculpture (London, Duckworth, 1907). The tendency to
multiply planes in relief, and to introduce the perspective effects and
the backgrounds of a picture, shows itself in some of the late Greek
or ‘Hellenistic’ reliefs, published by Schreiber under the title Die
Hellenistischen Reliefbilder (Wien, 1889 etc.), and more especially in
the smaller frieze from the altar base at Pergamon. Etruscan relief
sculpture is also affected by it. It is however in Roman work of the
early imperial period that we actually find the antique prototypes for
the kind of work that Vasari has in his mind. The reliefs on the tomb
of the Julii at St Rémy, of the age of Augustus, those on the Arch of
Titus, and most conspicuously the decorative sculpture connected
with the name of Trajan, are instances in point. They show
differences in plane, and occasionally a distinct effort after
perspective effects, and it is possible that the study of some of these
Roman examples by Brunelleschi and Donatello at the opening of the
fifteenth century may have contributed to the formation of the
picturesque tradition in Florentine relief sculpture of the period. This
style was however by no means universal in Roman work. The carved
sarcophagi are not much influenced by it, and these sarcophagi are of
special importance in the later history of sculpture, in that they were
the models used by Nicola Pisano and the other French and Italian
sculptors of the so-called ‘proto-Renaissance’ of the twelfth and
thirteenth centuries. In relation to antique sculpture as a whole the
pictorial style is quite abnormal. The genius of the classical Greek
relief is indeed totally opposed to that of the Italian reliefs
represented centrally by those of Ghiberti. The difference is
fundamentally one of material and technique. It is the same
distinction that was drawn by Michelangelo in his letter to Benedetto
Varchi, and noticed already in the Note on ‘The Nature of Sculpture,’
ante, p. 179, the distinction, that is, between sculpture that proceeds
by additions, ‘per via di porre,’ and that which advances by taking
material away, ‘per forza di levare.’ The normal Italian relief was in
cast bronze and was necessarily modelled work. The classical relief
was in marble and was essentially carved work, for the diversifying of
a flat surface. When the Greek relief was in baked clay it was
generally stamped from moulds and not modelled up by hand. The
cast bronze relief in classical Greece may be said only to have existed
in the form of small plaques for the decoration of vases or other
objects in metal. The normal bronze relief in the ancient world was
beaten up in sheet metal by the repoussé process.
In the case alike of the relief carved on the marble slab, or stamped
in clay from moulds, or beaten up in the sheet of metal, the nature of
the technique renders flatness of effect almost obligatory. The Greek
decorative relief cut ‘per forza di levare’ in a smooth marble slab, that
is most often one of the constructive stones of an edifice, naturally
sacrifices as little of the material as is possible, and in all Greek
reliefs, whether low or high, as much as possible of the work is kept
to the foremost plane, the original surface of the stone. Again, a
mould that is undercut, or at all deeply recessed, cannot be used for
stamping clay, while the difficulty of relief work in sheet metal is
greatly increased in proportion to the amount of salience of the
forms. Hence the general flatness of the antique relief, observable
even on the late Roman sarcophagi which served as models at the
first revival of Italian sculpture. Whether the field of the relief is
open or crowded, the objects all come together to the front.
How different are the conditions when the relief is modelled up by
hand in clay or wax! Here the starting point is the background, not as
in the carved relief the foreground, and the forms, worked ‘per via di
porre,’ can be made to stand out against this with an ease and
effectiveness which tempt the modeller to try all sorts of varieties of
relief in the same composition or the same figure, and to multiply
planes of distance till the objects on the foremost plane are starting
out clear of the ground. There is direct evidence (see ante, p. 194)
that in the first century B.C. the use of clay modelling as a preliminary
process in sculpture was greatly extended, and Roman pictorial
reliefs may themselves have been influenced from this side. There is
no question at any rate that the Italians of the Renaissance
surrendered themselves without hesitation to the fascination of this
kind of work, and the style of it dominates their later reliefs. The
contrast in this respect between Ghiberti’s second, or Old Testament,
Gates, and his earlier ones which adhered to the simpler style of
Andrea Pisano’s reliefs on the first of the three Baptistry Gates, is
most instructive. Andrea’s reliefs are in character mediaeval, and the
nearest parallel to them are the storied quatrefoils on the basement
of the western portals at Amiens. It is worthy of notice how classical
these are in style, and this is due to the fact that like Greek reliefs,
such as the frieze of the Parthenon, they were cut in the constructive
stones of the edifice in situ, and are in true stone technique.
This contrast of Greek and Italian reliefs furnishes a most
conspicuous object lesson on the importance of material and
technique in conditioning artistic practice. As was pointed out in the
Introductory Essay, these considerations have not in the past been
sufficiently emphasized in the scheme of education in design
recognized in our Schools of Art, though in several quarters now
there is a promise of better things.

THE PROCESSES OF THE BRONZE FOUNDER.

[§§ 55–69, ante, p. 158 ff.]

Vasari’s account of the processes attendant on casting in bronze is
intelligible and interesting, though he had himself little practical
acquaintance with the craft. Benvenuto Cellini on the other hand was
an expert bronze founder and the account he gives of the necessary
operations in the first three chapters of his treatise on Sculpture is
extremely graphic and detailed, and may be usefully employed to
amplify and explain Vasari’s notice. An expert knowledge of the
founder’s craft was not by any means universal among the Italian
sculptors of the Renaissance. Donatello did not possess it, nor did
Michelangelo. In the case of the former this is somewhat remarkable,
for Donatello was such a vigorous craftsman that we should have
expected to find him revelling in all the technical minutiae of the
foundry. We are expressly told however by Pomponius Gauricus that
Donatello lacked this knowledge, and never cast his own works but
always relied on the help of bell founders (Hans Semper, Donatello,
Wien, 1875, p. 317). Michelozzo on the other hand, who worked with
Donatello, could cast, and so could Ghiberti, A. Pollaiuolo, and
Verrocchio, while Alessandro Leopardi of Venice, who cast
Verrocchio’s Colleoni statue, was famed for his practical skill in this
department. It was customary, when expert help in casting was
required, to enlist the services of bell founders and makers of
cannon, but Cellini warns sculptors against trusting too much to
these mere mechanicians who lacked ingenuity and resource.
The following general sketch of the processes of casting will render
Vasari’s account more easy to follow. A successful cast in bronze
consists in a thin shell of the metal, representing on the exterior the
exact form of the model, or the complete design in the artist’s mind.
The best way to procure this is to provide first a similar shell, perfect
on its exterior surface, of wax, and then to melt away the wax and
replace it by molten bronze. For this to be possible the shell of wax
must be closely sealed between an outer envelope and an inner
packing or core. It can then be got rid of by melting, but care must
have been taken lest the core when it loses the support of the wax
should shift its position in relation to the envelope. To prevent this,
metal rods are run, skewer-fashion, through core and envelope to
retain them firmly in their relative positions. Molten bronze may
then be introduced into the space formerly occupied by the wax, and
this, when it is cold, and the envelope and core are both removed,
will be the cast required.
Attention has to be paid to secure that there shall be no moisture
and no remnant of wax in the parts where the molten bronze is to
come, otherwise steam may be generated and a dangerous explosion
follow. Similarly, air holes or vents must be provided, so that the air
may escape before the flowing metal. The cast when cold should,
theoretically, give a perfect result, but as a matter of fact, unless very
accomplished skill or great good fortune have presided over the
operations, the metal will be blistered or seamed or flawed in parts,
and these imperfections will have to be remedied by processes which
come under the head of chasing, and are described by Vasari at the
close of the chapter.
A direct and ingenious method of procuring the needful shell of
wax is that described by Vasari in § 67 as suitable for small figures
and reliefs. Over the model for such a small figure an envelope is
formed, in the shape of a hollow mould of fire-resisting material, so
constructed that it can be taken to pieces, the model extracted, and
the mould closed up again. The mould must now be cooled with cold
water and it is then filled with melted wax. Contact with the cold
sides of the mould chills the wax, which hardens all over in a sort of
crust or skin. The rest of the wax, still liquid, is then poured out and
the skin or crust suffered to harden. The interior is then filled with
clay of a kind that will stand heat. Rods or skewers are passed
through this and the envelope, the wax is melted out and its place
taken by the molten bronze.
This process, one of course only suitable for small objects,
presupposes the existence of a completely finished model to be
exactly reproduced in the metal. The simplest of all processes of
bronze casting dispenses with this model. Vasari does not describe
this simplest method but Cellini, who employed it both for his
‘Nymph of Fontainebleau’ and his ‘Perseus,’ gives an account of it
which is worth summarizing because the process is probably the one
adopted in most cases by the old Greek masters.
Cellini tells us that in making his large lunette-shaped relief of the
‘Nymph of Fontainebleau,’ now in the Louvre, he began by modelling
up the piece in fire-resisting clay, of course over a proper armature
or skeleton. He worked it out to full size and then let it dry till it had
shrunk about a finger’s breadth. He then covered it with a coating of
wax of rather less than this thickness, which he modelled with the
utmost care, finishing it in every detail so that it expressed to the full
his own idea for the finished work. This was then carefully covered in
successive layers with an envelope of fire-resisting material, which
would be properly tied by transverse rods to the core, and braced on
the exterior by an armature. The wax was then melted out, and the
core and envelope thoroughly dried, when the molten bronze was
poured in so as to reproduce the wax in every detail.
It is obvious that this is not only the most direct but the most
artistic method of work. The wax forms a complete unbroken surface
and receives and retains the exact impression in every detail of the
master’s hand. If the cast be thoroughly successful, the bronze will
reproduce the surface of the wax so perfectly that no further work
upon it will be needed, and an ‘untouched cast’ will be the result.
This method would suit the genius of the Greeks, and was no doubt
commonly employed by them, but it has the practical drawback that
if anything go wrong, and the bronze do not flow properly, the whole
work is spoilt, and will have to be built up again de novo from the
small study. Cellini tells us that his ‘Perseus,’ which he was casting in
this fashion, nearly came to grief from the cause just indicated, and
he accordingly recommends what he calls the second method, a
longer and less direct process, which has however the advantage that
a full-sized completed model is all the time preserved.
This process is the one described by Vasari. A full-sized clay model
is prepared and finished, and this is then covered with a plaster
envelope made in numerous sections, so that it can be taken to pieces
and put together again without the model, which may be preserved
for further use. The next step is to line the inside of the empty
envelope, or piece-mould, as it is called, with wax, and to fill up all
the rest of the interior with a core. The piece-mould is then removed,
and the surface of the wax is carefully gone over to secure that it shall
be perfect in every part. Over it is then laid in successive coats a fire-
resisting envelope between which and the core the wax is
hermetically sealed. The wax can then be melted out and replaced
with bronze. The piece-mould, which has been detached section by
section from the wax, will serve again for other reproductions. The
processes in which wax is employed are called cire perdue processes,
because the wax is got rid of in order to be replaced by the metal. The
usual process in vogue at the present day dispenses with any
employment of wax. The figure to be cast is piece-moulded and
reproduced in a suitable material, a certain thickness of which is in
every part pared away according to the thickness required for the
bronze. This core is then replaced with proper adjustments within a
fireproof mould, and the bronze is poured into the space prepared
for it.
OF PAINTING
 CHAPTER I. (XV.)
What Design is, and how good Pictures are made and known, and
concerning the invention of Compositions.

§ 74. The Nature and Materials of Design or Drawing.[186]

Seeing that Design, the parent of our three arts, Architecture,

Sculpture, and Painting, having its origin in the intellect, draws out
from many single things a general judgement, it is like a form or idea
of all the objects in nature, most marvellous in what it compasses, for
not only in the bodies of men and of animals but also in plants, in
buildings, in sculpture and in painting, design is cognizant of the
proportion of the whole to the parts and of the parts to each other
and to the whole. Seeing too that from this knowledge there arises a
certain conception and judgement, so that there is formed in the
mind that something which afterwards, when expressed by the
hands, is called design, we may conclude that design is not other
than a visible expression and declaration of our inner conception and
of that which others have imagined and given form to in their idea.
And from this, perhaps, arose the proverb among the ancients ‘ex
ungue leonem’ when a certain clever person, seeing carved in a stone
block the claw only of a lion, apprehended in his mind from its size
and form all the parts of the animal and then the whole together, just
as if he had had it present before his eyes. Some believe that accident
was the father of design and of the arts, and that use and experience
as foster-mother and schoolmaster, nourished it with the help of
knowledge and of reasoning, but I think that, with more truth,
accident may be said rather to have given the occasion for design,
than to be its father.
But let this be as it may, what design needs, when it has derived
from the judgement the mental image of anything, is that the hand,
through the study and practice of many years, may be free and apt to
draw and to express correctly, with the pen, the silver-point, the
charcoal, the chalk, or other instrument, whatever nature has
created. For when the intellect puts forth refined and judicious
conceptions, the hand which has practised design for many years,
exhibits the perfection and excellence of the arts as well as the
knowledge of the artist. And seeing that there are certain sculptors
who have not much practice in strokes and outlines, and
consequently cannot draw on paper, these work instead in clay or
wax, fashioning men, animals, and other things in relief, with
beautiful proportion and balance. Thus they effect the same thing as
does he who draws well on paper or other flat surface.
The masters who practise these arts have named or distinguished
the various kinds of design according to the description of the
drawing which they make. Those which are touched lightly and just
indicated with the pen or other instrument are called sketches, as
shall be explained in another place. Those, again, that have the first
lines encircling an object are called profiles or outlines.

§ 75. Use of Design (or Drawing) in the Various Arts.

All these, whether we call them profiles or otherwise, are as useful

to architecture and sculpture as to painting. Their chief use indeed is
in Architecture, because its designs are composed only of lines,
which so far as the architect is concerned, are nothing else than the
beginning and the end of his art, for all the rest, which is carried out
with the aid of models of wood formed from the said lines, is merely
the work of carvers and masons.[187]
In Sculpture, drawing is of service in the case of all the profiles,
because in going round from view to view the sculptor uses it when
he wishes to delineate the forms which please him best, or which he
intends to bring out in every dimension, whether in wax, or clay, or
marble, or wood, or other material.
In Painting, the lines are of service in many ways, but especially in
outlining every figure, because when they are well drawn, and made
correct and in proportion, the shadows and lights that are then
added give the strongest relief to the lines of the figure and the result
is all excellence and perfection. Hence it happens, that whoever
understands and manages these lines well, will, with the help of
practice and judgement, excel in each one of these arts. Therefore, he
who would learn thoroughly to express in drawing the conceptions of
the mind and anything else that pleases him, must after he has in
some degree trained his hand to make it more skilful in the arts,
exercise it in copying figures in relief either in marble or stone, or
else plaster casts taken from the life, or from some beautiful antique
statue, or even from models in relief of clay, which may either be
nude or clad in rags covered with clay to serve for clothing and
drapery. All these objects being motionless and without feeling,
greatly facilitate the work of the artist, because they stand still, which
does not happen in the case of live things that have movement. When
he has trained his hand by steady practice in drawing such objects,
let him begin to copy from nature and make a good and certain
practice herein, with all possible labour and diligence, for the things
studied from nature are really those which do honour to him who
strives to master them, since they have in themselves, besides a
certain grace and liveliness, that simple and easy sweetness which is
nature’s own, and which can only be learned perfectly from her, and
never to a sufficient degree from the things of art. Hold it moreover
for certain, that the practice that is acquired by many years of study
in drawing, as has been said above, is the true light of design and that
which makes men really proficient. Now, having discoursed long
enough on this subject let us go on to see what painting is.

§ 76. Of the Nature of Painting.

A painting, then, is a plane covered with patches of colour on the

surface of wood, wall, or canvas filling up the outlines spoken of
above, which, by virtue of a good design of encompassing lines,
surround the figure.[188] If the painter treat his flat surface with right
judgement, keeping the centre light and the edges and the
background dark and medium colour between the light and dark in
the intermediate spaces, the result of the combination of these three
fields of colour will be that everything between the one outline and
the other stands out and appears round and in relief. It is indeed true
that these three shades cannot suffice for every object treated in
detail, therefore it is necessary to divide every shade at least into two
half shades, making of the light two half tints, and of the dark two
lighter, and of the medium two other half tints which incline one to
the lighter and the other to the darker side. When these tints, being
of one colour only whatever it may be, are gradated, we see a
transition beginning with the light, and then the less light, and then a
little darker, so that little by little we find the pure black. Having then
made the mixtures, that is, these colours mixed together, and
wishing to work with oil or tempera or in fresco, we proceed to fill in
the outlines putting in their proper place the lights and darks, the
half tints and the lowered tones of the half tints and the lights. I
mean those tints mixed from the three first, light, medium and dark,
which lights and medium tints and darks and lower tones are copied
from the cartoon or other design which is made for any work before
we begin to put it into execution. It is necessary that the design be
carried out with good arrangement, firm drawing, and judgement
and invention, seeing that the composition in a picture is not other
than the parcelling out of the places where the figures come, so that
the spaces be not unshapely but in accordance with the judgement of
the eye, while the field is in one place well covered and in another
void. All this is the result of drawing and of having copied figures
from the life, or from models of figures made to represent anything
one wishes to make. Design cannot have a good origin if it have not
come from continual practice in copying natural objects, and from
the study of pictures by excellent masters and of ancient statues in
relief, as has been said many times. But above all, the best thing is to
draw men and women from the nude and thus fix in the memory by
constant exercise the muscles of the torso, back, legs, arms, and
knees, with the bones underneath. Then one may be sure that
through much study attitudes in any position can be drawn by help
of the imagination without one’s having the living forms in view.
Again having seen human bodies dissected one knows how the bones
lie, and the muscles and sinews, and all the order and conditions of
anatomy, so that it is possible with greater security and more
correctness to place the limbs and arrange the muscles of the body in
the figures we draw. And those who have this knowledge will
certainly draw the outlines of the figures perfectly, and these, when
drawn as they ought to be, show a pleasing grace and beautiful style.
He who studies good painting and sculpture, and at the same time
sees and understands the life, must necessarily have acquired a good
method in art. Hence springs the invention which groups figures in
fours, sixes, tens, twenties, in such a manner as to represent battles
and other great subjects of art. This invention demands an innate
propriety springing out of harmony and obedience; thus if a figure
move to greet another, the figure saluted having to respond should
not turn away. As with this example, so it is with all the rest. The
subject may offer many varied motives different one from another,
but the motives chosen must always bear relation to the work in
hand, and to what the artist is in process of representing. He ought to
distinguish between different movements and characteristics,
making the women with a sweet and beautiful air and also the
youths, but the old always grave of aspect, and especially the priests
and persons in authority. He must always take care however, that
everything is in relation to the work as a whole; so that when the
picture is looked at, one can recognize in it a harmonious unity,
wherein the passions strike terror, and the pleasing effects shed
sweetness, representing directly the intention of the painter, and not
the things he had no thought of. It is requisite therefore, for this
purpose, that he form the figures which have to be spirited with
movement and vigour, and that he make those which are distant to
retire from the principal figures by means of shade and colour that
gradually and softly become lower in tone. Thus the art will always
be associated with the grace of naturalness and of delicate charm of
colour, and the work be brought to perfection not with the stress of
cruel suffering, so that men who look at it have to endure pain on
account of the suffering which they see has been borne by the artist
in his work, but rather with rejoicing at his good fortune in that his
hand has received from heaven the lightness of movement which
shows his painting to be worked out with study and toil certainly, but
not with drudgery; so will it be that the figures, every one in its place,
will not appear dead to him who observes them, but alive and true.
Let painters avoid crudities, let it be their endeavour that the things
they are always producing shall not seem painted, but show
themselves alive and starting out of the canvas. This is the secret of
sound design and the true method recognized by him who has
painted as belonging to the pictures that are known and judged to be
good.
 CHAPTER II. (XVI.)
Of Sketches, Drawings, Cartoons, and Schemes of Perspective; how
they are made, and to what use they are put by the Painters.

§ 77. Sketches, Drawings, and Cartoons of different kinds.

Sketches, of which mention has been made above, are in artists’

language a sort of first drawing made to find out the manner of the
pose, and the first composition of the work. They are made in the
form of a blotch, and are put down by us only as a rough draft of the
whole. Out of the artist’s impetuous mood they are hastily thrown
off, with pen or other drawing instrument or with charcoal, only to
test the spirit of that which occurs to him, and for this reason we call
them sketches. From these come afterwards the drawings executed
in a more finished manner, in the doing of which the artist tries with
all possible diligence to copy from the life, if he do not feel himself
strong enough to be able to produce them from his own knowledge.
Later on, having measured them with the compasses or by the eye,
he enlarges from the small to a larger size according to the work in
hand. Drawings are made in various materials,[189] that is, either with
red chalk, which is a stone coming from the mountains of Germany,
soft enough to be easily sawn and reduced to a fine point suitable for
marking on leaves of paper in any way you wish; or with black chalk
that comes from the hills of France, which is of the same nature as
the red. Other drawings in light and shade are executed on tinted
paper which gives a middle shade; the pen marks the outlines, that
is, the contour or profile, and afterwards half-tone or shadow is given
with ink mixed with a little water which produces a delicate tint:
further, with a fine brush dipped in white lead mixed with gum, the
high lights are added. This method is very pictorial, and best shows
the scheme of colouring. Many work with the pen alone, leaving the
paper for the lights, which is difficult but in effect most masterly; and
innumerable other methods are practised in drawing, of which it is
not needful to make mention, because all represent the same thing,
that is drawing.
The designs having been made in this way, the artist who wishes to
work in fresco, that is, on the wall, must make cartoons; many
indeed prepare them even for working on panel. The cartoons are
made thus: sheets of paper, I mean square sheets, are fastened
together with paste made of flour and water cooked on the fire. They
are attached to the wall by this paste, which is spread two fingers’
breadth all round on the side next the wall, and are damped all over
by sprinkling cold water on them. In this moist state they are
stretched so that the creases are smoothed out in the drying. Then
when they are dry the artist proceeds, with a long rod, having a piece
of charcoal at the end, to transfer to the cartoon (in enlarged
proportions), to be judged of at a distance, all that in the small
drawing is shown on the small scale. In this manner little by little he
finishes, now one figure and now another. At this point the painters
go through all the processes of their art in reproducing their nudes
from the life, and the drapery from nature, and they draw the
perspectives in the same schemes that have been adopted on a small
scale in the first drawing, enlarging them in proportion.
If in these there should be perspective views, or buildings, these
are enlarged with the net, which is a lattice of small squares that are
made large on the cartoon, reproducing everything correctly, for of
course when the artist has drawn out the perspectives in the small
designs, taking them from the plan and setting up the elevations with
the right contours, and making the lines diminish and recede by
means of the intersections and the vanishing point, he must
reproduce them in proportion on the cartoon. But I do not wish to
speak further of the mode of drawing these out, because it is a
wearisome theme and difficult to explain. It is enough to say that
perspectives are beautiful in so far as they appear correct when
looked at, and diminish as they retire from the eye, and when they
are composed of a varied and beautiful scheme of buildings. The
painter must take care too, to make them diminish in proportion by
means of delicate gradations of colour that presuppose in the artist
correct discretion and good judgement.[190] The need of this is shown
in the difficulty of the many confused lines gathered from the plan,
the profile, and the intersection; but when covered with colour
everything becomes clear, and in consequence the artist gains a
reputation for skill and understanding and ingenuity in his art.
Many masters also before making the composition on the cartoon,
adopt the plan of fashioning a model in clay on a plane and of setting
up all the figures in the round to see the projections,[191] that is, the
shadows caused by a light being thrown on to the figures, which
projections correspond to the shadow cast by the sun, that more
sharply than any artificial light defines the figures by shade on the
ground; and so portraying the whole of the work, they have marked
the shadows that strike across now one figure, now another, whence
it comes that on account of the pains taken the cartoons as well as
the work reach the most finished perfection and strength, and stand
out from the paper in relief. All this shows the whole to be most
beautiful and highly finished.

§ 78. The Use of Cartoons in Mural and Panel Painting.

When these cartoons are used for fresco or wall painting, every day
at the junction with yesterday’s work a piece of the cartoon is cut off
and traced on the wall, which must be plastered afresh and perfectly
smoothed.[192] This piece of cartoon is put on the spot where the
figure is to be, and is marked; so that next day, when another piece
comes to be added, its exact place may be recognized, and no error
can arise. Afterwards, for transferring the outlines on to the said
piece, the artist proceeds to impress them with an iron stylus upon
the coat of plaster, which, being fresh, yields to the paper and thus
remains marked. He then removes the cartoon and by means of
those marks traced on the wall goes on to work with colours; this
then is how work in fresco or on the wall is carried out. The same
tracing is done on panels and on canvas, but in this case the cartoon
is all in one piece, the only difference being that it is necessary to rub
the back of the cartoon with charcoal or black powder, so that when
marked afterwards with the instrument it may transmit the outlines
and tracings to the canvas or panel. The cartoons are made in order
to secure that the work shall be carried out exactly and in due
proportion. There are many painters who for work in oil will omit all
this; but for fresco work it must be done and cannot be avoided.
Certainly the man who found out such an invention had a good
notion, since in the cartoons one sees the effect of the work as a
whole and these can be adjusted and altered until they are right,
which cannot be done on the work itself.
 CHAPTER III. (XVII.)
Of the Foreshortening of Figures looked at from beneath, and of
those on the Level.

§ 79. Foreshortenings.

Our artists have had the greatest skill in foreshortening figures,

that is, in making them appear larger than they really are; a
foreshortening being to us a thing drawn in shortened view, which
seeming to the eye to project forward has not the length or height
that it appears to have; however, the mass, outlines, shadows, and
lights make it seem to come forward and for this reason it is called
foreshortened. Never was there painter or draughtsman that did
better work of this sort than our Michelagnolo Buonarroti,[193] and
even yet no one has been able to surpass him, he has made his
figures stand out so marvellously. For this work he first made models
in clay or wax, and from these, because they remain stationary, he
took the outlines, the lights, and the shadows, rather than from the
living model. These foreshortenings give the greatest trouble to him
who does not understand them because his intelligence does not help
him to reach the depth of such a difficulty, to overcome which is a
more formidable task than any other in painting. Certainly our old
painters, as lovers of the art, found the solution of the difficulty by
using lines in perspective, a thing never done before, and made
therein so much progress that to-day there is true mastery in the
execution of foreshortenings. Those who censure the method of
foreshortening, I speak of our artists, are those who do not know how
to employ it; and for the sake of exalting themselves go on lowering
others. We have however a considerable number of master painters
who, although skilful, do not take pleasure in making foreshortened
figures, and yet when they see how beautiful they are and how
difficult, they not only do not censure but praise them highly. Of
these foreshortenings the moderns have given us some examples
which are to the point and difficult enough, as for instance in a vault
the figures which look upwards, are foreshortened and retire. We call
these foreshortenings ‘al di sotto in su’ (in the ‘up from below’ style),
and they have such force that they pierce the vaults. These cannot be
executed without study from the life, or from models at suitable
heights, else the attitudes and movements of such things cannot be
caught. And certainly the difficulty in this kind of work calls forth the
highest grace as well as great beauty, and results in something
stupendous in art. You will find, in the Lives of our Artists, that they
have given very great salience to works of this kind, and laboured to
complete them perfectly, whence they have obtained great praise.
The foreshortenings from beneath upwards (di sotto in su) are so
named because the object represented is elevated and looked at by
the eye raised upwards, and is not on the level line of the horizon:
wherefore because one must raise the head in the wish to see them,
and perceives first the soles of the feet and the other lower parts we
find the said name justly chosen.[194]
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com