Full download test bank at ebook textbookfull.

com

Aerial Robotic Manipulation

Research Development and

CLICK LINK TO DOWLOAD

https://textbookfull.com/product/aerial-
robotic-manipulation-research-development-
and-applications-anibal-ollero/

textbookfull
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Wearable Technology for Robotic Manipulation and

Learning Bin Fang

https://textbookfull.com/product/wearable-technology-for-robotic-
manipulation-and-learning-bin-fang/

9th International Conference on Robotic Vision Signal

Processing and Power Applications Empowering Research
and Innovation 1st Edition Haidi Ibrahim

https://textbookfull.com/product/9th-international-conference-on-
robotic-vision-signal-processing-and-power-applications-
empowering-research-and-innovation-1st-edition-haidi-ibrahim/

Unmanned Aerial Remote Sensing : UAS for Environmental

Applications 1st Edition David R. Green

https://textbookfull.com/product/unmanned-aerial-remote-sensing-
uas-for-environmental-applications-1st-edition-david-r-green/

Small Format Aerial Photography and UAS Imagery

Principles Techniques and Geoscience Applications 2nd
Edition James S. Aber

https://textbookfull.com/product/small-format-aerial-photography-
and-uas-imagery-principles-techniques-and-geoscience-
applications-2nd-edition-james-s-aber/
Quantitative Methods in Pharmaceutical Research and
Development Concepts and Applications 1st Edition Olga
V. Marchenko

https://textbookfull.com/product/quantitative-methods-in-
pharmaceutical-research-and-development-concepts-and-
applications-1st-edition-olga-v-marchenko/

Innovation Research and Development Management

Bobadilla

https://textbookfull.com/product/innovation-research-and-
development-management-bobadilla/

Knowledge Computing and Its Applications Knowledge

Manipulation and Processing Techniques Volume 1 1st
Edition S. Margret Anouncia

https://textbookfull.com/product/knowledge-computing-and-its-
applications-knowledge-manipulation-and-processing-techniques-
volume-1-1st-edition-s-margret-anouncia/

Aerial Warfare: The Battle for the Skies Frank Ledwidge

https://textbookfull.com/product/aerial-warfare-the-battle-for-
the-skies-frank-ledwidge/

Advances in Robotics Research: From Lab to Market:

ECHORD++: Robotic Science Supporting Innovation Antoni
Grau

https://textbookfull.com/product/advances-in-robotics-research-
from-lab-to-market-echord-robotic-science-supporting-innovation-
antoni-grau/
Springer Tracts in Advanced Robotics 129

Anibal Ollero
Bruno Siciliano Editors

Aerial Robotic
Manipulation
Research, Development and
Applications
Springer Tracts in Advanced Robotics

Volume 129

Series Editors
Bruno Siciliano, Dipartimento di Ingegneria Elettrica e Tecnologie
dell’Informazione, Università degli Studi di Napoli Federico II, Napoli, Italy
Oussama Khatib, Artificial Intelligence Laboratory, Department of Computer
Science, Stanford University, Stanford, CA, USA

Advisory Editors
Nancy Amato, Computer Science & Engineering, Texas A&M University, College
Station, TX, USA
Oliver Brock, Fakultät IV, TU Berlin, Berlin, Germany
Herman Bruyninckx, KU Leuven, Heverlee, Belgium
Wolfram Burgard, Institute of Computer Science, University of Freiburg, Freiburg,
Baden-Württemberg, Germany
Raja Chatila, ISIR, Paris cedex 05, France
Francois Chaumette, IRISA/INRIA, Rennes, Ardennes, France
Wan Kyun Chung, Robotics Laboratory, Mechanical Engineering, POSTECH,
Pohang, Korea (Republic of)
Peter Corke, Science and Engineering Faculty, Queensland University of
Technology, Brisbane, QLD, Australia
Paolo Dario, LEM, Scuola Superiore Sant’Anna, Pisa, Italy
Alessandro De Luca, DIAGAR, Sapienza Università di Roma, Roma, Italy
Rüdiger Dillmann, Humanoids and Intelligence Systems Lab, KIT - Karlsruher
Institut für Technologie, Karlsruhe, Germany
Ken Goldberg, University of California, Berkeley, CA, USA
John Hollerbach, School of Computing, University of Utah, Salt Lake, UT, USA
Lydia E. Kavraki, Department of Computer Science, Rice University, Houston, TX,
USA
Vijay Kumar, School of Engineering and Applied Mechanics, University of
Pennsylvania, Philadelphia, PA, USA
Bradley J. Nelson, Institute of Robotics and Intelligent Systems, ETH Zurich,
Zürich, Switzerland
Frank Chongwoo Park, Mechanical Engineering Department, Seoul National
University, Seoul, Korea (Republic of)
S. E. Salcudean, The University of British Columbia, Vancouver, BC, Canada
Roland Siegwart, LEE J205, ETH Zürich, Institute of Robotics & Autonomous
Systems Lab, Zürich, Switzerland
Gaurav S. Sukhatme, Department of Computer Science, University of Southern
California, Los Angeles, CA, USA
The Springer Tracts in Advanced Robotics (STAR) publish new developments and
advances in the fields of robotics research, rapidly and informally but with a high
quality. The intent is to cover all the technical contents, applications, and
multidisciplinary aspects of robotics, embedded in the fields of Mechanical
Engineering, Computer Science, Electrical Engineering, Mechatronics, Control, and
Life Sciences, as well as the methodologies behind them. Within the scope of the
series are monographs, lecture notes, selected contributions from specialized
conferences and workshops, as well as selected PhD theses.
Special offer: For all clients with a print standing order we offer free access to the
electronic volumes of the Series published in the current year.
Indexed by DBLP, Compendex, EI-Compendex, SCOPUS, Zentralblatt Math,
Ulrich’s, MathSciNet, Current Mathematical Publications, Mathematical Reviews,
MetaPress and Springerlink.

More information about this series at http://www.springer.com/series/5208

Anibal Ollero Bruno Siciliano
•

Editors

Aerial Robotic Manipulation

Research, Development and Applications

123
Editors
Anibal Ollero Bruno Siciliano
GRVC Robotics Lab Seville Dipartimento di Ingegneria Elettrica e delle
Universidad de Sevilla Tecnologie dell’Informazione
Seville, Spain Università di Napoli Federico II
Napoli, Italy

ISSN 1610-7438 ISSN 1610-742X (electronic)

Springer Tracts in Advanced Robotics
ISBN 978-3-030-12944-6 ISBN 978-3-030-12945-3 (eBook)
https://doi.org/10.1007/978-3-030-12945-3

Library of Congress Control Number: 2019932695

© Springer Nature Switzerland AG 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

Robotics is undergoing a major transformation in scope and dimension. From a

largely dominant industrial focus, robotics is rapidly expanding into human envi-
ronments and vigorously engaged in its new challenges. Interacting with, assisting,
serving and exploring with humans, the emerging robots will increasingly touch
people and their lives.
Beyond its impact on physical robots, the body of knowledge robotics has
produced is revealing a much wider range of applications reaching across diverse
research areas and scientific disciplines, such as biomechanics, haptics, neuro-
sciences, virtual simulation, animation, surgery and sensor networks, among others.
In return, the challenges of the new emerging areas are proving an abundant source
of stimulation and insights for the field of robotics. It is indeed at the intersection of
disciplines that the most striking advances happen.
The Springer Tracts in Advanced Robotics (STAR) is devoted to bringing to the
research community the latest advances in the robotics field on the basis of their
significance and quality. Through a wide and timely dissemination of critical
research developments in robotics, our objective with this series is to promote more
exchanges and collaborations among the researchers in the community and con-
tribute to further advancements in this rapidly growing field.
This book by Anibal Ollero and Bruno Siciliano offers unique insights into the
challenges of manipulation in aerial robotics. Based on developments pursued in
the framework of the European Commission’s funded projects, ARCAS and
AEROARMS, this volume covers a wide range of fundamental concepts and
advanced methodologies and technologies in this growing area in robotics. The
impressive work on experimental validation for major real-world applications
reveals the field’s increased maturity and expanded scope.
Rich by topics and authoritative contributors, these European projects culminate
with this unique reference on the current developments and new directions in aerial
robotics manipulation. A fine addition to the STAR series!

Stanford, California Oussama Khatib

January 2019 STAR Editor

v
Preface

Aerial robotics has experienced an exponential growth in the last years fuelled by
the technology development and the many application possibilities. Drones are
today very popular. Applications such as filming, monitoring, surveillance or
transportation, between others, are being performed almost every day, in addition to
the defence applications that were well known from many years ago.
Aerial robotics is the subject of particular scientific and technological events, but
currently, it is also one of the main areas in general robotics conferences, with many
sessions devoted to the aerial robotics topics.
Flight regulations, which have been published in the last years, are clarifying the
application possibilities by keeping the safety standards. New regulations will
appear based on risk analysis. Navigation at very low levels and the integration in
unmanned air traffic management systems will also favour the implementation of
new applications.
Aerial robotic manipulation is one of the topics that have attracted the attention
of researchers and that already have relevant applications. It can be considered as
manipulation with a flying base. It includes design and mechatronics aspects, which
are essential to achieve the required manipulation performance. Modelling and
control of aerial robots with one or more arms also pose relevant problems dealing
with kinematics, dynamics and aerodynamics of the flying robots physically
interacting with the environment. This book summarizes methods and technologies
in these topics.
The fundamental topics mentioned in the above paragraph are not the only ones
needed for aerial robotic manipulation. In fact, as in other robotics areas, perception
and planning are also very important ingredients. Then, this book also includes
research and technology development in perception and planning topics by using
new on-board computational capabilities.
Furthermore, this book includes applications in two different areas: (1) structure
assembly and (2) inspection and maintenance. These two applications have cur-
rently different technology readiness levels (TRL). Thus, inspection and mainte-
nance has higher TRL and has been already implemented in industrial scenarios for

vii
viii Preface

contact inspection with ultrasonic sensors to measure the wall thickness of pipes
and tanks.
The research and developments presented in this book were mainly conducted in
the framework of the ARCAS and AEROARMS projects funded by the European
Commission.
ARCAS (Aerial Robotics Cooperative Assembly system) was funded under the
Framework Programme 7. It started by the end of November 2011. The final review
was in January 2016. The ARCAS partners were Fundación Andaluza para el
Desarrollo Aeroespacial (FADA-CATEC), Universidad de Sevilla (US), Deutsches
Zentrum fr Luft-and Raumfahart (DLR), Università degli Studi di Napoli Federico II
(UNINA), Università degli Studi della Basilicata (UNIBAS), Università degli Studi
di Cassino e del Lazio Meridionale (UNICAS), Centre National de la Recherche
Scientifique (CNRS), Universitat Politècnica de Catalunya (UPC), Alstom (currently
General Electric) Inspection Robotics (AIR) and SpaceTech GmbH.
In ARCAS, the first worldwide aerial robotic manipulators including multirotors
and helicopters, operating indoors and outdoors with robotic arms with six and
seven degrees of freedom were demonstrated. The fundamental methods in control,
perception and planning in aerial robotics were developed in the context of the
ARCAS project. The controllers implemented both decentralized and centralized
techniques to cancel the effect of the motion of the arms on the hovering of the
aerial platform. The perception methods were based mainly on visual markers
combined with range-only radio technologies. Planning was mainly offline, but
reactivity approaches to avoid in real-time obstacles and other aerial robots were
also developed.
The main aerial robotic platforms and techniques developed in ARCAS are
included in Parts II, III, IV and V of this book.
The final demonstrations of ARCAS were performed indoors and outdoors.
Thus, a system integrated by FADA-CATEC performed indoor structure assembly
with several aerial manipulators, which is included in Part VI of this book. The
system integrated multirotors with very light six degrees of freedom arms enabled
with control techniques for single and cooperative aerial manipulators transporting
the same load. The control systems were developed by UNINA, UNIBAS,
UNICAS and FADA-CATEC. The perception techniques were developed by UPC
(visual) and US (range-only). The planning methods were developed by CNRS with
the collaboration of US for the obstacle detection and avoidance. It also included
the deployment of a crawler mock-up developed by the AIR.
The outdoor demonstrations included both the DLR Flettner helicopter equipped
with a seven degrees of freedom LWR (DLR-KUKA) robotic manipulator with
visual servoing and the multirotor of US with another robotic manipulator with
seven degrees of freedom.
ARCAS also included the demonstration of a space system for on-orbit flying
manipulation in satellite servicing. The demonstration was performed by DLR by
simulating the satellite space dynamics by using additional manipulators.
SpaceTech also participated in these demonstrations.
Preface ix

AEROARMS (AErial RObotic system integrating multiple ARMS and

advanced manipulation capabilities for inspection and maintenance) was funded
under the H2020 programme of the European Commission. The project started in
June 2015 and will conclude by the end of May of 2019. The partners are
Universidad de Sevilla (US), FADA-CATEC, CNRS, UNINA, UNIBAS,
UNICAS, TV NORD (TV), UPC, Elektra UAS GmbH (ELEK), General Electric
Inspection Robotics (GEIR) and SENSIMA.
AEROARMS has developed the first aerial robotic manipulators with multiple
arms and advanced manipulation capabilities to be applied in inspection and
maintenance in industrial plants, particularly in works at height that involve sig-
nificant risks for human operators and high costs. Special attention is paid to the
inspection and maintenance of oil and gas industries. The objectives of
AEROARMS are: (1) research and development in aerial robotic manipulation
including dual-arm manipulation systems for complex inspection and maintenance
tasks requiring dexterity and multidirectional thrust platforms, and (2) validation in
the industrial environment, including contact sensing while flying and deployment
of mobile robotic systems.
Particularly, AEROARMS has developed the first worldwide autonomous
dual-arm manipulator systems with four and five degrees of freedom arms,
including compliant arms, and also the first aerial manipulators with multidirec-
tional thrust platforms. Kinematics, dynamics and aerodynamic models, taking into
account the effects generated by nearby surfaces, are also included in Part II of this
book. The decentralized and centralized control techniques of these aerial manip-
ulators are also included in Part III, together with the kinematic behaviour-based
control and the visual servoing. These techniques were developed mainly by US,
CNRS, UNIBAS, UNICAS and UNINA.
The perception techniques were evolved and applied outdoors. Thus, new both
absolute localization and mapping and relative localization were developed by
using cameras, 3D lidar and radio-based radio. Moreover, outdoor robust perception
for detection, localization and grasping without visual markers were developed by
UPC and US. The results are shown in Section IV of the book.
New control-aware planning methods were developed by CNRS, UNIBAS and
UNICAS, and completed with on-line dynamics-aware local planning and reactivity
developed by US and included in Part V of this book.
The methods are being applied in industrial inspection and maintenance in oil
and gas plants by FADA-CATEC with the collaboration of the industrial partners
TV, GEIR and SENSIMA. The first results are included in Part VI. AEROARMS
has received the Overall ICT Innovation Radar Award 2017 of the European
Commission, between 25 finalists, for the contact inspection application of the
AEROX system developed by FADA-CATEC and included in Part VI.
It would not have been possible to produce this book without the help of many
people. We are grateful to all the persons involved in the ARCAS and
AEROARMS projects, which are not authors of the book chapters.
x Preface

We would like to thank Anne Bajart, Project Officer of ARCAS and

AEROARMS until October 2018. She had a very important role in the guidance of
both projects. Furthermore, we would like here to thank the excellent work of the
project reviewers. Paul Oh, José Neira, Uwe Nortman and Jeremy Gancet were
reviewers of ARCAS. Moreover, José Neira, Uwe Nortman and Tarek Hamel are
reviewers of AEROARMS. They all provided very useful comments that guided the
development of both projects.
In addition of ARCAS and AEROARMS, the work in this book has been
supported by the funding of other National projects such as the Spanish Ministry of
Economy and Competitiveness projects ARM-EXTEND (DPI2017-89790-R),
AEROCROS (DPI2015-71524-R), HuMoUR (TIN2017-90086-R) and
COLROBTRANSP (DPI2016-78957-R).

Seville, Spain Anibal Ollero

Napoli, Italy Bruno Siciliano
November 2018
Abstract

Aerial robotic manipulation integrates concepts and technologies coming from

unmanned aerial systems and robotics manipulation. It includes not only kinematic,
dynamics, aerodynamics and control but also perception, planning, design aspects,
mechatronics and cooperation between several aerial robotics manipulators. All
these topics are considered in this book in which the main research and develop-
ment approaches in aerial robotic manipulation are presented, including the
description of relevant systems. In addition of the research aspects, the book also
includes the deployment of real systems both indoors and outdoors, which is a
relevant characteristic of the book because most results of aerial robotic manipu-
lation have been validated only indoor using motion tracking systems. Moreover,
the book presents two relevant applications: structure assembly and inspection and
maintenance, which has started to be applied in the industry. The chapters of the
book will present the results of two main European Robotics Projects in aerial
robotics manipulation: FP7 ARCAS and H2020 AEROARMS. FP7 ARCAS
defined the basic concepts on aerial robotic manipulation, including cooperative
manipulation. The H2020 AEROARMS on aerial robot with multiple arms and
advanced manipulation capabilities for inspection and maintenance has two general
objectives: (1) development of advanced aerial robotic manipulation methods and
technologies, including manipulation with dual arms and multidirectional thrusters
aerial platforms; and (2) application to the inspection and maintenance.

xi
Contents

Part I Introduction
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Anibal Ollero and Bruno Siciliano

Part II Aerial Robots with Arms: Design, Modelling

and Mechatronics Aspects
Modeling and Design of Multirotors with Multi-joint Arms . . . . . . . . . . 15
Guillermo Heredia, Raul Cano, Antonio Jimenez-Cano and Anibal Ollero
Helicopter Based Aerial Manipulators . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Manuel Bejar, Anibal Ollero and Konstantin Kondak
Platforms with Multi-directional Total Thrust . . . . . . . . . . . . . . . . . . . . 53
Antonio Franchi
Multirotor Aerodynamic Effects in Aerial Manipulation . . . . . . . . . . . . 67
Pedro Sanchez-Cuevas, Guillermo Heredia and Anibal Ollero
Compliant Aerial Manipulators with Dual Arms . . . . . . . . . . . . . . . . . . 83
Alejandro Suarez, Guillermo Heredia and Anibal Ollero

Part III Control of Aerial Manipulators

Behavioral Coordinated Kinematic Control . . . . . . . . . . . . . . . . . . . . . . 101
Gianluca Antonelli
Centralized Control of Multirotors with Manipulators . . . . . . . . . . . . . . 119
Antonio Jimenez-Cano, Guillermo Heredia and Anibal Ollero
Centralized Control of Helicopters with Manipulators . . . . . . . . . . . . . . 127
Manuel Bejar, Anibal Ollero and Konstantin Kondak

xiii
xiv Contents

Decoupled Impedance and Passivity Control Methods . . . . . . . . . . . . . . 147

Francesco Pierri
Decentralized Control of Aerial Manipulators Through
a Momentum-Based Estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Fabio Ruggiero
Interaction Control of Platforms with Multi-directional
Total Thrust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Antonio Franchi
Visual Servoing of Aerial Manipulators . . . . . . . . . . . . . . . . . . . . . . . . . 191
Angel Santamaria-Navarro, Juan Andrade-Cetto and Vincenzo Lippiello
Coordinated Control of Multiple Aerial Manipulators . . . . . . . . . . . . . . 203
Fabrizio Caccavale

Part IV Perception for Aerial Robotic Manipulation

Odometry Estimation for Aerial Manipulators . . . . . . . . . . . . . . . . . . . . 219
A. Santamaria-Navarro, J. Solà and J. Andrade-Cetto
Absolute Localization and Mapping of Aerial Manipulators . . . . . . . . . 229
M. Polvillo, J. L. Paneque and J. R. Martinez-de Dios
Relative Localization for Aerial Manipulation with PL-SLAM . . . . . . . 239
A. Pumarola, Alexander Vakhitov, Antonio Agudo, F. Moreno-Noguer
and A. Sanfeliu
Precise Localization for Aerial Inspection Using Augmented
Reality Markers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
A. Amor-Martinez, A. Ruiz, F. Moreno-Noguer and A. Sanfeliu
Robust Perception for Aerial Inspection: Adaptive
and On-Line Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
M. Villamizar and A. Sanfeliu
Perception for Detection and Grasping . . . . . . . . . . . . . . . . . . . . . . . . . 275
E. Guerra, A. Pumarola, A. Grau and A. Sanfeliu
Object Detection and Probabilistic Object Representation
for Grasping with Two Arms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
P. Ramon Soria and B. C. Arrue

Part V Planning for Aerial Robotic Manipulation

Combining Assembly Planning and Geometric Task Planning . . . . . . . . 299
Raphaël Lallement, Juan Cortés, Mamoun Gharbi, Alexandre Boeuf,
Rachid Alami, Carmelo J. Fernandez-Agüera and Iván Maza
Contents xv

Motion Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

Alexandre Boeuf, Juan Cortés and Thierry Siméon
Reactivity and Dynamic Obstacle Avoidance . . . . . . . . . . . . . . . . . . . . . 333
Alvaro Caballero, Manuel Bejar, Angel Rodriguez-Castaño
and Anibal Ollero

Part VI Applications
Structure Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Miguel Angel Trujillo, Vincenzo Lippiello, Angel Santamaria
and Antidio Viguria
Inspection and Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Anibal Ollero, Antidio Viguria, Miguel Angel Trujillo, Moritz Oetiker
and Bernard Revaz

Part VII Conclusions and Future Directions

Conclusions and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Anibal Ollero and Bruno Siciliano
Acronyms

AR Auxiliary Robot in a Team of AROMAs

AROMA Aerial Robotic Manipulator
ARS-LRM Aerial Robotic System for Long-Reach Manipulation
ASP Assembly Sequence Planning
CAD Computer-Aided Design
CFD Computer Fluid Dynamics
CFRP Carbon Fibre Reinforced Plastic
COFP Collinearly Oriented Fixed propellers
COM Centre of Mass
CPU Central Processing Unit
CUDA Compute Unified Device Architecture
DGPS Differential Global Positioning System
DH Denavit-Hartenberg
DOF Degree of Freedom
DWA Dynamic Window Approach
ESC Electronic Speed Controller
FFT Fast Fourier Transform
GOFP Generically Oriented Fixed Propellers
GP Gaussian Process
GPIS Gaussian Process Implicit Surface
GPS Global Positioning System
GPU Graphics Processing Unit
GTP Geometric Task Planner
GWS Grasp Wrench Space
HATP Hierarchical Agent-based Task Planner
HR-LR High-Resolution Local Replanning
HR-LR-DA High-Resolution Local Replanning with Dynamics Awareness
HTN Hierarchical Task Network
MCVI Monte Carlo Value Iteration
ND Nearness Diagram

xvii
xviii Acronyms

OMPL Open Motion Planning Library

ORCA Optimal Reciprocal Collision Avoidance
PD Proportional-Derivative
PID Proportional-Integral-Derivative
POMDP Partially Observable Markov Decision Process
PRM Probabilistic Roadmap
PWM Pulse-Width Modulation
RPM Revolutions per Minute
RRT Rapidly Exploring Random Tree
RTK Real-Time Kinematic
SGP Symbolic-Geometric Planner
SITL Software-in-the-Loop
TCP Tool Centre Point
TCP/IP Transmission Control Protocol/Internet Protocol
TR Transporting Robot in a Team of AROMAs
UAV Unmanned Aerial Vehicle
VFH Vector Field Histogram
VPIB Variable Parameter Integral Backstepping
VTOL Vertical Take-Off and Landing
WGS84 World Geodetic System 1984
Symbols

N Number of aerial robotic manipulators (AROMAs)

NT Number of TRs in a team of AROMAs
NA Number of ARs in a team of AROMAs
D Number of DoFs of the AROMA; Di if more than one
AROMA
L Number of arms; Li if more than one AROMA
M Number of DoFs (joints) of the arm; Mj if more than one
arm in the AROMA; Mi;j if more than one AROMA
q ¼ ½q1 ; . . .; qM T Link-side angular position of the arm DoFs (joints). qi;j for
the jth joint of the ith arm if more than one arm in the
AROMA
q_ ¼ ½q_ 1 ; . . .; q_ M T q_ i Angular position of the arm DoFs (joint velocities). q_ i;j if
more than one arm in the AROMA
c ¼ ½c1 ; . . .; cM T Motor-side angular position of the compliant joints of an
arm
FW Inertial world-fixed coordinate frame
FB Coordinate frame attached to the centre of mass of the
vehicle’s body
FP i Coordinate frame attached to the base of the arm i; FP if
only one arm
FE i Coordinate frame attached to the end effector of the arm i;
FE if only one arm
Fa Absolute frame attached to an object grasped by the TRs or
located in a relevant point of the system (e.g. the centroid
of the TRs’ end effectors)
FC Coordinate frame attached to a camera mounted on the
AROMA
pb ¼ ½ x y z T Position of the aerial vehicle in the world reference frame,
i.e. FB with respect to FW
gb ¼ ½ u h w T Attitude of the aerial vehicle, roll, pitch, yaw Euler angles

xix
xx Symbols

Rb Rotation matrix expressing the rotation of FB with respect

to FW
p_ b Absolute linear velocity of the aerial vehicle
xb Rotational velocity of the aerial vehicle with respect to FW ,
expressed in FW
xbb ¼ ½xx ; xy ; xz  Rotational velocity of the aerial vehicle with respect to FW ,
expressed in FB
vc Velocity vector expressed in camera coordinate frame FC
Tb Transformation matrix between the time derivative of the
Euler angles gb and the angular velocity xb
Q Matrix that maps the time derivative of gb into xbb
m Mass of the aerial vehicle
mt Total mass of the AROMA, including the mass of the aerial
vehicle and the arms
Ib Constant inertia matrix of the UAV expressed with respect
to FB
f ¼ fu þ fe Total force input vector expressed in FW ; f b if expressed in
FB
fu Input rotor force vector expressed in FW
f bu ¼ ½ fx fy fz T Input rotor force vector expressed in FB
fe External force vector expressed in FW ; f be if expressed in
FB
s ¼ su þ se Total torque input vector expressed in FW ; sb if expressed
in FB
su Input rotor torque vector expressed in FW
sbu ¼ ½ s/ sh sw  T Input rotor torque vector expressed in FB
se External torque vector expressed in FB ; sbe if expressed in
FB
sA ¼ ½s1 ; . . .; sM T Input torque vector exerted on the arm DoFs (joints); si;j if
more than one arm in the AROMA
Ti Thrust force generated by rotor i in free air
xi Rotational velocity of rotor i
cT Rotor thrust constant
Qi Drag reaction torque for rotor i
cQ Drag factor parameter
G 2 R6n Matrix that relates control input forces and torques to the
squared rotational rotor velocities for a multidirectional
thrust aerial vehicle
Gq 2 R44 Matrix that relates control input forces and torques to the
squared rotational rotor velocities for a quadrotor
Gh 2 R46 Matrix that relates control input forces and torques to the
squared rotational rotor velocities for a hexarotor
Go 2 R48 Matrix that relates control input forces and torques to the
squared rotational rotor velocities for an octorotor
Symbols xxi

Gm 2 R66 Matrix that relates control input forces and torques to the
squared rotational rotor velocities for a multidirectional
thrust hexarotor
nv ¼ ½pb ; gb  Generalized configuration vector of an aerial vehicle
n ¼ ½pb ; gb ; qT Generalized configuration vector of an aerial manipulator
with an arm
ne ¼ ½pe ; ge  End effector pose in the inertial frame FW ; npe ¼ ½pbe ; gbe  if
expressed in FB
k Homogeneous transformation matrix for the k-th joint of the
k1 Ti
i-th arm
Tpeii Homogeneous transformation matrix between the arm base
FPi and end effector FEi frames for arm i
Je;p Jacobian matrix of the arm with respect to its base attached
to the vehicle; Jei ;pi if more than one arm
In 2 Rnn Identity matrix
On 2 Rnn Null matrix
 Part I
Introduction
Introduction

Anibal Ollero and Bruno Siciliano

Abstract This chapter is a general introduction of the book. First the motivation of
aerial robotic manipulation is presented. Next a very short introduction of unmanned
aerial vehicles and aerial robotics is included. The next section is devoted to intro-
duce unmanned aerial vehicles physically interacting with the environment with the
environment. Finally, aerial robotic manipulation while flying is introduced.

1 Why Aerial Robotic Manipulation?

Work at height is needed for many activities including construction, installation of

equipment, search and rescue, and others. These works are also needed for inspec-
tion and maintenance of a large number of assets including infrastructures, energy
generation and distribution systems, and industrial process plants.
Today the above works are performed by personnel by using scaffolding, ladders,
forklifts, ropes and climbing harnesses. These works involve significant risks and
there are many accidents due to falls. Also the costs are high and the works require
significant time.
In some cases manned aircrafts, and particularly helicopters, are used with human
operators performing activities from the aircraft. However, these activities are also
dangerous and costly.
The alternative is the application of aerial robotic manipulation technologies to
access high locations and perform the work. However, the application of these tech-
nologies is not easy. As will be pointed out in this book, it is not enough with the

A. Ollero (B)
GRVC Robotics Lab Seville, Universidad de Sevilla, Seville, Spain
e-mail: aollero@us.es
B. Siciliano
Department of Electrical Engineering and Information Technology,
University of Naples Federico II, Naples, Italy
e-mail: bruno.siciliano@unina.it

© Springer Nature Switzerland AG 2019 3

A. Ollero and B. Siciliano (eds.), Aerial Robotic Manipulation,
Springer Tracts in Advanced Robotics 129,
https://doi.org/10.1007/978-3-030-12945-3_1
4 A. Ollero and B. Siciliano

simple installation of devices and robotic arms in conventional aerial platforms as if

they were ground robotic platforms.
In this chapter we first introduce some concepts and terminology related to
Unmanned Aerial Vehicles and Aerial Robotics. Then, we consider aerial robots
physically interacting with the environment. Finally we introduce the main prob-
lems in aerial robotic manipulation.

2 Unmanned Aerial Vehicles and Aerial Robotics

Unmanned Aerial Vehicles (UAVs) are today not only well known in the scientific
and professional communities but also very popular. They are in the media, with
the name of “drones”, almost every day. They are a suitable alternative for many
applications [1] including filming, surveillance, environment monitoring, search and
rescue, agriculture, transportation, or even inspection and manufacturing, in addition
to the military that have dominated the UAV practical applications until recently.
Fixed wing UAVs (Fig. 1) have long flight endurance and long range of flight but
lower maneuverability. They cannot hover and are not able of vertical take-off and
landing, which are required for many applications. On the other hand, helicopters
and multi-rotor systems have these properties. Both type of aircrafts are considered
in this book.

Fig. 1 Fixed wing UAVs: a GRVC-Univ. Sevilla fixed wing with deployable ground robot under
the belly and 1 h flight endurance; b GRVC-Univ. Sevilla glider with 1 h flight endurance; c CATEC
X-vision with combustion engine, more than 4 h flight endurance, between 5 and 10 kg payload;
d CATEC Viewer with electrical brushless motor, more than 90 min flight endurance, up to 2 kg
payload
Introduction 5

Fig. 2 Helicopters: a INDRA Pelicano with combustion engine, 6 h flight endurance, 20 kg payload;
b Flying-Cam Sarah with electrical motor, 30 min flight endurance, 5 kg payload; c DLR Flettner
helicopter with 35 kg payload

Fig. 3 Multi-rotors: a GRVC-Univ. Sevilla multi-rotors with parallel axis in load transportation;
b CATEC multi-directional thrusters

Helicopters (Fig. 2) have usually better endurance and range than multi-rotors,
and also higher payload, which is also an important characteristic for many applica-
tions. Thus, there are many commercial unmanned helicopters that can carry tens of
kilograms.
However, the mechanical simplicity of multi-rotor systems (Fig. 3) has been a
key aspect for the popularity that currently have these aircrafts, which are usually
smaller and safer than conventional helicopters in which the size and energy of the
main rotor are relevant problems for many applications. The payload of commercial
multi-rotor systems is significantly lower, reaching only hundreds of grams or few
kilograms. Some configuration of multi-rotor systems, such as the multi-directional
thrusters (see Fig. 3b), increases the maneuverability but the aerodynamic efficiency
is lower decreasing the time of flight and payload.
A UAV consists of the airframe, the propulsion system and the autopilot, which
includes the guidance, navigation and low level control to generate the actuation with
the control surfaces and the propulsion system. The Unmanned Aerial System (UAS)
includes the UAV, the ground station and the payload to perform the applications.
The decisional autonomy of the UAS is a key aspect in many applications. There
are many possible levels of intervention of a human operator. The lowest level is the
6 A. Ollero and B. Siciliano

pilot of the conventional remotely piloted vehicles, who, by using conventional hand
controllers, provides control signals to the servo-controllers of the control surfaces or
of the propellers. The next level is the manual guidance. In this case the pilot provides
the direction of motion or linear velocities relying on the automatic stabilization of
the aircraft which greatly simplifies the control for the human operator. The next
abstraction level in the intervention hierarchy is the operator definition of trajectories
or sequences of way points in the ground station that are executed autonomously.
Finally, there are systems in which the operator only has to define missions. In this
case the plans are generated automatically by means of task and trajectory planning
software.
Aerial robots have some of the above mentioned autonomy levels. Thus, they have
environment perception, reactivity and planning capabilities to detect and avoid unex-
pected obstacles, and also to plan trajectories, tasks or even missions in complex envi-
ronments. This includes autonomous navigation and guidance in Global Navigation
Satellite Systems (GNSS) denied environments. Thus, Simultaneous Localization
And Mapping (SLAM) methods have been successfully applied with aerial robots.
Also autonomous target tracking, by using perception and trajectory generation,
eventually including obstacle detection and avoidance, has been implemented.
Multi-robot systems require the implementation of autonomous coordination to
share common resources, such as the aerial space, and cooperation to achieve a
common objective [2]. This includes intentional cooperation of multiple aerial robots
for applications such as surveillance, swarm intelligence based on the interaction with
the environment, formation control and even multiple robots with physical interaction
as required to transport jointly a single load.
The implementation of autonomous functionalities can be performed in different
ways involving, or not, the approval of an human operator.
The intervention of human pilots or operators could be very difficult without
helping aids. Thus, piloted flights near obstacles could be very hard if the pilot is not
very close the aircraft and there are not functionalities to help them.
The coordination or cooperation of multiple pilots in multi-robot systems is
another example of the difficulties involved in manual piloted or purely tele-operated
systems.

3 Physical Interaction of Unmanned Aerial Vehicles

The physical interaction of unmanned aerial vehicles while flying is relevant for
many applications [3]. The following kinds of interactions can be distinguished:
• Interaction with other flying objects such as needed for refueling (see Fig. 4a), joint
slung load transportation of a single object by using several aircrafts with small
payload capabilities [4] (Fig. 4b), capture of drones for security reasons, docking
of aircrafts and spacecraft, and on-orbit satellite maintenance and repairing.
• Interaction with ground objects including:
Introduction 7

Fig. 4 Physical interaction: a Air-to-Air refueling at GRVC-Univ. Sevilla; b Slung load joint load
transportation in the FP6 AWARE project; c GRVC-Univ. Sevilla picking mobile targets; d Landing
on a mobile platform with a tether in the FP7 EC-SAFEMOBIL project

– Interaction with ground mobile objects such as picking mobile targets (see
Fig. 4c), or landing on mobile platforms (Fig. 4d).
– Interaction with ground fixed objects involved in applications such as taking
samples, cleaning, contact inspection and manipulation in general.

The physical interaction of unmanned aerial vehicles while flying poses several
problems such as:

• Stability of the aircraft subject to forces and torques generated in the interactions.
• Accuracy in trajectory tracking, positioning with respect to the targets, grasping
and even to move in physical contact with objects.
• Consideration of aerodynamic perturbations due to the proximity of surfaces.
• Payload required to carry devices for physical interaction.
• Required flight time for moving in large areas, positioning and accurate
interventions.
• Reactivity to cancel the effect of wind perturbations flying close to objects.
• Planning taking into account the constraints involved in the physical interactions.

The above problems are not independent but there are relations between them.
Thus, for example, the aerodynamic perturbations also affect the stability of the
8 A. Ollero and B. Siciliano

aircraft and the accuracy; the payload also affect the time of flight and the agility
needed to provide reactivity in cluttered environment; also the planning can be related
to the dynamic properties, accuracy and aerodynamic effects.
In this book we are mainly concerned with aerial robots physically interacting
with ground fixed objects.

4 Robotic Manipulation While Flying

Robotic manipulation while flying can be considered as a particular case of flying

robots physically interacting with the environment.
The first research and development works were published by the beginning of
this decade and included the grasping and transportation of objects by using heli-
copters [5] and quadrotors [6, 7]. Most aerial robotic manipulations were performed
indoor. This was also the case of the AIROBOTS FP7 project that developed aerial
manipulator prototypes with few Degrees of Freedom (DoFs) [8] capable of exerting
forces on the environment.
In the ARCAS FP7 project (http://www.arcas-project.eu), devoted to Aerial
Robotics Cooperative Assembly Systems, the first aerial robots with 6 and 7 Degrees
of Freedom (DoF) robotic arms were developed and experimented both indoors and
outdoors [9].
Figure 5 shows the evolution of indoor aerial manipulators in ARCAS, from simple
platforms with 2 DoF arm to complex configurations with 6 DoF arms for assembly
tasks, which have been designed at CATEC to decrease the weight. Notice that multi-
rotor systems have 4 pairs of rotors. This configuration has lower efficiency than the
one with 8 rotors with parallel axis, but it is more compact.
Figure 6 shows the evolution of outdoor aerial robotic manipulators starting with
an electric helicopter with a simple grasping device continued by two lines of multi-
rotor systems at the University of Seville and helicopters at DLR. The arms in the
last multi-rotor configurations have 7 DoF and weight 2 kg. The first helicopter has a
conventional configuration while the second and third have the Flettner configuration
(see Fig. 2c). The helicopters have the Light Weight Robot (LWR) DLR-KUKA 7
DoF robotic arm.
The above mentioned DoFs provide accommodation and dexterity, needed to can-
cel unavoidable perturbations, present in aerial manipulation, particularly outdoors
due to wind perturbations.
The number of research and development works increased exponentially in the
last years with a large number of publications in main robotics and unmanned aerial
system conferences. Also the applications have started recently with very promising
results.
In this book we consider robotic manipulation in a broad sense. Thus, it integrates
concepts and technologies coming from unmanned aerial systems and robotic manip-
ulation including dexterous manipulation. It includes not only kinematic, dynamics
Introduction 9

Fig. 5 Evolution of indoor aerial robotic manipulators in FP7 ARCAS

Fig. 6 Evolution of outdoor aerial robotic manipulators, based on helicopters and multi-rotors, in
FP7 ARCAS

and control but also perception, planning, design aspects, mechatronics and cooper-
ation between several aerial robotics manipulators.
The book includes relevant research and development approaches in aerial robotic
manipulation coming from the H2020 European project on AErial RObots with
10 A. Ollero and B. Siciliano

multiple ARMS and advanced inspection capabilities for inspection and maintenance
(AEROARMS) (https://www.arcas-project.eu) and the above-mentioned ARCAS
project.
The objectives of AEROARMS are [10]: (1) the development of advanced aerial
manipulator systems with multiple arms, particularly dual arms for dexterous manip-
ulation, and multi-directional thrusters; (2) the application to inspection and main-
tenance, particularly the application to contact inspection of oil and gas plants.
In addition to the research aspects, the book also includes the deployment of
real systems. It presents applications where aerial robotic manipulation will have
an important impact in the short term: inspection and maintenance and structure
assembly. Thus, for example, an innovation on aerial manipulation for industrial
contact inspection obtained the Overall Innovation Radar Prize 2017 of the Euro-
pean Commission in the 2017 Information and Communication Technologies (ICT)
Congress.
The book integrates: novel designs of aerial robots (multi-rotors and helicopters)
with arms, including modeling (Part 2); control of aerial robotic manipulators includ-
ing centralized and decentralized control approaches of the aerial platform and the
arms, force control, visual servoing and coordinated control of several aerial manip-
ulators (Part 3); perception, including localization, simultaneous localization and
mapping, detection and grasping of objects, and cooperative perception with several
aerial robots (Part 4); planning including mission planning, task planning, off-line
and on-line motion planning, and reactivity (Part 5); and finally experimentation and
validation in structure assembly and inspection and maintenance (Part 6).

5 Conclusions

This chapter has introduced Aerial Robotic Manipulation. The chapter firstly intro-
duced Unmanned Aerial Vehicles and Unmanned Aerial Systems. Aerial robotics has
also been discussed by mentioning the main problems and advances. Then, aerial
robots physically interacting with the environment have been focused pointing out
several modalities of these interactions. Finally, aerial robotic manipulation has been
considered as a particular case of physical interaction.
After few years of accelerated evolution of aerial robotic manipulation, two ten-
dencies can be identified. The first consists of aerial robots with relatively simple
devices for grasping and other physical interactions with the environment. The lim-
itation in these simple devices to cancel perturbations while manipulating could be
compensated by using the flying base. Thus, for example, these perturbations can
be canceled by using the agility of omni-directional platforms provided by multi-
directional thrusters.
On the other hand, more complex arms with 6 or 7 DoFs can be used to cancel
the perturbations providing the required accommodation even with more conven-
tional aerial platforms. Additionally, these arms could also be used for dexterous
manipulation.
Introduction 11

The next chapters will detail particular solutions to implement aerial robotics
manipulation. Also the perception and planning functionalities for aerial manipula-
tion will be presented. These chapters will combine indoor and outdoor experiments
to show the validity of the proposed methods. Finally, the last part of the book,
analyses two relevant applications: structure assembly and contact inspections.

References

1. Valavanis, K.P., Vachtsevanos, G.J.: Handbook of Unmanned Aerial Vehicles. Springer Publish-
ing Company, Berlin (2014). Incorporated
2. Maza, I., Ollero, A., Casado, E., Scarlatti, D.: Classification of Multi-UAV Architectures. Hand-
book of Unmanned Aerial Vehicles. Springer, Netherlands (2015)
3. Kondak, K., Ollero, A., Maza, I., Krieger, K., Albu-Schaeffer, A., Schwarzbach, M., Laiacker,
M.: Unmanned aerial systems physically interacting with the environment: load transportation,
deployment, and aerial manipulation. Handbook of Unmanned Aerial Vehicles, pp. 2755–2785.
Springer, Netherlands (2015)
4. Bernard, M., Kondak, K., Maza, I., Ollero, A.: Autonomous transportation and deployment with
aerial robots for search and rescue missions. J. Field Robot. 28(6), 914–931 (2011)
5. Pounds, P.E.I., Bersak, D.R., Dollar, A.M.: Grasping from the air: hovering capture and load
stability. In: 2011 ICRA, May 2011, pp. 2491–2498 (2011)
6. Mellinger, D., Kumar, V.: Minimum snap trajectory generation and control for quadrotors. In:
Proceedings of the IEEE ICRA (2011)
7. Korpela, C., Orsag, M., Danko, T., Kobe, B., McNeil, C., Pisch, R., Oh, P.: Flight stability in
aerial redundant manipulators. In: 2012 ICRA, May 2012, pp. 3529–3530 (2012)
8. Fumagalli, M., Naldi, R., Macchelli, A., Forte, F., Keemink, A.Q.L., Stramigioli, S., Carloni,
R., Marconi, L.: Developing an aerial manipulator prototype: physical interaction with the envi-
ronment. IEEE Robot. Autom. Mag. 21(3), 41–50 (2014)
9. Jimenez-Cano, A.E., Martin, J., Heredia, G., Ollero, A., Cano, R.: Control of an aerial robot
with multi-link arm for assembly tasks. In: 2013 ICRA, May 2013, pp. 4916–4921 (2013)
10. Ollero, A., Heredia, G., Franchi, A., Antonelli, G., Kondak, K., Sanfeliu, A., Viguria, A.,
Martinez-de Dios, J.R., Pierri, F., Cortes, J., Santamaria-Navarro A., Trujillo Soto, M-A., Bal-
achandran, R., Andrade-Cetto, J., Rodriguez A.: The AEROARMS project: Aerial robots with
advanced manipulation capabilities for inspection and maintenance. IEEE Robot. Autom. Mag.
25(4), 12–23 (2018)
Another random document with
no related content on Scribd:
worse fortune, critical as well as pecuniary, in his later days. And in
the very year of the death there appears an egregious work—
extremely popular, maleficently powerful beyond all doubt throughout
the eighteenth century, and now chiefly known to non-experts in our
days by the humorous contradiction which gave its author’s name to
Shelley, and by the chance which made a literary connection,
towards the very end of its period of influence, between three such
extraordinarily assorted persons as Afra Behn, Bysshe himself, and
William Blake.[554]
Edward Bysshe’s Art of English Poetry[555] puts the eighteenth-
century theory of this art with a rigour and completeness which can
only be attributed either to something like genius, or to a wonderful
Bysshe’s Art and complete absence of it. His Rules for Making
of English English Verse are the first part of the book in order,
Poetry. but much the least in bulk. Then follow, first a
collection of “the most natural and sublime thoughts of the best
English poets,” or, in other words, an anthology, reasoned under
headings, from poets of the seventeenth century, extending to about
four hundred and fifty pages; and last a Dictionary of Rhymes. The
“best English poets” may be useful to give in a note.[556] The
Dictionary is preceded by a few prefatory remarks, including one
important historically, “Rhyme is by all allowed to be the chief
ornament of versification in the modern languages.” The killing frost
which had fallen on the flowers of Elizabethan poetry had killed one
weed at any rate—the craze against rhyme.
The Rules are preceded by a partly apologetic Preface, which
disclaims any wish to furnish tools to poetasters, and puts the work
“under the awful guard of the immortal Shakespeare, Milton [note
that this was before Addison’s critique], Dryden, &c.” The keynote is
struck, in the very first sentence of the text, with that
uncompromisingness which makes one rather admire Bysshe. “The
Structure of our verses, whether blank or in rhyme, consists in a
certain number of syllables; not in feet composed of long and short
syllables, as the verse of the Greeks and Romans.” And he adds
that, though some ingenious persons formerly puzzled themselves in
prescribing rules for the quantity of English syllables, and composed
verses by the measure of dactyls and spondees, yet that design is
now wholly exploded. In other words, he cannot conceive classical
feet without classical arrangement of feet.
“Our poetry admits, for the most part, of but three sorts of verses,
those of 10, 8, and 7 syllables. Those of 4, 6, 9, 11, 12, and 14 are
generally employed in masks and operas.” But 12 and 14 may be
used in Heroic verse with grace. Accent must be observed; and the
Pause must be at or near the middle, though in Heroics it may be at
the 3rd, 4th, 5th, 6th, or 7th syllable, determined by the seat of the
accent. Still, pauses at the 3rd and 7th must be used sparingly. The
2nd and 8th “can produce no true harmony”; and he seems to have
refused to contemplate anything so awful as a pause at the 1st or
9th. After decasyllables, octosyllables are commonest. As for lines of
9 and 11 syllables, “with the accent on the last [i.e., anapæstic
measures], the disagreeableness of their measure has wholly
excluded them from serious subjects.” The refining effected since the
days of Chaucer, Spenser, and other ancient poets consists
especially in the avoidance of the concourse of vowels and in the
rigid elision of the article, the contraction of preterperfect tenses
(“amaz’d,” not “amazed”), the rejection of alliteration (an instance in
Dryden is apologised for), of splitting words closely connected at the
end of a verse, and of polysyllables.
And a very large number of minute rules follow, the one guiding
principle of which is to reduce every line to its syllabic minimum,
never allowing trisyllabic substitution.
The book, base and mechanical as it may seem, is of the first
historical importance. It will be seen, even from these few extracts,
that the excellent Bysshe has no doubts, no half-lights. The idea,
which we have seen crystallising for a century and a half, that
English poetry is as strictly and inexorably syllabic as French, and
much more so than Greek or Latin, is here put in its baldest crudity.
Bysshe will have no feet at all: and no other division within the line
but at the pause, which is to be as centripetal as possible, like the
French cæsura. It follows from this that, except the feminine or
double ending, which is allowed ostensibly as a grace to rhymes,
though also in blank verse, nothing extra to the ten, the eight, or
whatever the line-norm may be, is permitted on any account.
Articles, prepositions that will stand it, pronouns, are to be rigidly
elided; weak or short syllables in the interior of words must be
slurred out. There is (only that Bysshe will not have even the name
of foot) no room for a trisyllabic foot anywhere, in what he equally
refuses to call iambic or trochaic verse.
But what is more startling still is that trisyllabic feet disappear, not
merely from the octosyllable and the heroic, but from English
prosody, or are admitted only to “Compositions for Musick and the
lowest sort of burlesque.” Dryden might have written, “After the
pangs of a desperate lover”; Prior might be writing “Dear Chloe, how
blubbered is that pretty face”: but Bysshe sternly averts his face from
them.
Now, if this astonishing impoverishment of English poetry had
been the isolated crotchet of a pedant or a poetaster, it would at
most deserve notice in a note. But it was nothing of the kind. “He,”
this insignificant person, “said it”: they went and did it. It expressed
the actual poetic practice of serious poets from Pope to Goldsmith:
and it expressed the deliberate theoretic creed of such a critic as
Johnson. The contrary practice of the great old poets was at best a
“licence,” at worst a “fault.” What had actually happened to French—
that it had been reduced to the iamb—what Gascoigne had lamented
and protested against, long before, was here threatened—or rather,
with bland ignoring, even of threat, laid down—as the unquestioned
and unquestionable law of English. The whole eighteenth century did
not, indeed, go the entire length of Bysshe. Prior—it is his
everlasting glory in English poetical history—took care of that, and
not only saved anapæstic cadence for us, but made it more popular
than ever. But the eighteenth century continued, charmingly as it
wrote them, to be a little ashamed of its anapæsts, to write them
affectedly as a relaxation, if not even a derogation—to indulge in
them (just as it might indulge in leap-frog with wig and long-skirted
coat laid aside) avowedly for a frolic. And about the decasyllable—
not quite so rigidly about the octosyllable—it accepted Bysshe
almost without a protest. All the infinite variety of true English
prosody, all the gliding or melting trochees, all the passion and throb
which trisyllabic feet give to iambic verse, were sacrificed, all
freedom of pause was relinquished, and the decasyllable tramped,
the octosyllable tripped, as regularly and as monotonously as a High
Dutch grenadier or a Low Dutch clock.
Bysshe had been frankly formal; it is not a small merit in him that
he knew what he had to do and did it: but persons who were little if
at all above him in taste or in intellect affected to despise him for this,
and Mr Charles Gildon in his Complete Art of Poetry,[557] published a
Gildon. few years later, is very high and mighty with Bysshe.
As for himself he does not think that Poetry consists
even in “colouring,” but in Design: and he hashes up his French
originals into some would-be modish dialogues, in which ladies of
fashion attack and defend poetry on the old lines, before he comes
to minuter recommendations. These differ chiefly from Bysshe’s in
that they are wordier, less peremptory, and given to substitute the
vagueness of the journalist for the precision of the schoolmaster. Nor
was this by any means Gildon’s only contribution to criticism. Among
the others perhaps the most interesting is an anonymous and
undated, but apparently not doubtful, rifacimento of Langbaine,[558]
which is curious as an example of peine du talion. Gildon (who has
employed his own or some other “careful hand” to give himself an
ingeniously, because not extravagantly, complimentary notice in the
Appendix) serves Langbaine in Langbaine’s own fashion; and, not
contented with reversing his judgments, indulges freely in such
phrases as “Mr Langbain mistakes,” “those scurrilous and digressory
remarks with which Mr Langbain has bespattered him [Dryden],” &c.
The book is in the main bibliographic and biographic rather than
critical.
A name which has something to do with criticism, and which
associates itself naturally with those of Dennis and Gildon in the
regiment of Pope’s victims, is that of Leonard Welsted, who in 1712
published a translation of Longinus, “with some remarks on the
Welsted. English Poets.” Welsted’s translation, whether made
directly from the Greek or not,[559] is readable
enough, and his alternative title, “A treatise on the Sovereign
Perfection of Writing,” is not unhappy. Neither are his Preface and
his appended “Remarks” contemptible. He can appreciate not merely
Milton but Spenser; is (how unlike Rymer!) transported with Othello,
and especially with its conclusion; and if he is not superior to others
in scorning “Latin rhymes,” at least has sufficient independence to be
very irreverent to Buchanan.
But there was a contemporary of Bysshe’s, more famous than
either Gildon or Welsted, whose soul was equally above mere
prosodic precept, and to whom, as it happens, Gildon himself pays a
compliment, as to a denizen of Grub Street, of whom Grub Street
could not but feel that he did it some honour by herding with its more
native and genuine population. Of him we must say something—not,
as we might almost have said it, in juxtaposition with the great poet
and critic whom he had earlier admired, but before coming to the
lesser, but still great, successors of Dryden, with whom he came into
collision in his evil days.
If John Dennis had been acquainted with the poetry of Tennyson
(at which he would probably have railed in his best manner, in which
he would certainly have detected plagiarisms from the classics), he
too might have applied to himself the words of Ulysses, “I am
Dennis. become a name.” Everybody who has the very
slightest knowledge of English literature knows, if
only in connection with Dryden, Addison, and Pope, the surly,
narrow, but not quite ignorant or incompetent critic, who in his
younger and more genial days admired the first, and in his soured
old age attacked the second and third. But it may be doubted
whether very many persons have an acquaintance, at all extensive,
with his works. They were never collected; the Select Works of John
Dennis[560] mainly consist of his utterly worthless verse. Much of the
criticism is hidden away in prefaces which were seldom reprinted,
and the original editions of which have become very rare. Even good
libraries frequently contain only two or three out of more than a
dozen or a score of separate documents: and though the British
Museum itself is well furnished, it is necessary to range through a
large number of publications to obtain a complete view of Dennis as
a critic.
That view, when obtained, may perhaps differ not a little from
those which have, in a certain general way, succeeded each other in
current literary judgment. During the reign of Pope and Addison, the
scurrilous assailant of the first, and the more courteous but in part
severe censor of the second, was naturally regarded as at best a
grumbling pedant, at worst a worthless Zoilus. The critics of the
Romantic school were not likely to be much attracted by Dennis.
More recently, something of a reaction has taken place in his favour;
and it has become not unusual to discover in him, if not exactly a
Longinus or a Coleridge, yet a serious and well-equipped critic, who
actually anticipated not a little that after-criticism has had to say.[561]
That this more charitable view is not entirely without foundation
On Rymer. may be at once admitted. As compared with Rymer,
in whose company he too often finds himself in
modern appreciation, Dennis shows, indeed, pretty well. He very
seldom—perhaps nowhere—exhibits that crass insensibility to poetry
which distinguishes “the worst critic who ever lived.” One of his
earliest and not his worst pieces, The Impartial Critic of 1693, is an
answer to Rymer himself, points out with acuteness and vigour that
“Tom the Second” would ruin the English stage if he had his way,
and even approaches the sole causeway of criticism across the deep
by advancing the argument that the circumstances of the Greek
drama were perfectly different from those of the English.[562] Yet
already there are danger-signals. That the piece (which includes a
Letter to a Friend and some dialogues) contains a great deal of
clumsy jocularity, does not much matter. But when we find Dennis
devoting some of this jocularity to Antigone’s lamentation over her
death unwedded, we feel sadly that the man who can write thus is
scarcely to be trusted on the spirit of poetry. And the admission that
Rymer’s censures of Shakespeare are “in most of the particulars
very sensible and just” is practically ruinous.[563]
Dennis’s answer to Collier is a little later,[564] but still earlier than
most of his better known work; and it is very characteristic of his
manner, which has not often, I think, been exactly described. As
elsewhere, so in this tract, which is entitled The Usefulness of the
Stage to the Happiness of Mankind, to Government and to Religion,
Dennis is uncompromisingly ethical; but he had here the excuse that
Collier, to whom he was replying, had taken the same line. There is
less, either here or elsewhere, for his method. This is to make a loud
clatter of assertions, arranged in a kind of pseudological order, which
seems to have really deceived the author, and may possibly have
deceived some of his readers, into believing it syllogistic and
conclusive. Dennis is very great at the word “must.” “As Poetry is an
Art it must be an imitation of nature”[565] and so forth; seldom shall
you find so many “musts” anywhere as in Dennis, save perhaps in
some of his modern analogues. Like all who argue in this fashion, he
becomes unable to distinguish fact and his own opinion. Collier, for
instance, had quoted (quite correctly) Seneca’s denunciation of the
Stage. To which Dennis replies, “It is not likely that Seneca should
condemn the drama, ... since ... he wrote plays himself.” That the
identity of the philosopher and the dramatist is not certain does not
matter: the characteristic thing is the setting of probability against
fact. But with Dennis hectoring assertion is everything. “It cannot
possibly be conceived that so reasonable a diversion as the drama
can encourage or incline men to so unreasonable a one as gaming
or so brutal a one as drunkenness.” With a man who thinks this an
argument, argument is impossible.
The fact is that, though he has, as has been admitted, a certain
advantage over Rymer, Lord Derby’s observation that “He never
knew whether it was John or Thomas who answered the bell” will too
On often apply here. Rymer himself was not ignorant;
Shakespeare. Dennis, especially in regard to ancient criticism, was
still better instructed: and though both were bad dramatists, with, in
consequence, a conscious or unconscious bias on dramatic matters,
Dennis was not so bad as Rymer. His devotion to Dryden does him
credit, though we may suspect that it was not the best part of Dryden
that he liked: and, amid the almost frantic spite and scurrility of his
later attacks on Pope, he not unfrequently hits a weak place in the
“young squab short gentleman’s” bright but not invulnerable armour.
Yet Dennis displays, as no really good critic could do, the
weaknesses of his time and school both in generals and particulars.
It is perfectly fair to compare him (giving weight for genius of course)
with Johnson, a critic whose general views (except on port and
claret) did not materially differ from his own. And, if we do so, we
shall find that while Johnson is generally, if not invariably, “too good
for such a breed,” Dennis almost as constantly shows its worst
features. He altered The Merry Wives of Windsor into The Comical
Gallant[566]—a most illaudable action certainly, yet great Dryden’s self
had done such things before. But he aggravated the crime by a
preface, in which he finds fault with the original as having “no less
than three actions” [would there were thirty-three!] by remarking that,
in the second part of Henry the Fourth, Falstaff “does nothing but
talk” [would he had talked so for five hundred acts instead of five!]
and by laying down ex cathedra such generalities as that “Humour,
not wit, is the business of comedy,” a statement as false as would be
its converse. In his Essay on the Genius of Shakespeare[567] he is not
so very far from Rymer himself in the drivelling arbitrariness of his
criticism. Shakespeare has actually made Aufidius, the general of
the Volscians, a base and profligate villain! Even Coriolanus himself
is allowed to be called a traitor by Aufidius, and nobody contradicts!
The rabble in Julius Cæsar and other such things “show want of Art,”
and there is a painful disregard of Poetical Justice. The same
hopeless wrong-headedness and (if I may so say) wrong-
mindedness appear in a very different work, the Remarks on the
Rape of the Lock.[568] I do not refer to Dennis’s mere scurrilities about
On “Ap—e” and the like. But part of the piece is quite
“Machines.” serious criticism. Few of us in modern times care
much for the “machinery” of this brilliantly artificial poem; but fewer
would think of objecting to it on Dennis’s grounds. Machines, it
seems, must be—
i. Taken from the religion of the Poet’s country.
ii. Allegorical in their application.
iii. Corresponding though opposed to each other.
iv. Justly subordinated and proportioned.
And Pope’s machines, we are told, fail in all these respects.
Now, putting the fourth ground aside as being a mere matter of
opinion (and some who are not fervent Papists think the machines of
the Rape very prettily and cleverly arranged in their puppet-show
way), one may ask Dennis “Who on earth told you so?” in respect of
all the others. And if he alleged (as he might) this or that sixteenth or
seventeenth century authority, “And who on earth told him so? and
what authority had the authority? Why should machines be taken
only from the religion of the country? Why should they be
allegorical? Why should Machine Dick on the one side invariably nod
to Machine Harry on the other?” And even if some sort of answer be
forthcoming, “Why should the poet not do as he please if he
succeeds thereby in giving the poetic pleasure?” To which last query
of course neither Dennis nor any of his school could return any
answer, except of the kind that requires bell, book, and candle.
Nor would he have hesitated to use this, for he is a rule-critic of
the very straitest kind, a “Tantivy” of poetic Divine Right. In his three
His general chief books of abstract criticism[569] he endeavours
theory of to elaborate, with Longinus in part for code, and with
Poetry. Milton for example, a noble, indeed, and creditable,
but utterly arbitrary and hopelessly narrow theory of poetry as
necessarily religious, and as having for its sole real end the
reformation of the mind, by a sort of enlarged Aristotelian katharsis
as to spirit, and by attention to the strict laws of the art in form.
Poetical Justice was a sort of mediate divinity to Dennis: as we have
seen, he upbraided Shakespeare for the want of it; he remonstrated,
in the Spectator, No. 548, and elsewhere, with Addison for taking too
little account of it; part at least of his enthusiasm for Milton comes
from Milton’s avowed intention to make his poem a theodicy.
A noble error! let it be repeated, with no hint or shadow of sarcasm
or of irreverence; but a fatal error as well. That Poetry, like all things
human, lives and moves and has its being in God, the present writer
believes as fervently and unhesitatingly as any Platonic philosopher
or any Patristic theologian; and he would cheerfully incur the wrath of
Savonarola by applying the epithet “divine,” in its fullest meaning, not
merely to tragedy and epic and hymn, but to song of wine and of
love. But this is not what Dennis meant at all. He meant that Poetry
is to have a definitely religious, definitely moral purpose—not that it
is and tends of itself necessarily ad majorem Dei gloriam, but that we
are to shape it according to what our theological and ethical ideas of
the glory of God are. This way easily comes bad poetry, not at all
easily good; and it excludes poetic varieties which may be as good
as the best written in obedience to it, and better. Moreover, putting
Dennis’s notion of the end of Poetry together with his notion of its
method or art (which latter is to be adjusted to some at least of the
straitest classical precepts), we can easily comprehend, and could
easily have anticipated, the narrow intolerance and the hectoring
pedantry which he shows towards all who follow not him. In a new
sense—not so very different from the old mediæval one, though put
with no mediæval glamour, and by an exponent full of eighteenth-
century prosaism, yet destitute of eighteenth-century neatness and
concinnity—Poetry becomes a part of theology; and the mere
irritableness of the man of letters is aggravated into the odium
theologicum. Bad poets (that is to say, bad according to Dennis) are
not merely faulty artists but wicked men; of this Dennis is sure. “And
when a man is sure,” as he himself somewhere naïvely observes,
“’tis his duty to speak with a modest assurance.” We know, from
examples more recent than poor Dennis, that, when a man is thus
minded, his assurance is very apt to eat up his modesty, taking his
charity, his good manners, and some other things, as condiments to
the meal.
Dennis and Addison, though the latter did not escape the absolute
impartiality of the former’s carping, were on terms of mutual respect
which, considering all things, were creditable to both. During the
Addison. latter part of his rather short lifetime Addison, it is
hardly necessary to say, enjoyed a sort of mild
dictatorship in Criticism as in other departments of literature; and his
right to it was scarcely disputed till near the close of the century,
though Johnson knew that he was not deep, and tells us that, in his
own last days, it was almost a fashion to look down on Addisonian
criticism. If, like others, he was displaced by the Romantic revival, he
received more lenient treatment than some, in virtue partly of his
own general moderation, partly of his championship of Milton. Yet
while his original literary gifts recovered high place during the
nineteenth century, his criticism has often been considered to
possess scarcely more than historic interest, and has sometimes
been rather roughly handled—for instance, by Mr Matthew Arnold.
But a recent writer,[570] by arguing that Addison’s treatment of the
Imagination, as a separate faculty, introduced a new principle into
criticism, has at any rate claimed for him a position which, if it could
be granted, would seat him among the very greatest masters of the
art, with Aristotle and Longinus among his own forerunners. As usual
let us, before discussing these various estimates, see what Addison
actually did as a critic.[571]
His début as such was not fortunate. He was, it is true, only three-
and-twenty when at “dearest Harry’s” request (that is to say Mr Harry
Sacheverell’s) he undertook an Account of the greatest English
The Account Poets.[572] In 1694 nobody, except Dryden, could be
of the Best expected to write very good verse, so that the
known poetical qualities of this verse-essay need not be
English Poets.
hardly dwelt upon, or indeed considered at all. We
may take it, as if it were prose, for the matter only. And thus
considered, it must surely be thought one of the worst examples of
the pert and tasteless ignorance of its school. Before Cowley nobody
but Chaucer and Spenser is mentioned at all, and the mentions of
these are simply grotesque. The lines convict Addison, almost
beyond appeal, of being at the time utterly ignorant of English literary
history up to 1600, and of having read Chaucer and Spenser
themselves, if he had read them at all, with his eyes shut. The
Chaucer section reads as if it were describing A C. Merry Tales or
the Jests of George Peele. Where Dryden, if he did not understand
Chaucer’s versification, and missed some of his poetry, could see
much even of that, and almost all the humour, the grace, the
sweetness, the “God’s plenty” of life and character that Chaucer has,
Addison sees nothing but a merry-andrew of the day before
yesterday.[573] So, too, the consummate art of Spenser, his exquisite
versification, his great ethical purpose, and yet his voluptuous
beauty, are quite hidden from Addison. He sees nothing but a
tedious allegory of improbable adventures, and objects to the “dull
moral” which “lies too plain below,” much as Temple had done before
him.[574] Cowley, Milton, and Waller are mentioned next, in at least
asserted chronological order. Cowley is “a mighty genius” full of
beauties and faults,

“Who more had pleased us had he pleased us less,”

but who is a perfect “milky way” of brilliancy, and has made Pindar
himself “take a nobler flight.” Milton alternately strikes Addison with
awe, rapture, and shock at his politics. He

“Betrays a bottom odious to the sight.”

So we turn to Waller, who is not only “courtly” but “moves our

passion,” (what a pity that he died too soon to "rehearse Maria’s
charms"!) to Roscommon, who “makes even rules a noble poetry,”
and Denham, whose Cooper’s Hill “we must,” of course, not “forget.”
“Great Dryden” is then, not unhappily, though not quite adequately,
celebrated, and the line on his Muse—

“She wears all dresses, and she charms in all,”

is not only neat, but very largely true. When Dryden shall decay,
luckily there is harmonious Congreve: and, if Addison were not tired
with rhyming, he would praise (he does so at some length) noble
Montague, who directs his artful muse to Dorset,

“In numbers such as Dorset’s self might use,”—

as to which all that can be said is that, if so, either the verses of
Montague or the verses of Dorset referred to are not those that have
come down to us under the names of the respective authors.
To dwell at all severely on this luckless production of a young
University wit would be not only unkind but uncritical. It shows that at
this time Addison knew next to nothing[575] about the English
literature not of his own day, and judged very badly of what he
pretended to know.
The prose works of his middle period, the Discourse on Medals
and the Remarks on Italy, are very fully illustrated from the Latin
poets—the division of literature that Addison knew best—but indulge
hardly at all in literary criticism. It was not till the launching of the
Tatler, by Steele and Swift, provided him with his natural medium of
utterance, that Addison became critical. This periodical itself, and the
less known ones that followed the Spectator, all contain exercises in
this character: but it is to the Spectator that men look, and look
rightly, for Addison’s credentials in the character of a critic. The Tatler
The Spectator Essays, such as the rather well known papers on
criticisms. Tom Folio and Ned Softly, those in the Guardian, the
good-natured puff of Tom D’Urfey, &c., are not so much serious and
deliberate literary criticisms, as applications, to subjects more or less
literary, of the peculiar method of gently malicious censorship, of
laughing castigation in manners and morals, which Addison carried
to such perfection in all the middle relations of life. Not only are the
Spectator articles far more numerous and far more weighty, but we
have his own authority for regarding them as, in some measure at
least, written on a deliberate system, and divisible into three groups.
The first of these groups consists of the early papers on True and
False Wit, and of essays on the stage. The second contains the
famous and elaborate criticism of Milton with other things; and the
third, the still later, still more serious, and still more ambitious, series
on the Pleasures of the Imagination. Addison is looking back from
the beginning of this last when he gives the general description,[576]
and it is quite possible that the complete trilogy was not in his mind
when he began the first group. But there is regular development in it,
and whether we agree or not with Mr Worsfold’s extremely high
estimate of the third division, it is quite certain that the whole
collection—of some thirty or forty essays—does clearly exhibit that
increasing sense of what criticism means, which is to be observed in
almost all good critics. For criticism is, on the one hand, an art in
which there are so few manuals or trustworthy short summaries—it
is one which depends so much more on reading and knowledge than
any creative art—and, above all, it is necessary to make so many
mistakes in it before one comes right, that, probably, not one single
example can be found of a critic of importance who was not a much
better critic when he left off than when he began.
In Group One[577] Addison is still animated by the slightly desultory
spirit of moral satire, which has been referred to above; and, though
fifteen or sixteen years have passed since the Account, he does not
seem to be so entirely free as we might wish from the crude sciolism,
On True and if not the sheer ignorance, of the earliest period. He
False Wit. is often admirable: his own humour, his taste, almost
perfect within its own narrow limits, and his good sense, made that
certain beforehand. But he has rather overloaded it with somewhat
artificial allegory, the ethical temper rather overpowers the literary,
and there is not a little of that arbitrary “blackmarking” of certain
literary things which is one of the worst faults of neo-classic criticism.
The Temple of Dulness is built (of course) “after the Gothic manner,”
and the image of the god is dressed “after the habit of a monk.”
Among the idolatrous rites and implements are not merely rebuses,
anagrams, verses arranged in artificial forms, and other things a little
childish, though perfectly harmless, but acrostics—trifles, perhaps,
yet trifles which can be made exquisitely graceful, and satisfying that
desire for mixing passion with playfulness which is not the worst
affection of the human heart.
He had led up to this batch, a few weeks earlier, by some cursory
remarks on Comedy, which form the tail of a more elaborate
examination of Tragedy, filling four or five numbers.[578] Readers who
On Tragedy. have already mastered the general drift of the
criticism of the time before him, will scarcely need
any long précis of his views, which, moreover, are in everybody’s
reach, and could not possibly be put more readably. Modern
tragedies, he thinks, excel those of Greece and Rome in the intricacy
and disposition of the fable, but fall short in the moral. He objects to
rhyme (except an end-couplet or two), and, though he thinks the
style of our tragedies superior to the sentiment, finds the former,
especially in Shakespeare, defaced by “sounding phrases, hard
metaphors, and forced expressions.” This is still more the case in
Lee. Otway is very “tender”: but it is a sad thing that the characters in
Venice Preserved should be traitors and rebels. Poetic justice (this
was what shocked Dennis), as generally understood, is rather
absurd, and quite unnecessary. And the tragi-comedy, which is the
product of the English theatre, is “one of the most monstrous
inventions that ever entered into a poet’s thought.” You “might as
well weave the adventures of Æneas and Hudibras into one poem”
[and, indeed, one might find some relief in this, as far as the
adventures of Æneas are concerned]. Tragedies are not even to
have a double plot. Rants, and especially impious rants, are bad.
Darkened stages, elaborate scenery and dresses, troops of supers,
&c., are as bad: bells, ghosts, thunder, and lightning still worse. “Of
all our methods of moving pity and terror, there is none so absurd
and barbarous as the dreadful butchering of one another,” though all
deaths on the stage are not to be forbidden.
Now, it is not difficult to characterise the criticism which appears in
this first group, strengthened, if anybody cares, by a few isolated
examples. It contains a great deal of common sense and good
ordinary taste; many of the things that it reprehends are really wrong,
and most of what it praises is good in a way. But the critic has as yet
no guiding theory, except what he thinks he has gathered from
Aristotle, and has certainly gathered from Horace, plus Common
Sense itself, with, as is the case with all English critics of this age, a
good deal from his French predecessors, especially Le Bossu and
Bouhours. Which borrowing, while it leads him into numerous minor
errors, leads him into two great ones—his denunciations of tragi-
comedy, and of the double plot. He is, moreover, essentially
arbitrary: his criticism will seldom stand the application of the “Why?”
the “Après?” and a harsh judge might, in some places, say that it is
not more arbitrary than ignorant.
The Second Group,[579] or Miltonic batch, with which may be taken
its “moon,” the partly playful but more largely serious examen of
Chevy Chase, is much the best known, and has been generally
ranked as the most important exhibition of Addison’s critical powers.
On Milton. It is not, however, out of paradox or desire to be
singular that it will be somewhat briefly discussed
here. By the student of Addison it cannot be too carefully studied; for
the historian of criticism it has indeed high importance, but
importance which can be very briefly summed up, and which
requires no extensive analysis of the eighteen distinct essays that
compose the Miltonic group, or the two on Chevy Chase. The critic
here takes for granted—and knows or assumes that his readers will
grant—two general positions:—
1. The Aristotelian-Horatian view of poetry, with a few of the more
commonplace utterances of Longinus, supplies the orthodox theory
of Poetics.
2. The ancients, especially Homer and Virgil, supply the most
perfect examples of the orthodox practice of poetry.
These things posed, he proceeds to examine Chevy Chase at
some, Paradise Lost at great, length by their aid; and discovers in
the ballad not a few, and in the epic very great and very numerous,
excellences. As Homer does this, so Milton does that: such a
passage in Virgil is a more or less exact analogue to such another in
Paradise Lost. Aristotle says this, Horace that, Longinus the third
thing; and you will find the dicta capitally exemplified in such and
such a place of Milton’s works. To men who accepted the principle—
as most, if not all, men did—the demonstration was no doubt both
interesting and satisfactory; and though it certainly did not start
general admiration of Milton, it stamped that admiration with a
comfortable seal of official orthodoxy. But it is actually more
antiquated than Dryden, in assuming that the question whether
Milton wrote according to Aristotle is coextensive with the question
whether he wrote good poetry.
The next batch is far more important.
What are the Pleasures of the Imagination? It is of the first
The moment to observe Addison’s exact definition.[580]
“Pleasures of Sight is the “sense which furnishes the imagination
the with its ideas; so that by the ‘Pleasures of the
Imagination.”
Imagination’ or Fancy, which I shall use
promiscuously, I here mean such as arise from visible objects, either
when we have them actually in our view, or when we call up their
ideas into our minds by paintings, statues, descriptions, or any the
like occasion.” We can have no images not thus furnished, though
they may be altered and compounded by imagination itself. To make
this quite sure, he repeats that he means only such pleasures as
thus arise. He then proceeds, at some length, to argue for the
innocence and refinement of such pleasures, their usefulness, and
so on; and further, to discuss the causes or origins of pleasure in
sight, which he finds to be three—greatness, uncommonness, and
beauty. The pleasantness of these is assigned to such and such
wise and good purposes of the Creator, with a reference to the great
modern discoveries of Mr Locke’s essay.
Addison then goes on to consider the sources of entertainment to
the imagination, and decides that, for the purpose, art is very inferior
to nature, though both rise in value as each borrows from the other.
He adduces, in illustration, an odd rococo mixture of scene-painting
and reflection of actual objects which he once saw (p. 404). Italian
and French gardens are next praised, in opposition to the old formal
English style, and naturally trained trees to the productions of the ars
topiaria; while a very long digression is made to greatness in
Architecture, illustrated by this remark (p. 409), “Let any one reflect
on the disposition of mind in which he finds himself at his first
entrance into the Pantheon at Rome, ... and consider how little in
proportion he is affected with the inside of a Gothic cathedral, though
it be five times larger than the other,” the reason being “the
greatness of the manner in the one, and the meanness in the other.”
So the “secondary” pleasures of the imagination—i.e., those
compounded and manufactured by memory—are illustrated by the
arts of sculpture and painting, with a good passage on description
generally, whence he turns to the Cartesian doctrine of the
association of ideas, and shows very ingeniously how the poet may
avail himself of this. Next comes a curious and often just analysis of
the reasons of pleasure in description—how, for instance, he likes
Milton’s Paradise better than his Hell, because brimstone and
sulphur are not so refreshing to the imagination as beds of flowers
and wildernesses of sweets. Or we may like things because they
“raise a secret ferment in the mind,” either directly, or so as to arouse
a feeling of relief by comparison, as when we read of tortures,
wounds, and deaths. Moreover, the poet may improve Nature. Let
oranges grow wild, and roses, woodbines, and jessamines flower at
the same time. As for “the fairy way of writing”[581]—that is to say, the
supernatural—it requires a very odd turn of mind. We do it better
than most other nations, because of our gloominess and melancholy
of temper. Shakespeare excels everybody else in touching “this
weak superstitious part” of his reader’s imagination. The glorifying of
the imagination, however, is by no means confined to the poet. In
good historians we “see” everything. None more gratify the
imagination than the authors of the new philosophy, astronomers,
microscopists. This (No. 420) is one of Addison’s most ambitious
passages of writing, and the whole ends (421) with a peroration
excellently hit off.
It is upon these papers mainly that Mr Worsfold[582] bases his high
eulogium of Addison as “the first genuine critic,” the first “who added
something to the last word of Hellenism,” the bringer of criticism “into
line with modern thought,” the establisher of “a new principle of
poetic appeal.” Let us, as uncontroversially as possible, and without
laying any undue stress on the fact that Mr Worsfold practically omits
Longinus altogether,[583] stick, in our humdrum way, to the facts.
In the first place, supposing for the moment that Addison uses
“imagination” in our full modern sense, and supposing, secondly, for
the moment also, that he assigns the appeal to the imagination as
the special engine of the poet, is this an original discovery of his? By
no means: there are many loci of former writers to negative this—
there is one that is fatal. And this is no more recondite a thing than
the famous Shakespearian description of

“The lunatic, the lover, and the poet,”

as

“Of imagination all compact,”

with what follows. But this is a mere question of property, plagiarism,

suggestion; and such questions are at best the exercises of literary
holiday-makers, at the worst the business of pedants and of fools.
A more important as well as a more dangerous question is this.
Does Addison make “the appeal to the imagination” the test of
poetry? It can only be answered that, by his own explicit words, he
does nothing of the kind. If he advances anything, it is that the
appeal to the imagination is the appeal of art generally—of prose
(even of scientific) literary art as well as of poetry, of painting,
sculpture, architecture, as well as of literature. In doing this he does
a good thing: he does something notable in the history of general
æsthetics; but in so far as literature, and especially poetry, is
concerned, he scarcely goes as far as Longinus in the well-known
passage,[584] though he works out his doctrine at much greater
length, and with assistance from Descartes and Locke.
But the most important and the most damaging question of all is
this, “Are not Addison and his panegyrist using words in equivocal
senses? Does Imagination in Addison’s mouth bear the meaning
which we, chiefly since Coleridge’s day, attach to the word? Does it
even mean what it meant to Longinus, much more what it meant to
Shakespeare?”
I have no hesitation in answering the two latter questions with an
absolute and unhesitating “No!”
It seems indeed extraordinary that, in face of Addison’s most
careful and explicit limitations, any one should delude himself into
thinking that even the Shakespearian and Addisonian Imaginations
are identical—much more that Addison’s Imagination is the supreme
faculty, creative, transcending Fancy,[585] superior to fact, not merely
compounding and refining upon, but altogether superseding and
almost scorning, ideas of sensation, which we mean by the word,
and which Philostratus or Apollonius[586] partly glimpsed. Addison
tells us—tells us over and over again—that all the ideas and
pleasures of the imagination are pleasures of sense, and, what is
more, that they are all pleasures of one sense—Sight. Why he
should have limited himself in this singular manner it is hard to say;
except that he was evidently full of Locke when he wrote, and,
indeed, almost entirely under the influence of the Essay. That he had
a contempt for music is elsewhere pretty evident; and this probably
explains his otherwise inexplicable omission of the supplies and
assistance given to Imagination by Hearing. His morality, as well as
old convention, excluded Touch, Taste, and Smell as low and gross,
though no candid philosophy could help acknowledging the immense
influence exercised upon Imagination by at least the first and the last
—Taste, because the most definite, being perhaps the least
imaginative of all. But the fact that he does exclude even these
senses, and still more rigidly excludes everything but Sense, is
insuperable, irremovable, ruthless. Addison may have been the first
modern critic to work out the appeal of art to the pleasures and ideas
furnished by the sense of sight. He is certainly nothing more.
But is he therefore to be ignored, or treated lightly, because of this
strange overvaluation of him? Certainly not. Though by no means a
His general critical very great critic, he is a useful, an
value. interesting, and a representative one. He
represents the classical attitude tempered, not merely by good sense
almost in quintessence, but by a large share of tolerance and
positive good taste, by freedom from the more utterly ridiculous
pseudo-Aristotelianisms, and by a wish to extend a concordat to
everything good even if it be not “faultless.” In his Account he is
evidently too crude to be very censurable: in his first group of essays
much of his censure is just. The elaborate vindication of Milton,
though now and for a long time past merely a curiosity, is again full
of good sense, displays (if not altogether according to knowledge) a
real liking for real poetic goodness, and had an inestimable effect in
keeping at least one poet of the better time privileged and popular
with readers throughout the Eighteenth Century. As for the essay on
the Pleasures of the Imagination, the fact that it has been wrongly
praised need not in the least interfere with a cordial estimate of its
real merits. It is not an epoch-making contribution to literary criticism;
it is rather one-sided, and strangely limited in range. But it is about
the first attempt at a general theory of æsthetics in English; it is a
most interesting, and a very early, example of that application of
common-sense philosophy to abstract subjects which Locke taught
to the English eighteenth century; and many of its remarks are
valuable and correct. Moreover, it did actually serve, for those who
could not, or who did not, read Longinus, as a corrective to pure
form-criticism, to Bysshe with his rigid ten syllables, to bare good
sense and conventional rule. Its Imagination was still only that which
supplies Images, and was strangely cramped besides; but it was
better than mere correctness, mere decency, mere stop-watch.
Between Addison and Pope, Steele, Atterbury, and Swift call for
Steele. notice. Steele has little for us.[587] There are few things
more curious than the almost entire abstinence from any
expression, in the slightest degree really critical, to be found in the
eulogy of Spenser, which he generously enough inserted in Sp. 540
to express “his passion for that charming author.” The numerous
friends whom he has so justly won for himself may perhaps insist
that there is criticism of the best in this very phrase; and that the
rather rash encomium on the poet’s “old words” as being “all truly
English” is balanced by the justice of the reference to his “exquisite
numbers.” But the fact is that Steele had neither the knowledge, nor
the patience, nor the coolness for critical work.
Atterbury gives rather more. He was himself a man of great
intellectual power, a scholar, an eloquent and delicate writer, and
possessed independent taste enough to admire Milton fervently at a
time when Addison had not yet made it wholly orthodox to admire
Atterbury. that poet at all, and when most Tories detested him.
But his observations on Waller[588] are the very
quintessence of pseudodoxy, as to that respectable person; and, by
a curious combination, though Waller is a rhymer confirmed and
complete, Atterbury joins with his admiration for him an antipathy to
rhyme—“this jingling kind of poetry,” “this troublesome bondage, as