Get ebook downloads in full at ebookmeta.

com

Concurrent Programming Algorithms Principles and

Foundations 1st Edition Michel Raynal

https://ebookmeta.com/product/concurrent-programming-
algorithms-principles-and-foundations-1st-edition-michel-
raynal/

OR CLICK BUTTON

DOWNLOAD NOW

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

Concurrent Programming:

Algorithms, Principles, and Foundations

Michel Raynal

Concurrent Programming:

Algorithms, Principles,

and Foundations

123

Michel Raynal

Institut Universitaire de France

IRISA-ISTIC

Université de Rennes 1

Rennes Cedex

France

ISBN 978-3-642-32026-2

ISBN 978-3-642-32027-9

(eBook)

DOI 10.1007/978-3-642-32027-9

Springer Heidelberg New York Dordrecht London

Library of Congress Control Number: 2012944394

ACM Computing Classification (1998): F.1, D.1, B.3

Springer-Verlag Berlin Heidelberg 2013

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. Exempted from
this legal reservation are brief excerpts in connection with reviews or
scholarly analysis or material supplied specifically for the purpose of
being entered and executed on a computer system, for exclusive use
by the purchaser of the work. Duplication of this publication or parts
thereof is permitted only under the provisions of the Copyright Law
of the Publisher’s location, in its current version, and permission for
use must always be obtained from Springer. Permissions for use may
be obtained through RightsLink at the Copyright Clearance Center.
Violations are liable to prosecution under the respective Copyright
Law.

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.

While the advice and information in this book are believed to be true
and accurate at the date of publication, neither the authors nor the
editors nor the publisher can accept any legal responsibility for any
errors or omissions that may be made. The publisher makes no
warranty, express or implied, with respect to the material contained
herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media
(www.springer.com)

Preface

As long as the grass grows and the rivers flow….

From American Indians

Homo sum: humani nihil a me alienum puto.

In Heautontimoroumenos, Publius Terencius (194–129 BC)

… Ce jour-là j’ai bien cru tenir quelque chose et que ma vie s’en
trouverait changée.

Mais rien de cette nature n’est définitivement acquis.

Comme une eau, le monde vous traverse et pour un temps vous

prête ses couleurs.

Puis se retire et vous replace devant ce vide qu’on porte en soi,

devant cette espèce d’insuffisance centrale de l’âme qu’il faut bien
apprendre à co

ˆ toyer, à combattre,

et qui, paradoxalement, est peut-être notre moteur le plus sûr.

In L’usage du monde (1963), Nicolas Bouvier (1929–1998)

vi

Preface

What synchronization is
A concurrent program is a program made up of several entities
(processes, peers, sensors, nodes, etc.) that cooperate to a common
goal. This cooperation is made

possible thanks to objects shared by the entities. These objects are

called con-

current objects. Let us observe that a concurrent object can be seen

as abstracting a service shared by clients (namely, the cooperating
entities).

A fundamental issue of computing science and computing

engineering consists

in the design and the implementation of concurrent objects. In order

that concurrent objects remain always consistent, the entities have
to synchronize their accesses to these objects. Ensuring correct
synchronization among a set of cooperating entities is far from being
a trivial task. We are no longer in the world of sequential

programming, and the approaches and methods used in sequential

computing are of

little help when one has to design concurrent programs. Concurrent

programming

requires not only great care but also knowledge of its scientific
foundations.

Moreover, concurrent programming becomes particularly difficult

when one has to

cope with failures of cooperating entities or concurrent objects.

Why this book?

Since the early work of E.W. Dijkstra (1965), who introduced the
mutual exclu-

sion problem, the concept of a process, the semaphore object, the

notion of a

weakest precondition, and guarded commands (among many other

contributions),

synchronization is no longer a catalog of tricks but a domain of

computing science with its own concepts, mechanisms, and
techniques whose results can be applied in many domains. This
means that process synchronization has to be a major topic of any
computer science curriculum.

This book is on synchronization and the implementation of

concurrent objects.

It presents in a uniform and comprehensive way the major results

that have been

produced and investigated in the past 30 years and have proved to

be useful from both theoretical and practical points of view. The
book has been written first for people who are not familiar with the
topic and the concepts that are presented.

These include mainly:

• Senior-level undergraduate students and graduate students in

computer science

or computer engineering, and graduate students in mathematics

who are inter-

ested in the foundations of process synchronization.

• Practitioners and engineers who want to be aware of the state-of-
the-art concepts, basic principles, mechanisms, and techniques
encountered in concurrent

programming and in the design of concurrent objects suited to

shared memory

systems.

Preface

vii

Prerequisites for this book include undergraduate courses on

algorithms and

base knowledge on operating systems. Selections of chapters for

undergraduate

and graduate courses are suggested in the section titled ‘‘How to

Use This Book’’

in the Afterword.

Content

As stressed by its title, this book is on algorithms, base principles,

and foundations of concurrent objects and synchronization in shared
memory systems, i.e., systems where the entities communicate by
reading and writing a common memory. (Such

a corpus of knowledge is becoming more and more important with

the advent of

new technologies such as multicore architectures.)

The book is composed of six parts. Three parts are more focused on
base
synchronization mechanisms and the construction of concurrent
objects, while the other three parts are more focused on the
foundations of synchronization. (A

noteworthy feature of the book is that nearly all the algorithms that
are presented are proved.)

• Part I is on lock-based synchronization, i.e., on well-known

synchronization

concepts, techniques, and mechanisms. It defines the most

important synchro-

nization problem in reliable asynchronous systems, namely the

mutual exclusion

problem (Chap. 1). It then presents several base approaches which

have been proposed to solve it with machine-level instructions
(Chap. 2). It also presents traditional approaches which have been
proposed at a higher abstraction level to solve synchronization
problems and implement concurrent objects, namely the

concept of a semaphore and, at an even more abstract level, the

concepts of

monitor and path expression (Chap. 3).

• After the reader has become familiar with base concepts and
mechanisms suited

to classical synchronization in reliable systems, Part II, which is

made up of a single chapter, addresses a fundamental concept of
synchronization; namely, it

presents and investigates the concept of atomicity and its properties.

This allows for the formalization of the notion of a correct execution
of a concurrent program in which processes cooperate by accessing
shared objects (Chap. 4).

• Part I has implicitly assumed that the cooperating processes do not

fail. Hence, the question: What does happen when cooperating
entities fail? This is the main

issue addressed in Part III (and all the rest of the book); namely, it
considers that cooperating entities can halt prematurely (crash
failure). To face the net effect of asynchrony and failures, it
introduces the notions of mutex-freedom and associated progress
conditions such as obstruction-freedom, non-blocking, and wait-

freedom (Chap. 5).

viii

Preface

The rest of Part III focuses on hybrid concurrent objects (Chap. 6),
wait-free implementations of paradigmatic concurrent objects such
as counters and store-collect objects (Chap. 7), snapshot objects
(Chap. 8), and renaming objects

(Chap. 9).

• Part IV, which is made up of a single chapter, is on software

transactional

memory systems. This is a relatively new approach whose aim is to

simplify the

job of programmers of concurrent applications. The idea is that

programmers

have to focus their efforts on which parts of their multiprocess

programs have to be executed atomically and not on the way
atomicity has to be realized (Chap. 10).

• Part V returns to the foundations side. It shows how reliable

atomic read/write registers (shared variables) can be built from non-
atomic bits. This part consists of three chapters. Chapter 11
introduces the notions of safe register, regular register, and atomic
register. Then, Chap. 12 shows how to build an atomic bit from a
safe bit. Finally, Chap. 13 shows how an atomic register of any size
can be built from safe and atomic bits.

This part shows that, while atomic read/write registers are easier to
use than safe read/write registers, they are not more powerful from
a computability point-of-view.

• Part VI, which concerns also the foundations side, is on the

computational

power of concurrent objects. It is made up of four chapters. It first

introduces the notion of a consensus object and shows that
consensus objects are universal

objects (Chap. 14). This means that, as soon as a system provides

us with atomic read/write registers and consensus objects, it is
possible to implement in a wait-free manner any object defined from
a sequential specification.

Part VI then introduces the notion of self-implementation and shows

how atomic

registers and consensus objects can be built from base objects of

the same type

which are not reliable (Chap. 15). Then, it presents the notion of a
consensus number and the associated consensus hierarchy which
allows the computability
power of concurrent objects to be ranked (Chap. 16). Finally, the last
chapter of the book focuses on the wait-free implementation of
consensus objects from

read/write registers and failure detectors (Chap. 17).

To have a more complete feeling of the spirit of this book, the reader
can also

consult the section ‘‘What Was the Aim of This Book’’ in the
Afterword) which

describes what it is hoped has been learned from this book. Each
chapter starts

with a short presentation of its content and a list of keywords; it

terminates with a summary of the main points that have explained
and developed. Each of the six

parts of the book is also introduced by a brief description of its aim

and its

technical content.

Preface

ix

Acknowledgments

This book originates from lecture notes for undergraduate and

graduate courses on process synchronization that I give at the
University of Rennes (France) and, as an invited professor, at several
universities all over the world. I would like to thank the students for
their questions that, in one way or another, have contributed to this
book.
I want to thank my colleagues Armando Castañeda (UNAM, MX),
Ajoy Datta

(UNLV, Nevada), Achour Mostéfaoui (Université de Nantes), and

François Taiani

(Lancaster University, UK) for their careful reading of chapters of this

book.

Thanks also to François Bonnet (JAIST, Kanazawa), Eli Gafni (UCLA),

Damien

Imbs (IRISA, Rennes), Segio Rajsbaum (UNAM, MX), Matthieu Roy

(LAAS,

Toulouse), and Corentin Travers (LABRI, Bordeaux) for long

discussions on wait-

freedom. Special thanks are due to Rachid Guerraoui (EPFL), with

whom I dis-

cussed numerous topics presented in this book (and many other

topics) during the past seven years. I would also like to thank Ph.
Louarn (IRISA, Rennes) who was

my Latex man when writing this book, and Ronan Nugent (Springer)
for his

support and his help in putting it all together.

Last but not least (and maybe most importantly), I also want to
thank all the

researchers whose results are presented in this book. Without their

work, this book would not exist (Since I typeset the entire text
myself (–

for the text and xfig

for figures–), any typesetting or technical errors that remain are my
responsibility.) Michel Raynal

Professeur des Universités

Institut Universitaire de France

IRISA-ISTIC, Université de Rennes 1

Campus de Beaulieu, 35042 Rennes, France

September–November 2005 and June–October 2011

Rennes, Mont-Louis (ARCHI’11), Gdanśk (SIROCCO’11), Saint-

Philibert,

Hong Kong (PolyU), Macau, Roma (DISC’11), Tirana (NBIS’11),

Grenoble (SSS’11), Saint-Grégoire, Douelle, Mexico City (UNAM).

Contents

Part I

Lock-Based Synchronization

The Mutual Exclusion Problem. . . . . . . . . . . . . . . . . . . . . . . . . .

1.1

Multiprocess Program . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.1.1
The Concept of a Sequential Process. . . . . . . . . . . .

1.1.2

The Concept of a Multiprocess Program . . . . . . . . .

1.2

Process Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . .

1.2.1

Processors and Processes . . . . . . . . . . . . . . . . . . . .

1.2.2

Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . .

1.2.3

Synchronization: Competition. . . . . . . . . . . . . . . . .

1.2.4

Synchronization: Cooperation . . . . . . . . . . . . . . . . .

7
1.2.5

The Aim of Synchronization

Is to Preserve Invariants . . . . . . . . . . . . . . . . . . . .

1.3

The Mutual Exclusion Problem . . . . . . . . . . . . . . . . . . . . .

1.3.1

The Mutual Exclusion Problem (Mutex) . . . . . . . . .

1.3.2

Lock Object . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

1.3.3

Three Families of Solutions . . . . . . . . . . . . . . . . . .

12

1.4

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13

1.5
Other documents randomly have
different content
 X
CHARLES ELIOT NORTON
 CHARLES ELIOT NORTON
It is a tradition in the city of Cambridge, Massachusetts, that
Howells used to exult, on arriving from his Western birthplace, in
having at length met for the first time, in Charles Eliot Norton, the
only man he had ever seen who had been cultivated up to the
highest point of which he was capable. To this the verdict of all
Cambridge readily assented. What the neighbors could not at that
time foresee was that the man thus praised would ever live to be an
octogenarian, or that in doing so he would share those attractions of
constantly increasing mildness and courtesy which are so often justly
claimed for advancing years. There was in him, at an earlier period,
a certain amount of visible self-will, and a certain impatience with
those who dissented from him,—he would not have been his father’s
son had it been otherwise. But these qualities diminished, and he
grew serener and more patient with others as the years went on.
Happy is he who has lived long enough to say with Goethe, “It is only
necessary to grow old to become more indulgent. I see no fault
committed which I have not committed myself.” This milder and more
genial spirit increased constantly as Norton grew older, until it served
at last only to make his high-bred nature more attractive.
He was born in Cambridge, November 16, 1827, and died in the
very house where he was born, October 21, 1908. He was
descended, like several other New England authors, from a line of
Puritan clergymen. He was the son of Professor Andrews Norton, of
Harvard University, who was descended from the Rev. John Norton,
born in Ipswich, Massachusetts, in 1651. The mother of the latter
was the daughter of Emanuel Downing, and the niece of Governor
John Winthrop. Mrs. Bradstreet, the well-known Puritan poetess,
was also an ancestress of Charles Norton. His mother, Mrs. Caroline
(Eliot) Norton, had also her ancestry among the most cultivated
families in New England, the name of Eliot having been prominent
for successive generations in connection with Harvard College. His
parents had a large and beautiful estate in Cambridge, and were (if
my memory serves me right) the one family in Cambridge that kept a
carriage,—a fact the more impressed upon remembrance because it
bore the initials “A. & C. N.” upon the panels, the only instance I
have ever seen in which the two joint proprietorships were thus
expressed. This, and the fact that I learned by heart in childhood
Wordsworth’s poem, “The White Doe of Rylstone, or The Fate of the
Nortons,” imparted to my youthful mind a slight feeling of romance
about the Cambridge household of that name, which was not
impaired by the fact that our parents on both sides were intimate
friends, that we lived in the same street (now called Kirkland Street),
and that I went to dancing-school at the Norton house. It is perhaps
humiliating to add that I disgraced myself on the very first day by
cutting off little Charlie’s front hair as a preliminary to the dancing
lesson.
The elder Professor Norton was one of the most marked
characters in Cambridge, and, although never a clergyman, was
professor in the Theological School. It was said of him by George
Ripley, with whom he had a bitter contest, that “He often expressed
rash and hasty judgments in regard to the labors of recent or
contemporary scholars, consulting his prejudices, as it would seem,
rather than competent authority. But in his own immediate
department of sacred learning he is entitled to the praise of sobriety
of thought and profoundness of investigation” (Frothingham’s
“Ripley,” 105). He was also a man of unusual literary tastes, and his
“Select Journal of Foreign Periodical Literature,” although too early
discontinued, took distinctly the lead of all American literary journals
up to that time.
The very beginning of Charles Norton’s career would seem at first
sight singularly in contrast with his later pursuits, and yet doubtless
had formed, in some respects, an excellent preparation for them.
Graduating at Harvard in 1846, and taking a fair rank at graduation,
he was soon after sent into a Boston counting-house to gain a
knowledge of the East India trade. In 1849 he went as supercargo on
a merchant ship bound for India, in which country he traveled
extensively, and returned home through Europe in 1851. There are
few more interesting studies in the development of literary
individuality than are to be found in the successive works bearing
Norton’s name, as one looks through the list of them in the Harvard
Library. The youth who entered upon literature anonymously, at the
age of twenty-five, as a compiler of hymns under the title of “Five
Christmas Hymns” in 1852, and followed this by “A Book of Hymns
for Young Persons” in 1854, did not even flinch from printing the
tragically Calvinistic verse which closes Addison’s famous hymn,
beginning “The Lord my pasture shall prepare,” with a conclusion so
formidable as death’s “gloomy horrors” and “dreadful shade.” In 1855
he edited, with Dr. Ezra Abbot, his father’s translations of the
Gospels with notes (2 vols.), and his “Evidences of the Genuineness
of the Gospels” (3 vols.). Charles Norton made further visits to
Europe in 1855-57, and again resided there from 1868 until 1873;
during which time his rapidly expanding literary acquaintanceships
quite weaned his mind from the early atmosphere of theology.
Although one of the writers in the very first number of the “Atlantic
Monthly,” he had no direct part in its planning. He wrote to me
(January 9, 1899), “I am sorry that I can tell you nothing about the
primordia of the ‘Atlantic.’ I was in Europe in 1856-57, whence I
brought home some MSS. for the new magazine.” It appears from
his later statement in the Anniversary Number that he had put all
these manuscripts by English authors in a trunk together, but that
this trunk and all the manuscripts were lost, except one accidentally
left unpacked, which was a prose paper by James Hannay on
Douglas Jerrold, “who is hardly,” as Norton justly says, “to be
reckoned among the immortals.” Hannay is yet more thoroughly
forgotten. But this inadequate service in respect to foreign material
was soon more than balanced, as one sees on tracing the list of
papers catalogued under Norton’s name in the Atlantic Index.
To appreciate the great variety and thorough preliminary
preparation of Norton’s mind, a student must take one of the early
volumes of the “Atlantic Monthly” and see how largely he was relied
upon for literary notices. If we examine, for instance, the fifth volume
(1860), we find in the first number a paper on Clough’s “Plutarch’s
Lives,” comprising ten pages of small print in double columns. There
then follow in the same volume papers on Hodson’s “Twelve Years of
a Soldier’s Life in India,” on “Friends in Council,” on Brooks’s
“Sermons,” on Trollope’s “West Indies and the Spanish Main,” on
“Captain John Brown,” on Vernon’s “Dante,” and one on “Model
Lodging-Houses in Boston.” When we remember that his “Notes of
Travel and Study in Italy” was also published in Boston that same
year, being reviewed by some one in a notice of two pages in this
same volume of the “Atlantic,” we may well ask who ever did more of
genuine literary work in the same amount of time. This was, of
course, before he became Professor in the college (1874), and his
preoccupation in that way, together with his continuous labor on his
translations of Dante, explains why there are comparatively few
entries under his name in Atlantic Indexes for later years. Again, he
and Lowell took charge of the “North American Review” in 1864, and
retained it until 1868, during which period Norton unquestionably
worked quite as hard as before, if we may judge by the collective
index to that periodical.
It is to be noticed, however, that his papers in the “North
American” are not merely graver and more prolonged, but less terse
and highly finished, than those in the “Atlantic”; while in the
development of his mind they show even greater freedom of
statement. He fearlessly lays down, for instance, the following
assertion, a very bold one for that period: “So far as the most
intelligent portion of society at the present day is concerned, the
Church in its actual constitution is an anachronism. Much of the
deepest and most religious life is led outside its wall, and there is a
constant and steady increase in those who not only find the claims of
the Church inconsistent with spiritual liberty, but also find its services
ill adapted to their wants.... It becomes more and more a simple
assemblage of persons gathered to go through with certain formal
ceremonies, the chief of which consists in listening to a man who is
seldom competent to teach.” It must be remembered that the
expression of such opinions to-day, when all his charges against the
actual Church may be found similarly stated by bishops and doctors
of divinity, must have produced a very different impression when
made forty years ago by a man of forty or thereabouts, who occupied
twenty pages in saying it, and rested in closing upon the calm basis,
“The true worship of God consists in the service of his children and
devotion to the common interests of men.” It may be that he who
wrote these words never held a regular pew in any church or
identified himself, on the other hand, with any public heretical
organization, even one so moderate as the Free Religious
Association. Yet the fact that he devoted his Sunday afternoons for
many years to talking and Scripture reading in a Hospital for
Incurables conducted by Roman Catholics perhaps showed that it
was safer to leave such a man to go on his own course and reach
the kingdom of heaven in his own way.
Norton never wrote about himself, if it could be avoided, unless his
recollections of early years, as read before the Cambridge Historical
Society, and reported in the second number of its proceedings, may
be regarded as an exception. Something nearest to this in literary
self-revelation is to be found, perhaps, in his work entitled “Letters of
John Ruskin,” published in 1904, and going back to his first invitation
from the elder Ruskin in 1855. This was on Norton’s first direct trip to
Europe, followed by a correspondence in which Ruskin writes to him,
February 25, 1861, “You have also done me no little good,” and
other phrases which show how this American, nine years younger
than himself, had already begun to influence that wayward mind.
Their correspondence was suspended, to be sure, by their difference
of attitude on the American Civil War; but it is pleasant to find that
after ten months of silence Ruskin wrote to Norton again, if bitterly.
Later still, we find successive letters addressed to Norton—now in
England again—in this loving gradation, “Dear Norton,” “My dearest
Norton,” “My dear Charles,” and “My dearest Charles,” and
thenceforth the contest is won. Not all completed, however, for in the
last years of life Ruskin addressed “Darling Charles,” and the last
words of his own writing traced in pencil “From your loving J. R.”
I have related especially this one touching tale of friendship,
because it was the climax of them all, and the best illustration of the
essential Americanism of Norton’s career.
He indeed afforded a peculiar and almost unique instance in New
England, not merely of a cultivated man who makes his home for life
in the house where he was born, but of one who has recognized for
life the peculiar associations of his boyhood and has found them still
the best. While Ruskin was pitying him for being doomed to wear out
his life in America, Norton with pleasure made his birthplace his
permanent abode, and fully recognized the attractions of the spot
where he was born. “What a fine microcosm,” he wrote to me
(January 9, 1899), “Cambridge and Boston and Concord made in the
40’s.” Norton affords in this respect a great contrast to his early
comrade, William Story, who shows himself in his letters wholly
detached from his native land, and finds nothing whatever in his
boyhood abode to attract him, although it was always found
attractive, not merely by Norton, but by Agassiz and Longfellow,
neither of whom was a native of Cambridge.
The only safeguard for a solitary literary workman lies in the
sequestered house without a telephone. This security belonged for
many years to Norton, until the needs of a growing family made him
a seller of land, a builder of a high-railed fence, and at last, but
reluctantly, a subscriber to the telephone. It needs but little study of
the cards bearing his name in the catalogue of the Harvard Library to
see on how enormous a scale his work has been done in this
seclusion. It is then only that one remembers his eight volumes of
delicately arranged scrap-books extending from 1861 to 1866, and
his six volumes of “Heart of Oak” selections for childhood. There
were comparatively few years of his maturer life during which he was
not editor of something, and there was also needed much
continuous labor in taking care of his personal library. When we
consider that he had the further responsibility of being practically the
literary executor or editor of several important men of letters, as of
Carlyle, Ruskin, Lowell, Curtis, and Clough; and that in each case
the work was done with absolute thoroughness; and that even in
summer he became the leading citizen of a country home and
personally engaged the public speakers who made his rural festals
famous, it is impossible not to draw the conclusion that no public
man in America surpassed the sequestered Norton in steadfastness
of labor.
It being made my duty in June, 1904, to read a poem before the
Harvard Phi Beta Kappa, I was tempted to include a few verses
about individual graduates, each of which was left, according to its
subject, for the audience to guess. The lines referring to Norton were
as follows:—

“There’s one I’ve watched from childhood, free of guile,

His man’s firm courage and his woman’s smile.
His portals open to the needy still,
He spreads calm sunshine over Shady Hill.”

The reference to the combined manly and womanly qualities of

Norton spoke for itself, and won applause even before the place of
residence was uttered; and I received from Norton this recognition of
the little tribute:—

Ashfield, 2 July, 1904.

My dear Higginson,—Your friendly words about me in
your Phi Beta poem give me so much pleasure that I cannot
refrain from thanking you for them. I care for them specially as
a memorial of our hereditary friendship. They bring to mind
my Mother’s affection for your Mother, and for Aunt Nancy,
who was as dear an Aunt to us children at Shady Hill as she
was to you and your brothers and sisters. What dear and
admirable women! What simple, happy lives they led! No
one’s heart will be more deeply touched by your poem than
mine.

One most agreeable result of Norton’s Cambridge boyhood has

not been generally recognized by those who have written about him.
His inherited estate was so large that he led a life absolutely free in
respect to the study of nature, and as Lowell, too, had the same
advantage, they could easily compare notes. In answer to a criticism
of mine with reference to Longfellow’s poem, “The Herons of
Elmwood,” on my theory that these herons merely flew over
Elmwood and only built their nests in what were then the dense
swamps east of Fresh Pond, he writes to me (January 4, 1899): “I
cannot swear that I ever saw a heron’s nest at Elmwood. But Lowell
told me of their nesting there, and only a few weeks ago Mrs. Burnett
told me of the years when they had built in the pines and of the time
of their final desertion of the place.” To this he adds in a note dated
five days later: “As to the night-herons lighting on pines, for many
years they were in the habit of lighting and staying for hours upon
mine and then flying off towards the [Chelsea] beach.” This taste
accounts for the immense zest and satisfaction with which Norton
edited a hitherto unknown manuscript of the poet Gray’s on natural
history, with admirable illustrations taken from the original book,
seeming almost incredibly accurate from any but a professional
naturalist, the book being entitled, “The Poet Gray as a Naturalist
with Selections from His Notes on the Systema Naturæ of Linnæus
with Facsimiles of Some of his Drawings.”
In the Charles Eliot Norton number of the “Harvard Graduates’
Magazine” commemorating his eightieth birthday, Professor Palmer,
with that singular felicity which characterizes him, says of Norton:
“He has been an epitome of the world’s best thought brought to our
own doors and opened for our daily use.” Edith Wharton with equal
felicity writes from Norton’s well-known dwelling at Ashfield, whose
very name, “High Pasture,” gives a signal for what follows:—

“Come up—come up; in the dim vale below

The autumn mist muffles the fading trees,
But on this keen hill-pasture, though the breeze
Has stretched the thwart boughs bare to meet the snow,
Night is not, autumn is not—but the flow
Of vast, ethereal and irradiate seas,
Poured from the far world’s flaming boundaries
In waxing tides of unimagined glow.

“And to that height illumined of the mind

He calls us still by the familiar way,
Leaving the sodden tracks of life behind,
Befogged in failure, chilled with love’s decay—
Showing us, as the night-mists upward wind,
How on the heights is day and still more day.”
 But I must draw to a close, and shall do this by reprinting the very
latest words addressed by this old friend to me; these being written
very near his last days. Having been away from Cambridge all
summer, I did not know that he had been at Cambridge or ill, and on
my writing to him received this cheerful and serene answer, wholly
illustrative of the man, although the very fact that it was dictated was
sadly ominous:—

Shady Hill, Cambridge, Mass., 6 October, 1908.

My dear Higginson,—Your letter the other day from
Ipswich gave me great pleasure....
It had never occurred to me that you were associated with
Ipswich through your Appleton relatives. My association with
the old town, whose charm has not wholly disappeared under
the hard hoof of the invader, begins still earlier than yours, for
the William Norton who landed there in 1636 was my direct
ancestor; and a considerable part of his pretty love story
seems to have been transacted there. I did not know the story
until I came upon it by accident, imbedded in some of the
volumes of the multifarious publications of our historical
society. It amused me to find that John Norton, whose
reputation is not for romance or for soft-heartedness, took an
active interest in pleading his brother’s cause with Governor
Winthrop, whose niece, Lucy Downing, had won the
susceptible heart of W. N.
My summer was a very peaceful and pleasant one here in
my old home till about six weeks ago, when I was struck down
... which has left me in a condition of extreme muscular
feebleness, but has not diminished my interest in the world
and its affairs. Happily my eyes are still good for reading, and
I have fallen back, as always on similar occasions, on
Shakespeare and Scott, but I have read one or two new
books also, the best of which, and a book of highest quality, is
the last volume of Morley’s essays.
 But I began meaning only to thank you for your pleasant
note and to send a cheer to you from my slower craft as your
gallant three-master goes by it with all sails set....
Always cordially yours,
C. E. Norton.
 XI
EDMUND CLARENCE STEDMAN
 EDMUND CLARENCE STEDMAN
The sudden death of Edmund Clarence Stedman at New York on
January 18, 1908, came with a strange pathos upon the readers of
his many writings, especially as following so soon upon that of his
life-long friend and compeer, Aldrich. Stedman had been for some
years an invalid, and had received, in his own phrase, his “three
calls,” that life would soon be ended. He was born at Hartford,
Connecticut, on October 8, 1833, and was the second son of Colonel
Edmund Burke Stedman and his wife Elizabeth Clement (Dodge)
Stedman. His great-grandfather was the Reverend Aaron Cleveland,
Jr., a Harvard graduate of 1735, and a man of great influence in his
day, who died in middle life under the hospitable roof of Benjamin
Franklin. Stedman’s mother was a woman of much literary talent,
and had great ultimate influence in the training of her son, although
she was early married again to the Honorable William B. Kinney,
who was afterwards the United States Minister to Turin. Her son,
being placed in charge of a great-uncle, spent his childhood in
Norwich, Connecticut, and entered Yale at sixteen, but did not
complete his course there, although in later life he was restored to
his class membership and received the degree of Master of Arts. He
went early into newspaper work in Norwich and then in New York,
going to the front for a time as newspaper correspondent during the
Civil War. He abandoned journalism after ten years or thereabouts,
and became a member of the New York Stock Exchange without
giving up his literary life, a combination apt to be of doubtful success.
He married, at twenty, Laura Hyde Woodworth, who died before him,
as did one of his sons, leaving only one son and a granddaughter as
his heirs. His funeral services took place at the Church of the
Messiah on January 21, 1908, conducted by the Reverend Dr.
Robert Collyer and the Reverend Dr. Henry van Dyke.
Those who happen to turn back to the number of the “Atlantic
Monthly” for January, 1898, will read with peculiar interest a
remarkable paper entitled “Our Two Most Honored Poets.” It bears
no author’s name, even in the Index, but is what we may venture to
call, after ten years, a singularly penetrating analysis of both Aldrich
and Stedman. Of the latter it is said: “His rhythmic sense is subtle,
and he often attains an aerial waywardness of melody which is of the
very essence of the lyric gift.” It also remarks most truly and sadly of
Stedman that he “is of those who have suffered the stress of the
day.” The critic adds: “Just now we felt grateful to Mr. Aldrich for
putting all this [that is, life’s tragedies] away in order that the clarity
and sweetness of his art might not suffer; now we feel something like
reverence for the man [Mr. Stedman] who, in conditions which make
for contentment and acquiescence, has not been able to escape
these large afflictions.” But these two gifted men have since passed
away, Aldrich from a career of singular contentment, Stedman after
ten years of almost constant business failure and a series of
calamities relating to those nearest and dearest.
One of the most prominent men in the New York literary
organizations, and one who knew Stedman intimately, writes me
thus in regard to the last years of his life: “As you probably know,
Stedman died poor. Only a few days ago he told me that after paying
all the debts hanging over him for years from the business losses
caused by ⸺’s mismanagement, he had not enough to live on, and
must keep on with his literary work. For this he had various plans, of
which our conversations developed only a possible rearrangement of
his past writings; an article now and then for the magazines (one, I
am told, he left completed); and reminiscences of his old friends
among men of letters—for which last he had, during eight months
past, been overhauling letters and papers, but had written nothing.
He was ailing, he said—had a serious heart affection which troubled
him for years, and he found it a daily struggle to keep up with the
daily claims on his time. You know what he was, in respect of letters,
—and letters. He could always say ‘No’ with animation; but in the
case of claims on his time by poets and other of the writing class, he
never could do the negative. He both liked the claims and didn’t. The
men who claimed were dear to him, partly because he knew them,
partly because he was glad to know them. He wore himself quite out.
His heart was exhausted by his brain. It was a genuine case of
heart-failure to do what the head required.”
 There lies before me a mass of private letters to me from
Stedman, dating back to November 2, 1873, when he greeted me for
the first time in a kinship we had just discovered. We had the same
great-grandfather, though each connection was through the mother,
we being alike great-grandchildren of the Reverend Aaron
Cleveland, Jr., from whom President Grover Cleveland was also
descended. At the time of this mutual discovery Stedman was
established in New York, and although I sometimes met him in
person, I can find no letters from him until after a period of more than
ten years, when he was engaged in editing his Library of American
Literature. He wrote to me afterwards, and often with quite cousinly
candor,—revealing frankly his cares, hopes, and sorrows, but never
with anything coarse or unmanly. All his enterprises were confided to
me so far as literature was concerned, and I, being nearly ten years
older, felt free to say what I thought of them. I wished, especially,
however, to see him carry out a project of translations from the
Greek pastoral poetry of Theocritus, Bion, and Moschus. The few
fragments given at the end of his volumes had always delighted me
and many other students, while his efforts at the “Agamemnon” of
Æschylus dealt with passages too formidable in their power for any
one but Edward FitzGerald to undertake.
After a few years of occasional correspondence, there came a lull.
Visiting New York rarely, I did not know of Stedman’s business
perplexities till they came upon me in the following letter, which was
apparently called out by one of mine written two months before.

71 West 54th Street, New York, July 12th, ’82.

My Dear Colonel,—I had gone over with “the majority”
[that is, to Europe], when your friendly card of May 9th was
written, and it finally reached me at Venice. In that city of light,
air, and heavenly noiselessness, my son and myself at last
had settled ourselves in ideal rooms, overlooking the Grand
Canal. We had seclusion, the Molo, the Lagoon, and a good
café, and pure and cheap Capri wine. Our books and papers
were unpacked for the first time, and I was ready to make an
end of the big and burdensome book which I ought to have
 finished a year ago. Dis aliter visum! The next morning I was
awakened to receive news, by wire, of a business loss which
brought me home, through the new Gothard tunnel and by the
first steamer. Here I am, patching up other people’s blunders,
with the thermometer in the nineties. I have lived through
worse troubles, but am in no very good humor. Let me renew
the amenities of life, by way of improving my disposition: and
I’ll begin by thanking you for calling my attention to the error in
re Palfrey—which, of course, I shall correct. Another friend
has written me to say that Lowell’s father was a Unitarian—
not a Congregationalist. But Lowell himself told me, the other
day, that his father never would call himself a Unitarian, and
that he was old-fashioned in his home tenets and discipline.
Mr. L. [Lowell] was under pretty heavy pressure, as you know,
when I saw him, but holding his own with some composure—
for a poet. Again thanking you, I am,
Always truly yrs.,
E. C. Stedman.

This must have been answered by some further expression of

solicitude, for this reply came, two months later,—

University Club, 370 Fifth Avenue, New York, Sunday,

Sept. 16, 1883.
My dear Higginson,—There is a good deal, say what you
will, in “moral support.” I have proved it during the last few
weeks: ’twould have been hard to get through with them, but
for just such words as yours. And I have had them in such
abundance that, despite rather poor displays of human nature
in a sample of my own manufacture, I am less than ever a
pessimist.
As for that which Sophocles pronounced the father of
meanness—πενία—both my wife and myself have been used
to it nearly all our lives, and probably shall have, now, to
renew our old acquaintance with it. Though somewhat
 demoralized by a few years of Philistine comfort—the
Persicos apparatus, &c.—I think we shall get along with
sufficient dignity.
We have suffered more, however, than the money-loss, bad
as that is. And hence we are doubly grateful to those who, like
yourself, send a cheery voice to us at just this time.
Ever sincerely yrs.,
Edmund C. Stedman.

During the next few years we had ample correspondence of a

wholly literary and cheerful tone. He became engaged upon his
Library of American Literature with a congenial fellow worker, Miss
Ellen Hutchinson, and I was only one of many who lent a hand or
made suggestions. He was working very hard, and once wrote that
he was going for a week to his boyhood home to rest. During all this
period there was, no doubt, the painful business entanglement in the
background, but there was also in the foreground the literary work
whose assuaging influence only one who has participated in it can
understand. Then came another blow in the death of his mother,
announced to me as follows:—

44 East 26th St., New York, Dec. 8th, 1889.

My dear Higginson,—Yes: I have been through a kind of
Holy Week, and have come out in so incorporeal a state that I
strive painfully, though most gratefully, to render thanks to
some, at least, of my beautiful mother’s friends and mine who
have taken note of her departure. I have always wished that
she and you could know more of each other—though nothing
of yours escaped her eager taste and judgment, for she was
not only a natural critic, but a very clanswoman, with a most
loyal faith in her blood and yours. Most of all, she was a
typical woman, an intensely human one, to the last, though
made of no common clay. She was of an age to die, and I am
glad that her fine intelligence was spared a season of
 dimness. Still, I have suffered a loss, and doubtless one that
will last a lifetime.
Sincerely yours,
E. C. Stedman.

The laborious volumes of literary selections having been

completed, there followed, still under the same pressure, another
series of books yet more ambitious. His “Victorian Poets” (1875,
thirteenth edition 1887) was followed by the “Poets of America”
(1885), “A Victorian Anthology” (1895), and “An American Anthology”
(1900). These books were what gave him his fame, the two former
being original studies of literature, made in prose; and the two latter
being collections of poetry from the two nations.
If we consider how vast a labor was represented in all those
volumes, it is interesting to revert to that comparison between
Stedman and his friend Aldrich with which this paper began. Their
literary lives led them apart; that of Aldrich tending always to
condensation, that of Stedman to expansion. As a consequence,
Aldrich seemed to grow younger and younger with years and
Stedman older; his work being always valuable, but often too
weighty, “living in thoughts, not breaths,” to adopt the delicate
distinction from Bailey’s “Festus.” There is a certain worth in all that
Stedman wrote, be it longer or shorter, but it needs a good deal of
literary power to retain the attention of readers so long as some of
his chapters demand. Opening at random his “Poets of America,”
one may find the author deep in a discussion of Lowell, for instance,
and complaining of that poet’s prose or verse. “Not compactly
moulded,” Stedman says, even of much of Lowell’s work. “He had a
way, moreover, of ‘dropping’ like his own bobolink, of letting down his
fine passages with odd conceits, mixed metaphors, and licenses
which, as a critic, he would not overlook in another. To all this add a
knack of coining uncouth words for special tints of meaning, when
there are good enough counters in the language for any poet’s
need.” These failings, Stedman says, “have perplexed the poet’s
friends and teased his reviewers.” Yet Lowell’s critic is more
chargeable with diffuseness than is Lowell himself in prose essays,
which is saying a good deal. Stedman devotes forty-five pages to
Lowell and thirty-nine even to Bayard Taylor, while he gives to
Thoreau but a few scattered lines and no pretense at a chapter.
There are, unquestionably, many fine passages scattered through
the book, as where he keenly points out that the first European
appreciation of American literature was “almost wholly due to
grotesque and humorous exploits—a welcome such as a prince in
his breathing-hour might give to a new-found jester or clown”; and
when he says, in reply to English criticism, that there is “something
worth an estimate in the division of an ocean gulf, that makes us like
the people of a new planet.”
Turning back to Stedman’s earlier book, the “Victorian Poets,” one
finds many a terse passage, as where he describes Landor as a
“royal Bohemian in art,” or compares the same author’s death in
Florence at ninety, a banished man, to “the death of some monarch
of the forest, most untamed when powerless.” Such passages
redeem a book from the danger of being forgotten, but they cannot in
the long run save it from the doom which awaits too great
diffuseness in words. During all this period of hard work, he found
room also for magazine articles, always thoroughly done. Nowhere is
there a finer analysis, on the whole, of the sources of difficulty in
Homeric translation than will be found in Stedman’s review of
Bryant’s translation of Homer, and nowhere a better vindication of a
serious and carefully executed book (“Atlantic Monthly,” May, 1872).
He wrote also an admirable volume of lectures on the “Nature and
Elements of Poetry” for delivery at Johns Hopkins University.
As years went on, our correspondence inevitably grew less close.
On March 10, 1893, he wrote, “I am so driven at this season, ‘let
alone’ financial worries, that I have to write letters when and where I
can.” Then follows a gap of seven years; in 1900 his granddaughter
writes on October 25, conveying affectionate messages from him;
two years after, April 2, 1903, he writes himself in the same key, then
adds, “Owing to difficulties absolutely beyond my control, I have
written scarcely a line for myself since the Yale bicentennial [1901]”;
and concludes, “I am very warmly your friend and kinsman.” It was a
full, easy, and natural communication, like his old letters; but it was
four years later when I heard from him again as follows, in a letter
which I will not withhold, in spite of what may be well regarded as its
over-sensitiveness and somewhat exaggerated tone.

2643 Broadway, New York City, Evening, March 20th,

1907.
My dear Kinsman,—Although I have given you no reason
to be assured of it, you are still just the same to me in my
honor and affection—you are never, and you never have
been, otherwise in my thoughts than my kinsman (by your first
recognition of our consanguinity) and my friend; yes, and
early teacher, for I long ago told you that it was your essays
that confirmed me, in my youth, in the course I chose for
myself.
I am going on to Aldrich’s funeral, and with a rather lone
and heavy heart, since I began life here in New York with him
before the Civil War, and had every expectation that he would
survive me: not wholly on the score of my seniority, but
because I have had my “three calls” and more, and because
he has ever been so strong and young and debonair. Health,
happiness, ease, travel, all “things waregan,” seemed his
natural right. If I, too, wished for a portion of his felicities, I
never envied one to whom they came by the very fitness of
things. And I grieve the more for his death, because it seems
to violate that fitness.
Now, I can’t think of meeting you on Friday without first
making this poor and inadequate attempt to set one thing
right. Your latest letter—I was, at least, moved by it to address
myself at once to a full reply, but was myself attacked that day
so sorely by the grippe that I went to bed before completing it
and was useless for weeks; the letter showed me that you
thought, as well you might, that I had been hurt or vexed by
something you had unwittingly done or written. I can say little
to-night but to confess that no act, word, or writing, of yours